User login
Patient Perceptions of Readmission Risk: An Exploratory Survey
Recent years have seen a proliferation of programs designed to prevent readmissions, including patient education initiatives, financial assistance programs, postdischarge services, and clinical personnel assigned to help patients navigate their posthospitalization clinical care. Although some strategies do not require direct patient participation (such as timely and effective handoffs between inpatient and outpatient care teams), many rely upon a commitment by the patient to participate in the postdischarge care plan. At our hospital, we have found that only about 2/3 of patients who are offered transitional interventions (such as postdischarge phone calls by nurses or home nursing through a “transition guide” program) receive the intended interventions, and those who do not receive them are more likely to be readmitted.1 While limited patient uptake may relate, in part, to factors that are difficult to overcome, such as inadequate housing or phone service, we have also encountered patients whose values, beliefs, or preferences about their care do not align with those of the care team. The purposes of this exploratory study were to (1) assess patient attitudes surrounding readmission, (2) ascertain whether these attitudes are associated with actual readmission, and (3) determine whether patients can estimate their own risk of readmission.
METHODS
From January 2014 to September 2016, we circulated surveys to patients on internal medicine nursing units who were being discharged home within 24 hours. Blank surveys were distributed to nursing units by the researchers. Unit clerks and support staff were educated on the purpose of the project and asked to distribute surveys to patients who were identified by unit case managers or nurses as slated for discharge. Staff members were not asked to help with or supervise survey completion. Surveys were generally filled out by patients, but we allowed family members to assist patients if needed, and to indicate so with a checkbox. There were no exclusion criteria. Because surveys were distributed by clinical staff, the received surveys can be considered a convenience sample. Patients were asked 5 questions with 4- or 5-point Likert scale responses:
(1) “How likely is it that you will be admitted to the hospital (have to stay in the hospital overnight) again within the next 30 days after you leave the hospital this time?” [answers ranging from “Very Unlikely (<5% chance)” to “Very Likely (>50% chance)”];
(2) “How would you feel about being rehospitalized in the next month?” [answers ranging from “Very sad, frustrated, or disappointed” to “Very happy or relieved”];
(3) “How much do you think that you personally can control whether or not you will be rehospitalized (based on what you do to take care of your body, take your medicines, and follow-up with your healthcare team)?” [answers ranging from “I have no control over whether I will be rehospitalized” to “I have complete control over whether I will be rehospitalized”];
(4) “Which of the options below best describes how you plan to follow the medical instructions after you leave the hospital?” [answers ranging from “I do NOT plan to do very much of what I am being asked to do by the doctors, nurses, therapists, and other members of the care team” to “I plan to do EVERYTHING I am being asked to do by the doctors, nurses, therapists and other members of the care team”]; and
(5) “Pick the item below that best describes YOUR OWN VIEW of the care team’s recommendations:” [answers ranging from “I DO NOT AGREE AT ALL that the best way to be healthy is to do exactly what I am being asked to do by the doctors, nurses, therapists, and other members of the care team” to “I FULLY AGREE that the best way to be healthy is to do exactly what I am being asked to do by the doctors, nurses, therapists, and other members of the care team”].
Responses were linked, based on discharge date and medical record number, to administrative data, including age, sex, race, payer, and clinical data. Subsequent hospitalizations to our hospital were ascertained from administrative data. We estimated expected risk of readmission using the all payer refined diagnosis related group coupled with the associated severity-of-illness (SOI) score, as we have reported previously.2-5 We restricted our analysis to patients who answered the question related to the likelihood of readmission. Logistic regression models were constructed using actual 30-day readmission as the dependent variable to determine whether patients could predict their own readmissions and whether patient attitudes and beliefs about their care were predictive of subsequent readmission. Patient survey responses were entered as continuous independent variables (ranging from 1-4 or 1-5, as appropriate). Multivariable logistic regression was used to determine whether patients could predict their readmissions independent of demographic variables and expected readmission rate (modeled continuously); we repeated this model after dichotomizing the patient’s estimate of the likelihood of readmission as either “unlikely” or “likely.” Patients with missing survey responses were excluded from individual models without imputation. The study was approved by the Johns Hopkins institutional review board.
RESULTS
Responses were obtained from 895 patients. Their median age was 56 years [interquartile range, 43-67], 51.4% were female, and 41.7% were white. Mean SOI was 2.53 (on a 1-4 scale), and median length-of-stay was representative for our medical service at 5.2 days (range, 1-66 days). Family members reported filling out the survey in 57 cases. The primary payer was Medicare in 40.7%, Medicaid in 24.9%, and other in 34.4%. A total of 138 patients (15.4%) were readmitted within 30 days. The Table shows survey responses and associated readmission rates. None of the attitudes related to readmission were predictive of actual readmission. However, patients were able to predict their own readmissions (P = .002 for linear trend). After adjustment for expected readmission rate, race, sex, age, and payer, the trend remained significant (P = .005). Other significant predictors of readmissions in this model included expected readmission rate (P = .002), age (P = .02), and payer (P = .002). After dichotomizing the patient estimate of readmission rate as “unlikely” (N = 581) or “likely” (N = 314), the unadjusted odds ratio associating a patient-estimated risk of readmission as “likely” with actual readmission was 1.8 (95% confidence interval, 1.2-2.5). The adjusted odds ratio (including the variables above) was 1.6 (1.1-2.4).
DISCUSSION
Our findings demonstrate that patients are able to quantify their own readmission risk. This was true even after adjustment for expected readmission rate, age, sex, race, and payer. However, we did not identify any patient attitudes, beliefs, or preferences related to readmission or discharge instructions that were associated with subsequent rehospitalization. Reassuringly, more than 80% of patients who responded to the survey indicated that they would be sad, frustrated, or disappointed should readmission occur. This suggests that most patients are invested in preventing rehospitalization. Also reassuring was that patients indicated that they agreed with the discharge care plan and intended to follow their discharge instructions.
The major limitation of this study is that it was a convenience sample. Surveys were distributed inconsistently by nursing unit staff, preventing us from calculating a response rate. Further, it is possible, if not likely, that those patients with higher levels of engagement were more likely to take the time to respond, enriching our sample with activated patients. Although we allowed family members to fill out surveys on behalf of patients, this was done in fewer than 10% of instances; as such, our data may have limited applicability to patients who are physically or cognitively unable to participate in the discharge process. Finally, in this study, we did not capture readmissions to other facilities.
We conclude that patients are able to predict their own readmissions, even after accounting for other potential predictors of readmission. However, we found no evidence to support the possibility that low levels of engagement, limited trust in the healthcare team, or nonchalance about being readmitted are associated with subsequent rehospitalization. Whether asking patients about their perceived risk of readmission might help target readmission prevention programs deserves further study.
Acknowledgments
Dr. Daniel J. Brotman had full access to the data in the study and takes responsibility for the integrity of the study data and the accuracy of the data analysis. The authors also thank the following individuals for their contributions: Drafting the manuscript (Brotman); revising the manuscript for important intellectual content (Brotman, Shihab, Tieu, Cheng, Bertram, Hoyer, Deutschendorf); acquiring the data (Brotman, Shihab, Tieu, Cheng, Bertram, Deutschendorf); interpreting the data (Brotman, Shihab, Tieu, Cheng, Bertram, Hoyer, Deutschendorf); and analyzing the data (Brotman). The authors thank nursing leadership and nursing unit staff for their assistance in distributing surveys.
Funding support: Johns Hopkins Hospitalist Scholars Program
Disclosures: The authors have declared no conflicts of interest.
1. Hoyer EH, Brotman DJ, Apfel A, et al. Improving outcomes after hospitalization: a prospective observational multi-center evaluation of care-coordination strategies on 30-day readmissions to Maryland hospitals. J Gen Int Med. 2017 (in press). PubMed
2. Oduyebo I, Lehmann CU, Pollack CE, et al. Association of self-reported hospital discharge handoffs with 30-day readmissions. JAMA Intern Med. 2013;173(8):624-629. PubMed
3. Hoyer EH, Needham DM, Atanelov L, Knox B, Friedman M, Brotman DJ. Association of impaired functional status at hospital discharge and subsequent rehospitalization. J Hosp Med. 2014;9(5):277-282. PubMed
4. Hoyer EH, Needham DM, Miller J, Deutschendorf A, Friedman M, Brotman DJ. Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951-1958. PubMed
5. Hoyer EH, Odonkor CA, Bhatia SN, Leung C, Deutschendorf A, Brotman DJ. Association between days to complete inpatient discharge summaries with all-payer hospital readmissions in Maryland. J Hosp Med. 2016;11(6):393-400. PubMed
Recent years have seen a proliferation of programs designed to prevent readmissions, including patient education initiatives, financial assistance programs, postdischarge services, and clinical personnel assigned to help patients navigate their posthospitalization clinical care. Although some strategies do not require direct patient participation (such as timely and effective handoffs between inpatient and outpatient care teams), many rely upon a commitment by the patient to participate in the postdischarge care plan. At our hospital, we have found that only about 2/3 of patients who are offered transitional interventions (such as postdischarge phone calls by nurses or home nursing through a “transition guide” program) receive the intended interventions, and those who do not receive them are more likely to be readmitted.1 While limited patient uptake may relate, in part, to factors that are difficult to overcome, such as inadequate housing or phone service, we have also encountered patients whose values, beliefs, or preferences about their care do not align with those of the care team. The purposes of this exploratory study were to (1) assess patient attitudes surrounding readmission, (2) ascertain whether these attitudes are associated with actual readmission, and (3) determine whether patients can estimate their own risk of readmission.
METHODS
From January 2014 to September 2016, we circulated surveys to patients on internal medicine nursing units who were being discharged home within 24 hours. Blank surveys were distributed to nursing units by the researchers. Unit clerks and support staff were educated on the purpose of the project and asked to distribute surveys to patients who were identified by unit case managers or nurses as slated for discharge. Staff members were not asked to help with or supervise survey completion. Surveys were generally filled out by patients, but we allowed family members to assist patients if needed, and to indicate so with a checkbox. There were no exclusion criteria. Because surveys were distributed by clinical staff, the received surveys can be considered a convenience sample. Patients were asked 5 questions with 4- or 5-point Likert scale responses:
(1) “How likely is it that you will be admitted to the hospital (have to stay in the hospital overnight) again within the next 30 days after you leave the hospital this time?” [answers ranging from “Very Unlikely (<5% chance)” to “Very Likely (>50% chance)”];
(2) “How would you feel about being rehospitalized in the next month?” [answers ranging from “Very sad, frustrated, or disappointed” to “Very happy or relieved”];
(3) “How much do you think that you personally can control whether or not you will be rehospitalized (based on what you do to take care of your body, take your medicines, and follow-up with your healthcare team)?” [answers ranging from “I have no control over whether I will be rehospitalized” to “I have complete control over whether I will be rehospitalized”];
(4) “Which of the options below best describes how you plan to follow the medical instructions after you leave the hospital?” [answers ranging from “I do NOT plan to do very much of what I am being asked to do by the doctors, nurses, therapists, and other members of the care team” to “I plan to do EVERYTHING I am being asked to do by the doctors, nurses, therapists and other members of the care team”]; and
(5) “Pick the item below that best describes YOUR OWN VIEW of the care team’s recommendations:” [answers ranging from “I DO NOT AGREE AT ALL that the best way to be healthy is to do exactly what I am being asked to do by the doctors, nurses, therapists, and other members of the care team” to “I FULLY AGREE that the best way to be healthy is to do exactly what I am being asked to do by the doctors, nurses, therapists, and other members of the care team”].
Responses were linked, based on discharge date and medical record number, to administrative data, including age, sex, race, payer, and clinical data. Subsequent hospitalizations to our hospital were ascertained from administrative data. We estimated expected risk of readmission using the all payer refined diagnosis related group coupled with the associated severity-of-illness (SOI) score, as we have reported previously.2-5 We restricted our analysis to patients who answered the question related to the likelihood of readmission. Logistic regression models were constructed using actual 30-day readmission as the dependent variable to determine whether patients could predict their own readmissions and whether patient attitudes and beliefs about their care were predictive of subsequent readmission. Patient survey responses were entered as continuous independent variables (ranging from 1-4 or 1-5, as appropriate). Multivariable logistic regression was used to determine whether patients could predict their readmissions independent of demographic variables and expected readmission rate (modeled continuously); we repeated this model after dichotomizing the patient’s estimate of the likelihood of readmission as either “unlikely” or “likely.” Patients with missing survey responses were excluded from individual models without imputation. The study was approved by the Johns Hopkins institutional review board.
RESULTS
Responses were obtained from 895 patients. Their median age was 56 years [interquartile range, 43-67], 51.4% were female, and 41.7% were white. Mean SOI was 2.53 (on a 1-4 scale), and median length-of-stay was representative for our medical service at 5.2 days (range, 1-66 days). Family members reported filling out the survey in 57 cases. The primary payer was Medicare in 40.7%, Medicaid in 24.9%, and other in 34.4%. A total of 138 patients (15.4%) were readmitted within 30 days. The Table shows survey responses and associated readmission rates. None of the attitudes related to readmission were predictive of actual readmission. However, patients were able to predict their own readmissions (P = .002 for linear trend). After adjustment for expected readmission rate, race, sex, age, and payer, the trend remained significant (P = .005). Other significant predictors of readmissions in this model included expected readmission rate (P = .002), age (P = .02), and payer (P = .002). After dichotomizing the patient estimate of readmission rate as “unlikely” (N = 581) or “likely” (N = 314), the unadjusted odds ratio associating a patient-estimated risk of readmission as “likely” with actual readmission was 1.8 (95% confidence interval, 1.2-2.5). The adjusted odds ratio (including the variables above) was 1.6 (1.1-2.4).
DISCUSSION
Our findings demonstrate that patients are able to quantify their own readmission risk. This was true even after adjustment for expected readmission rate, age, sex, race, and payer. However, we did not identify any patient attitudes, beliefs, or preferences related to readmission or discharge instructions that were associated with subsequent rehospitalization. Reassuringly, more than 80% of patients who responded to the survey indicated that they would be sad, frustrated, or disappointed should readmission occur. This suggests that most patients are invested in preventing rehospitalization. Also reassuring was that patients indicated that they agreed with the discharge care plan and intended to follow their discharge instructions.
The major limitation of this study is that it was a convenience sample. Surveys were distributed inconsistently by nursing unit staff, preventing us from calculating a response rate. Further, it is possible, if not likely, that those patients with higher levels of engagement were more likely to take the time to respond, enriching our sample with activated patients. Although we allowed family members to fill out surveys on behalf of patients, this was done in fewer than 10% of instances; as such, our data may have limited applicability to patients who are physically or cognitively unable to participate in the discharge process. Finally, in this study, we did not capture readmissions to other facilities.
We conclude that patients are able to predict their own readmissions, even after accounting for other potential predictors of readmission. However, we found no evidence to support the possibility that low levels of engagement, limited trust in the healthcare team, or nonchalance about being readmitted are associated with subsequent rehospitalization. Whether asking patients about their perceived risk of readmission might help target readmission prevention programs deserves further study.
Acknowledgments
Dr. Daniel J. Brotman had full access to the data in the study and takes responsibility for the integrity of the study data and the accuracy of the data analysis. The authors also thank the following individuals for their contributions: Drafting the manuscript (Brotman); revising the manuscript for important intellectual content (Brotman, Shihab, Tieu, Cheng, Bertram, Hoyer, Deutschendorf); acquiring the data (Brotman, Shihab, Tieu, Cheng, Bertram, Deutschendorf); interpreting the data (Brotman, Shihab, Tieu, Cheng, Bertram, Hoyer, Deutschendorf); and analyzing the data (Brotman). The authors thank nursing leadership and nursing unit staff for their assistance in distributing surveys.
Funding support: Johns Hopkins Hospitalist Scholars Program
Disclosures: The authors have declared no conflicts of interest.
Recent years have seen a proliferation of programs designed to prevent readmissions, including patient education initiatives, financial assistance programs, postdischarge services, and clinical personnel assigned to help patients navigate their posthospitalization clinical care. Although some strategies do not require direct patient participation (such as timely and effective handoffs between inpatient and outpatient care teams), many rely upon a commitment by the patient to participate in the postdischarge care plan. At our hospital, we have found that only about 2/3 of patients who are offered transitional interventions (such as postdischarge phone calls by nurses or home nursing through a “transition guide” program) receive the intended interventions, and those who do not receive them are more likely to be readmitted.1 While limited patient uptake may relate, in part, to factors that are difficult to overcome, such as inadequate housing or phone service, we have also encountered patients whose values, beliefs, or preferences about their care do not align with those of the care team. The purposes of this exploratory study were to (1) assess patient attitudes surrounding readmission, (2) ascertain whether these attitudes are associated with actual readmission, and (3) determine whether patients can estimate their own risk of readmission.
METHODS
From January 2014 to September 2016, we circulated surveys to patients on internal medicine nursing units who were being discharged home within 24 hours. Blank surveys were distributed to nursing units by the researchers. Unit clerks and support staff were educated on the purpose of the project and asked to distribute surveys to patients who were identified by unit case managers or nurses as slated for discharge. Staff members were not asked to help with or supervise survey completion. Surveys were generally filled out by patients, but we allowed family members to assist patients if needed, and to indicate so with a checkbox. There were no exclusion criteria. Because surveys were distributed by clinical staff, the received surveys can be considered a convenience sample. Patients were asked 5 questions with 4- or 5-point Likert scale responses:
(1) “How likely is it that you will be admitted to the hospital (have to stay in the hospital overnight) again within the next 30 days after you leave the hospital this time?” [answers ranging from “Very Unlikely (<5% chance)” to “Very Likely (>50% chance)”];
(2) “How would you feel about being rehospitalized in the next month?” [answers ranging from “Very sad, frustrated, or disappointed” to “Very happy or relieved”];
(3) “How much do you think that you personally can control whether or not you will be rehospitalized (based on what you do to take care of your body, take your medicines, and follow-up with your healthcare team)?” [answers ranging from “I have no control over whether I will be rehospitalized” to “I have complete control over whether I will be rehospitalized”];
(4) “Which of the options below best describes how you plan to follow the medical instructions after you leave the hospital?” [answers ranging from “I do NOT plan to do very much of what I am being asked to do by the doctors, nurses, therapists, and other members of the care team” to “I plan to do EVERYTHING I am being asked to do by the doctors, nurses, therapists and other members of the care team”]; and
(5) “Pick the item below that best describes YOUR OWN VIEW of the care team’s recommendations:” [answers ranging from “I DO NOT AGREE AT ALL that the best way to be healthy is to do exactly what I am being asked to do by the doctors, nurses, therapists, and other members of the care team” to “I FULLY AGREE that the best way to be healthy is to do exactly what I am being asked to do by the doctors, nurses, therapists, and other members of the care team”].
Responses were linked, based on discharge date and medical record number, to administrative data, including age, sex, race, payer, and clinical data. Subsequent hospitalizations to our hospital were ascertained from administrative data. We estimated expected risk of readmission using the all payer refined diagnosis related group coupled with the associated severity-of-illness (SOI) score, as we have reported previously.2-5 We restricted our analysis to patients who answered the question related to the likelihood of readmission. Logistic regression models were constructed using actual 30-day readmission as the dependent variable to determine whether patients could predict their own readmissions and whether patient attitudes and beliefs about their care were predictive of subsequent readmission. Patient survey responses were entered as continuous independent variables (ranging from 1-4 or 1-5, as appropriate). Multivariable logistic regression was used to determine whether patients could predict their readmissions independent of demographic variables and expected readmission rate (modeled continuously); we repeated this model after dichotomizing the patient’s estimate of the likelihood of readmission as either “unlikely” or “likely.” Patients with missing survey responses were excluded from individual models without imputation. The study was approved by the Johns Hopkins institutional review board.
RESULTS
Responses were obtained from 895 patients. Their median age was 56 years [interquartile range, 43-67], 51.4% were female, and 41.7% were white. Mean SOI was 2.53 (on a 1-4 scale), and median length-of-stay was representative for our medical service at 5.2 days (range, 1-66 days). Family members reported filling out the survey in 57 cases. The primary payer was Medicare in 40.7%, Medicaid in 24.9%, and other in 34.4%. A total of 138 patients (15.4%) were readmitted within 30 days. The Table shows survey responses and associated readmission rates. None of the attitudes related to readmission were predictive of actual readmission. However, patients were able to predict their own readmissions (P = .002 for linear trend). After adjustment for expected readmission rate, race, sex, age, and payer, the trend remained significant (P = .005). Other significant predictors of readmissions in this model included expected readmission rate (P = .002), age (P = .02), and payer (P = .002). After dichotomizing the patient estimate of readmission rate as “unlikely” (N = 581) or “likely” (N = 314), the unadjusted odds ratio associating a patient-estimated risk of readmission as “likely” with actual readmission was 1.8 (95% confidence interval, 1.2-2.5). The adjusted odds ratio (including the variables above) was 1.6 (1.1-2.4).
DISCUSSION
Our findings demonstrate that patients are able to quantify their own readmission risk. This was true even after adjustment for expected readmission rate, age, sex, race, and payer. However, we did not identify any patient attitudes, beliefs, or preferences related to readmission or discharge instructions that were associated with subsequent rehospitalization. Reassuringly, more than 80% of patients who responded to the survey indicated that they would be sad, frustrated, or disappointed should readmission occur. This suggests that most patients are invested in preventing rehospitalization. Also reassuring was that patients indicated that they agreed with the discharge care plan and intended to follow their discharge instructions.
The major limitation of this study is that it was a convenience sample. Surveys were distributed inconsistently by nursing unit staff, preventing us from calculating a response rate. Further, it is possible, if not likely, that those patients with higher levels of engagement were more likely to take the time to respond, enriching our sample with activated patients. Although we allowed family members to fill out surveys on behalf of patients, this was done in fewer than 10% of instances; as such, our data may have limited applicability to patients who are physically or cognitively unable to participate in the discharge process. Finally, in this study, we did not capture readmissions to other facilities.
We conclude that patients are able to predict their own readmissions, even after accounting for other potential predictors of readmission. However, we found no evidence to support the possibility that low levels of engagement, limited trust in the healthcare team, or nonchalance about being readmitted are associated with subsequent rehospitalization. Whether asking patients about their perceived risk of readmission might help target readmission prevention programs deserves further study.
Acknowledgments
Dr. Daniel J. Brotman had full access to the data in the study and takes responsibility for the integrity of the study data and the accuracy of the data analysis. The authors also thank the following individuals for their contributions: Drafting the manuscript (Brotman); revising the manuscript for important intellectual content (Brotman, Shihab, Tieu, Cheng, Bertram, Hoyer, Deutschendorf); acquiring the data (Brotman, Shihab, Tieu, Cheng, Bertram, Deutschendorf); interpreting the data (Brotman, Shihab, Tieu, Cheng, Bertram, Hoyer, Deutschendorf); and analyzing the data (Brotman). The authors thank nursing leadership and nursing unit staff for their assistance in distributing surveys.
Funding support: Johns Hopkins Hospitalist Scholars Program
Disclosures: The authors have declared no conflicts of interest.
1. Hoyer EH, Brotman DJ, Apfel A, et al. Improving outcomes after hospitalization: a prospective observational multi-center evaluation of care-coordination strategies on 30-day readmissions to Maryland hospitals. J Gen Int Med. 2017 (in press). PubMed
2. Oduyebo I, Lehmann CU, Pollack CE, et al. Association of self-reported hospital discharge handoffs with 30-day readmissions. JAMA Intern Med. 2013;173(8):624-629. PubMed
3. Hoyer EH, Needham DM, Atanelov L, Knox B, Friedman M, Brotman DJ. Association of impaired functional status at hospital discharge and subsequent rehospitalization. J Hosp Med. 2014;9(5):277-282. PubMed
4. Hoyer EH, Needham DM, Miller J, Deutschendorf A, Friedman M, Brotman DJ. Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951-1958. PubMed
5. Hoyer EH, Odonkor CA, Bhatia SN, Leung C, Deutschendorf A, Brotman DJ. Association between days to complete inpatient discharge summaries with all-payer hospital readmissions in Maryland. J Hosp Med. 2016;11(6):393-400. PubMed
1. Hoyer EH, Brotman DJ, Apfel A, et al. Improving outcomes after hospitalization: a prospective observational multi-center evaluation of care-coordination strategies on 30-day readmissions to Maryland hospitals. J Gen Int Med. 2017 (in press). PubMed
2. Oduyebo I, Lehmann CU, Pollack CE, et al. Association of self-reported hospital discharge handoffs with 30-day readmissions. JAMA Intern Med. 2013;173(8):624-629. PubMed
3. Hoyer EH, Needham DM, Atanelov L, Knox B, Friedman M, Brotman DJ. Association of impaired functional status at hospital discharge and subsequent rehospitalization. J Hosp Med. 2014;9(5):277-282. PubMed
4. Hoyer EH, Needham DM, Miller J, Deutschendorf A, Friedman M, Brotman DJ. Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951-1958. PubMed
5. Hoyer EH, Odonkor CA, Bhatia SN, Leung C, Deutschendorf A, Brotman DJ. Association between days to complete inpatient discharge summaries with all-payer hospital readmissions in Maryland. J Hosp Med. 2016;11(6):393-400. PubMed
© 2018 Society of Hospital Medicine
Reconsidering Hospital Readmission Measures
Hospital readmission rates are a consequential and contentious measure of hospital quality. Readmissions within 30 days of hospital discharge are part of the Centers for Medicare & Medicaid Services (CMS) Value-Based Purchasing Program and are publicly reported. Hospital-wide readmissions and condition-specific readmissions are heavily weighted by US News & World Report in its hospital rankings and in the new CMS Five-Star Quality Rating System.1 However, clinicians and researchers question the construct validity of current readmission measures.2,3
The focus on readmissions began in 2009 when Jencks et al.4 reported that 20% of Medicare patients were readmitted within 30 days after hospital discharge. Policy makers embraced readmission reduction, assuming that a hospital readmission so soon after discharge reflected poor quality of hospital care and that, with focused efforts, hospitals could reduce readmissions and save CMS money. In 2010, the Affordable Care Act introduced an initiative to reduce readmissions and, in 2012, the Hospital Readmission Reduction Program was implemented, financially penalizing hospitals with higher-than-expected readmission rates for patients hospitalized with principal diagnoses of heart failure, myocardial infarction, and pneumonia.5 Readmission measures have since proliferated and now include pay-for-performance metrics for hospitalizations for chronic obstructive pulmonary disease (COPD), coronary artery bypass grafting, and total hip or knee arthroplasty. Measures are also reported for stroke patients and for “hospital-wide readmissions,” a catch-all measure intended to capture readmission rates across most diagnoses, with various exclusions intended to prevent counting planned readmissions (eg, hospitalization for cholecystectomy following a hospitalization for cholecystitis). These measures use claims data to construct hierarchical regression models at the patient and hospital levels, assuming that variation among readmission rates are due to hospital quality effects. The goal of this approach is to level the playing field to avoid penalizing hospitals for caring for sicker patients who are at higher risk for readmission for reasons unrelated to hospital care. Yet hospital readmissions are influenced by a complex set of variables that go well beyond hospital care, some of which may be better captured by existing models than others. Below we review several potential biases in the hospital readmission measures and offer policy recommendations to improve the accuracy of these measures.
Variation in a quality measure is influenced by the quality of the underlying data, the mix of patients served, bias in the performance measure, and the degree of systemic or random error.6 Hospital readmission rates are subject to multiple sources of variation, and true differences in the quality of care are often a much smaller source of this variation. A recent analysis of patient readmissions following general surgery found that the majority were unrelated to suboptimal medical care.7 Consider 3 scenarios in which a patient with COPD is readmitted 22 days after discharge. In hospital 1, the patient was discharged without a prescription for a steroid inhaler. In hospital 2, the patient was discharged on a steroid inhaler, filled the prescription, and elected not to use it. In hospital 3, the patient was discharged on a steroid inhaler and was provided medical assistance to fill the prescription but still could not afford the $15 copay. In all 3 scenarios, the hospital would be equally culpable under the current readmission measures, suffering financial and reputational penalties.
Yet the hospitals in these scenarios are not equally culpable. Variation in the mix of patients and bias in the measure impacted performance. Hospital 1 should clearly be held accountable for the readmission. In the cases of hospitals 2 and 3, the situations are more nuanced. More education about COPD, financial investment by the hospital to cover a copay, or a different transitional care approach may have increased the likelihood of patient compliance, but, ultimately, hospitals 2 and 3 were impacted by personal health behaviors and access to public health services and financial assistance, and the readmissions were less within their control.8
To be valid, hospital readmission measures would need to ensure that all hospitals are similar in patient characteristics and in the need for an availability of public health services. Yet these factors vary among hospitals and cannot be accounted for by models that rely exclusively on patient-level variables, such as the nature and severity of illness. As a result, the existing readmission measures are biased against certain types of hospitals. Hospitals that treat a greater proportion of patients who are socioeconomically disadvantaged; who lack access to primary care, medical assistance, or public health programs; and who have substance abuse and mental health issues will have higher readmission rates. Hospitals that care for patients who fail initial treatments and require referral for complex care will also have higher readmission rates. These types of patients are not randomly distributed throughout our healthcare system. They are clustered at rural hospitals in underserved areas, certain urban health systems, safety net hospitals, and academic health centers. It is not surprising that readmission penalties have most severely impacted large academic hospitals that care for disadvantaged populations.2 These penalties may have unintended consequences, reducing a hospital’s willingness to care for disadvantaged populations.
While these biases may unfairly harm hospitals caring for disadvantaged patients, the readmission measures may also indirectly harm patients. Low hospital readmission rates are not associated with reduced mortality and, in some instances, track with higher mortality.9-11 This may result from measurement factors (patients who die cannot be readmitted), from neighborhood socioeconomic status (SES) factors that may impact readmissions more,12 or from actual patient harm (some patients need acute care following discharge and may have worse outcomes if that care is delayed).11 Doctors have long recognized this potential risk; empiric evidence now supports them. While mortality measures may also be impacted by sociodemographic variables,13 whether to adjust for SES should be defined by the purpose of the measure. If the measure is meant to evaluate hospital quality (or utilization in the case of readmissions), adjusting for SES is appropriate because it is unrealistic to expect a health system to reduce income inequality and provide safe housing. Failure to adjust for SES, which has a large impact on outcomes, may mask a quality of care issue. Conversely, if the purpose of a measure is for a community to improve population health, then it should not be adjusted for SES because the community could adjust for income inequality.
Despite the complex ethical challenges created by the efforts to reduce readmissions, there has been virtually no public dialogue with patients, physicians, and policy makers regarding how to balance the trade-offs between reducing readmission and maintaining safety. Patients would likely value increased survival more than reduced readmissions, yet the current CMS Five-Star Rating System for hospital quality weighs readmissions equally with mortality in its hospital rankings, potentially misinforming patients. For example, many well-known academic medical centers score well (4 or 5 stars) on mortality and poorly (1 or 2 stars) on readmissions, resulting in a low or average overall score, calling into question face validity and confounding consumers struggling to make decisions about where to seek care. The Medicare Payment Advisory Commission’s Report to the Congress14 highlights the multiple significant systematic and random errors with the hospital readmission data.
Revisiting the Hospital Readmission Measures
Given significant bias in the hospital readmission measures and the ethical challenges imposed by reducing readmissions, potentially at the expense of survival, we believe CMS needs to take action to remedy the problem. First, CMS should drop hospital readmissions as a quality measure from its hospital rankings. Other hospital-rating groups and insurers should do the same. When included in payment schemes, readmissions should not be construed as a quality measure but as a utilization measure, like length of stay.
Second, the Department of Health & Human Services (HHS) should invest in maturing the hospital readmission measures to ensure construct, content, and criterion validity and reliability. No doubt the risk adjustment is complex and may be inherently limited using Medicare claims data. In the case of SES adjustment, for example, limited numbers of SES measures can be constructed from current data sources.8,13 There are other approaches to address this recommendation. For example, HHS could define a preventable readmission as one linked to some process or outcome of hospital care, such as whether the patient was discharged on an inhaler. The National Quality Forum used this approach to define a preventable venous thromboembolic event as one occurring when a patient did not receive appropriate prophylaxis. In this way, only hospital 1 in the 3 scenarios for the patient with COPD would be penalized. However, we recognize that it is not always simple to define specific process measures (eg, prescribing an inhaler) that link to readmission outcomes and that there may be other important yet hard-to-measure interventions (eg, patient and family education) that are important components of patient-centered care and readmission prevention. This is why readmissions are so challenging as a quality measure. If experts cannot define clinician behaviors that have a strong theory of change or are causally related to reduced readmissions, it is hard to call readmissions a modifiable quality measure. Another potential strategy to level the playing field would be to compare readmission rates across peer institutions only. For instance, tertiary-care safety net hospitals would be compared to one another and rural community hospitals would be compared to one another.14 Lastly, new data sources could be added to account for the social, community-level, public health, and personal health factors that heavily influence a patient’s risk for readmission, in addition to hospital-level factors. Appropriate methods will be needed to develop statistical models for risk adjustment; however, this is a complex topic and beyond the scope of the current paper.
Third, HHS could continue to use the current readmission measures as population health measures while supporting multistakeholder teams to better understand how people and their communities, public health agencies, insurers, and healthcare providers can collaborate to help patients thrive and avoid readmissions by addressing true defects in care and care coordination.
While it is understandable why policy makers chose to focus on hospital readmissions, and while we recognize that concerns about the measures were unknown when they were created, emerging evidence demonstrates that the current readmission measures (particularly when used as a quality metric) lack construct validity, contain significant bias and systematic errors, and create ethical tension by rewarding hospitals both financially and reputationally for turning away sick and socially disadvantaged patients who may, consequently, have adverse outcomes. Current readmission measures need to be reconsidered.
Acknowledgments
The authors thank Christine G. Holzmueller, BLA, with the Armstrong Institute for Patient Safety and Quality, Johns Hopkins Medicine, for her assistance in editing the manuscript and preparing it for journal submission.
Disclosure
Dr. Pronovost errs on the side of full disclosure and reports receiving grant or contract support from the Agency for Healthcare Research and Quality, the Gordon and Betty Moore Foundation (research related to patient safety and quality of care), the National Institutes of Health (acute lung injury research), and the American Medical Association Inc. (improve blood pressure control); honoraria from various healthcare organizations for speaking on patient safety and quality (the Leigh Bureau manages engagements); book royalties from the Penguin Group for his book Safe Patients, Smart Hospitals; and was receiving stock and fees to serve as a director for Cantel Medical up until 24 months ago. Dr. Pronovost is a founder of Patient Doctor Technologies, a startup company that seeks to enhance the partnership between patients and clinicians with an application called Doctella. Dr. Brotman, Dr. Hoyer, and Ms. Deutschendorf report no relevant conflicts of interest.
1. Centers for Medicare & Medicaid Services. Five-star quality rating system. https://www.cms.gov/medicare/provider-enrollment-and-certification/certificationandcomplianc/fsqrs.html. Accessed October 11, 2016.
2. Joynt KE, Jha AK. Characteristics of hospitals receiving penalties under the Hospital Readmissions Reduction Program. JAMA. 2013;309(4):342-343. PubMed
3. Boozary AS, Manchin J, 3rd, Wicker RF. The Medicare Hospital Readmissions Reduction Program: time for reform. JAMA. 2015;314(4):347-348. PubMed
4. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360(14):1418-1428. PubMed
5. Centers for Medicare & Medicaid Services. Readmissions Reduction Program (HRRP). https://www.cms.gov/medicare/medicare-fee-for-service-payment/acuteinpatientpps/readmissions-reduction-program.html. Accessed April 12, 2017.
6. Parker C, Schwamm LH, Fonarow GC, Smith EE, Reeves MJ. Stroke quality metrics: systematic reviews of the relationships to patient-centered outcomes and impact of public reporting. Stroke. 2012;43(1):155-162. PubMed
7. McIntyre LK, Arbabi S, Robinson EF, Maier RV. Analysis of risk factors for patient readmission 30 days following discharge from general surgery. JAMA Surg. 2016;151(9):855-861. PubMed
8. Sheingold SH, Zuckerman R, Shartzer A. Understanding Medicare hospital readmission rates and differing penalties between safety-net and other hospitals. Health Aff (Millwood). 2016;35(1):124-131. PubMed
9. Brotman DJ, Hoyer EH, Leung C, Lepley D, Deutschendorf A. Associations between hospital-wide readmission rates and mortality measures at the hospital level: are hospital-wide readmissions a measure of quality? J Hosp Med. 2016;11(9):650-651. PubMed
10. Krumholz HM, Lin Z, Keenan PS, et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. JAMA. 2013;309(6):587-593. PubMed
11. Fan VS, Gaziano JM, Lew R, et al. A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial. Ann Intern Med. 2012;156(10):673-683. PubMed
12. Bikdeli B, Wayda B, Bao H, et al. Place of residence and outcomes of patients with heart failure: analysis from the Telemonitoring to Improve Heart Failure Outcomes Trial. Circ Cardiovasc Qual Outcomes. 2014;7(5):749-756. PubMed
13. Bernheim SM, Parzynski CS, Horwitz L, et al. Accounting for patients’ socioeconomic status does not change hospital readmission rates. Health Aff (Millwood). 2016;35(8):1461-1470. PubMed
14. Medicare Payment Advisory Commission. Refining the Hospital Readmissions Reduction Program. In: Report to the Congress: Medicare and the Health Care Delivery System, Chapter 4. June 2013. PubMed
Hospital readmission rates are a consequential and contentious measure of hospital quality. Readmissions within 30 days of hospital discharge are part of the Centers for Medicare & Medicaid Services (CMS) Value-Based Purchasing Program and are publicly reported. Hospital-wide readmissions and condition-specific readmissions are heavily weighted by US News & World Report in its hospital rankings and in the new CMS Five-Star Quality Rating System.1 However, clinicians and researchers question the construct validity of current readmission measures.2,3
The focus on readmissions began in 2009 when Jencks et al.4 reported that 20% of Medicare patients were readmitted within 30 days after hospital discharge. Policy makers embraced readmission reduction, assuming that a hospital readmission so soon after discharge reflected poor quality of hospital care and that, with focused efforts, hospitals could reduce readmissions and save CMS money. In 2010, the Affordable Care Act introduced an initiative to reduce readmissions and, in 2012, the Hospital Readmission Reduction Program was implemented, financially penalizing hospitals with higher-than-expected readmission rates for patients hospitalized with principal diagnoses of heart failure, myocardial infarction, and pneumonia.5 Readmission measures have since proliferated and now include pay-for-performance metrics for hospitalizations for chronic obstructive pulmonary disease (COPD), coronary artery bypass grafting, and total hip or knee arthroplasty. Measures are also reported for stroke patients and for “hospital-wide readmissions,” a catch-all measure intended to capture readmission rates across most diagnoses, with various exclusions intended to prevent counting planned readmissions (eg, hospitalization for cholecystectomy following a hospitalization for cholecystitis). These measures use claims data to construct hierarchical regression models at the patient and hospital levels, assuming that variation among readmission rates are due to hospital quality effects. The goal of this approach is to level the playing field to avoid penalizing hospitals for caring for sicker patients who are at higher risk for readmission for reasons unrelated to hospital care. Yet hospital readmissions are influenced by a complex set of variables that go well beyond hospital care, some of which may be better captured by existing models than others. Below we review several potential biases in the hospital readmission measures and offer policy recommendations to improve the accuracy of these measures.
Variation in a quality measure is influenced by the quality of the underlying data, the mix of patients served, bias in the performance measure, and the degree of systemic or random error.6 Hospital readmission rates are subject to multiple sources of variation, and true differences in the quality of care are often a much smaller source of this variation. A recent analysis of patient readmissions following general surgery found that the majority were unrelated to suboptimal medical care.7 Consider 3 scenarios in which a patient with COPD is readmitted 22 days after discharge. In hospital 1, the patient was discharged without a prescription for a steroid inhaler. In hospital 2, the patient was discharged on a steroid inhaler, filled the prescription, and elected not to use it. In hospital 3, the patient was discharged on a steroid inhaler and was provided medical assistance to fill the prescription but still could not afford the $15 copay. In all 3 scenarios, the hospital would be equally culpable under the current readmission measures, suffering financial and reputational penalties.
Yet the hospitals in these scenarios are not equally culpable. Variation in the mix of patients and bias in the measure impacted performance. Hospital 1 should clearly be held accountable for the readmission. In the cases of hospitals 2 and 3, the situations are more nuanced. More education about COPD, financial investment by the hospital to cover a copay, or a different transitional care approach may have increased the likelihood of patient compliance, but, ultimately, hospitals 2 and 3 were impacted by personal health behaviors and access to public health services and financial assistance, and the readmissions were less within their control.8
To be valid, hospital readmission measures would need to ensure that all hospitals are similar in patient characteristics and in the need for an availability of public health services. Yet these factors vary among hospitals and cannot be accounted for by models that rely exclusively on patient-level variables, such as the nature and severity of illness. As a result, the existing readmission measures are biased against certain types of hospitals. Hospitals that treat a greater proportion of patients who are socioeconomically disadvantaged; who lack access to primary care, medical assistance, or public health programs; and who have substance abuse and mental health issues will have higher readmission rates. Hospitals that care for patients who fail initial treatments and require referral for complex care will also have higher readmission rates. These types of patients are not randomly distributed throughout our healthcare system. They are clustered at rural hospitals in underserved areas, certain urban health systems, safety net hospitals, and academic health centers. It is not surprising that readmission penalties have most severely impacted large academic hospitals that care for disadvantaged populations.2 These penalties may have unintended consequences, reducing a hospital’s willingness to care for disadvantaged populations.
While these biases may unfairly harm hospitals caring for disadvantaged patients, the readmission measures may also indirectly harm patients. Low hospital readmission rates are not associated with reduced mortality and, in some instances, track with higher mortality.9-11 This may result from measurement factors (patients who die cannot be readmitted), from neighborhood socioeconomic status (SES) factors that may impact readmissions more,12 or from actual patient harm (some patients need acute care following discharge and may have worse outcomes if that care is delayed).11 Doctors have long recognized this potential risk; empiric evidence now supports them. While mortality measures may also be impacted by sociodemographic variables,13 whether to adjust for SES should be defined by the purpose of the measure. If the measure is meant to evaluate hospital quality (or utilization in the case of readmissions), adjusting for SES is appropriate because it is unrealistic to expect a health system to reduce income inequality and provide safe housing. Failure to adjust for SES, which has a large impact on outcomes, may mask a quality of care issue. Conversely, if the purpose of a measure is for a community to improve population health, then it should not be adjusted for SES because the community could adjust for income inequality.
Despite the complex ethical challenges created by the efforts to reduce readmissions, there has been virtually no public dialogue with patients, physicians, and policy makers regarding how to balance the trade-offs between reducing readmission and maintaining safety. Patients would likely value increased survival more than reduced readmissions, yet the current CMS Five-Star Rating System for hospital quality weighs readmissions equally with mortality in its hospital rankings, potentially misinforming patients. For example, many well-known academic medical centers score well (4 or 5 stars) on mortality and poorly (1 or 2 stars) on readmissions, resulting in a low or average overall score, calling into question face validity and confounding consumers struggling to make decisions about where to seek care. The Medicare Payment Advisory Commission’s Report to the Congress14 highlights the multiple significant systematic and random errors with the hospital readmission data.
Revisiting the Hospital Readmission Measures
Given significant bias in the hospital readmission measures and the ethical challenges imposed by reducing readmissions, potentially at the expense of survival, we believe CMS needs to take action to remedy the problem. First, CMS should drop hospital readmissions as a quality measure from its hospital rankings. Other hospital-rating groups and insurers should do the same. When included in payment schemes, readmissions should not be construed as a quality measure but as a utilization measure, like length of stay.
Second, the Department of Health & Human Services (HHS) should invest in maturing the hospital readmission measures to ensure construct, content, and criterion validity and reliability. No doubt the risk adjustment is complex and may be inherently limited using Medicare claims data. In the case of SES adjustment, for example, limited numbers of SES measures can be constructed from current data sources.8,13 There are other approaches to address this recommendation. For example, HHS could define a preventable readmission as one linked to some process or outcome of hospital care, such as whether the patient was discharged on an inhaler. The National Quality Forum used this approach to define a preventable venous thromboembolic event as one occurring when a patient did not receive appropriate prophylaxis. In this way, only hospital 1 in the 3 scenarios for the patient with COPD would be penalized. However, we recognize that it is not always simple to define specific process measures (eg, prescribing an inhaler) that link to readmission outcomes and that there may be other important yet hard-to-measure interventions (eg, patient and family education) that are important components of patient-centered care and readmission prevention. This is why readmissions are so challenging as a quality measure. If experts cannot define clinician behaviors that have a strong theory of change or are causally related to reduced readmissions, it is hard to call readmissions a modifiable quality measure. Another potential strategy to level the playing field would be to compare readmission rates across peer institutions only. For instance, tertiary-care safety net hospitals would be compared to one another and rural community hospitals would be compared to one another.14 Lastly, new data sources could be added to account for the social, community-level, public health, and personal health factors that heavily influence a patient’s risk for readmission, in addition to hospital-level factors. Appropriate methods will be needed to develop statistical models for risk adjustment; however, this is a complex topic and beyond the scope of the current paper.
Third, HHS could continue to use the current readmission measures as population health measures while supporting multistakeholder teams to better understand how people and their communities, public health agencies, insurers, and healthcare providers can collaborate to help patients thrive and avoid readmissions by addressing true defects in care and care coordination.
While it is understandable why policy makers chose to focus on hospital readmissions, and while we recognize that concerns about the measures were unknown when they were created, emerging evidence demonstrates that the current readmission measures (particularly when used as a quality metric) lack construct validity, contain significant bias and systematic errors, and create ethical tension by rewarding hospitals both financially and reputationally for turning away sick and socially disadvantaged patients who may, consequently, have adverse outcomes. Current readmission measures need to be reconsidered.
Acknowledgments
The authors thank Christine G. Holzmueller, BLA, with the Armstrong Institute for Patient Safety and Quality, Johns Hopkins Medicine, for her assistance in editing the manuscript and preparing it for journal submission.
Disclosure
Dr. Pronovost errs on the side of full disclosure and reports receiving grant or contract support from the Agency for Healthcare Research and Quality, the Gordon and Betty Moore Foundation (research related to patient safety and quality of care), the National Institutes of Health (acute lung injury research), and the American Medical Association Inc. (improve blood pressure control); honoraria from various healthcare organizations for speaking on patient safety and quality (the Leigh Bureau manages engagements); book royalties from the Penguin Group for his book Safe Patients, Smart Hospitals; and was receiving stock and fees to serve as a director for Cantel Medical up until 24 months ago. Dr. Pronovost is a founder of Patient Doctor Technologies, a startup company that seeks to enhance the partnership between patients and clinicians with an application called Doctella. Dr. Brotman, Dr. Hoyer, and Ms. Deutschendorf report no relevant conflicts of interest.
Hospital readmission rates are a consequential and contentious measure of hospital quality. Readmissions within 30 days of hospital discharge are part of the Centers for Medicare & Medicaid Services (CMS) Value-Based Purchasing Program and are publicly reported. Hospital-wide readmissions and condition-specific readmissions are heavily weighted by US News & World Report in its hospital rankings and in the new CMS Five-Star Quality Rating System.1 However, clinicians and researchers question the construct validity of current readmission measures.2,3
The focus on readmissions began in 2009 when Jencks et al.4 reported that 20% of Medicare patients were readmitted within 30 days after hospital discharge. Policy makers embraced readmission reduction, assuming that a hospital readmission so soon after discharge reflected poor quality of hospital care and that, with focused efforts, hospitals could reduce readmissions and save CMS money. In 2010, the Affordable Care Act introduced an initiative to reduce readmissions and, in 2012, the Hospital Readmission Reduction Program was implemented, financially penalizing hospitals with higher-than-expected readmission rates for patients hospitalized with principal diagnoses of heart failure, myocardial infarction, and pneumonia.5 Readmission measures have since proliferated and now include pay-for-performance metrics for hospitalizations for chronic obstructive pulmonary disease (COPD), coronary artery bypass grafting, and total hip or knee arthroplasty. Measures are also reported for stroke patients and for “hospital-wide readmissions,” a catch-all measure intended to capture readmission rates across most diagnoses, with various exclusions intended to prevent counting planned readmissions (eg, hospitalization for cholecystectomy following a hospitalization for cholecystitis). These measures use claims data to construct hierarchical regression models at the patient and hospital levels, assuming that variation among readmission rates are due to hospital quality effects. The goal of this approach is to level the playing field to avoid penalizing hospitals for caring for sicker patients who are at higher risk for readmission for reasons unrelated to hospital care. Yet hospital readmissions are influenced by a complex set of variables that go well beyond hospital care, some of which may be better captured by existing models than others. Below we review several potential biases in the hospital readmission measures and offer policy recommendations to improve the accuracy of these measures.
Variation in a quality measure is influenced by the quality of the underlying data, the mix of patients served, bias in the performance measure, and the degree of systemic or random error.6 Hospital readmission rates are subject to multiple sources of variation, and true differences in the quality of care are often a much smaller source of this variation. A recent analysis of patient readmissions following general surgery found that the majority were unrelated to suboptimal medical care.7 Consider 3 scenarios in which a patient with COPD is readmitted 22 days after discharge. In hospital 1, the patient was discharged without a prescription for a steroid inhaler. In hospital 2, the patient was discharged on a steroid inhaler, filled the prescription, and elected not to use it. In hospital 3, the patient was discharged on a steroid inhaler and was provided medical assistance to fill the prescription but still could not afford the $15 copay. In all 3 scenarios, the hospital would be equally culpable under the current readmission measures, suffering financial and reputational penalties.
Yet the hospitals in these scenarios are not equally culpable. Variation in the mix of patients and bias in the measure impacted performance. Hospital 1 should clearly be held accountable for the readmission. In the cases of hospitals 2 and 3, the situations are more nuanced. More education about COPD, financial investment by the hospital to cover a copay, or a different transitional care approach may have increased the likelihood of patient compliance, but, ultimately, hospitals 2 and 3 were impacted by personal health behaviors and access to public health services and financial assistance, and the readmissions were less within their control.8
To be valid, hospital readmission measures would need to ensure that all hospitals are similar in patient characteristics and in the need for an availability of public health services. Yet these factors vary among hospitals and cannot be accounted for by models that rely exclusively on patient-level variables, such as the nature and severity of illness. As a result, the existing readmission measures are biased against certain types of hospitals. Hospitals that treat a greater proportion of patients who are socioeconomically disadvantaged; who lack access to primary care, medical assistance, or public health programs; and who have substance abuse and mental health issues will have higher readmission rates. Hospitals that care for patients who fail initial treatments and require referral for complex care will also have higher readmission rates. These types of patients are not randomly distributed throughout our healthcare system. They are clustered at rural hospitals in underserved areas, certain urban health systems, safety net hospitals, and academic health centers. It is not surprising that readmission penalties have most severely impacted large academic hospitals that care for disadvantaged populations.2 These penalties may have unintended consequences, reducing a hospital’s willingness to care for disadvantaged populations.
While these biases may unfairly harm hospitals caring for disadvantaged patients, the readmission measures may also indirectly harm patients. Low hospital readmission rates are not associated with reduced mortality and, in some instances, track with higher mortality.9-11 This may result from measurement factors (patients who die cannot be readmitted), from neighborhood socioeconomic status (SES) factors that may impact readmissions more,12 or from actual patient harm (some patients need acute care following discharge and may have worse outcomes if that care is delayed).11 Doctors have long recognized this potential risk; empiric evidence now supports them. While mortality measures may also be impacted by sociodemographic variables,13 whether to adjust for SES should be defined by the purpose of the measure. If the measure is meant to evaluate hospital quality (or utilization in the case of readmissions), adjusting for SES is appropriate because it is unrealistic to expect a health system to reduce income inequality and provide safe housing. Failure to adjust for SES, which has a large impact on outcomes, may mask a quality of care issue. Conversely, if the purpose of a measure is for a community to improve population health, then it should not be adjusted for SES because the community could adjust for income inequality.
Despite the complex ethical challenges created by the efforts to reduce readmissions, there has been virtually no public dialogue with patients, physicians, and policy makers regarding how to balance the trade-offs between reducing readmission and maintaining safety. Patients would likely value increased survival more than reduced readmissions, yet the current CMS Five-Star Rating System for hospital quality weighs readmissions equally with mortality in its hospital rankings, potentially misinforming patients. For example, many well-known academic medical centers score well (4 or 5 stars) on mortality and poorly (1 or 2 stars) on readmissions, resulting in a low or average overall score, calling into question face validity and confounding consumers struggling to make decisions about where to seek care. The Medicare Payment Advisory Commission’s Report to the Congress14 highlights the multiple significant systematic and random errors with the hospital readmission data.
Revisiting the Hospital Readmission Measures
Given significant bias in the hospital readmission measures and the ethical challenges imposed by reducing readmissions, potentially at the expense of survival, we believe CMS needs to take action to remedy the problem. First, CMS should drop hospital readmissions as a quality measure from its hospital rankings. Other hospital-rating groups and insurers should do the same. When included in payment schemes, readmissions should not be construed as a quality measure but as a utilization measure, like length of stay.
Second, the Department of Health & Human Services (HHS) should invest in maturing the hospital readmission measures to ensure construct, content, and criterion validity and reliability. No doubt the risk adjustment is complex and may be inherently limited using Medicare claims data. In the case of SES adjustment, for example, limited numbers of SES measures can be constructed from current data sources.8,13 There are other approaches to address this recommendation. For example, HHS could define a preventable readmission as one linked to some process or outcome of hospital care, such as whether the patient was discharged on an inhaler. The National Quality Forum used this approach to define a preventable venous thromboembolic event as one occurring when a patient did not receive appropriate prophylaxis. In this way, only hospital 1 in the 3 scenarios for the patient with COPD would be penalized. However, we recognize that it is not always simple to define specific process measures (eg, prescribing an inhaler) that link to readmission outcomes and that there may be other important yet hard-to-measure interventions (eg, patient and family education) that are important components of patient-centered care and readmission prevention. This is why readmissions are so challenging as a quality measure. If experts cannot define clinician behaviors that have a strong theory of change or are causally related to reduced readmissions, it is hard to call readmissions a modifiable quality measure. Another potential strategy to level the playing field would be to compare readmission rates across peer institutions only. For instance, tertiary-care safety net hospitals would be compared to one another and rural community hospitals would be compared to one another.14 Lastly, new data sources could be added to account for the social, community-level, public health, and personal health factors that heavily influence a patient’s risk for readmission, in addition to hospital-level factors. Appropriate methods will be needed to develop statistical models for risk adjustment; however, this is a complex topic and beyond the scope of the current paper.
Third, HHS could continue to use the current readmission measures as population health measures while supporting multistakeholder teams to better understand how people and their communities, public health agencies, insurers, and healthcare providers can collaborate to help patients thrive and avoid readmissions by addressing true defects in care and care coordination.
While it is understandable why policy makers chose to focus on hospital readmissions, and while we recognize that concerns about the measures were unknown when they were created, emerging evidence demonstrates that the current readmission measures (particularly when used as a quality metric) lack construct validity, contain significant bias and systematic errors, and create ethical tension by rewarding hospitals both financially and reputationally for turning away sick and socially disadvantaged patients who may, consequently, have adverse outcomes. Current readmission measures need to be reconsidered.
Acknowledgments
The authors thank Christine G. Holzmueller, BLA, with the Armstrong Institute for Patient Safety and Quality, Johns Hopkins Medicine, for her assistance in editing the manuscript and preparing it for journal submission.
Disclosure
Dr. Pronovost errs on the side of full disclosure and reports receiving grant or contract support from the Agency for Healthcare Research and Quality, the Gordon and Betty Moore Foundation (research related to patient safety and quality of care), the National Institutes of Health (acute lung injury research), and the American Medical Association Inc. (improve blood pressure control); honoraria from various healthcare organizations for speaking on patient safety and quality (the Leigh Bureau manages engagements); book royalties from the Penguin Group for his book Safe Patients, Smart Hospitals; and was receiving stock and fees to serve as a director for Cantel Medical up until 24 months ago. Dr. Pronovost is a founder of Patient Doctor Technologies, a startup company that seeks to enhance the partnership between patients and clinicians with an application called Doctella. Dr. Brotman, Dr. Hoyer, and Ms. Deutschendorf report no relevant conflicts of interest.
1. Centers for Medicare & Medicaid Services. Five-star quality rating system. https://www.cms.gov/medicare/provider-enrollment-and-certification/certificationandcomplianc/fsqrs.html. Accessed October 11, 2016.
2. Joynt KE, Jha AK. Characteristics of hospitals receiving penalties under the Hospital Readmissions Reduction Program. JAMA. 2013;309(4):342-343. PubMed
3. Boozary AS, Manchin J, 3rd, Wicker RF. The Medicare Hospital Readmissions Reduction Program: time for reform. JAMA. 2015;314(4):347-348. PubMed
4. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360(14):1418-1428. PubMed
5. Centers for Medicare & Medicaid Services. Readmissions Reduction Program (HRRP). https://www.cms.gov/medicare/medicare-fee-for-service-payment/acuteinpatientpps/readmissions-reduction-program.html. Accessed April 12, 2017.
6. Parker C, Schwamm LH, Fonarow GC, Smith EE, Reeves MJ. Stroke quality metrics: systematic reviews of the relationships to patient-centered outcomes and impact of public reporting. Stroke. 2012;43(1):155-162. PubMed
7. McIntyre LK, Arbabi S, Robinson EF, Maier RV. Analysis of risk factors for patient readmission 30 days following discharge from general surgery. JAMA Surg. 2016;151(9):855-861. PubMed
8. Sheingold SH, Zuckerman R, Shartzer A. Understanding Medicare hospital readmission rates and differing penalties between safety-net and other hospitals. Health Aff (Millwood). 2016;35(1):124-131. PubMed
9. Brotman DJ, Hoyer EH, Leung C, Lepley D, Deutschendorf A. Associations between hospital-wide readmission rates and mortality measures at the hospital level: are hospital-wide readmissions a measure of quality? J Hosp Med. 2016;11(9):650-651. PubMed
10. Krumholz HM, Lin Z, Keenan PS, et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. JAMA. 2013;309(6):587-593. PubMed
11. Fan VS, Gaziano JM, Lew R, et al. A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial. Ann Intern Med. 2012;156(10):673-683. PubMed
12. Bikdeli B, Wayda B, Bao H, et al. Place of residence and outcomes of patients with heart failure: analysis from the Telemonitoring to Improve Heart Failure Outcomes Trial. Circ Cardiovasc Qual Outcomes. 2014;7(5):749-756. PubMed
13. Bernheim SM, Parzynski CS, Horwitz L, et al. Accounting for patients’ socioeconomic status does not change hospital readmission rates. Health Aff (Millwood). 2016;35(8):1461-1470. PubMed
14. Medicare Payment Advisory Commission. Refining the Hospital Readmissions Reduction Program. In: Report to the Congress: Medicare and the Health Care Delivery System, Chapter 4. June 2013. PubMed
1. Centers for Medicare & Medicaid Services. Five-star quality rating system. https://www.cms.gov/medicare/provider-enrollment-and-certification/certificationandcomplianc/fsqrs.html. Accessed October 11, 2016.
2. Joynt KE, Jha AK. Characteristics of hospitals receiving penalties under the Hospital Readmissions Reduction Program. JAMA. 2013;309(4):342-343. PubMed
3. Boozary AS, Manchin J, 3rd, Wicker RF. The Medicare Hospital Readmissions Reduction Program: time for reform. JAMA. 2015;314(4):347-348. PubMed
4. Jencks SF, Williams MV, Coleman EA. Rehospitalizations among patients in the Medicare fee-for-service program. N Engl J Med. 2009;360(14):1418-1428. PubMed
5. Centers for Medicare & Medicaid Services. Readmissions Reduction Program (HRRP). https://www.cms.gov/medicare/medicare-fee-for-service-payment/acuteinpatientpps/readmissions-reduction-program.html. Accessed April 12, 2017.
6. Parker C, Schwamm LH, Fonarow GC, Smith EE, Reeves MJ. Stroke quality metrics: systematic reviews of the relationships to patient-centered outcomes and impact of public reporting. Stroke. 2012;43(1):155-162. PubMed
7. McIntyre LK, Arbabi S, Robinson EF, Maier RV. Analysis of risk factors for patient readmission 30 days following discharge from general surgery. JAMA Surg. 2016;151(9):855-861. PubMed
8. Sheingold SH, Zuckerman R, Shartzer A. Understanding Medicare hospital readmission rates and differing penalties between safety-net and other hospitals. Health Aff (Millwood). 2016;35(1):124-131. PubMed
9. Brotman DJ, Hoyer EH, Leung C, Lepley D, Deutschendorf A. Associations between hospital-wide readmission rates and mortality measures at the hospital level: are hospital-wide readmissions a measure of quality? J Hosp Med. 2016;11(9):650-651. PubMed
10. Krumholz HM, Lin Z, Keenan PS, et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. JAMA. 2013;309(6):587-593. PubMed
11. Fan VS, Gaziano JM, Lew R, et al. A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial. Ann Intern Med. 2012;156(10):673-683. PubMed
12. Bikdeli B, Wayda B, Bao H, et al. Place of residence and outcomes of patients with heart failure: analysis from the Telemonitoring to Improve Heart Failure Outcomes Trial. Circ Cardiovasc Qual Outcomes. 2014;7(5):749-756. PubMed
13. Bernheim SM, Parzynski CS, Horwitz L, et al. Accounting for patients’ socioeconomic status does not change hospital readmission rates. Health Aff (Millwood). 2016;35(8):1461-1470. PubMed
14. Medicare Payment Advisory Commission. Refining the Hospital Readmissions Reduction Program. In: Report to the Congress: Medicare and the Health Care Delivery System, Chapter 4. June 2013. PubMed
© 2017 Society of Hospital Medicine
Readmission Rates and Mortality Measures
The Centers for Medicare & Medicaid Services (CMS) have sought to reduce readmissions in the 30 days following hospital discharge through penalties applied to hospitals with readmission rates that are higher than expected. Expected readmission rates for Medicare fee‐for‐service beneficiaries are calculated from models that use patient‐level administrative data to account for patient morbidities. Readmitted patients are defined as those who are discharged from the hospital alive and then rehospitalized at any acute care facility within 30 days of discharge. These models explicitly exclude sociodemographic variables that may impact quality of and access to outpatient care. Specific exclusions are also applied based on diagnosis codes so as to avoid penalizing hospitals for rehospitalizations that are likely to have been planned.
More recently, a hospital‐wide readmission measure has been developed, which seeks to provide a comprehensive view of each hospital's readmission rate by including the vast majority of Medicare patients. Like the condition‐specific readmission measures, the hospital‐wide readmission measure also excludes sociodemographic variables and incorporates specific condition‐based exclusions so as to avoid counting planned rehospitalizations (e.g., an admission for cholecystectomy following an admission for biliary sepsis). Although not currently used for pay‐for‐performance, this measure has been included in the CMS Star Report along with other readmission measures.[1] CMS does not currently disseminate a hospital‐wide mortality measure, but does disseminate hospital‐level adjusted 30‐day mortality rates for Medicare beneficiaries with discharge diagnoses of stroke, heart failure, myocardial infarction (MI), chronic obstructive pulmonary disease (COPD) and pneumonia, and principal procedure of coronary artery bypass grafting (CABG).
It is conceivable that aggressive efforts to reduce readmissions might delay life‐saving acute care in some scenarios,[2] and there is prior evidence that heart failure readmissions are inversely (but weakly) related to heart failure mortality.[3] It is also plausible that keeping tenuous patients alive until discharge might result in higher readmission rates. We sought to examine the relationship between hospital‐wide adjusted 30‐day readmissions and death rates across the acute care hospitals in the United States. Lacking a measure of hospital‐wide death rates, we examined the relation between hospital‐wide readmissions and each of the 6 condition‐specific mortality measures. For comparison, we also examined the relationships between condition‐specific readmission rates and mortality rates.
METHODS
We used publically available data published by CMS from July 1, 2011 through June 30, 2014.[4] These data are provided at the hospital level, without any patient‐level data. We included 4452 acute care facilities based on having hospital‐wide readmission rates, but not all facilities contributed data for each mortality measure. We excluded from analysis on a measure‐by‐measure basis those facilities for which outcomes were absent, without imputing missing outcome measures, because low volume of a given condition was the main reason for not reporting a measure. For each mortality measure, we constructed a logistic regression model to quantify the odds of performing in the lowest (best) mortality tertile as a function of hospital‐wide readmission tertile. To account for patient volumes, we included in each model the number of eligible patients at each hospital with the specified condition. We repeated these analyses using condition‐specific readmission rates (rather than the hospital‐wide readmission rates) as the independent variable. Specifications for CMS models for mortality and readmissions are publically available.[5]
RESULTS
After adjustment for patient volumes, hospitals in the highest hospital‐wide readmission tertile were more likely to perform in the lowest (best) mortality tertile for 3 of the 6 mortality measures: heart failure, COPD, and stroke (P < 0.001 for all). For MI, CABG and pneumonia, there was no significant association between high hospital‐wide readmission rates and low mortality (Table 1). Using condition‐specific readmission rates, there remained an inverse association between readmissions and mortality for heart failure and stroke, but not for COPD. In contrast, hospitals with the highest CABG‐specific readmission rates were significantly less likely to have low CABG‐specific mortality (P < 0.001).
| Hospital‐Wide Readmission Rate Tertile [Range of Adjusted Readmission Rates, %]* | |||
|---|---|---|---|
|
1st Tertile, n = 1359 [11.3%‐14.8%], Adjusted Odds Ratio (95% CI) |
2nd Tertile, n = 1785 [14.9%‐15.5%], Adjusted Odds Ratio (95% CI) |
3rd Tertile, n = 1308 [15.6%‐19.8%], Adjusted Odds Ratio (95% CI) |
|
| |||
| Mortality measure (no. of hospitals reporting) | |||
| Acute myocardial infarction (n = 2415) | 1.00 (referent) | 0.88 (0.711.09) | 1.02 (0.831.25) |
| Pneumonia (n = 4067) | 1.00 (referent) | 0.83 (0.710.98) | 1.11 (0.941.31) |
| Heart failure (n = 3668) | 1.00 (referent) | 1.21 (1.021.45) | 1.94 (1.632.30) |
| Stroke (n = 2754) | 1.00 (referent) | 1.13 (0.931.38) | 1.48 (1.221.79) |
| Chronic obstructive pulmonary disease (n = 3633) | 1.00 (referent) | 1.12 (0.951.33) | 1.73 (1.462.05) |
| Coronary artery bypass (n = 1058) | 1.00 (referent) | 0.87 (0.631.19) | 0.99 (0.741.34) |
| Condition‐specific readmission rate tertile | |||
| Mortality measure | |||
| Acute myocardial infarction | 1.00 (referent) | 0.88 (0.711.08) | 0.79 (0.640.99) |
| Pneumonia | 1.00 (referent) | 0.91 (0.781.07) | 0.89 (0.761.04) |
| Heart failure | 1.00 (referent) | 1.15 (0.961.36) | 1.56 (1.311.86) |
| Stroke | 1.00 (referent) | 1.65 (1.342.03) | 1.70 (1.232.35) |
| Chronic obstructive pulmonary disease | 1.00 (referent) | 0.83 (0.700.98) | 0.84 (0.710.99) |
| Coronary artery bypass | 1.00 (referent) | 0.59 (0.440.80) | 0.47 (0.340.64) |
DISCUSSION
We found that higher hospital‐wide readmission rates were associated with lower mortality at the hospital level for 3 of the 6 mortality measures we examined. The findings for heart failure parallel the findings of Krumholz and colleagues who examined 3 of these 6 measures (MI, pneumonia, and heart failure) in relation to readmissions for these specific populations.[3] This prior analysis, however, did not include the 3 more recently reported mortality measures (COPD, stroke, and CABG) and did not use hospital‐wide readmissions.
Causal mechanisms underlying the associations between mortality and readmission at the hospital level deserve further exploration. It is certainly possible that global efforts to keep patients out of the hospital might, in some instances, place patients at risk by delaying necessary acute care.[2] It is also possible that unmeasured variables, particularly access to hospice and palliative care services that might facilitate good deaths, could be associated with both reduced readmissions and higher death rates. Additionally, because deceased patients cannot be readmitted, one might expect that readmissions and mortality might be inversely associated, particularly for conditions with a high postdischarge mortality rate. Similarly, a hospital that does a particularly good job keeping chronically ill patients alive until discharge might exhibit a higher readmission rate than a hospital that is less adept at keeping tenuous patients alive until discharge.
Regardless of the mechanisms of these findings, we present these data to raise the concern that using readmission rates, particularly hospital‐wide readmission rates, as a measure of hospital quality is inherently problematic. It is particularly problematic that CMS has applied equal weight to readmissions and mortality in the Star Report.[1] High readmission rates may result from complications and poor handoffs, but may also stem from the legitimate need to care for chronically ill patients in a high‐intensity setting, particularly fragile patients who have been kept alive against the odds. In conclusion, caution is warranted in viewing readmissions as a quality metric until the associations we describe are better explained using patient‐level data and more robust adjustment than is possible with these publically available data.
Disclosures: Dr. Daniel J. Brotman had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. There was no financial support for this work. Contributions of the authors are as follows: drafting manuscript (Brotman), revision of manuscript for important intellectual content (brotman, Hoyer, Lepley, Deutschendorf, Leung), acquisition of data (Deutschendorf, Leung, Lepley), interpretation of data (Brotman, Hoyer, Lepley, Deutschendorf, Leung), data analysis (Brotman, Hoyer).
- Centers for Medicare and Medicaid Services. Available at: https://www.cms.gov/Outreach-and-Education/Outreach/NPC/Downloads/2015-08-13-Star-Ratings-Presentation.pdf. Accessed September 2015.
- , , , et al. A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial. Ann Intern Med. 2012;156(10):673–683.
- , , , et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. JAMA. 2013;309(6):587–593.
- Centers for Medicare and Medicaid Services. Hospital compare datasets. Available at: https://data.medicare.gov/data/hospital‐compare. Accessed September 2015.
- Centers for Medicare and Medicaid Services. Hospital quality initiative. Available at: https://www.cms.gov/Medicare/Quality‐Initiatives‐Patient‐Assessment‐Instruments/HospitalQualityInits. Accessed September 2015.
The Centers for Medicare & Medicaid Services (CMS) have sought to reduce readmissions in the 30 days following hospital discharge through penalties applied to hospitals with readmission rates that are higher than expected. Expected readmission rates for Medicare fee‐for‐service beneficiaries are calculated from models that use patient‐level administrative data to account for patient morbidities. Readmitted patients are defined as those who are discharged from the hospital alive and then rehospitalized at any acute care facility within 30 days of discharge. These models explicitly exclude sociodemographic variables that may impact quality of and access to outpatient care. Specific exclusions are also applied based on diagnosis codes so as to avoid penalizing hospitals for rehospitalizations that are likely to have been planned.
More recently, a hospital‐wide readmission measure has been developed, which seeks to provide a comprehensive view of each hospital's readmission rate by including the vast majority of Medicare patients. Like the condition‐specific readmission measures, the hospital‐wide readmission measure also excludes sociodemographic variables and incorporates specific condition‐based exclusions so as to avoid counting planned rehospitalizations (e.g., an admission for cholecystectomy following an admission for biliary sepsis). Although not currently used for pay‐for‐performance, this measure has been included in the CMS Star Report along with other readmission measures.[1] CMS does not currently disseminate a hospital‐wide mortality measure, but does disseminate hospital‐level adjusted 30‐day mortality rates for Medicare beneficiaries with discharge diagnoses of stroke, heart failure, myocardial infarction (MI), chronic obstructive pulmonary disease (COPD) and pneumonia, and principal procedure of coronary artery bypass grafting (CABG).
It is conceivable that aggressive efforts to reduce readmissions might delay life‐saving acute care in some scenarios,[2] and there is prior evidence that heart failure readmissions are inversely (but weakly) related to heart failure mortality.[3] It is also plausible that keeping tenuous patients alive until discharge might result in higher readmission rates. We sought to examine the relationship between hospital‐wide adjusted 30‐day readmissions and death rates across the acute care hospitals in the United States. Lacking a measure of hospital‐wide death rates, we examined the relation between hospital‐wide readmissions and each of the 6 condition‐specific mortality measures. For comparison, we also examined the relationships between condition‐specific readmission rates and mortality rates.
METHODS
We used publically available data published by CMS from July 1, 2011 through June 30, 2014.[4] These data are provided at the hospital level, without any patient‐level data. We included 4452 acute care facilities based on having hospital‐wide readmission rates, but not all facilities contributed data for each mortality measure. We excluded from analysis on a measure‐by‐measure basis those facilities for which outcomes were absent, without imputing missing outcome measures, because low volume of a given condition was the main reason for not reporting a measure. For each mortality measure, we constructed a logistic regression model to quantify the odds of performing in the lowest (best) mortality tertile as a function of hospital‐wide readmission tertile. To account for patient volumes, we included in each model the number of eligible patients at each hospital with the specified condition. We repeated these analyses using condition‐specific readmission rates (rather than the hospital‐wide readmission rates) as the independent variable. Specifications for CMS models for mortality and readmissions are publically available.[5]
RESULTS
After adjustment for patient volumes, hospitals in the highest hospital‐wide readmission tertile were more likely to perform in the lowest (best) mortality tertile for 3 of the 6 mortality measures: heart failure, COPD, and stroke (P < 0.001 for all). For MI, CABG and pneumonia, there was no significant association between high hospital‐wide readmission rates and low mortality (Table 1). Using condition‐specific readmission rates, there remained an inverse association between readmissions and mortality for heart failure and stroke, but not for COPD. In contrast, hospitals with the highest CABG‐specific readmission rates were significantly less likely to have low CABG‐specific mortality (P < 0.001).
| Hospital‐Wide Readmission Rate Tertile [Range of Adjusted Readmission Rates, %]* | |||
|---|---|---|---|
|
1st Tertile, n = 1359 [11.3%‐14.8%], Adjusted Odds Ratio (95% CI) |
2nd Tertile, n = 1785 [14.9%‐15.5%], Adjusted Odds Ratio (95% CI) |
3rd Tertile, n = 1308 [15.6%‐19.8%], Adjusted Odds Ratio (95% CI) |
|
| |||
| Mortality measure (no. of hospitals reporting) | |||
| Acute myocardial infarction (n = 2415) | 1.00 (referent) | 0.88 (0.711.09) | 1.02 (0.831.25) |
| Pneumonia (n = 4067) | 1.00 (referent) | 0.83 (0.710.98) | 1.11 (0.941.31) |
| Heart failure (n = 3668) | 1.00 (referent) | 1.21 (1.021.45) | 1.94 (1.632.30) |
| Stroke (n = 2754) | 1.00 (referent) | 1.13 (0.931.38) | 1.48 (1.221.79) |
| Chronic obstructive pulmonary disease (n = 3633) | 1.00 (referent) | 1.12 (0.951.33) | 1.73 (1.462.05) |
| Coronary artery bypass (n = 1058) | 1.00 (referent) | 0.87 (0.631.19) | 0.99 (0.741.34) |
| Condition‐specific readmission rate tertile | |||
| Mortality measure | |||
| Acute myocardial infarction | 1.00 (referent) | 0.88 (0.711.08) | 0.79 (0.640.99) |
| Pneumonia | 1.00 (referent) | 0.91 (0.781.07) | 0.89 (0.761.04) |
| Heart failure | 1.00 (referent) | 1.15 (0.961.36) | 1.56 (1.311.86) |
| Stroke | 1.00 (referent) | 1.65 (1.342.03) | 1.70 (1.232.35) |
| Chronic obstructive pulmonary disease | 1.00 (referent) | 0.83 (0.700.98) | 0.84 (0.710.99) |
| Coronary artery bypass | 1.00 (referent) | 0.59 (0.440.80) | 0.47 (0.340.64) |
DISCUSSION
We found that higher hospital‐wide readmission rates were associated with lower mortality at the hospital level for 3 of the 6 mortality measures we examined. The findings for heart failure parallel the findings of Krumholz and colleagues who examined 3 of these 6 measures (MI, pneumonia, and heart failure) in relation to readmissions for these specific populations.[3] This prior analysis, however, did not include the 3 more recently reported mortality measures (COPD, stroke, and CABG) and did not use hospital‐wide readmissions.
Causal mechanisms underlying the associations between mortality and readmission at the hospital level deserve further exploration. It is certainly possible that global efforts to keep patients out of the hospital might, in some instances, place patients at risk by delaying necessary acute care.[2] It is also possible that unmeasured variables, particularly access to hospice and palliative care services that might facilitate good deaths, could be associated with both reduced readmissions and higher death rates. Additionally, because deceased patients cannot be readmitted, one might expect that readmissions and mortality might be inversely associated, particularly for conditions with a high postdischarge mortality rate. Similarly, a hospital that does a particularly good job keeping chronically ill patients alive until discharge might exhibit a higher readmission rate than a hospital that is less adept at keeping tenuous patients alive until discharge.
Regardless of the mechanisms of these findings, we present these data to raise the concern that using readmission rates, particularly hospital‐wide readmission rates, as a measure of hospital quality is inherently problematic. It is particularly problematic that CMS has applied equal weight to readmissions and mortality in the Star Report.[1] High readmission rates may result from complications and poor handoffs, but may also stem from the legitimate need to care for chronically ill patients in a high‐intensity setting, particularly fragile patients who have been kept alive against the odds. In conclusion, caution is warranted in viewing readmissions as a quality metric until the associations we describe are better explained using patient‐level data and more robust adjustment than is possible with these publically available data.
Disclosures: Dr. Daniel J. Brotman had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. There was no financial support for this work. Contributions of the authors are as follows: drafting manuscript (Brotman), revision of manuscript for important intellectual content (brotman, Hoyer, Lepley, Deutschendorf, Leung), acquisition of data (Deutschendorf, Leung, Lepley), interpretation of data (Brotman, Hoyer, Lepley, Deutschendorf, Leung), data analysis (Brotman, Hoyer).
The Centers for Medicare & Medicaid Services (CMS) have sought to reduce readmissions in the 30 days following hospital discharge through penalties applied to hospitals with readmission rates that are higher than expected. Expected readmission rates for Medicare fee‐for‐service beneficiaries are calculated from models that use patient‐level administrative data to account for patient morbidities. Readmitted patients are defined as those who are discharged from the hospital alive and then rehospitalized at any acute care facility within 30 days of discharge. These models explicitly exclude sociodemographic variables that may impact quality of and access to outpatient care. Specific exclusions are also applied based on diagnosis codes so as to avoid penalizing hospitals for rehospitalizations that are likely to have been planned.
More recently, a hospital‐wide readmission measure has been developed, which seeks to provide a comprehensive view of each hospital's readmission rate by including the vast majority of Medicare patients. Like the condition‐specific readmission measures, the hospital‐wide readmission measure also excludes sociodemographic variables and incorporates specific condition‐based exclusions so as to avoid counting planned rehospitalizations (e.g., an admission for cholecystectomy following an admission for biliary sepsis). Although not currently used for pay‐for‐performance, this measure has been included in the CMS Star Report along with other readmission measures.[1] CMS does not currently disseminate a hospital‐wide mortality measure, but does disseminate hospital‐level adjusted 30‐day mortality rates for Medicare beneficiaries with discharge diagnoses of stroke, heart failure, myocardial infarction (MI), chronic obstructive pulmonary disease (COPD) and pneumonia, and principal procedure of coronary artery bypass grafting (CABG).
It is conceivable that aggressive efforts to reduce readmissions might delay life‐saving acute care in some scenarios,[2] and there is prior evidence that heart failure readmissions are inversely (but weakly) related to heart failure mortality.[3] It is also plausible that keeping tenuous patients alive until discharge might result in higher readmission rates. We sought to examine the relationship between hospital‐wide adjusted 30‐day readmissions and death rates across the acute care hospitals in the United States. Lacking a measure of hospital‐wide death rates, we examined the relation between hospital‐wide readmissions and each of the 6 condition‐specific mortality measures. For comparison, we also examined the relationships between condition‐specific readmission rates and mortality rates.
METHODS
We used publically available data published by CMS from July 1, 2011 through June 30, 2014.[4] These data are provided at the hospital level, without any patient‐level data. We included 4452 acute care facilities based on having hospital‐wide readmission rates, but not all facilities contributed data for each mortality measure. We excluded from analysis on a measure‐by‐measure basis those facilities for which outcomes were absent, without imputing missing outcome measures, because low volume of a given condition was the main reason for not reporting a measure. For each mortality measure, we constructed a logistic regression model to quantify the odds of performing in the lowest (best) mortality tertile as a function of hospital‐wide readmission tertile. To account for patient volumes, we included in each model the number of eligible patients at each hospital with the specified condition. We repeated these analyses using condition‐specific readmission rates (rather than the hospital‐wide readmission rates) as the independent variable. Specifications for CMS models for mortality and readmissions are publically available.[5]
RESULTS
After adjustment for patient volumes, hospitals in the highest hospital‐wide readmission tertile were more likely to perform in the lowest (best) mortality tertile for 3 of the 6 mortality measures: heart failure, COPD, and stroke (P < 0.001 for all). For MI, CABG and pneumonia, there was no significant association between high hospital‐wide readmission rates and low mortality (Table 1). Using condition‐specific readmission rates, there remained an inverse association between readmissions and mortality for heart failure and stroke, but not for COPD. In contrast, hospitals with the highest CABG‐specific readmission rates were significantly less likely to have low CABG‐specific mortality (P < 0.001).
| Hospital‐Wide Readmission Rate Tertile [Range of Adjusted Readmission Rates, %]* | |||
|---|---|---|---|
|
1st Tertile, n = 1359 [11.3%‐14.8%], Adjusted Odds Ratio (95% CI) |
2nd Tertile, n = 1785 [14.9%‐15.5%], Adjusted Odds Ratio (95% CI) |
3rd Tertile, n = 1308 [15.6%‐19.8%], Adjusted Odds Ratio (95% CI) |
|
| |||
| Mortality measure (no. of hospitals reporting) | |||
| Acute myocardial infarction (n = 2415) | 1.00 (referent) | 0.88 (0.711.09) | 1.02 (0.831.25) |
| Pneumonia (n = 4067) | 1.00 (referent) | 0.83 (0.710.98) | 1.11 (0.941.31) |
| Heart failure (n = 3668) | 1.00 (referent) | 1.21 (1.021.45) | 1.94 (1.632.30) |
| Stroke (n = 2754) | 1.00 (referent) | 1.13 (0.931.38) | 1.48 (1.221.79) |
| Chronic obstructive pulmonary disease (n = 3633) | 1.00 (referent) | 1.12 (0.951.33) | 1.73 (1.462.05) |
| Coronary artery bypass (n = 1058) | 1.00 (referent) | 0.87 (0.631.19) | 0.99 (0.741.34) |
| Condition‐specific readmission rate tertile | |||
| Mortality measure | |||
| Acute myocardial infarction | 1.00 (referent) | 0.88 (0.711.08) | 0.79 (0.640.99) |
| Pneumonia | 1.00 (referent) | 0.91 (0.781.07) | 0.89 (0.761.04) |
| Heart failure | 1.00 (referent) | 1.15 (0.961.36) | 1.56 (1.311.86) |
| Stroke | 1.00 (referent) | 1.65 (1.342.03) | 1.70 (1.232.35) |
| Chronic obstructive pulmonary disease | 1.00 (referent) | 0.83 (0.700.98) | 0.84 (0.710.99) |
| Coronary artery bypass | 1.00 (referent) | 0.59 (0.440.80) | 0.47 (0.340.64) |
DISCUSSION
We found that higher hospital‐wide readmission rates were associated with lower mortality at the hospital level for 3 of the 6 mortality measures we examined. The findings for heart failure parallel the findings of Krumholz and colleagues who examined 3 of these 6 measures (MI, pneumonia, and heart failure) in relation to readmissions for these specific populations.[3] This prior analysis, however, did not include the 3 more recently reported mortality measures (COPD, stroke, and CABG) and did not use hospital‐wide readmissions.
Causal mechanisms underlying the associations between mortality and readmission at the hospital level deserve further exploration. It is certainly possible that global efforts to keep patients out of the hospital might, in some instances, place patients at risk by delaying necessary acute care.[2] It is also possible that unmeasured variables, particularly access to hospice and palliative care services that might facilitate good deaths, could be associated with both reduced readmissions and higher death rates. Additionally, because deceased patients cannot be readmitted, one might expect that readmissions and mortality might be inversely associated, particularly for conditions with a high postdischarge mortality rate. Similarly, a hospital that does a particularly good job keeping chronically ill patients alive until discharge might exhibit a higher readmission rate than a hospital that is less adept at keeping tenuous patients alive until discharge.
Regardless of the mechanisms of these findings, we present these data to raise the concern that using readmission rates, particularly hospital‐wide readmission rates, as a measure of hospital quality is inherently problematic. It is particularly problematic that CMS has applied equal weight to readmissions and mortality in the Star Report.[1] High readmission rates may result from complications and poor handoffs, but may also stem from the legitimate need to care for chronically ill patients in a high‐intensity setting, particularly fragile patients who have been kept alive against the odds. In conclusion, caution is warranted in viewing readmissions as a quality metric until the associations we describe are better explained using patient‐level data and more robust adjustment than is possible with these publically available data.
Disclosures: Dr. Daniel J. Brotman had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. There was no financial support for this work. Contributions of the authors are as follows: drafting manuscript (Brotman), revision of manuscript for important intellectual content (brotman, Hoyer, Lepley, Deutschendorf, Leung), acquisition of data (Deutschendorf, Leung, Lepley), interpretation of data (Brotman, Hoyer, Lepley, Deutschendorf, Leung), data analysis (Brotman, Hoyer).
- Centers for Medicare and Medicaid Services. Available at: https://www.cms.gov/Outreach-and-Education/Outreach/NPC/Downloads/2015-08-13-Star-Ratings-Presentation.pdf. Accessed September 2015.
- , , , et al. A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial. Ann Intern Med. 2012;156(10):673–683.
- , , , et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. JAMA. 2013;309(6):587–593.
- Centers for Medicare and Medicaid Services. Hospital compare datasets. Available at: https://data.medicare.gov/data/hospital‐compare. Accessed September 2015.
- Centers for Medicare and Medicaid Services. Hospital quality initiative. Available at: https://www.cms.gov/Medicare/Quality‐Initiatives‐Patient‐Assessment‐Instruments/HospitalQualityInits. Accessed September 2015.
- Centers for Medicare and Medicaid Services. Available at: https://www.cms.gov/Outreach-and-Education/Outreach/NPC/Downloads/2015-08-13-Star-Ratings-Presentation.pdf. Accessed September 2015.
- , , , et al. A comprehensive care management program to prevent chronic obstructive pulmonary disease hospitalizations: a randomized, controlled trial. Ann Intern Med. 2012;156(10):673–683.
- , , , et al. Relationship between hospital readmission and mortality rates for patients hospitalized with acute myocardial infarction, heart failure, or pneumonia. JAMA. 2013;309(6):587–593.
- Centers for Medicare and Medicaid Services. Hospital compare datasets. Available at: https://data.medicare.gov/data/hospital‐compare. Accessed September 2015.
- Centers for Medicare and Medicaid Services. Hospital quality initiative. Available at: https://www.cms.gov/Medicare/Quality‐Initiatives‐Patient‐Assessment‐Instruments/HospitalQualityInits. Accessed September 2015.
Discharge Summaries and Readmissions
Across the continuum of care, the discharge summary is a critical tool for communication among care providers.[1] In the United States, the Joint Commission policies mandate that all hospital providers complete a discharge summary for patients with specific components to foster effective communication with future providers.[2] Because outpatient providers and emergency physicians rely on clinical information in the discharge summary to ensure appropriate postdischarge continuity of care, timely documentation is potentially an essential aspect of readmission reduction initiatives.[3, 4, 5] Prior reports indicate that poor discharge documentation of follow‐up plan‐of‐care increases the risk of hospitalization, whereas structured instructions, patient education, and direct communications with primary care physicians (PCPs) reduce repeat hospital visits.[6, 7, 8, 9] However, the current literature is limited in its narrow focus on the contents of discharge summaries, considered only same‐hospital readmissions, or considered readmissions within 3 months of discharge.[10, 11, 12, 13] Moreover, some prior research has suggested no association between discharge summary timeliness with readmission,[12, 13, 14] whereas another study did find a relationship,[15] hence the need to study this further is important. Filling this gap in knowledge could provide an avenue to track and improve quality of patient care, as delays in discharge summaries have been linked with pot‐discharge adverse outcomes and patient safety concerns.[15, 16, 17, 18] Because readmissions often occur soon after discharge, having timely discharge summaries may be particularly important to outcomes.[19, 20]
This research began under the framework of evaluating a bundle of care coordination strategies that were implemented at the Johns Hopkins Health System. These strategies were informed by the early Centers for Medicare and Medicaid Services (CMS) demonstration projects and other best practices that have been documented in the literature to improve utilization and improve communication during transitions of care.[21, 22, 23, 24, 25] Later they were augmented through a contract with the Center of Medicare and Medicaid Innovation to improve access to healthcare services and improve patient outcomes through improved care coordination processes. One of the domains our institution has increased efforts to improve is in provider handoffs. Toward that goal, we have worked to disentangle the effects of different factors of provider‐to‐provider communication that may influence readmissions.[26] For example, effective written provider handoffs in the form of accurate and timely discharge summaries was considered a key care coordination component of this program, but there was institutional resistance to endorsing an expectation that discharge summary turnaround should be shortened. To build a case for this concept, we sought to test the hypothesis that, at our hospital, longer time to complete hospital discharge summaries was associated with increased readmission rates. Unique to this analysis is that, in the state of Maryland, there is statewide reporting of readmissions, so we were able to account for intra‐ and interhospital readmissions for an all‐payer population. The authors anticipated that findings from this study would help inform discharge quality‐improvement initiatives and reemphasize the importance of timely discharge documentation across all disciplines as part of quality patient care.
METHODS
Study Population and Setting
We conducted a single‐center, retrospective cohort study of 87,994 consecutive patients discharged from Johns Hopkins Hospital, which is a 1000‐bed, tertiary academic medical center in Baltimore, Maryland between January 1, 2013 and December 31, 2014. One thousand ninety‐three (1.2%) of the records on days to complete the discharge summary were missing and were excluded from the analysis.
Data Source and Covariates
Data were derived from several sources. The Johns Hopkins Hospital data mart financial database, used for mandatory reporting to the State of Maryland, provided the following patient data: age, gender, race/ethnicity, payer (Medicare, Medicaid, and other) as a proxy for socioeconomic status,[27] hospital service prior to discharge (gynecologyobstetrics, medicine, neurosciences, oncology, pediatrics, and surgical sciences), hospital length of stay (LOS) prior to discharge, Agency for Healthcare Research and Quality (AHRQ) Comorbidity Index (which is an update to the original Elixhauser methodology[28]), and all‐payerrefined diagnosis‐related group (APRDRG) and severity of illness (SOI) combinations (a tool to group patients into clinically comparable disease and SOI categories expected to use similar resources and experience similar outcomes). The Health Services Cost Review Commission (HSCRC) in Maryland provided the observed readmission rate in Maryland for each APRDRG‐SOI combination and served as an expected readmission rate. This risk stratification methodology is similar to the approach used in previous studies.[26, 29] Discharge summary turnaround time was obtained from institutional administrative databases used to track compliance with discharge summary completion. Discharge location (home, facility, home with homecare or hospice, or other) was obtained from Curaspan databases (Curaspan Health Group, Inc., Newton, MA).
Primary Outcome: 30‐Day Readmission
The primary outcome was unplanned rehospitalizations to an acute care hospital in Maryland within 30 days of discharge from Johns Hopkins Hospital. This was as defined by the Maryland HSCRC using an algorithm to exclude readmissions that were likely to be scheduled, as defined by the index admission diagnosis and readmission diagnosis; this algorithm is updated based on the CMS all‐cause readmission algorithm.[30, 31]
Primary Exposure: Days to Complete the Discharge Summary
Discharge summary completion time was defined as the date when the discharge attending physician electronically signs the discharge summary. At our institution, an auto‐fax system sends documents (eg, discharge summaries, clinic notes) to linked providers (eg, primary care providers) shortly after midnight from the day the document is signed by an attending physician. During the period of the project, the policy for discharge summaries at the Johns Hopkins Hospital went from requiring them to be completed within 30 days to 14 days, and we were hoping to use our analyses to inform decision makers why this was important. To emphasize the need for timely completion of discharge summaries, we dichotomized the number of days to complete the discharge summary into >3 versus 3 days (20th percentile cutoff) and modeled it as a continuous variable (per 3‐day increase in days to complete the discharge summary).
Statistical Analysis
To evaluate differences in patient characteristics by readmission status, analysis of variance and 2 tests were used for continuous and dichotomous variables, respectively. Logistic regression was used to evaluate the association between days to complete the discharge summary >3 days and readmission status, adjusting for potentially confounding variables. Before inclusion in the logistic regression model, we confirmed a lack of multicollinearity in the multivariable regression model using variance inflation factors. We evaluated residual versus predicted value plots and residual versus fitted value plots with a locally weighted scatterplot smoothing line. In a sensitivity analysis we evaluated the association between readmission status and different cutoffs (>8 days, 50th percentile; and >14 days, 70% percentile). In a separate analysis, we used interaction terms to test whether the association between the association between days to complete the discharge summary >3 days and hospital readmission varied by the covariates in the analysis (age, sex, race, payer, hospital service, discharge location, LOS, APRDRG‐SOI expected readmission rate, and AHRQ Comorbidity Index). We observed a significant interaction between 30‐day readmission and days to complete the discharge summary >3 days by hospital service. Hence, we separately calculated the adjusted mean readmission rates separately for each hospital service using the least squared means method for the multivariable logistic regression analysis and adjusting for the previously mentioned covariates. In a separate analysis, we used linear regression to evaluate the association between LOS and days to complete the discharge summary, adjusting for potentially confounding variables. Statistical significance was defined as a 2‐sided P < 0.05. Data were analyzed with R (version 2.15.0; R Foundation for Statistical Computing, Vienna, Austria;
RESULTS
Readmitted Patients
In the study period, 14,248 out of 87,994 (16.2%) consecutive eligible patients were readmitted to a hospital in Maryland from patients discharged from Johns Hopkins Hospital between January 1, 2013 and December 31, 2014. A total of 11,027 (77.4%) of the readmissions were back to Johns Hopkins Hospital. Table 1 compares characteristics of readmitted versus nonreadmitted patients, with the following variables being significantly different between these patient groups: age, gender, healthcare payer, hospital service, discharge location, length of stay expected readmission rate, AHRQ Comorbidity Index, and days to complete inpatient discharge summary.
| Characteristics | All Patients, N = 87,994 | Not Readmitted, N = 73,746 | Readmitted, N = 14,248 | P Value |
|---|---|---|---|---|
| ||||
| Age, y | 42.1 (25.1) | 41.3 (25.4) | 46.4 (23.1) | <0.001 |
| Male | 43,210 (49.1%) | 35,851 (48.6%) | 7,359 (51.6%) | <0.001 |
| Race | <0.001 | |||
| Caucasian | 45,705 (51.9%) | 3,8661 (52.4%) | 7,044 (49.4%) | |
| African American | 32,777 (37.2%) | 2,6841 (36.4%) | 5,936 (41.7%) | |
| Other | 9,512 (10.8%) | 8,244 (11.2%) | 1,268 (8.9%) | |
| Payer | <0.001 | |||
| Medicare | 22,345 (25.4%) | 17,614 (23.9%) | 4,731 (33.2%) | |
| Medicaid | 24,080 (27.4%) | 20,100 (27.3%) | 3,980 (27.9%) | |
| Other | 41,569 (47.2%) | 36,032 (48.9%) | 5,537 (38.9%) | |
| Hospital service | <0.001 | |||
| Gynecologyobstetrics | 9,299 (10.6%) | 8,829 (12.0%) | 470 (3.3%) | |
| Medicine | 26,036 (29.6%) | 20,069 (27.2%) | 5,967 (41.9%) | |
| Neurosciences | 8,269 (9.4%) | 7,331 (9.9%) | 938 (6.6%) | |
| Oncology | 5,222 (5.9%) | 3,898 (5.3%) | 1,324 (9.3%) | |
| Pediatrics | 17,029 (19.4%) | 14,684 (19.9%) | 2,345 (16.5%) | |
| Surgical sciences | 22,139 (25.2%) | 18,935 (25.7%) | 3,204 (22.5%) | |
| Discharge location | <0.001 | |||
| Home | 65,478 (74.4%) | 56,359 (76.4%) | 9,119 (64.0%) | |
| Home with homecare or hospice | 9,524 (10.8%) | 7,440 (10.1%) | 2,084 (14.6%) | |
| Facility (SNF, rehabilitation facility) | 5,398 (6.1%) | 4,131 (5.6%) | 1,267 (8.9%) | |
| Other | 7,594 (8.6%) | 5,816 (7.9%) | 1,778 (12.5%) | |
| Length of stay, d | 5.5 (8.6) | 5.1 (7.8) | 7.5 (11.6) | <0.001 |
| APRDRG‐SOI Expected Readmission Rate, % | 14.4 (9.5) | 13.3 (9.2) | 20.1 (9.0) | <0.001 |
| AHRQ Comorbidity Index (1 point) | 2.5 (1.4) | 2.4 (1.4) | 3.0 (1.8) | <0.001 |
| Discharge summary completed >3 days | 66,242 (75.3%) | 55,329 (75.0%) | 10,913 (76.6%) | <0.001 |
Association Between Days to Complete the Discharge Summary and Readmission
After hospital discharge, median (IQR) number of days to complete discharge summaries was 8 (416) days. After hospital discharge, median (IQR) number of days to complete discharge summaries and the number of days from discharge to readmission was 8 (416) and 11 (519) days, respectively (P < 0.001). Six thousand one hundred one patients (42.8%) were readmitted before their discharge summary was completed. The median (IQR) days to complete discharge summaries by hospital service in order from shortest to longest was: oncology, 6 (212) days; surgical sciences, 6 (312) days; pediatrics, 7 (315) days; gynecologyobstetrics, 8 (415) days; medicine, 9 (420) days; neurosciences, 12 (621) days.
When we divided the number of days to complete the discharge summary into deciles (02, 2.13, 3.14, 4.16, 6.18, 8.210, 10.114, 14.119, 19.130, >30), a longer number of days to complete discharge summaries had higher unadjusted and adjusted readmission rates (Figure 1). In unadjusted analysis, Table 2 shows that older age, male sex, African American race, oncological versus medicine hospital service, discharge location, longer LOS, higher APRDRG‐SOI expected readmission rate, and higher AHRQ Comorbidity Index were associated with readmission. Days to complete the discharge summary >3 days versus 3 days was associated with a higher readmission rate, with an unadjusted odds ratio (OR) and 95% confidence interval (CI) of 1.09 (95% CI: 1.04 to 1.13, P < 0.001).
| Characteristic | Bivariable Analysis* | Multivariable Analysis* | ||
|---|---|---|---|---|
| OR (95% CI) | P Value | OR (95% CI) | P Value | |
| ||||
| Age, 10 y | 1.09 (1.08 to 1.09) | <0.001 | 0.97 (0.95 to 0.98) | <0.001 |
| Male | 1.13 (1.09 to 1.17) | <0.001 | 1.01 (0.97 to 1.05) | 0.76 |
| Race | ||||
| Caucasian | Referent | Referent | ||
| African American | 1.21 (1.17 to 1.26) | <0.001 | 1.01 (0.96 to 1.05) | 0.74 |
| Other | 0.84 (0.79 to 0.90) | <0.001 | 0.92 (0.86 to 0.98) | 0.01 |
| Payer | ||||
| Medicare | Referent | Referent | ||
| Medicaid | 0.74 (0.70 to 0.77) | <0.001 | 1.03 (0.97 to 1.09) | 0.42 |
| Other | 0.57 (0.55 to 0.60) | <0.001 | 0.86 (0.82 to 0.91) | <0.001 |
| Hospital service | ||||
| Medicine | Referent | Referent | ||
| Gynecologyobstetrics | 0.18 (0.16 to 0.20) | <0.001 | 0.50 (0.45 to 0.56) | <0.001 |
| Neurosciences | 0.43 (0.40 to 0.46) | <0.001 | 0.76 (0.70 to 0.82) | <0.001 |
| Oncology | 1.14 (1.07 to 1.22) | <0.001 | 1.18 (1.10 to 1.28) | <0.001 |
| Pediatrics | 0.54 (0.51 to 0.57) | <0.001 | 0.77 (0.71 to 0.83) | <0.001 |
| Surgical sciences | 0.57 (0.54 to 0.60) | <0.001 | 0.92 (0.87 to 0.97) | 0.002 |
| Discharge location | ||||
| Home | Referent | |||
| Facility (SNF, rehabilitation facility) | 1.90 (1.77 to 2.03) | <0.001 | 1.11 (1.02 to 1.19) | 0.009 |
| Home with homecare or hospice | 1.73 (1.64 to 1.83) | <0.001 | 1.26 (1.19 to 1.34) | <0.001 |
| Other | 1.89 (1.78 to 2.00) | <0.001 | 1.25 (1.18 to 1.34) | <0.001 |
| Length of stay, d | 1.03 (1.02 to 1.03) | <0.001 | 1.00 (1.00 to 1.01) | <0.001 |
| APRDRG‐SOI expected readmission rate, % | 1.08 (1.07 to 1.08) | <0.001 | 1.06 (1.06 to 1.06) | <0.001 |
| AHRQ Comorbidity Index (1 point) | 1.27 (1.26 to 1.28) | <0.001 | 1.11 (1.09 to 1.12) | <0.001 |
| Discharge summary completed >3 days | 1.09 (1.04 to 1.14) | <0.001 | 1.09 (1.05 to 1.14) | <0.001 |
Multivariable and Secondary Analyses
In adjusted analysis (Table 2), patients discharged from an oncologic service relative to a medicine hospital service (OR: 1.19, 95% CI: 1.10 to 1.28, P < 0.001), patients discharged to a facility, home with homecare or hospice, or other location compared to home (facility OR: 1.11, 95% CI: 1.02 to 1.19, P = 0.009; home with homecare or hospice OR: 1.26, 95% CI: 1.19 to 1.34, P < 0.001; other OR: 1.25, 95% CI: 1.18 to 1.34, P < 0.001), patients with longer LOS (OR: 1.11 per day, 95% CI: 1.10 to 1.12, P < 0.001), patients with a higher expected readmission rates (OR: 1.01 per percent, 95% CI: 1.00 to 1.01, P < 0.001), and patients with a higher AHRQ comorbidity index (OR: 1.06 per 1 point, 95% CI: 1.06 to 1.06, P < 0.001) had higher 30‐day readmission rates. Overall, days to complete the discharge summary >3 days versus 3 days was associated with a higher readmission rate (OR: 1.09, 95% CI: 1.05 to 1.14, P < 0.001).
In a sensitivity analysis, discharge summary completion >8 days (median) versus 8 days was associated with higher unadjusted readmission rate (OR: 1.11, 95% CI: 1.07 to 1.15, P < 0.001) and a higher adjusted readmission rate (OR: 1.06, 95% CI: 1.02 to 1.10, P < 0.001). Discharge summary completion >14 days (70th percentile) versus 14 days was also associated with higher unadjusted readmission rate (OR: 1.15, 95% CI: 1.08 to 1.21, P < 0.001) and a higher adjusted readmission rate (OR: 1.09, 95% CI: 1.02 to 1.16, P = 0.008). The association between days to complete the discharge summary >3 days and readmissions was found to vary significantly by hospital service (P = 0.03). For comparing days to complete the discharge summary >3 versus 3 days, Table 3 shows that neurosciences, pediatrics, oncology, and medicine hospital services were associated with significantly increased adjusted mean readmission rates. Additionally, when days to complete the discharge summary was modeled as a continuous variable, we found that for every 3 days the odds of readmission increased by 1% (OR: 1.01, 95% CI: 1.00 to 1.01, P < 0.001).
| Days to Complete Discharge Summary by Hospital Service | Adjusted Mean Readmission Rate (95% CI)* | P Value |
|---|---|---|
| ||
| Gynecologyobstetrics | 0.30 | |
| 03 days, n = 1,792 | 5.4 (4.1 to 6.7) | |
| >3 days, n = 7,507 | 6.0 (4.9 to 7.0) | |
| Medicine | 0.04 | |
| 03 days, n = 6,137 | 21.1 (20.0 to 22.3) | |
| >3 days, n = 19,899 | 22.4 (21.6 to 23.2) | |
| Neurosciences | 0.02 | |
| 03 days, n = 1,116 | 10.1 (8.2 to 12.1) | |
| >3 days, n = 7,153 | 12.5 (11.6 to 13.5) | |
| Oncology | 0.01 | |
| 03 days, n = 1,885 | 25.0 (22.6 to 27.4) | |
| >3 days, n = 3,337 | 28.2 (26.6 to 30.2) | |
| Pediatrics | 0.001 | |
| 03 days, n = 4,561 | 9.5 (6.9 to 12.2) | |
| >3 days, n = 12,468 | 11.4 (8.9 to 13.9) | |
| Surgical sciences | 0.89 | |
| 03 days, n = 6,261 | 15.2 (14.2 to 16.1) | |
| >3 days, n = 15,878 | 15.1 (14.4 to 15.8) | |
In an unadjusted analysis, we found that the relationship between LOS and days to complete the discharge summary was not significant ( coefficient and 95% CI:, 0.01, 0.02 to 0.00, P = 0.20). However, we found a small but significant relationship in our multivariable analysis, such that each hospitalization day was associated with a 0.01 (95% CI: 0.00 to 0.02, P = 0.03) increase in days to complete the discharge summary.
DISCUSSION
In this single‐center retrospective analysis, the number of days to complete the discharge summary was significantly associated with readmissions after hospitalization. This association was independent of age, gender, comorbidity index, payer, discharge location, length of hospital stay, expected readmission rate based on diagnosis and severity of illness, and all hospital services. The odds of readmission for patients with delayed discharge summaries was small but significant. This is important in the current landscape of readmissions, particularly for institutions who are challenged to reduce readmission rates, and a small relative difference in readmissions may be the difference between getting penalized or not. In the context of prior studies, the results highlight the role of timely discharge summary as an under‐recognized metric, which may be a valid litmus test for care coordination. The findings also emphasize the potential of early summaries to expedite communication and to help facilitate quality of patient care. Hence, the study results extend the literature examining the relationship of delay in discharge summary with unfavorable patient outcomes.[15, 32]
In contrast to prior reports with limited focus on same‐hospital readmissions,[18, 33, 34, 35] readmissions beyond 30 days,[12] or focused on a specific patient population,[13, 36] this study evaluates both intra‐ and interhospital 30‐day readmissions in Maryland in an all‐payer, multi‐institution, diverse patient population. Additionally, prior research is conflicting with respect to whether timely discharges summaries are significantly associated with increased hospital readmissions.[12, 13, 14, 15] Although it is not surprising that inadequate care during hospitalization could result in readmissions, the role of discharge summaries remain underappreciated. Having a timely discharge summary may not always prevent readmissions, but our study showed that 43% of readmission occurred before the discharge summary completion. Not having a completed discharge summary at the time of readmission may have been a driver for the positive association between timely completion and 30‐day readmission we observed. This study highlights that delay in the discharge summary could be a marker of poor transitions of care, because suboptimal dissemination of critical information to care providers may result in discontinuity of patient care posthospitalization.
A plausible mechanism of the association between discharge summary delays and readmissions could be the provision of collateral information, which may potentially alter the threshold for readmissions. For example, in the emergency room/emergency department (ER/ED) setting, discharge summaries may help with preventable readmissions. For patients who present repeatedly with the same complaint, timely summaries to ER/ED providers may help reframe the patient complaints, such as patient has concern X, which was previously identified to be related to diagnosis Y. As others have shown, the content of discharge summaries, format, and accessibility (electronic vs paper chart), as well as timely distribution of summaries, are key factors that impact quality outcomes.[2, 12, 15, 37, 38] By detailing prior hospital information (ie, discharge medications, prior presentations, tests completed), summaries could help prevent errors in medication dosing, reduce unnecessary testing, and help facilitate admission triage. Summaries may have information regarding a new diagnosis such as the results of an endoscopic evaluation that revealed the source of occult gastrointestinal bleeding, which could help contextualize a complaint of repeat melena and redirect goals of care. Discussions of goals of care in the discharge summary may guide primary providers in continued care management plans.
Our study findings underscore a positive correlation between late discharge summaries and readmissions. However, the extent that this is a causal relationship is unclear; the association of delay in days to complete the discharge summary with readmission may be an epiphenomenon related to processes related to quality of clinical care. For example, delays in discharge summary completion could be a marker of other system issues, such as a stressed work environment. It is possible that providers who fail to complete timely discharge summaries may also fail to do other important functions related to transitions of care and care coordination. However, even if this is so, timely discharge summaries could become a focal point for discussion for optimization of care transitions. A discharge summary could be delayed because the patient has already been readmitted before the summary was distributed, thus making that original summary less relevant. Delays could also be a reflection of the data complexity for patients with longer hospital stays. This is supported by the small but significant relationship between LOS and days to complete the discharge summary in this study. Lastly, delays in discharge summary completion may also be a proxy of provider communication and can reflect the culture of communication at the institution.
Although unplanned hospital readmission is an important outcome, many readmissions may be related to other factors such as disease progression, rather than late summaries or the lack of postdischarge communication. For instance, prior reports did not find any association between the PCP seeing the discharge summaries or direct communications with the PCP and 30‐day clinical outcomes for readmission and death.[26, 39] However, these studies were limited in their use of self‐reported handoffs, did not measure quality of information transfer, and failed to capture a broader audience beyond the PCP, such as ED physicians or specialists.
Our results suggest that the relationship between days to complete discharge summaries and 30‐day readmissions may vary depending on whether the hospitalization is primarily surgical/procedural versus medical treatment. A recent study found that most readmissions after surgery were associated with new complications related to the procedure and not exacerbation of prior index hospitalization complications.[40] Hence, treatment for common causes of hospital readmissions after surgical or gynecological procedures, such as wound infections, acute anemia, ileus, or dehydration, may not necessarily require a completed discharge summary for appropriate management. However, we caution extending this finding to clinical practice before further studies are conducted on specific procedures and in different clinical settings.
Results from this study also support institutional policies that specify the need for practitioners to complete discharge summaries contemporaneously, such as at the time of discharge or within a couple of days. Unlike other forms of communication that are optional, discharge summaries are required, so we recommend that practitioners be held accountable for short turnaround times. For example, providers could be graded and rated on timely completions of discharge summaries, among other performance variables. Anecdotally at our institutions, we have heard from practitioners that it takes less time to complete them when you do them on the day of discharge, because the hospitalization course is fresher in their mind and they have to wade through less information in the medical record to complete an accurate discharge summary. To this point, a barrier to on‐time completion is that providers may have misconceptions about what is really vital information to convey to the next provider. In agreement with past research and in the era of the electronic medical record system, we recommend that the discharge summary should be a quick synthesis of key findings that incorporates only the important elements, such as why the patient was hospitalized, what were key findings and key responses to therapy, what is pending at the time of discharge, what medications the patient is currently taking, and what are the follow‐up plans, rather than a lengthy expose of all the findings.[13, 36, 41, 42]
Lastly, our study results should be taken in the context of its limitations. As a single‐center study, findings may lack generalizability. In particular, the results may not generalize to hospitals that lack access to statewide reporting. We were also not able to assess readmission for patients who may have been readmitted to a hospital outside of Maryland. Although we adjusted for pertinent variables such as age, gender, healthcare payer, hospital service, comorbidity index, discharge location, LOS, and expected readmission rates, there may be other relevant confounders that we failed to capture or measure optimally. Median days to complete the discharge summary in this study was 8 days, which is longer than practices at other institutions, and may also limit this study's generalizability.[15, 36, 42] However, prior research supports our findings,[15] and a systematic review found that only 29% and 52% of discharge summaries were completed by 2 weeks and 4 weeks, respectively.[9] Finally, as noted above and perhaps most important, it is possible that discharge summary turnaround time does not in itself causally impact readmissions, but rather reflects an underlying commitment of the inpatient team to effectively coordinate care following hospital discharge.
CONCLUSION
In sum, this study delineates an underappreciated but important relationship of timely discharge summary completion and readmission outcomes. The discharge summary may be a relevant metric reflecting quality of patient care. Healthcare providers may begin to target timely discharge summaries as a potential focal point of quality‐improvement projects with the goal to facilitate better patient outcomes.
Disclosures
The authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated, and, if applicable, the authors certify that all financial and material support for this research (eg, Centers for Medicare and Medicaid Services, National Institutes of Health, or National Health Service grants) and work are clearly identified. This study was supported by funding opportunity, number CMS‐1C1‐12‐0001, from the Centers for Medicare and Medicaid Services and Center for Medicare and Medicaid Innovation. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Department of Health and Human Services or any of its agencies.
- , , , et al. Development and sustainability of an inpatient‐to‐outpatient discharge handoff tool: a quality improvement project. Jt Comm J Qual Patient Saf. 2014;40(5):219–227.
- , , , , , . Documentation of mandated discharge summary components in transitions from acute to subacute care. In: Henriksen K, Battles JB, Keyes MA, et al., eds. Advances in Patient Safety: New Directions and Alternative Approaches. Vol. 2. Culture and Redesign. Rockville, MD: Agency for Healthcare Research and Quality; 2008.
- , , , . Better transitions: improving comprehension of discharge instructions. Front Health Serv Manage. 2009;25(3):11–32.
- , , , . Structured discharge education improves early outcome in orthopedic patients. Int J Orthop Trauma Nurs. 2010;14(2):66–74.
- , , , et al. Hospital discharge documentation and risk of rehospitalisation. BMJ Qual Saf. 2011;20(9):773–778.
- , , . The hospital discharge: a review of a high risk care transition with highlights of a reengineered discharge process. J Patient Saf. 2007;3(2):97–106.
- , , , , . Interventions to reduce 30‐day rehospitalization: a systematic review. Ann Intern Med. 2011;155(8):520–528.
- , , , , , . Effect of hospital follow‐up appointment on clinical event outcomes and mortality. Arch Intern Med. 2010;170(11):955–960.
- , , , , , . Deficits in communication and information transfer between hospital‐based and primary care physicians: implications for patient safety and continuity of care. JAMA. 2007;297(8):831–841.
- , , , , . Provider characteristics, clinical‐work processes and their relationship to discharge summary quality for sub‐acute care patients. J Gen Intern Med. 2012;27(1):78–84.
- , , , et al. Contemporary evidence about hospital strategies for reducing 30‐day readmissions: a national study. J Am Coll Cardiol. 2012;60(7):607–614.
- , , , . Effect of discharge summary availability during post‐discharge visits on hospital readmission. J Gen Intern Med. 2002;17(3):186–192.
- , , , et al. Association of discharge summary quality with readmission risk for patients hospitalized with heart failure exacerbation. Circ Cardiovasc Qual Outcomes. 2015;8(1):109–111.
- , , , et al. The independent association of provider and information continuity on outcomes after hospital discharge: implications for hospitalists. J Hosp Med. 2010;5(7):398–405.
- , , , , . Timeliness in discharge summary dissemination is associated with patients' clinical outcomes. J Eval Clin Pract. 2013;19(1):76–79.
- , , , et al. Communication and information deficits in patients discharged to rehabilitation facilities: an evaluation of five acute care hospitals. J Hosp Med. 2009;4(8):E28–E33.
- , , . Preventing readmissions through comprehensive discharge planning. Prof Case Manag. 2013;18(2):56–63; quiz 64–65.
- , , , et al. Effect of a postdischarge virtual ward on readmission or death for high‐risk patients: a randomized clinical trial. JAMA. 2014;312(13):1305–1312..
- , , . Risk factors for early unplanned hospital readmission in the elderly. J Gen Intern Med. 1991;6(3):223–228.
- , , , , . Differences between early and late readmissions among patients: a cohort study. Ann Intern Med. 2015;162(11):741–749.
- , , , et al. Post‐acute care payment reform demonstration report to congress supplement‐interim report. Centers for Medicare 14(3):1–14; quiz 88–89.
- , , , et al. Comprehensive discharge planning and home follow‐up of hospitalized elders: a randomized clinical trial. JAMA. 1999;281(7):613–620.
- , , , . Posthospital care transitions: patterns, complications, and risk identification. Health Serv Res. 2004;39(5):1449–1465.
- , , , et al. Transitions of care consensus policy statement: American College of Physicians, Society of General Internal Medicine, Society of Hospital Medicine, American Geriatrics Society, American College of Emergency Physicians, and Society for Academic Emergency Medicine. J Hosp Med. 2009;4(6):364–370.
- , , , et al. Association of self‐reported hospital discharge handoffs with 30‐day readmissions. JAMA Intern Med. 2013;173(8):624–629.
- , . Socioeconomic disparities in health: pathways and policies. Health Aff (Millwood). 2002;21(2):60–76.
- , , , . Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27.
- , , , , , . Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951–1958.
- Centers for Medicare 35(10):1044–1059.
- , , , et al. Understanding and execution of discharge instructions. Am J Med Qual. 2013;28(5):383–391.
- , , , , . Beyond the hospital gates: elucidating the interactive association of social support, depressive symptoms, and physical function with 30‐day readmissions. Am J Phys Med Rehabil. 2015;94(7):555–567.
- , , , et al. Improving the discharge process by embedding a discharge facilitator in a resident team. J Hosp Med. 2011;6(9):494–500.
- , , , et al. Hospital variation in quality of discharge summaries for patients hospitalized with heart failure exacerbation. Circ Cardiovasc Qual Outcomes. 2015;8(1):77–86
- , , , . Addressing the business of discharge: building a case for an electronic discharge summary. J Hosp Med. 2011;6(1):37–42.
- , , , , . Joint commission requirements for discharge instructions in patients with heart failure: is understanding important for preventing readmissions? J Card Fail. 2014;20(9):641–649.
- , , , et al. Association of communication between hospital‐based physicians and primary care providers with patient outcomes. J Gen Intern Med. 2009;24(3):381–386.
- , , , et al. Underlying reasons associated with hospital readmission following surgery in the united states. JAMA. 2015;313(5):483–495.
- , , , , . Assessing quality and efficiency of discharge summaries. Am J Med Qual. 2005;20(6):337–343.
- , , , et al. Comprehensive quality of discharge summaries at an academic medical center. J Hosp Med. 2013;8(8):436–443.
Across the continuum of care, the discharge summary is a critical tool for communication among care providers.[1] In the United States, the Joint Commission policies mandate that all hospital providers complete a discharge summary for patients with specific components to foster effective communication with future providers.[2] Because outpatient providers and emergency physicians rely on clinical information in the discharge summary to ensure appropriate postdischarge continuity of care, timely documentation is potentially an essential aspect of readmission reduction initiatives.[3, 4, 5] Prior reports indicate that poor discharge documentation of follow‐up plan‐of‐care increases the risk of hospitalization, whereas structured instructions, patient education, and direct communications with primary care physicians (PCPs) reduce repeat hospital visits.[6, 7, 8, 9] However, the current literature is limited in its narrow focus on the contents of discharge summaries, considered only same‐hospital readmissions, or considered readmissions within 3 months of discharge.[10, 11, 12, 13] Moreover, some prior research has suggested no association between discharge summary timeliness with readmission,[12, 13, 14] whereas another study did find a relationship,[15] hence the need to study this further is important. Filling this gap in knowledge could provide an avenue to track and improve quality of patient care, as delays in discharge summaries have been linked with pot‐discharge adverse outcomes and patient safety concerns.[15, 16, 17, 18] Because readmissions often occur soon after discharge, having timely discharge summaries may be particularly important to outcomes.[19, 20]
This research began under the framework of evaluating a bundle of care coordination strategies that were implemented at the Johns Hopkins Health System. These strategies were informed by the early Centers for Medicare and Medicaid Services (CMS) demonstration projects and other best practices that have been documented in the literature to improve utilization and improve communication during transitions of care.[21, 22, 23, 24, 25] Later they were augmented through a contract with the Center of Medicare and Medicaid Innovation to improve access to healthcare services and improve patient outcomes through improved care coordination processes. One of the domains our institution has increased efforts to improve is in provider handoffs. Toward that goal, we have worked to disentangle the effects of different factors of provider‐to‐provider communication that may influence readmissions.[26] For example, effective written provider handoffs in the form of accurate and timely discharge summaries was considered a key care coordination component of this program, but there was institutional resistance to endorsing an expectation that discharge summary turnaround should be shortened. To build a case for this concept, we sought to test the hypothesis that, at our hospital, longer time to complete hospital discharge summaries was associated with increased readmission rates. Unique to this analysis is that, in the state of Maryland, there is statewide reporting of readmissions, so we were able to account for intra‐ and interhospital readmissions for an all‐payer population. The authors anticipated that findings from this study would help inform discharge quality‐improvement initiatives and reemphasize the importance of timely discharge documentation across all disciplines as part of quality patient care.
METHODS
Study Population and Setting
We conducted a single‐center, retrospective cohort study of 87,994 consecutive patients discharged from Johns Hopkins Hospital, which is a 1000‐bed, tertiary academic medical center in Baltimore, Maryland between January 1, 2013 and December 31, 2014. One thousand ninety‐three (1.2%) of the records on days to complete the discharge summary were missing and were excluded from the analysis.
Data Source and Covariates
Data were derived from several sources. The Johns Hopkins Hospital data mart financial database, used for mandatory reporting to the State of Maryland, provided the following patient data: age, gender, race/ethnicity, payer (Medicare, Medicaid, and other) as a proxy for socioeconomic status,[27] hospital service prior to discharge (gynecologyobstetrics, medicine, neurosciences, oncology, pediatrics, and surgical sciences), hospital length of stay (LOS) prior to discharge, Agency for Healthcare Research and Quality (AHRQ) Comorbidity Index (which is an update to the original Elixhauser methodology[28]), and all‐payerrefined diagnosis‐related group (APRDRG) and severity of illness (SOI) combinations (a tool to group patients into clinically comparable disease and SOI categories expected to use similar resources and experience similar outcomes). The Health Services Cost Review Commission (HSCRC) in Maryland provided the observed readmission rate in Maryland for each APRDRG‐SOI combination and served as an expected readmission rate. This risk stratification methodology is similar to the approach used in previous studies.[26, 29] Discharge summary turnaround time was obtained from institutional administrative databases used to track compliance with discharge summary completion. Discharge location (home, facility, home with homecare or hospice, or other) was obtained from Curaspan databases (Curaspan Health Group, Inc., Newton, MA).
Primary Outcome: 30‐Day Readmission
The primary outcome was unplanned rehospitalizations to an acute care hospital in Maryland within 30 days of discharge from Johns Hopkins Hospital. This was as defined by the Maryland HSCRC using an algorithm to exclude readmissions that were likely to be scheduled, as defined by the index admission diagnosis and readmission diagnosis; this algorithm is updated based on the CMS all‐cause readmission algorithm.[30, 31]
Primary Exposure: Days to Complete the Discharge Summary
Discharge summary completion time was defined as the date when the discharge attending physician electronically signs the discharge summary. At our institution, an auto‐fax system sends documents (eg, discharge summaries, clinic notes) to linked providers (eg, primary care providers) shortly after midnight from the day the document is signed by an attending physician. During the period of the project, the policy for discharge summaries at the Johns Hopkins Hospital went from requiring them to be completed within 30 days to 14 days, and we were hoping to use our analyses to inform decision makers why this was important. To emphasize the need for timely completion of discharge summaries, we dichotomized the number of days to complete the discharge summary into >3 versus 3 days (20th percentile cutoff) and modeled it as a continuous variable (per 3‐day increase in days to complete the discharge summary).
Statistical Analysis
To evaluate differences in patient characteristics by readmission status, analysis of variance and 2 tests were used for continuous and dichotomous variables, respectively. Logistic regression was used to evaluate the association between days to complete the discharge summary >3 days and readmission status, adjusting for potentially confounding variables. Before inclusion in the logistic regression model, we confirmed a lack of multicollinearity in the multivariable regression model using variance inflation factors. We evaluated residual versus predicted value plots and residual versus fitted value plots with a locally weighted scatterplot smoothing line. In a sensitivity analysis we evaluated the association between readmission status and different cutoffs (>8 days, 50th percentile; and >14 days, 70% percentile). In a separate analysis, we used interaction terms to test whether the association between the association between days to complete the discharge summary >3 days and hospital readmission varied by the covariates in the analysis (age, sex, race, payer, hospital service, discharge location, LOS, APRDRG‐SOI expected readmission rate, and AHRQ Comorbidity Index). We observed a significant interaction between 30‐day readmission and days to complete the discharge summary >3 days by hospital service. Hence, we separately calculated the adjusted mean readmission rates separately for each hospital service using the least squared means method for the multivariable logistic regression analysis and adjusting for the previously mentioned covariates. In a separate analysis, we used linear regression to evaluate the association between LOS and days to complete the discharge summary, adjusting for potentially confounding variables. Statistical significance was defined as a 2‐sided P < 0.05. Data were analyzed with R (version 2.15.0; R Foundation for Statistical Computing, Vienna, Austria;
RESULTS
Readmitted Patients
In the study period, 14,248 out of 87,994 (16.2%) consecutive eligible patients were readmitted to a hospital in Maryland from patients discharged from Johns Hopkins Hospital between January 1, 2013 and December 31, 2014. A total of 11,027 (77.4%) of the readmissions were back to Johns Hopkins Hospital. Table 1 compares characteristics of readmitted versus nonreadmitted patients, with the following variables being significantly different between these patient groups: age, gender, healthcare payer, hospital service, discharge location, length of stay expected readmission rate, AHRQ Comorbidity Index, and days to complete inpatient discharge summary.
| Characteristics | All Patients, N = 87,994 | Not Readmitted, N = 73,746 | Readmitted, N = 14,248 | P Value |
|---|---|---|---|---|
| ||||
| Age, y | 42.1 (25.1) | 41.3 (25.4) | 46.4 (23.1) | <0.001 |
| Male | 43,210 (49.1%) | 35,851 (48.6%) | 7,359 (51.6%) | <0.001 |
| Race | <0.001 | |||
| Caucasian | 45,705 (51.9%) | 3,8661 (52.4%) | 7,044 (49.4%) | |
| African American | 32,777 (37.2%) | 2,6841 (36.4%) | 5,936 (41.7%) | |
| Other | 9,512 (10.8%) | 8,244 (11.2%) | 1,268 (8.9%) | |
| Payer | <0.001 | |||
| Medicare | 22,345 (25.4%) | 17,614 (23.9%) | 4,731 (33.2%) | |
| Medicaid | 24,080 (27.4%) | 20,100 (27.3%) | 3,980 (27.9%) | |
| Other | 41,569 (47.2%) | 36,032 (48.9%) | 5,537 (38.9%) | |
| Hospital service | <0.001 | |||
| Gynecologyobstetrics | 9,299 (10.6%) | 8,829 (12.0%) | 470 (3.3%) | |
| Medicine | 26,036 (29.6%) | 20,069 (27.2%) | 5,967 (41.9%) | |
| Neurosciences | 8,269 (9.4%) | 7,331 (9.9%) | 938 (6.6%) | |
| Oncology | 5,222 (5.9%) | 3,898 (5.3%) | 1,324 (9.3%) | |
| Pediatrics | 17,029 (19.4%) | 14,684 (19.9%) | 2,345 (16.5%) | |
| Surgical sciences | 22,139 (25.2%) | 18,935 (25.7%) | 3,204 (22.5%) | |
| Discharge location | <0.001 | |||
| Home | 65,478 (74.4%) | 56,359 (76.4%) | 9,119 (64.0%) | |
| Home with homecare or hospice | 9,524 (10.8%) | 7,440 (10.1%) | 2,084 (14.6%) | |
| Facility (SNF, rehabilitation facility) | 5,398 (6.1%) | 4,131 (5.6%) | 1,267 (8.9%) | |
| Other | 7,594 (8.6%) | 5,816 (7.9%) | 1,778 (12.5%) | |
| Length of stay, d | 5.5 (8.6) | 5.1 (7.8) | 7.5 (11.6) | <0.001 |
| APRDRG‐SOI Expected Readmission Rate, % | 14.4 (9.5) | 13.3 (9.2) | 20.1 (9.0) | <0.001 |
| AHRQ Comorbidity Index (1 point) | 2.5 (1.4) | 2.4 (1.4) | 3.0 (1.8) | <0.001 |
| Discharge summary completed >3 days | 66,242 (75.3%) | 55,329 (75.0%) | 10,913 (76.6%) | <0.001 |
Association Between Days to Complete the Discharge Summary and Readmission
After hospital discharge, median (IQR) number of days to complete discharge summaries was 8 (416) days. After hospital discharge, median (IQR) number of days to complete discharge summaries and the number of days from discharge to readmission was 8 (416) and 11 (519) days, respectively (P < 0.001). Six thousand one hundred one patients (42.8%) were readmitted before their discharge summary was completed. The median (IQR) days to complete discharge summaries by hospital service in order from shortest to longest was: oncology, 6 (212) days; surgical sciences, 6 (312) days; pediatrics, 7 (315) days; gynecologyobstetrics, 8 (415) days; medicine, 9 (420) days; neurosciences, 12 (621) days.
When we divided the number of days to complete the discharge summary into deciles (02, 2.13, 3.14, 4.16, 6.18, 8.210, 10.114, 14.119, 19.130, >30), a longer number of days to complete discharge summaries had higher unadjusted and adjusted readmission rates (Figure 1). In unadjusted analysis, Table 2 shows that older age, male sex, African American race, oncological versus medicine hospital service, discharge location, longer LOS, higher APRDRG‐SOI expected readmission rate, and higher AHRQ Comorbidity Index were associated with readmission. Days to complete the discharge summary >3 days versus 3 days was associated with a higher readmission rate, with an unadjusted odds ratio (OR) and 95% confidence interval (CI) of 1.09 (95% CI: 1.04 to 1.13, P < 0.001).
| Characteristic | Bivariable Analysis* | Multivariable Analysis* | ||
|---|---|---|---|---|
| OR (95% CI) | P Value | OR (95% CI) | P Value | |
| ||||
| Age, 10 y | 1.09 (1.08 to 1.09) | <0.001 | 0.97 (0.95 to 0.98) | <0.001 |
| Male | 1.13 (1.09 to 1.17) | <0.001 | 1.01 (0.97 to 1.05) | 0.76 |
| Race | ||||
| Caucasian | Referent | Referent | ||
| African American | 1.21 (1.17 to 1.26) | <0.001 | 1.01 (0.96 to 1.05) | 0.74 |
| Other | 0.84 (0.79 to 0.90) | <0.001 | 0.92 (0.86 to 0.98) | 0.01 |
| Payer | ||||
| Medicare | Referent | Referent | ||
| Medicaid | 0.74 (0.70 to 0.77) | <0.001 | 1.03 (0.97 to 1.09) | 0.42 |
| Other | 0.57 (0.55 to 0.60) | <0.001 | 0.86 (0.82 to 0.91) | <0.001 |
| Hospital service | ||||
| Medicine | Referent | Referent | ||
| Gynecologyobstetrics | 0.18 (0.16 to 0.20) | <0.001 | 0.50 (0.45 to 0.56) | <0.001 |
| Neurosciences | 0.43 (0.40 to 0.46) | <0.001 | 0.76 (0.70 to 0.82) | <0.001 |
| Oncology | 1.14 (1.07 to 1.22) | <0.001 | 1.18 (1.10 to 1.28) | <0.001 |
| Pediatrics | 0.54 (0.51 to 0.57) | <0.001 | 0.77 (0.71 to 0.83) | <0.001 |
| Surgical sciences | 0.57 (0.54 to 0.60) | <0.001 | 0.92 (0.87 to 0.97) | 0.002 |
| Discharge location | ||||
| Home | Referent | |||
| Facility (SNF, rehabilitation facility) | 1.90 (1.77 to 2.03) | <0.001 | 1.11 (1.02 to 1.19) | 0.009 |
| Home with homecare or hospice | 1.73 (1.64 to 1.83) | <0.001 | 1.26 (1.19 to 1.34) | <0.001 |
| Other | 1.89 (1.78 to 2.00) | <0.001 | 1.25 (1.18 to 1.34) | <0.001 |
| Length of stay, d | 1.03 (1.02 to 1.03) | <0.001 | 1.00 (1.00 to 1.01) | <0.001 |
| APRDRG‐SOI expected readmission rate, % | 1.08 (1.07 to 1.08) | <0.001 | 1.06 (1.06 to 1.06) | <0.001 |
| AHRQ Comorbidity Index (1 point) | 1.27 (1.26 to 1.28) | <0.001 | 1.11 (1.09 to 1.12) | <0.001 |
| Discharge summary completed >3 days | 1.09 (1.04 to 1.14) | <0.001 | 1.09 (1.05 to 1.14) | <0.001 |
Multivariable and Secondary Analyses
In adjusted analysis (Table 2), patients discharged from an oncologic service relative to a medicine hospital service (OR: 1.19, 95% CI: 1.10 to 1.28, P < 0.001), patients discharged to a facility, home with homecare or hospice, or other location compared to home (facility OR: 1.11, 95% CI: 1.02 to 1.19, P = 0.009; home with homecare or hospice OR: 1.26, 95% CI: 1.19 to 1.34, P < 0.001; other OR: 1.25, 95% CI: 1.18 to 1.34, P < 0.001), patients with longer LOS (OR: 1.11 per day, 95% CI: 1.10 to 1.12, P < 0.001), patients with a higher expected readmission rates (OR: 1.01 per percent, 95% CI: 1.00 to 1.01, P < 0.001), and patients with a higher AHRQ comorbidity index (OR: 1.06 per 1 point, 95% CI: 1.06 to 1.06, P < 0.001) had higher 30‐day readmission rates. Overall, days to complete the discharge summary >3 days versus 3 days was associated with a higher readmission rate (OR: 1.09, 95% CI: 1.05 to 1.14, P < 0.001).
In a sensitivity analysis, discharge summary completion >8 days (median) versus 8 days was associated with higher unadjusted readmission rate (OR: 1.11, 95% CI: 1.07 to 1.15, P < 0.001) and a higher adjusted readmission rate (OR: 1.06, 95% CI: 1.02 to 1.10, P < 0.001). Discharge summary completion >14 days (70th percentile) versus 14 days was also associated with higher unadjusted readmission rate (OR: 1.15, 95% CI: 1.08 to 1.21, P < 0.001) and a higher adjusted readmission rate (OR: 1.09, 95% CI: 1.02 to 1.16, P = 0.008). The association between days to complete the discharge summary >3 days and readmissions was found to vary significantly by hospital service (P = 0.03). For comparing days to complete the discharge summary >3 versus 3 days, Table 3 shows that neurosciences, pediatrics, oncology, and medicine hospital services were associated with significantly increased adjusted mean readmission rates. Additionally, when days to complete the discharge summary was modeled as a continuous variable, we found that for every 3 days the odds of readmission increased by 1% (OR: 1.01, 95% CI: 1.00 to 1.01, P < 0.001).
| Days to Complete Discharge Summary by Hospital Service | Adjusted Mean Readmission Rate (95% CI)* | P Value |
|---|---|---|
| ||
| Gynecologyobstetrics | 0.30 | |
| 03 days, n = 1,792 | 5.4 (4.1 to 6.7) | |
| >3 days, n = 7,507 | 6.0 (4.9 to 7.0) | |
| Medicine | 0.04 | |
| 03 days, n = 6,137 | 21.1 (20.0 to 22.3) | |
| >3 days, n = 19,899 | 22.4 (21.6 to 23.2) | |
| Neurosciences | 0.02 | |
| 03 days, n = 1,116 | 10.1 (8.2 to 12.1) | |
| >3 days, n = 7,153 | 12.5 (11.6 to 13.5) | |
| Oncology | 0.01 | |
| 03 days, n = 1,885 | 25.0 (22.6 to 27.4) | |
| >3 days, n = 3,337 | 28.2 (26.6 to 30.2) | |
| Pediatrics | 0.001 | |
| 03 days, n = 4,561 | 9.5 (6.9 to 12.2) | |
| >3 days, n = 12,468 | 11.4 (8.9 to 13.9) | |
| Surgical sciences | 0.89 | |
| 03 days, n = 6,261 | 15.2 (14.2 to 16.1) | |
| >3 days, n = 15,878 | 15.1 (14.4 to 15.8) | |
In an unadjusted analysis, we found that the relationship between LOS and days to complete the discharge summary was not significant ( coefficient and 95% CI:, 0.01, 0.02 to 0.00, P = 0.20). However, we found a small but significant relationship in our multivariable analysis, such that each hospitalization day was associated with a 0.01 (95% CI: 0.00 to 0.02, P = 0.03) increase in days to complete the discharge summary.
DISCUSSION
In this single‐center retrospective analysis, the number of days to complete the discharge summary was significantly associated with readmissions after hospitalization. This association was independent of age, gender, comorbidity index, payer, discharge location, length of hospital stay, expected readmission rate based on diagnosis and severity of illness, and all hospital services. The odds of readmission for patients with delayed discharge summaries was small but significant. This is important in the current landscape of readmissions, particularly for institutions who are challenged to reduce readmission rates, and a small relative difference in readmissions may be the difference between getting penalized or not. In the context of prior studies, the results highlight the role of timely discharge summary as an under‐recognized metric, which may be a valid litmus test for care coordination. The findings also emphasize the potential of early summaries to expedite communication and to help facilitate quality of patient care. Hence, the study results extend the literature examining the relationship of delay in discharge summary with unfavorable patient outcomes.[15, 32]
In contrast to prior reports with limited focus on same‐hospital readmissions,[18, 33, 34, 35] readmissions beyond 30 days,[12] or focused on a specific patient population,[13, 36] this study evaluates both intra‐ and interhospital 30‐day readmissions in Maryland in an all‐payer, multi‐institution, diverse patient population. Additionally, prior research is conflicting with respect to whether timely discharges summaries are significantly associated with increased hospital readmissions.[12, 13, 14, 15] Although it is not surprising that inadequate care during hospitalization could result in readmissions, the role of discharge summaries remain underappreciated. Having a timely discharge summary may not always prevent readmissions, but our study showed that 43% of readmission occurred before the discharge summary completion. Not having a completed discharge summary at the time of readmission may have been a driver for the positive association between timely completion and 30‐day readmission we observed. This study highlights that delay in the discharge summary could be a marker of poor transitions of care, because suboptimal dissemination of critical information to care providers may result in discontinuity of patient care posthospitalization.
A plausible mechanism of the association between discharge summary delays and readmissions could be the provision of collateral information, which may potentially alter the threshold for readmissions. For example, in the emergency room/emergency department (ER/ED) setting, discharge summaries may help with preventable readmissions. For patients who present repeatedly with the same complaint, timely summaries to ER/ED providers may help reframe the patient complaints, such as patient has concern X, which was previously identified to be related to diagnosis Y. As others have shown, the content of discharge summaries, format, and accessibility (electronic vs paper chart), as well as timely distribution of summaries, are key factors that impact quality outcomes.[2, 12, 15, 37, 38] By detailing prior hospital information (ie, discharge medications, prior presentations, tests completed), summaries could help prevent errors in medication dosing, reduce unnecessary testing, and help facilitate admission triage. Summaries may have information regarding a new diagnosis such as the results of an endoscopic evaluation that revealed the source of occult gastrointestinal bleeding, which could help contextualize a complaint of repeat melena and redirect goals of care. Discussions of goals of care in the discharge summary may guide primary providers in continued care management plans.
Our study findings underscore a positive correlation between late discharge summaries and readmissions. However, the extent that this is a causal relationship is unclear; the association of delay in days to complete the discharge summary with readmission may be an epiphenomenon related to processes related to quality of clinical care. For example, delays in discharge summary completion could be a marker of other system issues, such as a stressed work environment. It is possible that providers who fail to complete timely discharge summaries may also fail to do other important functions related to transitions of care and care coordination. However, even if this is so, timely discharge summaries could become a focal point for discussion for optimization of care transitions. A discharge summary could be delayed because the patient has already been readmitted before the summary was distributed, thus making that original summary less relevant. Delays could also be a reflection of the data complexity for patients with longer hospital stays. This is supported by the small but significant relationship between LOS and days to complete the discharge summary in this study. Lastly, delays in discharge summary completion may also be a proxy of provider communication and can reflect the culture of communication at the institution.
Although unplanned hospital readmission is an important outcome, many readmissions may be related to other factors such as disease progression, rather than late summaries or the lack of postdischarge communication. For instance, prior reports did not find any association between the PCP seeing the discharge summaries or direct communications with the PCP and 30‐day clinical outcomes for readmission and death.[26, 39] However, these studies were limited in their use of self‐reported handoffs, did not measure quality of information transfer, and failed to capture a broader audience beyond the PCP, such as ED physicians or specialists.
Our results suggest that the relationship between days to complete discharge summaries and 30‐day readmissions may vary depending on whether the hospitalization is primarily surgical/procedural versus medical treatment. A recent study found that most readmissions after surgery were associated with new complications related to the procedure and not exacerbation of prior index hospitalization complications.[40] Hence, treatment for common causes of hospital readmissions after surgical or gynecological procedures, such as wound infections, acute anemia, ileus, or dehydration, may not necessarily require a completed discharge summary for appropriate management. However, we caution extending this finding to clinical practice before further studies are conducted on specific procedures and in different clinical settings.
Results from this study also support institutional policies that specify the need for practitioners to complete discharge summaries contemporaneously, such as at the time of discharge or within a couple of days. Unlike other forms of communication that are optional, discharge summaries are required, so we recommend that practitioners be held accountable for short turnaround times. For example, providers could be graded and rated on timely completions of discharge summaries, among other performance variables. Anecdotally at our institutions, we have heard from practitioners that it takes less time to complete them when you do them on the day of discharge, because the hospitalization course is fresher in their mind and they have to wade through less information in the medical record to complete an accurate discharge summary. To this point, a barrier to on‐time completion is that providers may have misconceptions about what is really vital information to convey to the next provider. In agreement with past research and in the era of the electronic medical record system, we recommend that the discharge summary should be a quick synthesis of key findings that incorporates only the important elements, such as why the patient was hospitalized, what were key findings and key responses to therapy, what is pending at the time of discharge, what medications the patient is currently taking, and what are the follow‐up plans, rather than a lengthy expose of all the findings.[13, 36, 41, 42]
Lastly, our study results should be taken in the context of its limitations. As a single‐center study, findings may lack generalizability. In particular, the results may not generalize to hospitals that lack access to statewide reporting. We were also not able to assess readmission for patients who may have been readmitted to a hospital outside of Maryland. Although we adjusted for pertinent variables such as age, gender, healthcare payer, hospital service, comorbidity index, discharge location, LOS, and expected readmission rates, there may be other relevant confounders that we failed to capture or measure optimally. Median days to complete the discharge summary in this study was 8 days, which is longer than practices at other institutions, and may also limit this study's generalizability.[15, 36, 42] However, prior research supports our findings,[15] and a systematic review found that only 29% and 52% of discharge summaries were completed by 2 weeks and 4 weeks, respectively.[9] Finally, as noted above and perhaps most important, it is possible that discharge summary turnaround time does not in itself causally impact readmissions, but rather reflects an underlying commitment of the inpatient team to effectively coordinate care following hospital discharge.
CONCLUSION
In sum, this study delineates an underappreciated but important relationship of timely discharge summary completion and readmission outcomes. The discharge summary may be a relevant metric reflecting quality of patient care. Healthcare providers may begin to target timely discharge summaries as a potential focal point of quality‐improvement projects with the goal to facilitate better patient outcomes.
Disclosures
The authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated, and, if applicable, the authors certify that all financial and material support for this research (eg, Centers for Medicare and Medicaid Services, National Institutes of Health, or National Health Service grants) and work are clearly identified. This study was supported by funding opportunity, number CMS‐1C1‐12‐0001, from the Centers for Medicare and Medicaid Services and Center for Medicare and Medicaid Innovation. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Department of Health and Human Services or any of its agencies.
Across the continuum of care, the discharge summary is a critical tool for communication among care providers.[1] In the United States, the Joint Commission policies mandate that all hospital providers complete a discharge summary for patients with specific components to foster effective communication with future providers.[2] Because outpatient providers and emergency physicians rely on clinical information in the discharge summary to ensure appropriate postdischarge continuity of care, timely documentation is potentially an essential aspect of readmission reduction initiatives.[3, 4, 5] Prior reports indicate that poor discharge documentation of follow‐up plan‐of‐care increases the risk of hospitalization, whereas structured instructions, patient education, and direct communications with primary care physicians (PCPs) reduce repeat hospital visits.[6, 7, 8, 9] However, the current literature is limited in its narrow focus on the contents of discharge summaries, considered only same‐hospital readmissions, or considered readmissions within 3 months of discharge.[10, 11, 12, 13] Moreover, some prior research has suggested no association between discharge summary timeliness with readmission,[12, 13, 14] whereas another study did find a relationship,[15] hence the need to study this further is important. Filling this gap in knowledge could provide an avenue to track and improve quality of patient care, as delays in discharge summaries have been linked with pot‐discharge adverse outcomes and patient safety concerns.[15, 16, 17, 18] Because readmissions often occur soon after discharge, having timely discharge summaries may be particularly important to outcomes.[19, 20]
This research began under the framework of evaluating a bundle of care coordination strategies that were implemented at the Johns Hopkins Health System. These strategies were informed by the early Centers for Medicare and Medicaid Services (CMS) demonstration projects and other best practices that have been documented in the literature to improve utilization and improve communication during transitions of care.[21, 22, 23, 24, 25] Later they were augmented through a contract with the Center of Medicare and Medicaid Innovation to improve access to healthcare services and improve patient outcomes through improved care coordination processes. One of the domains our institution has increased efforts to improve is in provider handoffs. Toward that goal, we have worked to disentangle the effects of different factors of provider‐to‐provider communication that may influence readmissions.[26] For example, effective written provider handoffs in the form of accurate and timely discharge summaries was considered a key care coordination component of this program, but there was institutional resistance to endorsing an expectation that discharge summary turnaround should be shortened. To build a case for this concept, we sought to test the hypothesis that, at our hospital, longer time to complete hospital discharge summaries was associated with increased readmission rates. Unique to this analysis is that, in the state of Maryland, there is statewide reporting of readmissions, so we were able to account for intra‐ and interhospital readmissions for an all‐payer population. The authors anticipated that findings from this study would help inform discharge quality‐improvement initiatives and reemphasize the importance of timely discharge documentation across all disciplines as part of quality patient care.
METHODS
Study Population and Setting
We conducted a single‐center, retrospective cohort study of 87,994 consecutive patients discharged from Johns Hopkins Hospital, which is a 1000‐bed, tertiary academic medical center in Baltimore, Maryland between January 1, 2013 and December 31, 2014. One thousand ninety‐three (1.2%) of the records on days to complete the discharge summary were missing and were excluded from the analysis.
Data Source and Covariates
Data were derived from several sources. The Johns Hopkins Hospital data mart financial database, used for mandatory reporting to the State of Maryland, provided the following patient data: age, gender, race/ethnicity, payer (Medicare, Medicaid, and other) as a proxy for socioeconomic status,[27] hospital service prior to discharge (gynecologyobstetrics, medicine, neurosciences, oncology, pediatrics, and surgical sciences), hospital length of stay (LOS) prior to discharge, Agency for Healthcare Research and Quality (AHRQ) Comorbidity Index (which is an update to the original Elixhauser methodology[28]), and all‐payerrefined diagnosis‐related group (APRDRG) and severity of illness (SOI) combinations (a tool to group patients into clinically comparable disease and SOI categories expected to use similar resources and experience similar outcomes). The Health Services Cost Review Commission (HSCRC) in Maryland provided the observed readmission rate in Maryland for each APRDRG‐SOI combination and served as an expected readmission rate. This risk stratification methodology is similar to the approach used in previous studies.[26, 29] Discharge summary turnaround time was obtained from institutional administrative databases used to track compliance with discharge summary completion. Discharge location (home, facility, home with homecare or hospice, or other) was obtained from Curaspan databases (Curaspan Health Group, Inc., Newton, MA).
Primary Outcome: 30‐Day Readmission
The primary outcome was unplanned rehospitalizations to an acute care hospital in Maryland within 30 days of discharge from Johns Hopkins Hospital. This was as defined by the Maryland HSCRC using an algorithm to exclude readmissions that were likely to be scheduled, as defined by the index admission diagnosis and readmission diagnosis; this algorithm is updated based on the CMS all‐cause readmission algorithm.[30, 31]
Primary Exposure: Days to Complete the Discharge Summary
Discharge summary completion time was defined as the date when the discharge attending physician electronically signs the discharge summary. At our institution, an auto‐fax system sends documents (eg, discharge summaries, clinic notes) to linked providers (eg, primary care providers) shortly after midnight from the day the document is signed by an attending physician. During the period of the project, the policy for discharge summaries at the Johns Hopkins Hospital went from requiring them to be completed within 30 days to 14 days, and we were hoping to use our analyses to inform decision makers why this was important. To emphasize the need for timely completion of discharge summaries, we dichotomized the number of days to complete the discharge summary into >3 versus 3 days (20th percentile cutoff) and modeled it as a continuous variable (per 3‐day increase in days to complete the discharge summary).
Statistical Analysis
To evaluate differences in patient characteristics by readmission status, analysis of variance and 2 tests were used for continuous and dichotomous variables, respectively. Logistic regression was used to evaluate the association between days to complete the discharge summary >3 days and readmission status, adjusting for potentially confounding variables. Before inclusion in the logistic regression model, we confirmed a lack of multicollinearity in the multivariable regression model using variance inflation factors. We evaluated residual versus predicted value plots and residual versus fitted value plots with a locally weighted scatterplot smoothing line. In a sensitivity analysis we evaluated the association between readmission status and different cutoffs (>8 days, 50th percentile; and >14 days, 70% percentile). In a separate analysis, we used interaction terms to test whether the association between the association between days to complete the discharge summary >3 days and hospital readmission varied by the covariates in the analysis (age, sex, race, payer, hospital service, discharge location, LOS, APRDRG‐SOI expected readmission rate, and AHRQ Comorbidity Index). We observed a significant interaction between 30‐day readmission and days to complete the discharge summary >3 days by hospital service. Hence, we separately calculated the adjusted mean readmission rates separately for each hospital service using the least squared means method for the multivariable logistic regression analysis and adjusting for the previously mentioned covariates. In a separate analysis, we used linear regression to evaluate the association between LOS and days to complete the discharge summary, adjusting for potentially confounding variables. Statistical significance was defined as a 2‐sided P < 0.05. Data were analyzed with R (version 2.15.0; R Foundation for Statistical Computing, Vienna, Austria;
RESULTS
Readmitted Patients
In the study period, 14,248 out of 87,994 (16.2%) consecutive eligible patients were readmitted to a hospital in Maryland from patients discharged from Johns Hopkins Hospital between January 1, 2013 and December 31, 2014. A total of 11,027 (77.4%) of the readmissions were back to Johns Hopkins Hospital. Table 1 compares characteristics of readmitted versus nonreadmitted patients, with the following variables being significantly different between these patient groups: age, gender, healthcare payer, hospital service, discharge location, length of stay expected readmission rate, AHRQ Comorbidity Index, and days to complete inpatient discharge summary.
| Characteristics | All Patients, N = 87,994 | Not Readmitted, N = 73,746 | Readmitted, N = 14,248 | P Value |
|---|---|---|---|---|
| ||||
| Age, y | 42.1 (25.1) | 41.3 (25.4) | 46.4 (23.1) | <0.001 |
| Male | 43,210 (49.1%) | 35,851 (48.6%) | 7,359 (51.6%) | <0.001 |
| Race | <0.001 | |||
| Caucasian | 45,705 (51.9%) | 3,8661 (52.4%) | 7,044 (49.4%) | |
| African American | 32,777 (37.2%) | 2,6841 (36.4%) | 5,936 (41.7%) | |
| Other | 9,512 (10.8%) | 8,244 (11.2%) | 1,268 (8.9%) | |
| Payer | <0.001 | |||
| Medicare | 22,345 (25.4%) | 17,614 (23.9%) | 4,731 (33.2%) | |
| Medicaid | 24,080 (27.4%) | 20,100 (27.3%) | 3,980 (27.9%) | |
| Other | 41,569 (47.2%) | 36,032 (48.9%) | 5,537 (38.9%) | |
| Hospital service | <0.001 | |||
| Gynecologyobstetrics | 9,299 (10.6%) | 8,829 (12.0%) | 470 (3.3%) | |
| Medicine | 26,036 (29.6%) | 20,069 (27.2%) | 5,967 (41.9%) | |
| Neurosciences | 8,269 (9.4%) | 7,331 (9.9%) | 938 (6.6%) | |
| Oncology | 5,222 (5.9%) | 3,898 (5.3%) | 1,324 (9.3%) | |
| Pediatrics | 17,029 (19.4%) | 14,684 (19.9%) | 2,345 (16.5%) | |
| Surgical sciences | 22,139 (25.2%) | 18,935 (25.7%) | 3,204 (22.5%) | |
| Discharge location | <0.001 | |||
| Home | 65,478 (74.4%) | 56,359 (76.4%) | 9,119 (64.0%) | |
| Home with homecare or hospice | 9,524 (10.8%) | 7,440 (10.1%) | 2,084 (14.6%) | |
| Facility (SNF, rehabilitation facility) | 5,398 (6.1%) | 4,131 (5.6%) | 1,267 (8.9%) | |
| Other | 7,594 (8.6%) | 5,816 (7.9%) | 1,778 (12.5%) | |
| Length of stay, d | 5.5 (8.6) | 5.1 (7.8) | 7.5 (11.6) | <0.001 |
| APRDRG‐SOI Expected Readmission Rate, % | 14.4 (9.5) | 13.3 (9.2) | 20.1 (9.0) | <0.001 |
| AHRQ Comorbidity Index (1 point) | 2.5 (1.4) | 2.4 (1.4) | 3.0 (1.8) | <0.001 |
| Discharge summary completed >3 days | 66,242 (75.3%) | 55,329 (75.0%) | 10,913 (76.6%) | <0.001 |
Association Between Days to Complete the Discharge Summary and Readmission
After hospital discharge, median (IQR) number of days to complete discharge summaries was 8 (416) days. After hospital discharge, median (IQR) number of days to complete discharge summaries and the number of days from discharge to readmission was 8 (416) and 11 (519) days, respectively (P < 0.001). Six thousand one hundred one patients (42.8%) were readmitted before their discharge summary was completed. The median (IQR) days to complete discharge summaries by hospital service in order from shortest to longest was: oncology, 6 (212) days; surgical sciences, 6 (312) days; pediatrics, 7 (315) days; gynecologyobstetrics, 8 (415) days; medicine, 9 (420) days; neurosciences, 12 (621) days.
When we divided the number of days to complete the discharge summary into deciles (02, 2.13, 3.14, 4.16, 6.18, 8.210, 10.114, 14.119, 19.130, >30), a longer number of days to complete discharge summaries had higher unadjusted and adjusted readmission rates (Figure 1). In unadjusted analysis, Table 2 shows that older age, male sex, African American race, oncological versus medicine hospital service, discharge location, longer LOS, higher APRDRG‐SOI expected readmission rate, and higher AHRQ Comorbidity Index were associated with readmission. Days to complete the discharge summary >3 days versus 3 days was associated with a higher readmission rate, with an unadjusted odds ratio (OR) and 95% confidence interval (CI) of 1.09 (95% CI: 1.04 to 1.13, P < 0.001).
| Characteristic | Bivariable Analysis* | Multivariable Analysis* | ||
|---|---|---|---|---|
| OR (95% CI) | P Value | OR (95% CI) | P Value | |
| ||||
| Age, 10 y | 1.09 (1.08 to 1.09) | <0.001 | 0.97 (0.95 to 0.98) | <0.001 |
| Male | 1.13 (1.09 to 1.17) | <0.001 | 1.01 (0.97 to 1.05) | 0.76 |
| Race | ||||
| Caucasian | Referent | Referent | ||
| African American | 1.21 (1.17 to 1.26) | <0.001 | 1.01 (0.96 to 1.05) | 0.74 |
| Other | 0.84 (0.79 to 0.90) | <0.001 | 0.92 (0.86 to 0.98) | 0.01 |
| Payer | ||||
| Medicare | Referent | Referent | ||
| Medicaid | 0.74 (0.70 to 0.77) | <0.001 | 1.03 (0.97 to 1.09) | 0.42 |
| Other | 0.57 (0.55 to 0.60) | <0.001 | 0.86 (0.82 to 0.91) | <0.001 |
| Hospital service | ||||
| Medicine | Referent | Referent | ||
| Gynecologyobstetrics | 0.18 (0.16 to 0.20) | <0.001 | 0.50 (0.45 to 0.56) | <0.001 |
| Neurosciences | 0.43 (0.40 to 0.46) | <0.001 | 0.76 (0.70 to 0.82) | <0.001 |
| Oncology | 1.14 (1.07 to 1.22) | <0.001 | 1.18 (1.10 to 1.28) | <0.001 |
| Pediatrics | 0.54 (0.51 to 0.57) | <0.001 | 0.77 (0.71 to 0.83) | <0.001 |
| Surgical sciences | 0.57 (0.54 to 0.60) | <0.001 | 0.92 (0.87 to 0.97) | 0.002 |
| Discharge location | ||||
| Home | Referent | |||
| Facility (SNF, rehabilitation facility) | 1.90 (1.77 to 2.03) | <0.001 | 1.11 (1.02 to 1.19) | 0.009 |
| Home with homecare or hospice | 1.73 (1.64 to 1.83) | <0.001 | 1.26 (1.19 to 1.34) | <0.001 |
| Other | 1.89 (1.78 to 2.00) | <0.001 | 1.25 (1.18 to 1.34) | <0.001 |
| Length of stay, d | 1.03 (1.02 to 1.03) | <0.001 | 1.00 (1.00 to 1.01) | <0.001 |
| APRDRG‐SOI expected readmission rate, % | 1.08 (1.07 to 1.08) | <0.001 | 1.06 (1.06 to 1.06) | <0.001 |
| AHRQ Comorbidity Index (1 point) | 1.27 (1.26 to 1.28) | <0.001 | 1.11 (1.09 to 1.12) | <0.001 |
| Discharge summary completed >3 days | 1.09 (1.04 to 1.14) | <0.001 | 1.09 (1.05 to 1.14) | <0.001 |
Multivariable and Secondary Analyses
In adjusted analysis (Table 2), patients discharged from an oncologic service relative to a medicine hospital service (OR: 1.19, 95% CI: 1.10 to 1.28, P < 0.001), patients discharged to a facility, home with homecare or hospice, or other location compared to home (facility OR: 1.11, 95% CI: 1.02 to 1.19, P = 0.009; home with homecare or hospice OR: 1.26, 95% CI: 1.19 to 1.34, P < 0.001; other OR: 1.25, 95% CI: 1.18 to 1.34, P < 0.001), patients with longer LOS (OR: 1.11 per day, 95% CI: 1.10 to 1.12, P < 0.001), patients with a higher expected readmission rates (OR: 1.01 per percent, 95% CI: 1.00 to 1.01, P < 0.001), and patients with a higher AHRQ comorbidity index (OR: 1.06 per 1 point, 95% CI: 1.06 to 1.06, P < 0.001) had higher 30‐day readmission rates. Overall, days to complete the discharge summary >3 days versus 3 days was associated with a higher readmission rate (OR: 1.09, 95% CI: 1.05 to 1.14, P < 0.001).
In a sensitivity analysis, discharge summary completion >8 days (median) versus 8 days was associated with higher unadjusted readmission rate (OR: 1.11, 95% CI: 1.07 to 1.15, P < 0.001) and a higher adjusted readmission rate (OR: 1.06, 95% CI: 1.02 to 1.10, P < 0.001). Discharge summary completion >14 days (70th percentile) versus 14 days was also associated with higher unadjusted readmission rate (OR: 1.15, 95% CI: 1.08 to 1.21, P < 0.001) and a higher adjusted readmission rate (OR: 1.09, 95% CI: 1.02 to 1.16, P = 0.008). The association between days to complete the discharge summary >3 days and readmissions was found to vary significantly by hospital service (P = 0.03). For comparing days to complete the discharge summary >3 versus 3 days, Table 3 shows that neurosciences, pediatrics, oncology, and medicine hospital services were associated with significantly increased adjusted mean readmission rates. Additionally, when days to complete the discharge summary was modeled as a continuous variable, we found that for every 3 days the odds of readmission increased by 1% (OR: 1.01, 95% CI: 1.00 to 1.01, P < 0.001).
| Days to Complete Discharge Summary by Hospital Service | Adjusted Mean Readmission Rate (95% CI)* | P Value |
|---|---|---|
| ||
| Gynecologyobstetrics | 0.30 | |
| 03 days, n = 1,792 | 5.4 (4.1 to 6.7) | |
| >3 days, n = 7,507 | 6.0 (4.9 to 7.0) | |
| Medicine | 0.04 | |
| 03 days, n = 6,137 | 21.1 (20.0 to 22.3) | |
| >3 days, n = 19,899 | 22.4 (21.6 to 23.2) | |
| Neurosciences | 0.02 | |
| 03 days, n = 1,116 | 10.1 (8.2 to 12.1) | |
| >3 days, n = 7,153 | 12.5 (11.6 to 13.5) | |
| Oncology | 0.01 | |
| 03 days, n = 1,885 | 25.0 (22.6 to 27.4) | |
| >3 days, n = 3,337 | 28.2 (26.6 to 30.2) | |
| Pediatrics | 0.001 | |
| 03 days, n = 4,561 | 9.5 (6.9 to 12.2) | |
| >3 days, n = 12,468 | 11.4 (8.9 to 13.9) | |
| Surgical sciences | 0.89 | |
| 03 days, n = 6,261 | 15.2 (14.2 to 16.1) | |
| >3 days, n = 15,878 | 15.1 (14.4 to 15.8) | |
In an unadjusted analysis, we found that the relationship between LOS and days to complete the discharge summary was not significant ( coefficient and 95% CI:, 0.01, 0.02 to 0.00, P = 0.20). However, we found a small but significant relationship in our multivariable analysis, such that each hospitalization day was associated with a 0.01 (95% CI: 0.00 to 0.02, P = 0.03) increase in days to complete the discharge summary.
DISCUSSION
In this single‐center retrospective analysis, the number of days to complete the discharge summary was significantly associated with readmissions after hospitalization. This association was independent of age, gender, comorbidity index, payer, discharge location, length of hospital stay, expected readmission rate based on diagnosis and severity of illness, and all hospital services. The odds of readmission for patients with delayed discharge summaries was small but significant. This is important in the current landscape of readmissions, particularly for institutions who are challenged to reduce readmission rates, and a small relative difference in readmissions may be the difference between getting penalized or not. In the context of prior studies, the results highlight the role of timely discharge summary as an under‐recognized metric, which may be a valid litmus test for care coordination. The findings also emphasize the potential of early summaries to expedite communication and to help facilitate quality of patient care. Hence, the study results extend the literature examining the relationship of delay in discharge summary with unfavorable patient outcomes.[15, 32]
In contrast to prior reports with limited focus on same‐hospital readmissions,[18, 33, 34, 35] readmissions beyond 30 days,[12] or focused on a specific patient population,[13, 36] this study evaluates both intra‐ and interhospital 30‐day readmissions in Maryland in an all‐payer, multi‐institution, diverse patient population. Additionally, prior research is conflicting with respect to whether timely discharges summaries are significantly associated with increased hospital readmissions.[12, 13, 14, 15] Although it is not surprising that inadequate care during hospitalization could result in readmissions, the role of discharge summaries remain underappreciated. Having a timely discharge summary may not always prevent readmissions, but our study showed that 43% of readmission occurred before the discharge summary completion. Not having a completed discharge summary at the time of readmission may have been a driver for the positive association between timely completion and 30‐day readmission we observed. This study highlights that delay in the discharge summary could be a marker of poor transitions of care, because suboptimal dissemination of critical information to care providers may result in discontinuity of patient care posthospitalization.
A plausible mechanism of the association between discharge summary delays and readmissions could be the provision of collateral information, which may potentially alter the threshold for readmissions. For example, in the emergency room/emergency department (ER/ED) setting, discharge summaries may help with preventable readmissions. For patients who present repeatedly with the same complaint, timely summaries to ER/ED providers may help reframe the patient complaints, such as patient has concern X, which was previously identified to be related to diagnosis Y. As others have shown, the content of discharge summaries, format, and accessibility (electronic vs paper chart), as well as timely distribution of summaries, are key factors that impact quality outcomes.[2, 12, 15, 37, 38] By detailing prior hospital information (ie, discharge medications, prior presentations, tests completed), summaries could help prevent errors in medication dosing, reduce unnecessary testing, and help facilitate admission triage. Summaries may have information regarding a new diagnosis such as the results of an endoscopic evaluation that revealed the source of occult gastrointestinal bleeding, which could help contextualize a complaint of repeat melena and redirect goals of care. Discussions of goals of care in the discharge summary may guide primary providers in continued care management plans.
Our study findings underscore a positive correlation between late discharge summaries and readmissions. However, the extent that this is a causal relationship is unclear; the association of delay in days to complete the discharge summary with readmission may be an epiphenomenon related to processes related to quality of clinical care. For example, delays in discharge summary completion could be a marker of other system issues, such as a stressed work environment. It is possible that providers who fail to complete timely discharge summaries may also fail to do other important functions related to transitions of care and care coordination. However, even if this is so, timely discharge summaries could become a focal point for discussion for optimization of care transitions. A discharge summary could be delayed because the patient has already been readmitted before the summary was distributed, thus making that original summary less relevant. Delays could also be a reflection of the data complexity for patients with longer hospital stays. This is supported by the small but significant relationship between LOS and days to complete the discharge summary in this study. Lastly, delays in discharge summary completion may also be a proxy of provider communication and can reflect the culture of communication at the institution.
Although unplanned hospital readmission is an important outcome, many readmissions may be related to other factors such as disease progression, rather than late summaries or the lack of postdischarge communication. For instance, prior reports did not find any association between the PCP seeing the discharge summaries or direct communications with the PCP and 30‐day clinical outcomes for readmission and death.[26, 39] However, these studies were limited in their use of self‐reported handoffs, did not measure quality of information transfer, and failed to capture a broader audience beyond the PCP, such as ED physicians or specialists.
Our results suggest that the relationship between days to complete discharge summaries and 30‐day readmissions may vary depending on whether the hospitalization is primarily surgical/procedural versus medical treatment. A recent study found that most readmissions after surgery were associated with new complications related to the procedure and not exacerbation of prior index hospitalization complications.[40] Hence, treatment for common causes of hospital readmissions after surgical or gynecological procedures, such as wound infections, acute anemia, ileus, or dehydration, may not necessarily require a completed discharge summary for appropriate management. However, we caution extending this finding to clinical practice before further studies are conducted on specific procedures and in different clinical settings.
Results from this study also support institutional policies that specify the need for practitioners to complete discharge summaries contemporaneously, such as at the time of discharge or within a couple of days. Unlike other forms of communication that are optional, discharge summaries are required, so we recommend that practitioners be held accountable for short turnaround times. For example, providers could be graded and rated on timely completions of discharge summaries, among other performance variables. Anecdotally at our institutions, we have heard from practitioners that it takes less time to complete them when you do them on the day of discharge, because the hospitalization course is fresher in their mind and they have to wade through less information in the medical record to complete an accurate discharge summary. To this point, a barrier to on‐time completion is that providers may have misconceptions about what is really vital information to convey to the next provider. In agreement with past research and in the era of the electronic medical record system, we recommend that the discharge summary should be a quick synthesis of key findings that incorporates only the important elements, such as why the patient was hospitalized, what were key findings and key responses to therapy, what is pending at the time of discharge, what medications the patient is currently taking, and what are the follow‐up plans, rather than a lengthy expose of all the findings.[13, 36, 41, 42]
Lastly, our study results should be taken in the context of its limitations. As a single‐center study, findings may lack generalizability. In particular, the results may not generalize to hospitals that lack access to statewide reporting. We were also not able to assess readmission for patients who may have been readmitted to a hospital outside of Maryland. Although we adjusted for pertinent variables such as age, gender, healthcare payer, hospital service, comorbidity index, discharge location, LOS, and expected readmission rates, there may be other relevant confounders that we failed to capture or measure optimally. Median days to complete the discharge summary in this study was 8 days, which is longer than practices at other institutions, and may also limit this study's generalizability.[15, 36, 42] However, prior research supports our findings,[15] and a systematic review found that only 29% and 52% of discharge summaries were completed by 2 weeks and 4 weeks, respectively.[9] Finally, as noted above and perhaps most important, it is possible that discharge summary turnaround time does not in itself causally impact readmissions, but rather reflects an underlying commitment of the inpatient team to effectively coordinate care following hospital discharge.
CONCLUSION
In sum, this study delineates an underappreciated but important relationship of timely discharge summary completion and readmission outcomes. The discharge summary may be a relevant metric reflecting quality of patient care. Healthcare providers may begin to target timely discharge summaries as a potential focal point of quality‐improvement projects with the goal to facilitate better patient outcomes.
Disclosures
The authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated, and, if applicable, the authors certify that all financial and material support for this research (eg, Centers for Medicare and Medicaid Services, National Institutes of Health, or National Health Service grants) and work are clearly identified. This study was supported by funding opportunity, number CMS‐1C1‐12‐0001, from the Centers for Medicare and Medicaid Services and Center for Medicare and Medicaid Innovation. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Department of Health and Human Services or any of its agencies.
- , , , et al. Development and sustainability of an inpatient‐to‐outpatient discharge handoff tool: a quality improvement project. Jt Comm J Qual Patient Saf. 2014;40(5):219–227.
- , , , , , . Documentation of mandated discharge summary components in transitions from acute to subacute care. In: Henriksen K, Battles JB, Keyes MA, et al., eds. Advances in Patient Safety: New Directions and Alternative Approaches. Vol. 2. Culture and Redesign. Rockville, MD: Agency for Healthcare Research and Quality; 2008.
- , , , . Better transitions: improving comprehension of discharge instructions. Front Health Serv Manage. 2009;25(3):11–32.
- , , , . Structured discharge education improves early outcome in orthopedic patients. Int J Orthop Trauma Nurs. 2010;14(2):66–74.
- , , , et al. Hospital discharge documentation and risk of rehospitalisation. BMJ Qual Saf. 2011;20(9):773–778.
- , , . The hospital discharge: a review of a high risk care transition with highlights of a reengineered discharge process. J Patient Saf. 2007;3(2):97–106.
- , , , , . Interventions to reduce 30‐day rehospitalization: a systematic review. Ann Intern Med. 2011;155(8):520–528.
- , , , , , . Effect of hospital follow‐up appointment on clinical event outcomes and mortality. Arch Intern Med. 2010;170(11):955–960.
- , , , , , . Deficits in communication and information transfer between hospital‐based and primary care physicians: implications for patient safety and continuity of care. JAMA. 2007;297(8):831–841.
- , , , , . Provider characteristics, clinical‐work processes and their relationship to discharge summary quality for sub‐acute care patients. J Gen Intern Med. 2012;27(1):78–84.
- , , , et al. Contemporary evidence about hospital strategies for reducing 30‐day readmissions: a national study. J Am Coll Cardiol. 2012;60(7):607–614.
- , , , . Effect of discharge summary availability during post‐discharge visits on hospital readmission. J Gen Intern Med. 2002;17(3):186–192.
- , , , et al. Association of discharge summary quality with readmission risk for patients hospitalized with heart failure exacerbation. Circ Cardiovasc Qual Outcomes. 2015;8(1):109–111.
- , , , et al. The independent association of provider and information continuity on outcomes after hospital discharge: implications for hospitalists. J Hosp Med. 2010;5(7):398–405.
- , , , , . Timeliness in discharge summary dissemination is associated with patients' clinical outcomes. J Eval Clin Pract. 2013;19(1):76–79.
- , , , et al. Communication and information deficits in patients discharged to rehabilitation facilities: an evaluation of five acute care hospitals. J Hosp Med. 2009;4(8):E28–E33.
- , , . Preventing readmissions through comprehensive discharge planning. Prof Case Manag. 2013;18(2):56–63; quiz 64–65.
- , , , et al. Effect of a postdischarge virtual ward on readmission or death for high‐risk patients: a randomized clinical trial. JAMA. 2014;312(13):1305–1312..
- , , . Risk factors for early unplanned hospital readmission in the elderly. J Gen Intern Med. 1991;6(3):223–228.
- , , , , . Differences between early and late readmissions among patients: a cohort study. Ann Intern Med. 2015;162(11):741–749.
- , , , et al. Post‐acute care payment reform demonstration report to congress supplement‐interim report. Centers for Medicare 14(3):1–14; quiz 88–89.
- , , , et al. Comprehensive discharge planning and home follow‐up of hospitalized elders: a randomized clinical trial. JAMA. 1999;281(7):613–620.
- , , , . Posthospital care transitions: patterns, complications, and risk identification. Health Serv Res. 2004;39(5):1449–1465.
- , , , et al. Transitions of care consensus policy statement: American College of Physicians, Society of General Internal Medicine, Society of Hospital Medicine, American Geriatrics Society, American College of Emergency Physicians, and Society for Academic Emergency Medicine. J Hosp Med. 2009;4(6):364–370.
- , , , et al. Association of self‐reported hospital discharge handoffs with 30‐day readmissions. JAMA Intern Med. 2013;173(8):624–629.
- , . Socioeconomic disparities in health: pathways and policies. Health Aff (Millwood). 2002;21(2):60–76.
- , , , . Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27.
- , , , , , . Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951–1958.
- Centers for Medicare 35(10):1044–1059.
- , , , et al. Understanding and execution of discharge instructions. Am J Med Qual. 2013;28(5):383–391.
- , , , , . Beyond the hospital gates: elucidating the interactive association of social support, depressive symptoms, and physical function with 30‐day readmissions. Am J Phys Med Rehabil. 2015;94(7):555–567.
- , , , et al. Improving the discharge process by embedding a discharge facilitator in a resident team. J Hosp Med. 2011;6(9):494–500.
- , , , et al. Hospital variation in quality of discharge summaries for patients hospitalized with heart failure exacerbation. Circ Cardiovasc Qual Outcomes. 2015;8(1):77–86
- , , , . Addressing the business of discharge: building a case for an electronic discharge summary. J Hosp Med. 2011;6(1):37–42.
- , , , , . Joint commission requirements for discharge instructions in patients with heart failure: is understanding important for preventing readmissions? J Card Fail. 2014;20(9):641–649.
- , , , et al. Association of communication between hospital‐based physicians and primary care providers with patient outcomes. J Gen Intern Med. 2009;24(3):381–386.
- , , , et al. Underlying reasons associated with hospital readmission following surgery in the united states. JAMA. 2015;313(5):483–495.
- , , , , . Assessing quality and efficiency of discharge summaries. Am J Med Qual. 2005;20(6):337–343.
- , , , et al. Comprehensive quality of discharge summaries at an academic medical center. J Hosp Med. 2013;8(8):436–443.
- , , , et al. Development and sustainability of an inpatient‐to‐outpatient discharge handoff tool: a quality improvement project. Jt Comm J Qual Patient Saf. 2014;40(5):219–227.
- , , , , , . Documentation of mandated discharge summary components in transitions from acute to subacute care. In: Henriksen K, Battles JB, Keyes MA, et al., eds. Advances in Patient Safety: New Directions and Alternative Approaches. Vol. 2. Culture and Redesign. Rockville, MD: Agency for Healthcare Research and Quality; 2008.
- , , , . Better transitions: improving comprehension of discharge instructions. Front Health Serv Manage. 2009;25(3):11–32.
- , , , . Structured discharge education improves early outcome in orthopedic patients. Int J Orthop Trauma Nurs. 2010;14(2):66–74.
- , , , et al. Hospital discharge documentation and risk of rehospitalisation. BMJ Qual Saf. 2011;20(9):773–778.
- , , . The hospital discharge: a review of a high risk care transition with highlights of a reengineered discharge process. J Patient Saf. 2007;3(2):97–106.
- , , , , . Interventions to reduce 30‐day rehospitalization: a systematic review. Ann Intern Med. 2011;155(8):520–528.
- , , , , , . Effect of hospital follow‐up appointment on clinical event outcomes and mortality. Arch Intern Med. 2010;170(11):955–960.
- , , , , , . Deficits in communication and information transfer between hospital‐based and primary care physicians: implications for patient safety and continuity of care. JAMA. 2007;297(8):831–841.
- , , , , . Provider characteristics, clinical‐work processes and their relationship to discharge summary quality for sub‐acute care patients. J Gen Intern Med. 2012;27(1):78–84.
- , , , et al. Contemporary evidence about hospital strategies for reducing 30‐day readmissions: a national study. J Am Coll Cardiol. 2012;60(7):607–614.
- , , , . Effect of discharge summary availability during post‐discharge visits on hospital readmission. J Gen Intern Med. 2002;17(3):186–192.
- , , , et al. Association of discharge summary quality with readmission risk for patients hospitalized with heart failure exacerbation. Circ Cardiovasc Qual Outcomes. 2015;8(1):109–111.
- , , , et al. The independent association of provider and information continuity on outcomes after hospital discharge: implications for hospitalists. J Hosp Med. 2010;5(7):398–405.
- , , , , . Timeliness in discharge summary dissemination is associated with patients' clinical outcomes. J Eval Clin Pract. 2013;19(1):76–79.
- , , , et al. Communication and information deficits in patients discharged to rehabilitation facilities: an evaluation of five acute care hospitals. J Hosp Med. 2009;4(8):E28–E33.
- , , . Preventing readmissions through comprehensive discharge planning. Prof Case Manag. 2013;18(2):56–63; quiz 64–65.
- , , , et al. Effect of a postdischarge virtual ward on readmission or death for high‐risk patients: a randomized clinical trial. JAMA. 2014;312(13):1305–1312..
- , , . Risk factors for early unplanned hospital readmission in the elderly. J Gen Intern Med. 1991;6(3):223–228.
- , , , , . Differences between early and late readmissions among patients: a cohort study. Ann Intern Med. 2015;162(11):741–749.
- , , , et al. Post‐acute care payment reform demonstration report to congress supplement‐interim report. Centers for Medicare 14(3):1–14; quiz 88–89.
- , , , et al. Comprehensive discharge planning and home follow‐up of hospitalized elders: a randomized clinical trial. JAMA. 1999;281(7):613–620.
- , , , . Posthospital care transitions: patterns, complications, and risk identification. Health Serv Res. 2004;39(5):1449–1465.
- , , , et al. Transitions of care consensus policy statement: American College of Physicians, Society of General Internal Medicine, Society of Hospital Medicine, American Geriatrics Society, American College of Emergency Physicians, and Society for Academic Emergency Medicine. J Hosp Med. 2009;4(6):364–370.
- , , , et al. Association of self‐reported hospital discharge handoffs with 30‐day readmissions. JAMA Intern Med. 2013;173(8):624–629.
- , . Socioeconomic disparities in health: pathways and policies. Health Aff (Millwood). 2002;21(2):60–76.
- , , , . Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27.
- , , , , , . Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951–1958.
- Centers for Medicare 35(10):1044–1059.
- , , , et al. Understanding and execution of discharge instructions. Am J Med Qual. 2013;28(5):383–391.
- , , , , . Beyond the hospital gates: elucidating the interactive association of social support, depressive symptoms, and physical function with 30‐day readmissions. Am J Phys Med Rehabil. 2015;94(7):555–567.
- , , , et al. Improving the discharge process by embedding a discharge facilitator in a resident team. J Hosp Med. 2011;6(9):494–500.
- , , , et al. Hospital variation in quality of discharge summaries for patients hospitalized with heart failure exacerbation. Circ Cardiovasc Qual Outcomes. 2015;8(1):77–86
- , , , . Addressing the business of discharge: building a case for an electronic discharge summary. J Hosp Med. 2011;6(1):37–42.
- , , , , . Joint commission requirements for discharge instructions in patients with heart failure: is understanding important for preventing readmissions? J Card Fail. 2014;20(9):641–649.
- , , , et al. Association of communication between hospital‐based physicians and primary care providers with patient outcomes. J Gen Intern Med. 2009;24(3):381–386.
- , , , et al. Underlying reasons associated with hospital readmission following surgery in the united states. JAMA. 2015;313(5):483–495.
- , , , , . Assessing quality and efficiency of discharge summaries. Am J Med Qual. 2005;20(6):337–343.
- , , , et al. Comprehensive quality of discharge summaries at an academic medical center. J Hosp Med. 2013;8(8):436–443.
© 2016 Society of Hospital Medicine
Promoting Mobility and Reducing LOS
Annually, more than 35 million patients are hospitalized in the United States, with many experiencing hospital‐acquired impairments in physical functioning during their in‐patient stay.[1, 2, 3, 4] Such impairments include difficulties performing basic activities of daily living, such as rising from a chair, toileting, or ambulating. This functional decline may result in increased length of stay (LOS), nursing home placement, and decreased mobility and participation in community activities even years after hospitalization.[1, 2, 3, 5, 6, 7] Ameliorating this hospital‐acquired functional impairment is important to improving patient outcomes and reducing healthcare utilization. Even the sickest hospitalized patients (eg, those in the intensive care unit [ICU]), can safely and feasibly benefit from early mobilization.[6, 8, 9, 10, 11] In the non‐ICU setting there is also evidence that patient mobilization reduces LOS and hospital costs, while improving patient satisfaction and physical and psychological outcomes.[12, 13, 14, 15, 16] These studies are, however, difficult to replicate as part of routine clinical care, because they often do not present the details of how early mobility was incorporated into daily practice, require additional hospital resources (eg, specially trained providers or additional staff), or are focused only on a select patient population.
The Johns Hopkins medical ICU started early rehabilitation quality‐improvement (QI) work in 2007, which has demonstrated ongoing reductions in LOS and been transformative in terms of helping to foster a culture of mobility at our institution. Previous research suggests that ICU‐based rehabilitation interventions are often not carried over to the ward setting, even in post‐ICU patients.[17] Moreover, trends for sicker patients being admitted in our general medicine units,[18] growing reports of patients spending most of their time in bed,[2, 19, 20] and healthcare policies emphasizing the importance of improving inpatient outcomes motivated the need for QI to improve patient mobility in this setting. Experience from the medical ICU‐based early rehabilitation program helped drive multidisciplinary collaboration of stakeholders to develop this nurse‐driven, mobility promotion QI project on 2 general medicine hospital units. The main goals of the project were to see whether a QI framework can be used in a general medicine setting to increase patient mobility and reduce LOS.[21, 22]
METHODS
Overview of Project
Mobility, for this project, was defined as a patient getting out of bed (eg, sitting out of bed, toileting at bedside commode or bathroom, standing, and ambulating). We aimed to increase patient mobility using preexisting unit staffing ratios of clinicians and support staff. This project was reported in accordance with the SQUIRE (Standards for QUality Improvement Reporting Excellence) guidelines and used a structured QI model that had been used to successfully promote early mobility in the intensive care unit.[21, 23, 24, 25] The planning phase of the QI project began in spring 2012, with initiation of the 12‐month project on March 1, 2013. During the 12‐month QI period, prospective collection of mobility status occurred for all patients, with no exclusions based on patient characteristics.
Setting
The QI project setting was 2, 24‐bed, general medicine units at the Johns Hopkins Hospital, a large academic medical center located in Baltimore, Maryland.
QI Process
The primary goals of the QI project were to mobilize patients 3 times daily, quantify and document the mobility of the patients, set daily goals to increase mobility (eg, move up 1 step on the scale today), and standardize the description of patient mobility across all hospital staff. We used a structured QI model that that has been used to implement an early mobility program in a medical ICU at our institution[21, 22, 24] (see Supporting Information, Appendix, in the online version of this article). At a programmatic level, we involved key stakeholders (nurses, physicians, rehabilitation therapists, administrators) in the QI project team, we identified local barriers to implementation through team meetings as well as a survey tool to identify perceived barriers,[26] and we developed a scale (the Johns Hopkins Highest Level of Mobility [JH‐HLM]) to document mobility. The JH‐HLM is an 8‐point ordinal scale that captures mobility milestones, where 1 = only lying, 2 = bed activities, 3 = sit at edge of bed, 4 = transfer to chair/commode, 5 = standing for 1 minute, 6 = walking 10+ steps, 7 = walking 25+ feet, and 8 = walking 250+ feet (see Supporting Information, Appendix and Supporting Figure 1, in the online version of this article for additional information on the JH‐HLM scale).
The 12‐month QI project was characterized by several phases and milestones and involved a number of intervention components. During the first 4 months (ramp‐up phase), nurses received education in the form of unit‐based presentations, hands‐on‐training, and online education modules. On a 5‐times weekly basis, nurses met with rehabilitation therapists for unit‐based huddles to discuss baseline patient mobility, current patient mobility levels, barriers to mobilizing patients, and daily goals to progress mobility. Mobility levels were included on daily nursing report sheets to facilitate communication with subsequent shifts. Discussion of JH‐HLM scores also occurred during daily unit‐based care‐coordination meetings of the nurses, physicians, and social‐workers to address barriers to mobilizing patients, such as optimizing pain control, facilitating discharge location planning, and expediting physician consultation with physical and occupational therapy for appropriate patients. Audit and feedback from huddles and care‐coordination rounds resulted in improved nurse attendance and engagement during these meetings. Nurses were expected to document patient mobility scores using the JH‐HLM 3 times daily in the patient medical record. On the fourth month, reports on JH‐HLM scores and documentation compliance were available to nurse managers, champions, and unit staff. Via twice‐monthly meetings with the units and quarterly meetings with hospital leadership and administration, problems arising during the QI intervention were evaluated and resolved on a timely basis. Seven months after project execution started, educational sessions were repeated to all staff, and feedback was provided based on the data collected, such as documentation compliance rates and patient mobility levels, and nurse champions presented the project during an American Nurses Credentialing Center magnet recognition program visit. Lastly, mobility scores and documentation compliance were continually assessed for 4 months after the project completion to determine sustainability of the intervention. Additional details of the QI project implementation are provided in the Supporting Information, Appendix, in the online version of this article.
Data Sources and Covariates for Project Evaluation
The Sunrise Clinical Manager system (Allscripts Healthcare Solutions Inc., Chicago, IL) was used to document and extract nursing‐documented JH‐HLM scores. The Johns Hopkins Hospital Datamart financial database, used for mandatory reporting to the State of Maryland, provided data on LOS, age, sex, race (white, black, other), payer (Medicare, Medicaid, other), primary admission diagnosis, and comorbidity index using Agency for Healthcare Research and Quality (AHRQ) methodology.[27] Expected LOS was calculated using the risk adjustment method developed by the University Health System Consortium (UHC).[28] This calculation uses a combination of the Diagnostic‐Related Group grouper and the Sachs Complication Profiler[29] in conjunction with data on specific patient characteristics (age, sex, urgency of admission, payer category) to construct risk‐adjustment regression models that assign expected values for LOS, and is not based on actual LOS.[28] The databases were linked at the patient level using the patient's medical record and unique admission record number.
Outcome Measures
Two functional outcome measures were based on daily JH‐HLM scores, which frequently occurred several times on each patient‐day: (1) the maximum daily JH‐HLM scores for each patient‐day during hospitalization, and (2) the intrapatient change in JH‐HLM scores between the maximum JH‐HLM score within 24 hours of hospital admission and 24 hours before discharge for all patients who were on the unit >48 hours. We also compared the mean LOS during the 12‐month QI project versus the 12‐months prior so we could more accurately address seasonal differences.[30, 31, 32, 33, 34, 35] Lastly, because the perception of increased falls was an important barrier to address in the QI process, we compared the rate of injurious falls between the QI period and 12‐months prior.
Statistical Analysis
To evaluate changes in the percent of ambulatory patients (JH‐HLM 6), we compared the initial 4 months of the QI project (ramp‐up phase) with the same 4‐month period occurring immediately after project completion (post‐QI phase) using generalized estimating equations to account for clustering at the patient‐level. This test was also used to evaluate changes in documentation compliance rates between the 2 phases, with compliance defined as at least 1 instance of JH‐HLM documentation per day, excluding the day of admission and discharge. To evaluate if improved JH‐HLM results were driven by improved documentation compliance rates over time, we performed a sensitivity analysis by imputing a JH‐HLM score of 6 (ambulate 10+ steps) for any missing daily maximum JH‐HLM scores.
To assess unadjusted changes in LOS during the 12‐month QI project versus the same period 1 year earlier, we compared mean and median LOS using a t test and Wilcoxon rank sum test, respectively. We used a multivariable linear regression model to estimate the change (expressed in days) in adjusted median LOS comparing the project months (March 2013March 2014) with 12 months prior (March 2012March 2013). The model adjusted for age, gender, race, payer, admission diagnostic category, UHC expected LOS, and AHRQ comorbidity index. We confirmed a lack of multicollinearity in the multivariable regression model using variance inflation factors. We evaluated residual versus predicted value plots and residual versus fitted value plots with a locally weighted scatterplot smoothing line to confirm model fit. P values are reported from the test of the null hypothesis that the change in adjusted median LOS is the same comparing the QI project months versus 12 months prior. Separate models estimated and tested the change in adjusted median LOS by tertiles of expected LOS (<4, 47, and >7 days). Lastly, we compared the rate of injurious falls (the number of injurious falls by total patient‐days) between the QI period and 12 months prior using an exact Poisson method.[36] Statistical significance was defined as a 2‐sided P < 0.05. Statistical analyses were conducted using R (version 3.1.0; The R Foundation for Statistical Computing, Vienna, Austria;
RESULTS
During the QI project period, 3352 patients were admitted to the 2 general medicine units. Twelve (0.4%) patients expired on the units, but their data were retained in the analysis. Mean (standard deviation [SD]) age of the patients was 54.4 (18.3) years, with 47% male, and 54% African American. A total of 1896 of 6654 (28%) patients on the QI units were 65 years old. Patient characteristics were similar during the QI period versus 12 months prior (Table 1).
| Characteristics | Comparison Period, March 2012March 2013, N = 3,302 | QI Period, March 2013March 2014, N = 3,352 |
|---|---|---|
| ||
| Age, y | 53.3 (17.8) | 54.4 (18.3) |
| Male | 1467 (44%) | 1569 (47%) |
| Race | ||
| African American | 1883 (57%) | 1809 (54%) |
| Caucasian | 1269 (38%) | 1348 (40%) |
| Other | 150 (5%) | 195 (6%) |
| Payer | ||
| Medicare | 1310 (40%) | 1470 (44%) |
| Medicaid | 1015 (31%) | 925 (28%) |
| Other | 977 (30%) | 957 (29%) |
| Admission diagnostic category | ||
| Infectious disease | 579 (18%) | 629 (19%) |
| Pulmonary | 519 (16%) | 559 (17%) |
| Gastrointestinal | 535 (16%) | 494 (15%) |
| Cardiovascular | 410 (12%) | 405 (12%) |
| Hematologic | 199 (6%) | 195 (6%) |
| Renal | 220 (7%) | 205 (6%) |
| Other | 840 (25%) | 865 (26%) |
| UHC expected length of stay, d | 5.5 (3.3) | 5.3 (3.2) |
| AHRQ comorbidity index | 3.3 (1.7) | 3.5 (1.8) |
During the 12‐month QI project, there were a total of 13,815 patient‐days of documented mobility data and the median (interquartile range [IQR]) number of days of documentation for each hospital admission was 3 (25) days. Compliance with daily documentation of JH‐HLM was 85.0% over the entire 12‐month QI project. Documentation compliance started at 83% during the ramp‐up phase and increased to 89% during the last 4 months of the project (late‐QI phase, P < 0.001).
Comparing the ramp‐up phase versus post‐QI phase, the percentage of patient‐days in which patients ambulated (JH‐HLM 6) increased from 43% to 70% (P < 0.001), and the percentage of patients who experienced an improvement in their mobility scores between admission and discharge increased from 32% to 45% (P < 0.001), as shown in Table 2. In the sensitivity analysis imputing missing daily JH‐HLM scores and comparing the ramp‐up versus post‐QI phases, the results were similar to the primary analysis; the percent of patient‐days where patients ambulated increased from 60% to 78% (P < 0.001), and the percent of patients who experienced an improvement in their mobility scores increased from 26% to 48% (P < 0.001).
| JH‐HLM Category | Ramp‐up Phase, March 1, 2013 June 30, 2013, n = 4,649 | Late‐QI Phase, November 1, 2013February 28, 2013, n = 4,515 | Post‐QI Phase, March 1, 2014 June 30, 2014, n = 4,298 |
|---|---|---|---|
| Change in Mobility (Admission Versus Discharge) | Ramp‐up Phase, March 1, 2013June 30, 2013, n = 968 | Late‐QI Phase, November 1, 2013February 28, 2013, n = 893 | Post‐QI Phase, March 1, 2014 June 30, 2014, n = 834 |
| |||
| Walk (JH‐HLM = 6, 7, or 8) | 1,994 (43) | 3,430 (76) | 2,986 (70) |
| Stand/chair (JH‐HLM = 4 or 5) | 1,772 (38) | 488 (10) | 511 (12) |
| Bed (JH‐HLM = 1, 2, or 3) | 883 (19) | 597 (13) | 801 (19) |
| Improved | 305 (32) | 392 (44) | 379 (45) |
| No change | 512 (53) | 428 (48) | 386 (46) |
| Declined | 151 (16) | 73 (8) | 69 (8) |
LOS during the 12‐month QI project versus the 12‐months immediately prior was shorter (Table 3), with an unadjusted median (IQR) LOS of 3 (26) versus 4 (27) days (P < 0.001) and an unadjusted mean (SD) LOS of 5.1 (5.6) versus 6.0 (7.6) (P < 0.001).
Adjusted Median LOS, d | Absolute Change in Adjusted Median LOS (95% CI), d | P Value | ||
|---|---|---|---|---|
| 12 Months Prior | QI Project Months | |||
| ||||
| All patients | 6.01 | 5.61 | 0.40 (0.57 to 0.21), N = 4,411 | <0.001 |
| Subgroups by ELOS | ||||
| ELOS <4 days | 4.68 | 4.77 | 0.09 (0.13 to 0.32), N = 1,357 | 0.42 |
| ELOS 47 days | 5.68 | 5.38 | 0.30 (0.57 to 0.01), N = 1,509 | 0.04 |
| ELOS >7 days | 8.07 | 6.96 | 1.11 (1.53 to 0.65), N = 1,545 | <0.001 |
Table 3 displays the change in adjusted median LOS for the project months versus the 12 months prior among the QI units. We found that for all patients, there was an overall reduction in adjusted median LOS of 0.40 (95% confidence interval [CI]: 0.57 to 0.21, P<0.001) days. When we divided patients into tertiles based on their UHC expected LOS (ELOS), we observed that patients with longer ELOS had greater reductions in adjusted median LOS. Patients on the QI units with ELOS <4 days (lowest tertile) did not show a significant reduction in adjusted median LOS (0.09 days, 95% CI: 0.13 to 0.32, P = 0.42); however, patients with UHC ELOS 4 to 7 days (middle tertile) and ELOS >7 days (highest tertile) had a significant reduction in adjusted median LOS by 0.30 (95% CI: 0.57 to 0.01, P = 0.04) and 1.11 (95% CI: 1.53 to 0.65, P < 0.001) days during the QI project versus 12 months prior, respectively.
Lastly, we found that there was no difference in the rate of injurious falls on the QI units during QI period compared to 12 months prior (QI: 0.34 per 1000 patient‐days versus 12 months prior: 0.48 per 1000 patient‐days, P = 0.73).
DISCUSSION
We conducted a nurse‐driven, multidisciplinary mobility promotion QI project on 2 general medicine units at a large teaching hospital. The 12‐month QI project, conducted between March 1, 2013 and February 28, 2014, was associated with patients ambulating more frequently, with improved mobility status between hospital admission and discharge. These improvements in mobility were not associated with increased rates of injurious falls, and were sustained for at least 4 months after project completion. The QI project was associated with overall significant reduction in LOS for more complex patients with longer expected LOS (4 days or longer). Hence, such QI efforts may be important for maintaining or improving patients' functional status during hospitalization in a safe and cost‐effective manner.
Our findings are consistent with previous studies showing that mobility promotion in the acute hospital setting is feasible, can reduce length of stay, and can be applied to a diverse population including vulnerable medical patients with multiple comorbidities and the elderly.[12, 16, 37, 38, 39, 40, 41, 42] These studies provide valuable evidence of the benefits of mobility promotion; however, it is difficult to translate these prior results into routine clinical practice because they used specially trained staff to mobilize patients, focused on a select patient population, or did not specify how the mobility intervention was delivered within daily clinical workflows. Research in the medical ICU at our institution has previously described the use of a structured QI model to successfully implement an early rehabilitation program.[22, 24] Here, we successfully adapted the same QI framework to a general medicine setting. Hence, our study contributes to the literature with respect to (1) use of a structured QI framework to develop a successful patient mobility program in a general medicine patient population, and (2) sharing best practices from 1 clinical setting, such as the ICU, as a source of learning and knowledge translation for other care settings, with the addition of novel tools, such as the JH‐HLM scale.
There may have been several factors that contributed to shorter stays in the hospital we observed during the QI project. First, we increased the number of ambulatory patient‐days, which may have helped prevent physiological complications of bed rest, such as muscle weakness, atelectasis, insulin resistance, vascular dysfunction, contractures, and pressure ulcers.[43] As such, mobility promotion has been associated with reduced rates of other hospital‐acquired complications, such as deep venous thrombosis, pneumonia, and delirium.[44, 45, 46] In our study, we saw the greatest LOS reduction in more complex patients who were expected to spend a longer time in the hospital and are at greater risk of developing complications from bed rest. Second, our early mobility project may have had a direct impact on care‐coordination processes as reported in prior studies.[47, 48, 49] An important component of our intervention was incorporating functional status into multidisciplinary discussions, either through nurse‐to‐therapist huddles or care‐coordination rounds between nurses, therapists, physicians, social workers, and case managers. During care‐coordination rounds, JH‐HLM scores were reported to expedite appropriate physical and occupational therapy consultations and assist in determining appropriate discharge location. During the QI project, we transitioned from a unit‐based daily huddle between nursing and rehabilitation therapists to a system where mobility status was discussed primarily during care coordination rounds 5 times per week. We saw that mobility scores were maintained after QI project completion, suggesting that reporting on patient function in a multidisciplinary setting is a potentially sustainable mechanism to improve care‐coordination processes that are affected by functional status.
Our study has several potential limitations. First, this is a single‐site study in 2 general medicine units of a large academic hospital. Further research is needed to determine if this structured QI intervention and its benefits can be generalized to different settings and different patient populations. Second, because the documentation was initially an optional element in the electronic medical record system, we observed higher rates of missing documentation during the first 4 months of the project versus the comparison period at 4 months after project completion. However, a sensitivity analysis conducted of these missing data demonstrated similar results to our primary analysis. Third, our nonrandomized pre‐post study design does not allow us to conclude a direct cause‐and‐effect relationship between our intervention and increased mobility and reduced LOS. Although patient characteristics were similar between the 2 periods and adjusted for in our multivariable regression analysis, we cannot rule out the possibility of secular trends in LOS on the project units and that broader QI efforts at our institution also contributed to reduction in LOS. Fourth, we do not have data on 30‐day readmissions and discharge location. Future studies should explore the impact of hospital‐based mobility interventions on these outcomes.[50] Fifth, although nurses consistently documented the highest level of mobility on a daily basis, these data did not capture other potentially important information about patient mobility such as the daily frequency that patients were mobilized, the length of time a patient was engaged in a mobility event (ie, number of hours sitting in a chair), or the mobility that occurred during physical therapy or occupational therapy sessions. Hence, although we used JH‐HLM as a marker of improved mobility during our QI project it is likely that our data cannot fully describe the total mobility and activity that patients experienced during hospitalization. Lastly, although the front‐line staff and QI team found the JH‐HLM scale to be a useful tool to measure and advance patient mobility, further studies are needed to evaluate the reliability and validity of this scale.
CONCLUSION
A structured QI process can improve patient mobility and may contribute to reduction in LOS, particularly for more complex patients in this setting. Active prevention of decline in physical function that commonly occurs during hospitalization may prove valuable for improving patient outcomes and reducing healthcare resource utilization.
Disclosures
The authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated. The authors report no conflicts of interest.
- , , , et al. Loss of independence in activities of daily living in older adults hospitalized with medical illnesses: increased vulnerability with age. J Am Geriatr Soc. 2003;51(4):451–458.
- , , . Prevalence and outcomes of low mobility in hospitalized older patients. J Am Geriatr Soc. 2004;52(8):1263–1270.
- , , , , , . Trajectories of life‐space mobility after hospitalization. Ann Intern Med. 2009;150(6):372–378.
- , , . Hospitalization‐associated disability: “She was probably able to ambulate, but I'm not sure”. JAMA. 2011;306(16):1782–1793.
- , , , , , . Physical fitness and all‐cause mortality. A prospective study of healthy men and women. JAMA. 1989;262(17):2395–2401.
- . Mobilizing patients in the intensive care unit: improving neuromuscular weakness and physical function. JAMA. 2008;300(14):1685–1690.
- , . Mobility limitation in the older patient: a clinical review. JAMA. 2013;310(11):1168–1177.
- , , , et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373(9678):1874–1882.
- , , . Technology to enhance physical rehabilitation of critically ill patients. Crit Care Med. 2009;37(10 suppl):S436–S441.
- , , , et al. Receiving early mobility during an intensive care unit admission is a predictor of improved outcomes in acute respiratory failure. Am J Med Sci. 2011;341(5):373–377.
- . Physiotherapy in intensive care: an updated systematic review. Chest. 2013;144(3):825–847.
- , , . Exercise for acutely hospitalised older medical patients. Cochrane Database Syst Rev. 2007;(1):CD005955.
- , , . Extra physical therapy reduces patient length of stay and improves functional outcomes and quality of life in people with acute or subacute conditions: a systematic review. Arch Phys Med Rehabil. 2011;92(9):1490–1500.
- , . Impact of early mobilization protocol on the medical‐surgical inpatient population: an integrated review of literature. Clin Nurse Spec. 2012;26(2):87–94.
- , , . Outcomes of inpatient mobilization: a literature review. J Clin Nurs. 2014;23(11–12):1486–1501.
- , , , et al. The early mobility bundle: a simple enhancement of therapy which may reduce incidence of hospital‐acquired pneumonia and length of hospital stay. J Hosp Infect. 2014;88(1):34–39.
- , , , , . Physical therapy on the wards after early physical activity and mobility in the intensive care unit. Phys Ther. 2012;92(12):1518–1523.
- , , , . Impact of hospital variables on case mix index as a marker of disease severity. Popul Health Manag. 2014;17(1):28–34.
- , , , , . Frequency of hallway ambulation by hospitalized older adults on medical units of an academic hospital. Geriatr Nurs. 2004;25(4):212–217.
- , , . Activity level of hospital medical inpatients: an observational study. Arch Gerontol Geriatr. 2012;55(2):417–421.
- , , . Translating evidence into practice: a model for large scale knowledge translation. BMJ. 2008;337:a1714.
- , , , et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil. 2010;91(4):536–542.
- , , , , ; SQUIRE development group. Publication guidelines for quality improvement studies in health care: evolution of the SQUIRE project. BMJ. 2009;338:a3152.
- , . Rehabilitation quality improvement in an intensive care unit setting: implementation of a quality improvement model. Top Stroke Rehabil. 2010;17(4):271–281.
- , , , . ICU early mobilization: from recommendation to implementation at three medical centers. Crit Care Med. 2013;41(9 suppl 1):S69–S80.
- , , , . Barriers to early mobility of hospitalized general medicine patients: survey development and results. Am J Phys Med Rehabil. 2015;94(4):304–312.
- , , , . Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27.
- UHC Clinical Information Management Risk Adjustment of the UHC Clinical Data Base. Chicago, IL: University HealthSystem Consortium; 1998.
- Sachs Complications Profiler, Version 1.0, User's Guide. Evanston, IL: Sachs Group; 1995.
- , , . Assessing hospital‐associated deaths from discharge data. The role of length of stay and comorbidities. JAMA. 1988;260(15):2240–2246.
- , , , et al. Annual rates of admission and seasonal variations in hospitalizations for heart failure. Eur J Heart Fail. 2002;4(6):779–786.
- , , , . Regional and seasonal variation in the length of hospital stay for chronic obstructive pulmonary disease in Finland. Int J Circumpolar Health. 2002;61(2):131–135.
- , , , , , ; Canadian CABG Surgery Quality Indicator Consensus Panel. The identification and development of canadian coronary artery bypass graft surgery quality indicators. J Thorac Cardiovasc Surg. 2005;130(5):1257.
- , , , . Quality of care and length of hospital stay among patients with stroke. Med Care. 2009;47(5):575–582.
- . A systematic review of outcomes and quality measures in adult patients cared for by hospitalists vs nonhospitalists. Mayo Clin Proc. 2009;84(3):248–254.
- . Confidence intervals that match Fisher's exact or Blaker's exact tests. Biostatistics. 2010;11(2):373–374.
- , , , et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med. 1999;340(9):669–676.
- , , , , . Early mobilization of patients hospitalized with community‐acquired pneumonia. Chest. 2003;124(3):883–889.
- , , , et al. Effects of a walking intervention on fatigue‐related experiences of hospitalized acute myelogenous leukemia patients undergoing chemotherapy: a randomized controlled trial. J Pain Symptom Manage. 2008;35(5):524–534.
- , , , , . Early ambulation and length of stay in older adults hospitalized for acute illness. Arch Intern Med. 2010;170(21):1942–1943.
- , , . Impact of a nurse‐driven mobility protocol on functional decline in hospitalized older adults. J Nurs Care Qual. 2009;24(4):325–331.
- , , . Exercising body and mind: an integrated approach to functional independence in hospitalized older people. J Am Geriatr Soc. 2008;56(4):630–635.
- . Consequences of bed rest. Crit Care Med. 2009;37(10 suppl):S422–S428.
- , , , , . Time to ambulation after hip fracture surgery: relation to hospitalization outcomes. J Gerontol A Biol Sci Med Sci. 2003;58(11):M1042–M1045.
- , , , . Early mobilization after total knee replacement reduces the incidence of deep venous thrombosis. ANZ J Surg. 2009;79(7–8):526–529.
- , , , . Efficacy and safety of postoperative early mobilization for chronic subdural hematoma in elderly patients. Acta Neurochir (Wien). 2010;152(7):1171–1174.
- , , , et al. Impact of relational coordination on quality of care, postoperative pain and functioning, and length of stay: a nine‐hospital study of surgical patients. Med Care. 2000;38(8):807–819.
- Care coordination cuts admissions, ED visits, LOS. Hosp Case Manag. 2013;21(5):67–68.
- , . A heart failure initiative to reduce the length of stay and readmission rates. Prof Case Manag. 2014;19(6):276–284.
- , , , , , . Association of impaired functional status at hospital discharge and subsequent rehospitalization. J Hosp Med. 2014;9(5):277–282.
Annually, more than 35 million patients are hospitalized in the United States, with many experiencing hospital‐acquired impairments in physical functioning during their in‐patient stay.[1, 2, 3, 4] Such impairments include difficulties performing basic activities of daily living, such as rising from a chair, toileting, or ambulating. This functional decline may result in increased length of stay (LOS), nursing home placement, and decreased mobility and participation in community activities even years after hospitalization.[1, 2, 3, 5, 6, 7] Ameliorating this hospital‐acquired functional impairment is important to improving patient outcomes and reducing healthcare utilization. Even the sickest hospitalized patients (eg, those in the intensive care unit [ICU]), can safely and feasibly benefit from early mobilization.[6, 8, 9, 10, 11] In the non‐ICU setting there is also evidence that patient mobilization reduces LOS and hospital costs, while improving patient satisfaction and physical and psychological outcomes.[12, 13, 14, 15, 16] These studies are, however, difficult to replicate as part of routine clinical care, because they often do not present the details of how early mobility was incorporated into daily practice, require additional hospital resources (eg, specially trained providers or additional staff), or are focused only on a select patient population.
The Johns Hopkins medical ICU started early rehabilitation quality‐improvement (QI) work in 2007, which has demonstrated ongoing reductions in LOS and been transformative in terms of helping to foster a culture of mobility at our institution. Previous research suggests that ICU‐based rehabilitation interventions are often not carried over to the ward setting, even in post‐ICU patients.[17] Moreover, trends for sicker patients being admitted in our general medicine units,[18] growing reports of patients spending most of their time in bed,[2, 19, 20] and healthcare policies emphasizing the importance of improving inpatient outcomes motivated the need for QI to improve patient mobility in this setting. Experience from the medical ICU‐based early rehabilitation program helped drive multidisciplinary collaboration of stakeholders to develop this nurse‐driven, mobility promotion QI project on 2 general medicine hospital units. The main goals of the project were to see whether a QI framework can be used in a general medicine setting to increase patient mobility and reduce LOS.[21, 22]
METHODS
Overview of Project
Mobility, for this project, was defined as a patient getting out of bed (eg, sitting out of bed, toileting at bedside commode or bathroom, standing, and ambulating). We aimed to increase patient mobility using preexisting unit staffing ratios of clinicians and support staff. This project was reported in accordance with the SQUIRE (Standards for QUality Improvement Reporting Excellence) guidelines and used a structured QI model that had been used to successfully promote early mobility in the intensive care unit.[21, 23, 24, 25] The planning phase of the QI project began in spring 2012, with initiation of the 12‐month project on March 1, 2013. During the 12‐month QI period, prospective collection of mobility status occurred for all patients, with no exclusions based on patient characteristics.
Setting
The QI project setting was 2, 24‐bed, general medicine units at the Johns Hopkins Hospital, a large academic medical center located in Baltimore, Maryland.
QI Process
The primary goals of the QI project were to mobilize patients 3 times daily, quantify and document the mobility of the patients, set daily goals to increase mobility (eg, move up 1 step on the scale today), and standardize the description of patient mobility across all hospital staff. We used a structured QI model that that has been used to implement an early mobility program in a medical ICU at our institution[21, 22, 24] (see Supporting Information, Appendix, in the online version of this article). At a programmatic level, we involved key stakeholders (nurses, physicians, rehabilitation therapists, administrators) in the QI project team, we identified local barriers to implementation through team meetings as well as a survey tool to identify perceived barriers,[26] and we developed a scale (the Johns Hopkins Highest Level of Mobility [JH‐HLM]) to document mobility. The JH‐HLM is an 8‐point ordinal scale that captures mobility milestones, where 1 = only lying, 2 = bed activities, 3 = sit at edge of bed, 4 = transfer to chair/commode, 5 = standing for 1 minute, 6 = walking 10+ steps, 7 = walking 25+ feet, and 8 = walking 250+ feet (see Supporting Information, Appendix and Supporting Figure 1, in the online version of this article for additional information on the JH‐HLM scale).
The 12‐month QI project was characterized by several phases and milestones and involved a number of intervention components. During the first 4 months (ramp‐up phase), nurses received education in the form of unit‐based presentations, hands‐on‐training, and online education modules. On a 5‐times weekly basis, nurses met with rehabilitation therapists for unit‐based huddles to discuss baseline patient mobility, current patient mobility levels, barriers to mobilizing patients, and daily goals to progress mobility. Mobility levels were included on daily nursing report sheets to facilitate communication with subsequent shifts. Discussion of JH‐HLM scores also occurred during daily unit‐based care‐coordination meetings of the nurses, physicians, and social‐workers to address barriers to mobilizing patients, such as optimizing pain control, facilitating discharge location planning, and expediting physician consultation with physical and occupational therapy for appropriate patients. Audit and feedback from huddles and care‐coordination rounds resulted in improved nurse attendance and engagement during these meetings. Nurses were expected to document patient mobility scores using the JH‐HLM 3 times daily in the patient medical record. On the fourth month, reports on JH‐HLM scores and documentation compliance were available to nurse managers, champions, and unit staff. Via twice‐monthly meetings with the units and quarterly meetings with hospital leadership and administration, problems arising during the QI intervention were evaluated and resolved on a timely basis. Seven months after project execution started, educational sessions were repeated to all staff, and feedback was provided based on the data collected, such as documentation compliance rates and patient mobility levels, and nurse champions presented the project during an American Nurses Credentialing Center magnet recognition program visit. Lastly, mobility scores and documentation compliance were continually assessed for 4 months after the project completion to determine sustainability of the intervention. Additional details of the QI project implementation are provided in the Supporting Information, Appendix, in the online version of this article.
Data Sources and Covariates for Project Evaluation
The Sunrise Clinical Manager system (Allscripts Healthcare Solutions Inc., Chicago, IL) was used to document and extract nursing‐documented JH‐HLM scores. The Johns Hopkins Hospital Datamart financial database, used for mandatory reporting to the State of Maryland, provided data on LOS, age, sex, race (white, black, other), payer (Medicare, Medicaid, other), primary admission diagnosis, and comorbidity index using Agency for Healthcare Research and Quality (AHRQ) methodology.[27] Expected LOS was calculated using the risk adjustment method developed by the University Health System Consortium (UHC).[28] This calculation uses a combination of the Diagnostic‐Related Group grouper and the Sachs Complication Profiler[29] in conjunction with data on specific patient characteristics (age, sex, urgency of admission, payer category) to construct risk‐adjustment regression models that assign expected values for LOS, and is not based on actual LOS.[28] The databases were linked at the patient level using the patient's medical record and unique admission record number.
Outcome Measures
Two functional outcome measures were based on daily JH‐HLM scores, which frequently occurred several times on each patient‐day: (1) the maximum daily JH‐HLM scores for each patient‐day during hospitalization, and (2) the intrapatient change in JH‐HLM scores between the maximum JH‐HLM score within 24 hours of hospital admission and 24 hours before discharge for all patients who were on the unit >48 hours. We also compared the mean LOS during the 12‐month QI project versus the 12‐months prior so we could more accurately address seasonal differences.[30, 31, 32, 33, 34, 35] Lastly, because the perception of increased falls was an important barrier to address in the QI process, we compared the rate of injurious falls between the QI period and 12‐months prior.
Statistical Analysis
To evaluate changes in the percent of ambulatory patients (JH‐HLM 6), we compared the initial 4 months of the QI project (ramp‐up phase) with the same 4‐month period occurring immediately after project completion (post‐QI phase) using generalized estimating equations to account for clustering at the patient‐level. This test was also used to evaluate changes in documentation compliance rates between the 2 phases, with compliance defined as at least 1 instance of JH‐HLM documentation per day, excluding the day of admission and discharge. To evaluate if improved JH‐HLM results were driven by improved documentation compliance rates over time, we performed a sensitivity analysis by imputing a JH‐HLM score of 6 (ambulate 10+ steps) for any missing daily maximum JH‐HLM scores.
To assess unadjusted changes in LOS during the 12‐month QI project versus the same period 1 year earlier, we compared mean and median LOS using a t test and Wilcoxon rank sum test, respectively. We used a multivariable linear regression model to estimate the change (expressed in days) in adjusted median LOS comparing the project months (March 2013March 2014) with 12 months prior (March 2012March 2013). The model adjusted for age, gender, race, payer, admission diagnostic category, UHC expected LOS, and AHRQ comorbidity index. We confirmed a lack of multicollinearity in the multivariable regression model using variance inflation factors. We evaluated residual versus predicted value plots and residual versus fitted value plots with a locally weighted scatterplot smoothing line to confirm model fit. P values are reported from the test of the null hypothesis that the change in adjusted median LOS is the same comparing the QI project months versus 12 months prior. Separate models estimated and tested the change in adjusted median LOS by tertiles of expected LOS (<4, 47, and >7 days). Lastly, we compared the rate of injurious falls (the number of injurious falls by total patient‐days) between the QI period and 12 months prior using an exact Poisson method.[36] Statistical significance was defined as a 2‐sided P < 0.05. Statistical analyses were conducted using R (version 3.1.0; The R Foundation for Statistical Computing, Vienna, Austria;
RESULTS
During the QI project period, 3352 patients were admitted to the 2 general medicine units. Twelve (0.4%) patients expired on the units, but their data were retained in the analysis. Mean (standard deviation [SD]) age of the patients was 54.4 (18.3) years, with 47% male, and 54% African American. A total of 1896 of 6654 (28%) patients on the QI units were 65 years old. Patient characteristics were similar during the QI period versus 12 months prior (Table 1).
| Characteristics | Comparison Period, March 2012March 2013, N = 3,302 | QI Period, March 2013March 2014, N = 3,352 |
|---|---|---|
| ||
| Age, y | 53.3 (17.8) | 54.4 (18.3) |
| Male | 1467 (44%) | 1569 (47%) |
| Race | ||
| African American | 1883 (57%) | 1809 (54%) |
| Caucasian | 1269 (38%) | 1348 (40%) |
| Other | 150 (5%) | 195 (6%) |
| Payer | ||
| Medicare | 1310 (40%) | 1470 (44%) |
| Medicaid | 1015 (31%) | 925 (28%) |
| Other | 977 (30%) | 957 (29%) |
| Admission diagnostic category | ||
| Infectious disease | 579 (18%) | 629 (19%) |
| Pulmonary | 519 (16%) | 559 (17%) |
| Gastrointestinal | 535 (16%) | 494 (15%) |
| Cardiovascular | 410 (12%) | 405 (12%) |
| Hematologic | 199 (6%) | 195 (6%) |
| Renal | 220 (7%) | 205 (6%) |
| Other | 840 (25%) | 865 (26%) |
| UHC expected length of stay, d | 5.5 (3.3) | 5.3 (3.2) |
| AHRQ comorbidity index | 3.3 (1.7) | 3.5 (1.8) |
During the 12‐month QI project, there were a total of 13,815 patient‐days of documented mobility data and the median (interquartile range [IQR]) number of days of documentation for each hospital admission was 3 (25) days. Compliance with daily documentation of JH‐HLM was 85.0% over the entire 12‐month QI project. Documentation compliance started at 83% during the ramp‐up phase and increased to 89% during the last 4 months of the project (late‐QI phase, P < 0.001).
Comparing the ramp‐up phase versus post‐QI phase, the percentage of patient‐days in which patients ambulated (JH‐HLM 6) increased from 43% to 70% (P < 0.001), and the percentage of patients who experienced an improvement in their mobility scores between admission and discharge increased from 32% to 45% (P < 0.001), as shown in Table 2. In the sensitivity analysis imputing missing daily JH‐HLM scores and comparing the ramp‐up versus post‐QI phases, the results were similar to the primary analysis; the percent of patient‐days where patients ambulated increased from 60% to 78% (P < 0.001), and the percent of patients who experienced an improvement in their mobility scores increased from 26% to 48% (P < 0.001).
| JH‐HLM Category | Ramp‐up Phase, March 1, 2013 June 30, 2013, n = 4,649 | Late‐QI Phase, November 1, 2013February 28, 2013, n = 4,515 | Post‐QI Phase, March 1, 2014 June 30, 2014, n = 4,298 |
|---|---|---|---|
| Change in Mobility (Admission Versus Discharge) | Ramp‐up Phase, March 1, 2013June 30, 2013, n = 968 | Late‐QI Phase, November 1, 2013February 28, 2013, n = 893 | Post‐QI Phase, March 1, 2014 June 30, 2014, n = 834 |
| |||
| Walk (JH‐HLM = 6, 7, or 8) | 1,994 (43) | 3,430 (76) | 2,986 (70) |
| Stand/chair (JH‐HLM = 4 or 5) | 1,772 (38) | 488 (10) | 511 (12) |
| Bed (JH‐HLM = 1, 2, or 3) | 883 (19) | 597 (13) | 801 (19) |
| Improved | 305 (32) | 392 (44) | 379 (45) |
| No change | 512 (53) | 428 (48) | 386 (46) |
| Declined | 151 (16) | 73 (8) | 69 (8) |
LOS during the 12‐month QI project versus the 12‐months immediately prior was shorter (Table 3), with an unadjusted median (IQR) LOS of 3 (26) versus 4 (27) days (P < 0.001) and an unadjusted mean (SD) LOS of 5.1 (5.6) versus 6.0 (7.6) (P < 0.001).
Adjusted Median LOS, d | Absolute Change in Adjusted Median LOS (95% CI), d | P Value | ||
|---|---|---|---|---|
| 12 Months Prior | QI Project Months | |||
| ||||
| All patients | 6.01 | 5.61 | 0.40 (0.57 to 0.21), N = 4,411 | <0.001 |
| Subgroups by ELOS | ||||
| ELOS <4 days | 4.68 | 4.77 | 0.09 (0.13 to 0.32), N = 1,357 | 0.42 |
| ELOS 47 days | 5.68 | 5.38 | 0.30 (0.57 to 0.01), N = 1,509 | 0.04 |
| ELOS >7 days | 8.07 | 6.96 | 1.11 (1.53 to 0.65), N = 1,545 | <0.001 |
Table 3 displays the change in adjusted median LOS for the project months versus the 12 months prior among the QI units. We found that for all patients, there was an overall reduction in adjusted median LOS of 0.40 (95% confidence interval [CI]: 0.57 to 0.21, P<0.001) days. When we divided patients into tertiles based on their UHC expected LOS (ELOS), we observed that patients with longer ELOS had greater reductions in adjusted median LOS. Patients on the QI units with ELOS <4 days (lowest tertile) did not show a significant reduction in adjusted median LOS (0.09 days, 95% CI: 0.13 to 0.32, P = 0.42); however, patients with UHC ELOS 4 to 7 days (middle tertile) and ELOS >7 days (highest tertile) had a significant reduction in adjusted median LOS by 0.30 (95% CI: 0.57 to 0.01, P = 0.04) and 1.11 (95% CI: 1.53 to 0.65, P < 0.001) days during the QI project versus 12 months prior, respectively.
Lastly, we found that there was no difference in the rate of injurious falls on the QI units during QI period compared to 12 months prior (QI: 0.34 per 1000 patient‐days versus 12 months prior: 0.48 per 1000 patient‐days, P = 0.73).
DISCUSSION
We conducted a nurse‐driven, multidisciplinary mobility promotion QI project on 2 general medicine units at a large teaching hospital. The 12‐month QI project, conducted between March 1, 2013 and February 28, 2014, was associated with patients ambulating more frequently, with improved mobility status between hospital admission and discharge. These improvements in mobility were not associated with increased rates of injurious falls, and were sustained for at least 4 months after project completion. The QI project was associated with overall significant reduction in LOS for more complex patients with longer expected LOS (4 days or longer). Hence, such QI efforts may be important for maintaining or improving patients' functional status during hospitalization in a safe and cost‐effective manner.
Our findings are consistent with previous studies showing that mobility promotion in the acute hospital setting is feasible, can reduce length of stay, and can be applied to a diverse population including vulnerable medical patients with multiple comorbidities and the elderly.[12, 16, 37, 38, 39, 40, 41, 42] These studies provide valuable evidence of the benefits of mobility promotion; however, it is difficult to translate these prior results into routine clinical practice because they used specially trained staff to mobilize patients, focused on a select patient population, or did not specify how the mobility intervention was delivered within daily clinical workflows. Research in the medical ICU at our institution has previously described the use of a structured QI model to successfully implement an early rehabilitation program.[22, 24] Here, we successfully adapted the same QI framework to a general medicine setting. Hence, our study contributes to the literature with respect to (1) use of a structured QI framework to develop a successful patient mobility program in a general medicine patient population, and (2) sharing best practices from 1 clinical setting, such as the ICU, as a source of learning and knowledge translation for other care settings, with the addition of novel tools, such as the JH‐HLM scale.
There may have been several factors that contributed to shorter stays in the hospital we observed during the QI project. First, we increased the number of ambulatory patient‐days, which may have helped prevent physiological complications of bed rest, such as muscle weakness, atelectasis, insulin resistance, vascular dysfunction, contractures, and pressure ulcers.[43] As such, mobility promotion has been associated with reduced rates of other hospital‐acquired complications, such as deep venous thrombosis, pneumonia, and delirium.[44, 45, 46] In our study, we saw the greatest LOS reduction in more complex patients who were expected to spend a longer time in the hospital and are at greater risk of developing complications from bed rest. Second, our early mobility project may have had a direct impact on care‐coordination processes as reported in prior studies.[47, 48, 49] An important component of our intervention was incorporating functional status into multidisciplinary discussions, either through nurse‐to‐therapist huddles or care‐coordination rounds between nurses, therapists, physicians, social workers, and case managers. During care‐coordination rounds, JH‐HLM scores were reported to expedite appropriate physical and occupational therapy consultations and assist in determining appropriate discharge location. During the QI project, we transitioned from a unit‐based daily huddle between nursing and rehabilitation therapists to a system where mobility status was discussed primarily during care coordination rounds 5 times per week. We saw that mobility scores were maintained after QI project completion, suggesting that reporting on patient function in a multidisciplinary setting is a potentially sustainable mechanism to improve care‐coordination processes that are affected by functional status.
Our study has several potential limitations. First, this is a single‐site study in 2 general medicine units of a large academic hospital. Further research is needed to determine if this structured QI intervention and its benefits can be generalized to different settings and different patient populations. Second, because the documentation was initially an optional element in the electronic medical record system, we observed higher rates of missing documentation during the first 4 months of the project versus the comparison period at 4 months after project completion. However, a sensitivity analysis conducted of these missing data demonstrated similar results to our primary analysis. Third, our nonrandomized pre‐post study design does not allow us to conclude a direct cause‐and‐effect relationship between our intervention and increased mobility and reduced LOS. Although patient characteristics were similar between the 2 periods and adjusted for in our multivariable regression analysis, we cannot rule out the possibility of secular trends in LOS on the project units and that broader QI efforts at our institution also contributed to reduction in LOS. Fourth, we do not have data on 30‐day readmissions and discharge location. Future studies should explore the impact of hospital‐based mobility interventions on these outcomes.[50] Fifth, although nurses consistently documented the highest level of mobility on a daily basis, these data did not capture other potentially important information about patient mobility such as the daily frequency that patients were mobilized, the length of time a patient was engaged in a mobility event (ie, number of hours sitting in a chair), or the mobility that occurred during physical therapy or occupational therapy sessions. Hence, although we used JH‐HLM as a marker of improved mobility during our QI project it is likely that our data cannot fully describe the total mobility and activity that patients experienced during hospitalization. Lastly, although the front‐line staff and QI team found the JH‐HLM scale to be a useful tool to measure and advance patient mobility, further studies are needed to evaluate the reliability and validity of this scale.
CONCLUSION
A structured QI process can improve patient mobility and may contribute to reduction in LOS, particularly for more complex patients in this setting. Active prevention of decline in physical function that commonly occurs during hospitalization may prove valuable for improving patient outcomes and reducing healthcare resource utilization.
Disclosures
The authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated. The authors report no conflicts of interest.
Annually, more than 35 million patients are hospitalized in the United States, with many experiencing hospital‐acquired impairments in physical functioning during their in‐patient stay.[1, 2, 3, 4] Such impairments include difficulties performing basic activities of daily living, such as rising from a chair, toileting, or ambulating. This functional decline may result in increased length of stay (LOS), nursing home placement, and decreased mobility and participation in community activities even years after hospitalization.[1, 2, 3, 5, 6, 7] Ameliorating this hospital‐acquired functional impairment is important to improving patient outcomes and reducing healthcare utilization. Even the sickest hospitalized patients (eg, those in the intensive care unit [ICU]), can safely and feasibly benefit from early mobilization.[6, 8, 9, 10, 11] In the non‐ICU setting there is also evidence that patient mobilization reduces LOS and hospital costs, while improving patient satisfaction and physical and psychological outcomes.[12, 13, 14, 15, 16] These studies are, however, difficult to replicate as part of routine clinical care, because they often do not present the details of how early mobility was incorporated into daily practice, require additional hospital resources (eg, specially trained providers or additional staff), or are focused only on a select patient population.
The Johns Hopkins medical ICU started early rehabilitation quality‐improvement (QI) work in 2007, which has demonstrated ongoing reductions in LOS and been transformative in terms of helping to foster a culture of mobility at our institution. Previous research suggests that ICU‐based rehabilitation interventions are often not carried over to the ward setting, even in post‐ICU patients.[17] Moreover, trends for sicker patients being admitted in our general medicine units,[18] growing reports of patients spending most of their time in bed,[2, 19, 20] and healthcare policies emphasizing the importance of improving inpatient outcomes motivated the need for QI to improve patient mobility in this setting. Experience from the medical ICU‐based early rehabilitation program helped drive multidisciplinary collaboration of stakeholders to develop this nurse‐driven, mobility promotion QI project on 2 general medicine hospital units. The main goals of the project were to see whether a QI framework can be used in a general medicine setting to increase patient mobility and reduce LOS.[21, 22]
METHODS
Overview of Project
Mobility, for this project, was defined as a patient getting out of bed (eg, sitting out of bed, toileting at bedside commode or bathroom, standing, and ambulating). We aimed to increase patient mobility using preexisting unit staffing ratios of clinicians and support staff. This project was reported in accordance with the SQUIRE (Standards for QUality Improvement Reporting Excellence) guidelines and used a structured QI model that had been used to successfully promote early mobility in the intensive care unit.[21, 23, 24, 25] The planning phase of the QI project began in spring 2012, with initiation of the 12‐month project on March 1, 2013. During the 12‐month QI period, prospective collection of mobility status occurred for all patients, with no exclusions based on patient characteristics.
Setting
The QI project setting was 2, 24‐bed, general medicine units at the Johns Hopkins Hospital, a large academic medical center located in Baltimore, Maryland.
QI Process
The primary goals of the QI project were to mobilize patients 3 times daily, quantify and document the mobility of the patients, set daily goals to increase mobility (eg, move up 1 step on the scale today), and standardize the description of patient mobility across all hospital staff. We used a structured QI model that that has been used to implement an early mobility program in a medical ICU at our institution[21, 22, 24] (see Supporting Information, Appendix, in the online version of this article). At a programmatic level, we involved key stakeholders (nurses, physicians, rehabilitation therapists, administrators) in the QI project team, we identified local barriers to implementation through team meetings as well as a survey tool to identify perceived barriers,[26] and we developed a scale (the Johns Hopkins Highest Level of Mobility [JH‐HLM]) to document mobility. The JH‐HLM is an 8‐point ordinal scale that captures mobility milestones, where 1 = only lying, 2 = bed activities, 3 = sit at edge of bed, 4 = transfer to chair/commode, 5 = standing for 1 minute, 6 = walking 10+ steps, 7 = walking 25+ feet, and 8 = walking 250+ feet (see Supporting Information, Appendix and Supporting Figure 1, in the online version of this article for additional information on the JH‐HLM scale).
The 12‐month QI project was characterized by several phases and milestones and involved a number of intervention components. During the first 4 months (ramp‐up phase), nurses received education in the form of unit‐based presentations, hands‐on‐training, and online education modules. On a 5‐times weekly basis, nurses met with rehabilitation therapists for unit‐based huddles to discuss baseline patient mobility, current patient mobility levels, barriers to mobilizing patients, and daily goals to progress mobility. Mobility levels were included on daily nursing report sheets to facilitate communication with subsequent shifts. Discussion of JH‐HLM scores also occurred during daily unit‐based care‐coordination meetings of the nurses, physicians, and social‐workers to address barriers to mobilizing patients, such as optimizing pain control, facilitating discharge location planning, and expediting physician consultation with physical and occupational therapy for appropriate patients. Audit and feedback from huddles and care‐coordination rounds resulted in improved nurse attendance and engagement during these meetings. Nurses were expected to document patient mobility scores using the JH‐HLM 3 times daily in the patient medical record. On the fourth month, reports on JH‐HLM scores and documentation compliance were available to nurse managers, champions, and unit staff. Via twice‐monthly meetings with the units and quarterly meetings with hospital leadership and administration, problems arising during the QI intervention were evaluated and resolved on a timely basis. Seven months after project execution started, educational sessions were repeated to all staff, and feedback was provided based on the data collected, such as documentation compliance rates and patient mobility levels, and nurse champions presented the project during an American Nurses Credentialing Center magnet recognition program visit. Lastly, mobility scores and documentation compliance were continually assessed for 4 months after the project completion to determine sustainability of the intervention. Additional details of the QI project implementation are provided in the Supporting Information, Appendix, in the online version of this article.
Data Sources and Covariates for Project Evaluation
The Sunrise Clinical Manager system (Allscripts Healthcare Solutions Inc., Chicago, IL) was used to document and extract nursing‐documented JH‐HLM scores. The Johns Hopkins Hospital Datamart financial database, used for mandatory reporting to the State of Maryland, provided data on LOS, age, sex, race (white, black, other), payer (Medicare, Medicaid, other), primary admission diagnosis, and comorbidity index using Agency for Healthcare Research and Quality (AHRQ) methodology.[27] Expected LOS was calculated using the risk adjustment method developed by the University Health System Consortium (UHC).[28] This calculation uses a combination of the Diagnostic‐Related Group grouper and the Sachs Complication Profiler[29] in conjunction with data on specific patient characteristics (age, sex, urgency of admission, payer category) to construct risk‐adjustment regression models that assign expected values for LOS, and is not based on actual LOS.[28] The databases were linked at the patient level using the patient's medical record and unique admission record number.
Outcome Measures
Two functional outcome measures were based on daily JH‐HLM scores, which frequently occurred several times on each patient‐day: (1) the maximum daily JH‐HLM scores for each patient‐day during hospitalization, and (2) the intrapatient change in JH‐HLM scores between the maximum JH‐HLM score within 24 hours of hospital admission and 24 hours before discharge for all patients who were on the unit >48 hours. We also compared the mean LOS during the 12‐month QI project versus the 12‐months prior so we could more accurately address seasonal differences.[30, 31, 32, 33, 34, 35] Lastly, because the perception of increased falls was an important barrier to address in the QI process, we compared the rate of injurious falls between the QI period and 12‐months prior.
Statistical Analysis
To evaluate changes in the percent of ambulatory patients (JH‐HLM 6), we compared the initial 4 months of the QI project (ramp‐up phase) with the same 4‐month period occurring immediately after project completion (post‐QI phase) using generalized estimating equations to account for clustering at the patient‐level. This test was also used to evaluate changes in documentation compliance rates between the 2 phases, with compliance defined as at least 1 instance of JH‐HLM documentation per day, excluding the day of admission and discharge. To evaluate if improved JH‐HLM results were driven by improved documentation compliance rates over time, we performed a sensitivity analysis by imputing a JH‐HLM score of 6 (ambulate 10+ steps) for any missing daily maximum JH‐HLM scores.
To assess unadjusted changes in LOS during the 12‐month QI project versus the same period 1 year earlier, we compared mean and median LOS using a t test and Wilcoxon rank sum test, respectively. We used a multivariable linear regression model to estimate the change (expressed in days) in adjusted median LOS comparing the project months (March 2013March 2014) with 12 months prior (March 2012March 2013). The model adjusted for age, gender, race, payer, admission diagnostic category, UHC expected LOS, and AHRQ comorbidity index. We confirmed a lack of multicollinearity in the multivariable regression model using variance inflation factors. We evaluated residual versus predicted value plots and residual versus fitted value plots with a locally weighted scatterplot smoothing line to confirm model fit. P values are reported from the test of the null hypothesis that the change in adjusted median LOS is the same comparing the QI project months versus 12 months prior. Separate models estimated and tested the change in adjusted median LOS by tertiles of expected LOS (<4, 47, and >7 days). Lastly, we compared the rate of injurious falls (the number of injurious falls by total patient‐days) between the QI period and 12 months prior using an exact Poisson method.[36] Statistical significance was defined as a 2‐sided P < 0.05. Statistical analyses were conducted using R (version 3.1.0; The R Foundation for Statistical Computing, Vienna, Austria;
RESULTS
During the QI project period, 3352 patients were admitted to the 2 general medicine units. Twelve (0.4%) patients expired on the units, but their data were retained in the analysis. Mean (standard deviation [SD]) age of the patients was 54.4 (18.3) years, with 47% male, and 54% African American. A total of 1896 of 6654 (28%) patients on the QI units were 65 years old. Patient characteristics were similar during the QI period versus 12 months prior (Table 1).
| Characteristics | Comparison Period, March 2012March 2013, N = 3,302 | QI Period, March 2013March 2014, N = 3,352 |
|---|---|---|
| ||
| Age, y | 53.3 (17.8) | 54.4 (18.3) |
| Male | 1467 (44%) | 1569 (47%) |
| Race | ||
| African American | 1883 (57%) | 1809 (54%) |
| Caucasian | 1269 (38%) | 1348 (40%) |
| Other | 150 (5%) | 195 (6%) |
| Payer | ||
| Medicare | 1310 (40%) | 1470 (44%) |
| Medicaid | 1015 (31%) | 925 (28%) |
| Other | 977 (30%) | 957 (29%) |
| Admission diagnostic category | ||
| Infectious disease | 579 (18%) | 629 (19%) |
| Pulmonary | 519 (16%) | 559 (17%) |
| Gastrointestinal | 535 (16%) | 494 (15%) |
| Cardiovascular | 410 (12%) | 405 (12%) |
| Hematologic | 199 (6%) | 195 (6%) |
| Renal | 220 (7%) | 205 (6%) |
| Other | 840 (25%) | 865 (26%) |
| UHC expected length of stay, d | 5.5 (3.3) | 5.3 (3.2) |
| AHRQ comorbidity index | 3.3 (1.7) | 3.5 (1.8) |
During the 12‐month QI project, there were a total of 13,815 patient‐days of documented mobility data and the median (interquartile range [IQR]) number of days of documentation for each hospital admission was 3 (25) days. Compliance with daily documentation of JH‐HLM was 85.0% over the entire 12‐month QI project. Documentation compliance started at 83% during the ramp‐up phase and increased to 89% during the last 4 months of the project (late‐QI phase, P < 0.001).
Comparing the ramp‐up phase versus post‐QI phase, the percentage of patient‐days in which patients ambulated (JH‐HLM 6) increased from 43% to 70% (P < 0.001), and the percentage of patients who experienced an improvement in their mobility scores between admission and discharge increased from 32% to 45% (P < 0.001), as shown in Table 2. In the sensitivity analysis imputing missing daily JH‐HLM scores and comparing the ramp‐up versus post‐QI phases, the results were similar to the primary analysis; the percent of patient‐days where patients ambulated increased from 60% to 78% (P < 0.001), and the percent of patients who experienced an improvement in their mobility scores increased from 26% to 48% (P < 0.001).
| JH‐HLM Category | Ramp‐up Phase, March 1, 2013 June 30, 2013, n = 4,649 | Late‐QI Phase, November 1, 2013February 28, 2013, n = 4,515 | Post‐QI Phase, March 1, 2014 June 30, 2014, n = 4,298 |
|---|---|---|---|
| Change in Mobility (Admission Versus Discharge) | Ramp‐up Phase, March 1, 2013June 30, 2013, n = 968 | Late‐QI Phase, November 1, 2013February 28, 2013, n = 893 | Post‐QI Phase, March 1, 2014 June 30, 2014, n = 834 |
| |||
| Walk (JH‐HLM = 6, 7, or 8) | 1,994 (43) | 3,430 (76) | 2,986 (70) |
| Stand/chair (JH‐HLM = 4 or 5) | 1,772 (38) | 488 (10) | 511 (12) |
| Bed (JH‐HLM = 1, 2, or 3) | 883 (19) | 597 (13) | 801 (19) |
| Improved | 305 (32) | 392 (44) | 379 (45) |
| No change | 512 (53) | 428 (48) | 386 (46) |
| Declined | 151 (16) | 73 (8) | 69 (8) |
LOS during the 12‐month QI project versus the 12‐months immediately prior was shorter (Table 3), with an unadjusted median (IQR) LOS of 3 (26) versus 4 (27) days (P < 0.001) and an unadjusted mean (SD) LOS of 5.1 (5.6) versus 6.0 (7.6) (P < 0.001).
Adjusted Median LOS, d | Absolute Change in Adjusted Median LOS (95% CI), d | P Value | ||
|---|---|---|---|---|
| 12 Months Prior | QI Project Months | |||
| ||||
| All patients | 6.01 | 5.61 | 0.40 (0.57 to 0.21), N = 4,411 | <0.001 |
| Subgroups by ELOS | ||||
| ELOS <4 days | 4.68 | 4.77 | 0.09 (0.13 to 0.32), N = 1,357 | 0.42 |
| ELOS 47 days | 5.68 | 5.38 | 0.30 (0.57 to 0.01), N = 1,509 | 0.04 |
| ELOS >7 days | 8.07 | 6.96 | 1.11 (1.53 to 0.65), N = 1,545 | <0.001 |
Table 3 displays the change in adjusted median LOS for the project months versus the 12 months prior among the QI units. We found that for all patients, there was an overall reduction in adjusted median LOS of 0.40 (95% confidence interval [CI]: 0.57 to 0.21, P<0.001) days. When we divided patients into tertiles based on their UHC expected LOS (ELOS), we observed that patients with longer ELOS had greater reductions in adjusted median LOS. Patients on the QI units with ELOS <4 days (lowest tertile) did not show a significant reduction in adjusted median LOS (0.09 days, 95% CI: 0.13 to 0.32, P = 0.42); however, patients with UHC ELOS 4 to 7 days (middle tertile) and ELOS >7 days (highest tertile) had a significant reduction in adjusted median LOS by 0.30 (95% CI: 0.57 to 0.01, P = 0.04) and 1.11 (95% CI: 1.53 to 0.65, P < 0.001) days during the QI project versus 12 months prior, respectively.
Lastly, we found that there was no difference in the rate of injurious falls on the QI units during QI period compared to 12 months prior (QI: 0.34 per 1000 patient‐days versus 12 months prior: 0.48 per 1000 patient‐days, P = 0.73).
DISCUSSION
We conducted a nurse‐driven, multidisciplinary mobility promotion QI project on 2 general medicine units at a large teaching hospital. The 12‐month QI project, conducted between March 1, 2013 and February 28, 2014, was associated with patients ambulating more frequently, with improved mobility status between hospital admission and discharge. These improvements in mobility were not associated with increased rates of injurious falls, and were sustained for at least 4 months after project completion. The QI project was associated with overall significant reduction in LOS for more complex patients with longer expected LOS (4 days or longer). Hence, such QI efforts may be important for maintaining or improving patients' functional status during hospitalization in a safe and cost‐effective manner.
Our findings are consistent with previous studies showing that mobility promotion in the acute hospital setting is feasible, can reduce length of stay, and can be applied to a diverse population including vulnerable medical patients with multiple comorbidities and the elderly.[12, 16, 37, 38, 39, 40, 41, 42] These studies provide valuable evidence of the benefits of mobility promotion; however, it is difficult to translate these prior results into routine clinical practice because they used specially trained staff to mobilize patients, focused on a select patient population, or did not specify how the mobility intervention was delivered within daily clinical workflows. Research in the medical ICU at our institution has previously described the use of a structured QI model to successfully implement an early rehabilitation program.[22, 24] Here, we successfully adapted the same QI framework to a general medicine setting. Hence, our study contributes to the literature with respect to (1) use of a structured QI framework to develop a successful patient mobility program in a general medicine patient population, and (2) sharing best practices from 1 clinical setting, such as the ICU, as a source of learning and knowledge translation for other care settings, with the addition of novel tools, such as the JH‐HLM scale.
There may have been several factors that contributed to shorter stays in the hospital we observed during the QI project. First, we increased the number of ambulatory patient‐days, which may have helped prevent physiological complications of bed rest, such as muscle weakness, atelectasis, insulin resistance, vascular dysfunction, contractures, and pressure ulcers.[43] As such, mobility promotion has been associated with reduced rates of other hospital‐acquired complications, such as deep venous thrombosis, pneumonia, and delirium.[44, 45, 46] In our study, we saw the greatest LOS reduction in more complex patients who were expected to spend a longer time in the hospital and are at greater risk of developing complications from bed rest. Second, our early mobility project may have had a direct impact on care‐coordination processes as reported in prior studies.[47, 48, 49] An important component of our intervention was incorporating functional status into multidisciplinary discussions, either through nurse‐to‐therapist huddles or care‐coordination rounds between nurses, therapists, physicians, social workers, and case managers. During care‐coordination rounds, JH‐HLM scores were reported to expedite appropriate physical and occupational therapy consultations and assist in determining appropriate discharge location. During the QI project, we transitioned from a unit‐based daily huddle between nursing and rehabilitation therapists to a system where mobility status was discussed primarily during care coordination rounds 5 times per week. We saw that mobility scores were maintained after QI project completion, suggesting that reporting on patient function in a multidisciplinary setting is a potentially sustainable mechanism to improve care‐coordination processes that are affected by functional status.
Our study has several potential limitations. First, this is a single‐site study in 2 general medicine units of a large academic hospital. Further research is needed to determine if this structured QI intervention and its benefits can be generalized to different settings and different patient populations. Second, because the documentation was initially an optional element in the electronic medical record system, we observed higher rates of missing documentation during the first 4 months of the project versus the comparison period at 4 months after project completion. However, a sensitivity analysis conducted of these missing data demonstrated similar results to our primary analysis. Third, our nonrandomized pre‐post study design does not allow us to conclude a direct cause‐and‐effect relationship between our intervention and increased mobility and reduced LOS. Although patient characteristics were similar between the 2 periods and adjusted for in our multivariable regression analysis, we cannot rule out the possibility of secular trends in LOS on the project units and that broader QI efforts at our institution also contributed to reduction in LOS. Fourth, we do not have data on 30‐day readmissions and discharge location. Future studies should explore the impact of hospital‐based mobility interventions on these outcomes.[50] Fifth, although nurses consistently documented the highest level of mobility on a daily basis, these data did not capture other potentially important information about patient mobility such as the daily frequency that patients were mobilized, the length of time a patient was engaged in a mobility event (ie, number of hours sitting in a chair), or the mobility that occurred during physical therapy or occupational therapy sessions. Hence, although we used JH‐HLM as a marker of improved mobility during our QI project it is likely that our data cannot fully describe the total mobility and activity that patients experienced during hospitalization. Lastly, although the front‐line staff and QI team found the JH‐HLM scale to be a useful tool to measure and advance patient mobility, further studies are needed to evaluate the reliability and validity of this scale.
CONCLUSION
A structured QI process can improve patient mobility and may contribute to reduction in LOS, particularly for more complex patients in this setting. Active prevention of decline in physical function that commonly occurs during hospitalization may prove valuable for improving patient outcomes and reducing healthcare resource utilization.
Disclosures
The authors certify that no party having a direct interest in the results of the research supporting this article has or will confer a benefit on us or on any organization with which we are associated. The authors report no conflicts of interest.
- , , , et al. Loss of independence in activities of daily living in older adults hospitalized with medical illnesses: increased vulnerability with age. J Am Geriatr Soc. 2003;51(4):451–458.
- , , . Prevalence and outcomes of low mobility in hospitalized older patients. J Am Geriatr Soc. 2004;52(8):1263–1270.
- , , , , , . Trajectories of life‐space mobility after hospitalization. Ann Intern Med. 2009;150(6):372–378.
- , , . Hospitalization‐associated disability: “She was probably able to ambulate, but I'm not sure”. JAMA. 2011;306(16):1782–1793.
- , , , , , . Physical fitness and all‐cause mortality. A prospective study of healthy men and women. JAMA. 1989;262(17):2395–2401.
- . Mobilizing patients in the intensive care unit: improving neuromuscular weakness and physical function. JAMA. 2008;300(14):1685–1690.
- , . Mobility limitation in the older patient: a clinical review. JAMA. 2013;310(11):1168–1177.
- , , , et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373(9678):1874–1882.
- , , . Technology to enhance physical rehabilitation of critically ill patients. Crit Care Med. 2009;37(10 suppl):S436–S441.
- , , , et al. Receiving early mobility during an intensive care unit admission is a predictor of improved outcomes in acute respiratory failure. Am J Med Sci. 2011;341(5):373–377.
- . Physiotherapy in intensive care: an updated systematic review. Chest. 2013;144(3):825–847.
- , , . Exercise for acutely hospitalised older medical patients. Cochrane Database Syst Rev. 2007;(1):CD005955.
- , , . Extra physical therapy reduces patient length of stay and improves functional outcomes and quality of life in people with acute or subacute conditions: a systematic review. Arch Phys Med Rehabil. 2011;92(9):1490–1500.
- , . Impact of early mobilization protocol on the medical‐surgical inpatient population: an integrated review of literature. Clin Nurse Spec. 2012;26(2):87–94.
- , , . Outcomes of inpatient mobilization: a literature review. J Clin Nurs. 2014;23(11–12):1486–1501.
- , , , et al. The early mobility bundle: a simple enhancement of therapy which may reduce incidence of hospital‐acquired pneumonia and length of hospital stay. J Hosp Infect. 2014;88(1):34–39.
- , , , , . Physical therapy on the wards after early physical activity and mobility in the intensive care unit. Phys Ther. 2012;92(12):1518–1523.
- , , , . Impact of hospital variables on case mix index as a marker of disease severity. Popul Health Manag. 2014;17(1):28–34.
- , , , , . Frequency of hallway ambulation by hospitalized older adults on medical units of an academic hospital. Geriatr Nurs. 2004;25(4):212–217.
- , , . Activity level of hospital medical inpatients: an observational study. Arch Gerontol Geriatr. 2012;55(2):417–421.
- , , . Translating evidence into practice: a model for large scale knowledge translation. BMJ. 2008;337:a1714.
- , , , et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil. 2010;91(4):536–542.
- , , , , ; SQUIRE development group. Publication guidelines for quality improvement studies in health care: evolution of the SQUIRE project. BMJ. 2009;338:a3152.
- , . Rehabilitation quality improvement in an intensive care unit setting: implementation of a quality improvement model. Top Stroke Rehabil. 2010;17(4):271–281.
- , , , . ICU early mobilization: from recommendation to implementation at three medical centers. Crit Care Med. 2013;41(9 suppl 1):S69–S80.
- , , , . Barriers to early mobility of hospitalized general medicine patients: survey development and results. Am J Phys Med Rehabil. 2015;94(4):304–312.
- , , , . Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27.
- UHC Clinical Information Management Risk Adjustment of the UHC Clinical Data Base. Chicago, IL: University HealthSystem Consortium; 1998.
- Sachs Complications Profiler, Version 1.0, User's Guide. Evanston, IL: Sachs Group; 1995.
- , , . Assessing hospital‐associated deaths from discharge data. The role of length of stay and comorbidities. JAMA. 1988;260(15):2240–2246.
- , , , et al. Annual rates of admission and seasonal variations in hospitalizations for heart failure. Eur J Heart Fail. 2002;4(6):779–786.
- , , , . Regional and seasonal variation in the length of hospital stay for chronic obstructive pulmonary disease in Finland. Int J Circumpolar Health. 2002;61(2):131–135.
- , , , , , ; Canadian CABG Surgery Quality Indicator Consensus Panel. The identification and development of canadian coronary artery bypass graft surgery quality indicators. J Thorac Cardiovasc Surg. 2005;130(5):1257.
- , , , . Quality of care and length of hospital stay among patients with stroke. Med Care. 2009;47(5):575–582.
- . A systematic review of outcomes and quality measures in adult patients cared for by hospitalists vs nonhospitalists. Mayo Clin Proc. 2009;84(3):248–254.
- . Confidence intervals that match Fisher's exact or Blaker's exact tests. Biostatistics. 2010;11(2):373–374.
- , , , et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med. 1999;340(9):669–676.
- , , , , . Early mobilization of patients hospitalized with community‐acquired pneumonia. Chest. 2003;124(3):883–889.
- , , , et al. Effects of a walking intervention on fatigue‐related experiences of hospitalized acute myelogenous leukemia patients undergoing chemotherapy: a randomized controlled trial. J Pain Symptom Manage. 2008;35(5):524–534.
- , , , , . Early ambulation and length of stay in older adults hospitalized for acute illness. Arch Intern Med. 2010;170(21):1942–1943.
- , , . Impact of a nurse‐driven mobility protocol on functional decline in hospitalized older adults. J Nurs Care Qual. 2009;24(4):325–331.
- , , . Exercising body and mind: an integrated approach to functional independence in hospitalized older people. J Am Geriatr Soc. 2008;56(4):630–635.
- . Consequences of bed rest. Crit Care Med. 2009;37(10 suppl):S422–S428.
- , , , , . Time to ambulation after hip fracture surgery: relation to hospitalization outcomes. J Gerontol A Biol Sci Med Sci. 2003;58(11):M1042–M1045.
- , , , . Early mobilization after total knee replacement reduces the incidence of deep venous thrombosis. ANZ J Surg. 2009;79(7–8):526–529.
- , , , . Efficacy and safety of postoperative early mobilization for chronic subdural hematoma in elderly patients. Acta Neurochir (Wien). 2010;152(7):1171–1174.
- , , , et al. Impact of relational coordination on quality of care, postoperative pain and functioning, and length of stay: a nine‐hospital study of surgical patients. Med Care. 2000;38(8):807–819.
- Care coordination cuts admissions, ED visits, LOS. Hosp Case Manag. 2013;21(5):67–68.
- , . A heart failure initiative to reduce the length of stay and readmission rates. Prof Case Manag. 2014;19(6):276–284.
- , , , , , . Association of impaired functional status at hospital discharge and subsequent rehospitalization. J Hosp Med. 2014;9(5):277–282.
- , , , et al. Loss of independence in activities of daily living in older adults hospitalized with medical illnesses: increased vulnerability with age. J Am Geriatr Soc. 2003;51(4):451–458.
- , , . Prevalence and outcomes of low mobility in hospitalized older patients. J Am Geriatr Soc. 2004;52(8):1263–1270.
- , , , , , . Trajectories of life‐space mobility after hospitalization. Ann Intern Med. 2009;150(6):372–378.
- , , . Hospitalization‐associated disability: “She was probably able to ambulate, but I'm not sure”. JAMA. 2011;306(16):1782–1793.
- , , , , , . Physical fitness and all‐cause mortality. A prospective study of healthy men and women. JAMA. 1989;262(17):2395–2401.
- . Mobilizing patients in the intensive care unit: improving neuromuscular weakness and physical function. JAMA. 2008;300(14):1685–1690.
- , . Mobility limitation in the older patient: a clinical review. JAMA. 2013;310(11):1168–1177.
- , , , et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373(9678):1874–1882.
- , , . Technology to enhance physical rehabilitation of critically ill patients. Crit Care Med. 2009;37(10 suppl):S436–S441.
- , , , et al. Receiving early mobility during an intensive care unit admission is a predictor of improved outcomes in acute respiratory failure. Am J Med Sci. 2011;341(5):373–377.
- . Physiotherapy in intensive care: an updated systematic review. Chest. 2013;144(3):825–847.
- , , . Exercise for acutely hospitalised older medical patients. Cochrane Database Syst Rev. 2007;(1):CD005955.
- , , . Extra physical therapy reduces patient length of stay and improves functional outcomes and quality of life in people with acute or subacute conditions: a systematic review. Arch Phys Med Rehabil. 2011;92(9):1490–1500.
- , . Impact of early mobilization protocol on the medical‐surgical inpatient population: an integrated review of literature. Clin Nurse Spec. 2012;26(2):87–94.
- , , . Outcomes of inpatient mobilization: a literature review. J Clin Nurs. 2014;23(11–12):1486–1501.
- , , , et al. The early mobility bundle: a simple enhancement of therapy which may reduce incidence of hospital‐acquired pneumonia and length of hospital stay. J Hosp Infect. 2014;88(1):34–39.
- , , , , . Physical therapy on the wards after early physical activity and mobility in the intensive care unit. Phys Ther. 2012;92(12):1518–1523.
- , , , . Impact of hospital variables on case mix index as a marker of disease severity. Popul Health Manag. 2014;17(1):28–34.
- , , , , . Frequency of hallway ambulation by hospitalized older adults on medical units of an academic hospital. Geriatr Nurs. 2004;25(4):212–217.
- , , . Activity level of hospital medical inpatients: an observational study. Arch Gerontol Geriatr. 2012;55(2):417–421.
- , , . Translating evidence into practice: a model for large scale knowledge translation. BMJ. 2008;337:a1714.
- , , , et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil. 2010;91(4):536–542.
- , , , , ; SQUIRE development group. Publication guidelines for quality improvement studies in health care: evolution of the SQUIRE project. BMJ. 2009;338:a3152.
- , . Rehabilitation quality improvement in an intensive care unit setting: implementation of a quality improvement model. Top Stroke Rehabil. 2010;17(4):271–281.
- , , , . ICU early mobilization: from recommendation to implementation at three medical centers. Crit Care Med. 2013;41(9 suppl 1):S69–S80.
- , , , . Barriers to early mobility of hospitalized general medicine patients: survey development and results. Am J Phys Med Rehabil. 2015;94(4):304–312.
- , , , . Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27.
- UHC Clinical Information Management Risk Adjustment of the UHC Clinical Data Base. Chicago, IL: University HealthSystem Consortium; 1998.
- Sachs Complications Profiler, Version 1.0, User's Guide. Evanston, IL: Sachs Group; 1995.
- , , . Assessing hospital‐associated deaths from discharge data. The role of length of stay and comorbidities. JAMA. 1988;260(15):2240–2246.
- , , , et al. Annual rates of admission and seasonal variations in hospitalizations for heart failure. Eur J Heart Fail. 2002;4(6):779–786.
- , , , . Regional and seasonal variation in the length of hospital stay for chronic obstructive pulmonary disease in Finland. Int J Circumpolar Health. 2002;61(2):131–135.
- , , , , , ; Canadian CABG Surgery Quality Indicator Consensus Panel. The identification and development of canadian coronary artery bypass graft surgery quality indicators. J Thorac Cardiovasc Surg. 2005;130(5):1257.
- , , , . Quality of care and length of hospital stay among patients with stroke. Med Care. 2009;47(5):575–582.
- . A systematic review of outcomes and quality measures in adult patients cared for by hospitalists vs nonhospitalists. Mayo Clin Proc. 2009;84(3):248–254.
- . Confidence intervals that match Fisher's exact or Blaker's exact tests. Biostatistics. 2010;11(2):373–374.
- , , , et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med. 1999;340(9):669–676.
- , , , , . Early mobilization of patients hospitalized with community‐acquired pneumonia. Chest. 2003;124(3):883–889.
- , , , et al. Effects of a walking intervention on fatigue‐related experiences of hospitalized acute myelogenous leukemia patients undergoing chemotherapy: a randomized controlled trial. J Pain Symptom Manage. 2008;35(5):524–534.
- , , , , . Early ambulation and length of stay in older adults hospitalized for acute illness. Arch Intern Med. 2010;170(21):1942–1943.
- , , . Impact of a nurse‐driven mobility protocol on functional decline in hospitalized older adults. J Nurs Care Qual. 2009;24(4):325–331.
- , , . Exercising body and mind: an integrated approach to functional independence in hospitalized older people. J Am Geriatr Soc. 2008;56(4):630–635.
- . Consequences of bed rest. Crit Care Med. 2009;37(10 suppl):S422–S428.
- , , , , . Time to ambulation after hip fracture surgery: relation to hospitalization outcomes. J Gerontol A Biol Sci Med Sci. 2003;58(11):M1042–M1045.
- , , , . Early mobilization after total knee replacement reduces the incidence of deep venous thrombosis. ANZ J Surg. 2009;79(7–8):526–529.
- , , , . Efficacy and safety of postoperative early mobilization for chronic subdural hematoma in elderly patients. Acta Neurochir (Wien). 2010;152(7):1171–1174.
- , , , et al. Impact of relational coordination on quality of care, postoperative pain and functioning, and length of stay: a nine‐hospital study of surgical patients. Med Care. 2000;38(8):807–819.
- Care coordination cuts admissions, ED visits, LOS. Hosp Case Manag. 2013;21(5):67–68.
- , . A heart failure initiative to reduce the length of stay and readmission rates. Prof Case Manag. 2014;19(6):276–284.
- , , , , , . Association of impaired functional status at hospital discharge and subsequent rehospitalization. J Hosp Med. 2014;9(5):277–282.
© 2016 Society of Hospital Medicine
FIM at Discharge and Rehospitalization
Federally mandated pay‐for‐performance initiatives promote minimizing 30‐day hospital readmissions to improve healthcare quality and reduce costs. Although the reasons for readmissions are multifactorial, many patients are readmitted for a condition other than their initial hospital admitting diagnosis.[1] Impairments in functional status experienced during acute care hospitalization contribute to patients being discharged in a debilitated state and being vulnerable to postdischarge complications and potentially hospital readmission.[2] As such, decreased functional status may be an important and potentially modifiable risk factor for acute care hospital readmission.[3]
Previous studies have suggested that impaired functional status may be an important predictor of rehospitalization.[4, 5, 6, 7] However, inferences from existing studies are limited because they did not consider functional status as their primary focus, they only considered specific patient populations (eg, stroke) or readmissions occurring well beyond the 30‐day period defined by federal pay‐for‐performance standards.[4, 5, 6, 8, 9, 10] Our objective was to evaluate the association between functional status near the time of discharge from acute care hospital and 30‐day readmission for patients admitted to an acute inpatient rehabilitation facility. As a secondary objective, we sought to investigate the relationship between functional status and readmission by diagnostic category (medical, neurologic, or orthopedic).
METHODS
Study Population and Setting
We conducted a single‐center, retrospective study of patients admitted to an inpatient rehabilitation facility at a community hospital between July 1, 2006 and December 31, 2012. This facility provides intensive rehabilitation consisting of 3 hours of therapy per day, skilled nursing care on a 24‐hour basis, and medical care by a physiatrist. We excluded patients who died during inpatient rehabilitation (n=15, 0.2%) and patients not admitted directly from an acute care setting (n=178, 2.0%).
Data Source and Covariates
Data were derived from the Uniform Data System for Medical Rehabilitation (UDSMR), which is an administrative database providing the following data upon admission to an inpatient rehabilitation facility[11, 12, 13]: age, gender, race/ethnicity, marital status, the discharge setting, the admission Functional Independence Measure (FIM) score (details further below), and admission diagnostic category as defined by the primary discharge diagnosis from the acute care hospital and grouped by functional related groups (a case‐mix system for medical rehabilitation).[12, 14] The 3M ClinTrac management software (3M, St. Paul, MN), used for mandatory reporting to the State of Maryland, provided all‐payerrefined diagnosis related group (APRDRG) and severity of illness (SOI) combinations (a tool to group patients into clinically comparable disease and severity‐of‐illness categories expected to use similar resources and experience similar outcomes). The University HealthSystem Consortium (UHC) database provided national readmission rates for all APRDRG‐SOI combinations using a methodology that has been previously described.[15, 16] Expected readmission rates for APRDRG‐SOI combinations served as a patient risk stratification tool based on clinical logic that evaluates age, comorbidities, principal diagnosis during hospitalization, and procedures conducted during hospitalization.[17]
Primary Outcome: Acute Care Readmission
The primary outcome was all‐cause acute care readmission, defined as patient transfer to an acute care hospital during inpatient rehabilitation within 30 days from admission to inpatient rehabilitation. The care model for our inpatient rehabilitation unit is such that when patients become sick or develop a complication, they are admitted directly to a clinical unit (eg, intensive care unit) at the community hospital through a rapid‐response intervention, or the physiatrist arranges with an admitting inpatient attending to accept the patient directly to his or her service.
Primary Exposure: Functional Independence Measure
Functional status was measured using the FIM score.[18] The FIM score is an 18‐item measure of functional status, with each item scored on a scale from 1 to 7 (dependent to independent). Various aspects of motor function and cognitive function are assessed. The FIM has been validated and shown to be reliable and reproducible.[13, 19, 20] By definition for the FIM instrument, admission FIM scores are assessed by trained multidisciplinary personnel first over the 72 hours of the rehabilitation stay, and for this study served as a proxy for patient functional status upon discharge from the acute care setting in our analysis. This 72‐hour time window allows for full assessment by therapists and nurses; however, in clinical practice at the inpatient rehabilitation unit involved in this study, much of the FIM assessment occurs within the first 24 hours of the rehabilitation stay. For our analysis, we divided FIM scores into low, medium, and high functional groups. The thresholds for these groups were based on total FIM score tertiles from a prior study<60, 60 to 76, and >76.[16] As a secondary analysis we created 6 subscales of the overall FIM score based on previous research. These subscales included: transfers (transfer to chair/wheelchair, toilet, and tub/shower), locomotion (walking and stairs), self‐care (eating, grooming, bathing, dressing, and toileting), sphincter control (bladder and bowel management), communication (comprehension and expression), and social cognition (social interaction, problem solving, and memory).[21]
Statistical Analysis
To evaluate differences in patient characteristics by diagnostic category, analysis of variance and 2 tests were used for continuous and dichotomous variables, respectively. Logistic regression was used to evaluate the association between FIM score category and readmission status, adjusting for potentially confounding variables available from the UDSMR and UHC databases. We used interaction terms to test whether the association between the FIM score and readmissions varied significantly across diagnostic categories and by age. As a secondary analysis, we modeled FIM score as a continuous variable. We expressed the odds ratio in this analysis per 10‐point change in FIM, because this represents a clinically relevant change in function.[22] Logistic regression was also used to evaluate the association between FIM subscale scores (transfers, locomotion, self‐care, sphincter control, communication, and social cognition) and readmission status. Statistical significance was defined as a 2‐sided P<0.05. Data were analyzed with R (version 2.15.0;
RESULTS
Readmitted Patients and Diagnostic Categories
A total of 9405 consecutive eligible patients were admitted to the acute inpatient rehabilitation facility between July 1, 2006 and December 31, 2012. A total of 1182 (13%) patients were readmitted back to an acute care hospital from inpatient rehabilitation. Median (interquartile range) time to readmission from acute care hospital discharge was 6 days (310 days), and median length of stay for patients who were discharged to the community from inpatient rehabilitation was 8 days (612 days).
Table 1 shows characteristics of all inpatient rehabilitation patients by diagnostic category. For the neurologic category, the most common primary diagnoses were stroke and spinal cord injury; for the medical category, infection, renal failure, congestive heart failure, and chronic obstructive pulmonary disease; and for the orthopedic category, spinal arthrodesis, knee and hip replacements. Mean FIM scores were lowest and highest for patients admitted with a primarily neurologic and orthopedic diagnosis, respectively.
| Characteristic | All Patients, N=9405 | Diagnostic Category | |||
|---|---|---|---|---|---|
| Neurologic, n=3706 | Medical, n=2135 | Orthopedic, n=3564 | P Valueb | ||
| |||||
| Age, y | 67.8 (14.2) | 66.7 (15.3) | 67.0 (14.9) | 69.3 (12.4) | <0.001 |
| Male | 4,068 (43%) | 1,816 (49%) | 1,119 (52%) | 1,133 (32%) | <0.001 |
| Race | <0.001 | ||||
| Caucasian | 6,106 (65%) | 2344 (63%) | 1,320 (62%) | 2,442 (69%) | |
| African American | 2,501 (27%) | 984 (27%) | 658 (31%) | 859 (24%) | |
| Other | 798 (8%) | 378 (10%) | 157 (7%) | 263 (7%) | |
| Married | 4,330 (46%) | 1,683 (45%) | 931 (44%) | 1,716 (48%) | 0.002 |
| APRDRG‐SOI expected readmission rate | 18.0 (7.4) | 20.5 (6.8) | 21.3 (7.5) | 13.5 (5.6) | <0.001 |
| Total admission FIM score | 68.7 (17.2) | 60.4 (18.6) | 69.1 (15.5) | 77.2 (11.7) | <0.001 |
FIM Score Category and Risk of Readmission
Figure 1 shows that patients in the low admission FIM score category had the highest unadjusted rate of readmission for each diagnostic category. In unadjusted analysis, Table 2 shows that younger age, male sex, APDRG‐SOI expected readmission rate, and orthopedic and medical diagnostic categories were associated with readmission. As a continuous variable, FIM scores were linearly associated with readmission (Figure 2), with an unadjusted odds ratio (OR) and 95% confidence interval (CI) of 1.4 (1.4‐1.4, P<0.001) for a 10‐point decrease in FIM. Compared to patients with high admission FIM scores, patients with low and middle FIM scores had higher unadjusted odds of readmission (OR: 4.0; 95% CI: 3.4‐4.7; P<0.001 and OR: 1.8; 95% CI: 1.5‐2.1; P<0.001, respectively). Mean FIM subscale scores for patients readmitted versus not readmitted were transfers (5.3 vs 7.0, P<0.001), locomotion (1.6 vs 2.3, P<0.001), self‐care (17.0 vs 20.8, P<0.001), communication (10.6 vs 11.5, P<0.001), and social cognition (15.1 vs 16.6, P<0.001).
| Bivariable Analysisb | Multivariable Analysisb | |||||
|---|---|---|---|---|---|---|
| Characteristic | All Patients, N=9405 | Readmitted, n=1,182 | OR (95% CI) | P Value | OR (95% CI) | P Value |
| ||||||
| Age, y | 68.0 (14.2) | 66.4 (14.5) | 0.9 (0.91.0) | <0.001 | 0.9 (0.91.0) | <0.001 |
| Male | 3,431 (42%) | 637 (54%) | 1.6 (1.41.8) | <0.001 | 1.3 (1.11.5) | < 0.001 |
| Race | ||||||
| Caucasian | 5,340 (65%) | 766 (65%) | 1.0 | 1.0 | ||
| African American | 2,177 (26%) | 324 (27%) | 1.0 (0.91.2) | 0.60 | 1.0 (0.81.1) | 0.75 |
| Other | 706 (9%) | 92 (8%) | 0.9 (0.71.1) | 0.41 | 0.8 (0.61.0) | 0.12 |
| Married | 3,775 (46%) | 555 (47%) | 1.0 (0.91.2) | 0.50 | 1.0 (0.91.2) | 0.67 |
| Admission diagnosis category | ||||||
| Neurologic | 3,205 (39%) | 501 (42%) | 1.0 | 1.0 | ||
| Medical | 1,726 (21%) | 409 (35%) | 1.5 (1.31.7) | <0.001 | 1.8 (1.62.1) | < 0.001 |
| Orthopedic | 3,292 (40%) | 272 (23%) | 0.5 (0.50.6) | <0.001 | 1.3 (1.11.6) | 0.005 |
| APDRG‐SOI expected readmission rate | 17.4 (7.1%) | 22.2 (8.0%) | 1.1 (1.11.1) | <0.001 | 1.1 (1.01.1) | < 0.001 |
| Total FIM score category | ||||||
| High FIM, >76 points | 3,517 (43%) | 257 (22%) | 1.0 | 1.0 | ||
| Middle FIM, 60points | 2,742 (33%) | 353 (30%) | 1.8 (1.52.1) | <0.001 | 1.5 (1.31.8) | < 0.001 |
| Low FIM, <60 points | 1,964 (24%) | 572 (48%) | 4.0 (3.44.7) | <0.001 | 3.0 (2.53.6) | < 0.001 |
Multivariable and Subset Analyses
Patients with a primary medical diagnosis had higher odds of readmission to the hospital, (OR: 1.8; 95% CI: 1.6‐2.1, P<0.001), relative to patients with a neurologic or orthopedic diagnosis (Table 2). Across all diagnoses, the adjusted odds ratios (95% CIs) for the low and middle versus high FIM score category were 3.0 (2.5‐3.6; P<0.001) and 1.5 (1.3‐1.8; P<0.001) respectively (Table 2). When modeled as a continuous variable, a 10‐point decrease in FIM score was associated with a significantly increased adjusted readmission rate (OR: 1.4; 95% CI: 1.3‐1.4; P<0.001). In adjusted analysis including all subscales of the FIM, only the physical subscales, transfers (P<0.001), locomotion (P=0.002), and self‐care (P<0.001), were significantly associated with readmission. For each diagnostic category, there were similar significant associations between admission FIM score group and readmission status (Table 3). The odds of readmission by FIM score did not differ significantly across the 3 major diagnostic categories (P=0.20 for interaction term), suggesting that the effect of functional status was similar across various types of patients. We also did not observe a statistical interaction between age and FIM score group in predicting readmission (P=0.58). Patients in the lowest FIM group with a medical diagnosis had the highest adjusted readmission rate of 28.7% (Table 3).
| Multivariable Analysisa | Adjusted Readmission Ratesb | |||
|---|---|---|---|---|
| No. | OR (95% CI) | P Value | % (95% CI) | |
| ||||
| Neurologic | ||||
| High FIM (>76 points) | 755 | 1.0 | 7.3 (4.710.0) | |
| Middle FIM (6076 points) | 1,283 | 1.4 (1.02.1) | 0.06 | 9.1 (7.011.1) |
| Low FIM (<60 points) | 1,668 | 3.3 (2.34.7) | <0.001 | 18.7 (16.820.6) |
| Medical | ||||
| High FIM (>76 points) | 807 | 1.0 | 11.2 (8.114.3) | |
| Middle FIM (6076 points) | 766 | 1.8 (1.32.4) | <0.001 | 17.7 (14.520.9) |
| Low FIM (<60 points) | 562 | 3.2 (2.44.3) | <0.001 | 28.7 (25.132.4) |
| Orthopedic | ||||
| High FIM (>76 points) | 2,212 | 1.0 | 6.1 (4.77.6) | |
| Middle FIM (6076 points) | 1,046 | 1.4 (1.11.9) | 0.02 | 8.3 (6.410.1) |
| Low FIM (<60 points) | 306 | 2.2 (1.53.3) | <0.001 | 13.5 (10.416.7) |
DISCUSSION
In this study of 9405 consecutive patients admitted from acute care hospitals to a single inpatient rehabilitation facility, we investigated the association between functional status and readmission to an acute care hospital. We found that low functional status near the time of acute care hospital discharge was strongly associated with higher readmission rates. This relationship was consistently observed across major patient diagnostic categories, with low functioning medical patients having the highest rate of readmission (28.7%). Efforts to maintain or improve functional status during acute care hospitalization may be an important modifiable risk factor for acute care hospital readmission.
Previous studies have suggested that functional status may serve as an indicator of physiological reserve, and therefore vulnerability to medical complications and readmission.[6, 16, 23, 24, 25] Physiologic reserve refers to a person's ability to endure acute illness and is influenced by a number of factors, such as the adequacy of oxygen delivery to tissues, cardiovascular health, immune state, and nutritional status.[26] We found that motor subscales of the FIM score (transfers, locomotion, and self‐care), but not the other subscales, were independently associated with readmissions, which may suggest that lower motor scores are a stronger marker of physiologic reserve.[10, 16, 27] Although not our primary focus, we did note in our multivariable models that after adjusting for functional status, patients in a medical diagnostic category had higher readmission rates compared to patients with a primary neurologic or orthopedic diagnosis, but the impact of FIM score was consistent across all these diagnostic categories. We speculate that medical conditions that result in hospitalization, such as sepsis or acute kidney failure, may be more likely to result in multiorgan dysfunction that may impair physiological reserve and increase susceptibility to medical complications.[28, 29, 30, 31] In comparison, acute neurologic and orthopedic diagnoses, such as stroke or hip arthroplasty, directly impair gross motor function,[32, 33, 34, 35] with relative sparing of overall physiologic reserve.
The association between low functional status and readmissions is supported by previous studies across multiple hospital settings.[4, 5, 7, 8, 9, 27, 36] Despite this finding, routine inpatient medical practice may not fully address functional impairments. For instance, systematic measurement and documentation of functional status on admission and during hospitalization are not routine and may be a barrier to identifying medical patients at high risk for readmission.[37, 38, 39] Moreover, without recognition of functional impairment and its implications, current clinical practice may suboptimally prevent and treat physical impairments during inpatient care. However, such barriers can be surmounted. For example, in the medical intensive care unit setting, there is growing recognition that proactive and aggressive management of hospital‐acquired functional impairments through early rehabilitation is safe and feasible, improving patient outcomes while reducing hospital costs and readmissions.[3, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51] Moreover, 2 recent meta‐analyses have shown that physical therapy hospital‐based exercise programs can improve length of stay, overall hospital costs, and rates of discharge to home.[52, 53] Finally, a randomized trial has demonstrated that an individualized exercise regimen started in the acute hospital setting with long‐term telephone follow‐up can significantly reduce emergency hospital readmissions and improve quality of life in older adults.[54] Therefore, decreased functional status likely represents a modifiable risk factor for hospital readmission, and further research is necessary to more systematically identify low‐functioning patients and implement early mobility and activity programs to reduce hospital‐acquired functional impairment.[2, 49, 55]
Our analysis has potential limitations. First, this was an observational study and we are unable to demonstrate a direct cause‐and‐effect relationship between functional status and readmission. However, our results are consistent with prior literature in this field. Second, our cohort only included patients who were discharged from an acute hospital to a rehabilitation facility, which may limit its generalizability. However, we included a large patient sample size with a broad range of admission FIM scores, and our findings are consistent with other studies conducted in different clinical settings. Third, although 1 of our goals was to evaluate how readmission rates differed by diagnostic category, it is possible that individual diagnoses within each category may have different risks for readmission, and future larger studies could evaluate more detailed diagnostic grouping approaches. Fourth, we also recognize that although FIM score assessment has been validated, admission assessment occurs over a 72‐hour time period, during which patients' function could potentially change a clinically meaningful degree. Fifth, there may be residual confounding because of limitations in available data within our administrative dataset; however, we did account for severity of illness using a standardized measure, and prior research has demonstrated that the relationship between functional status and readmissions may be minimally confounded by demographic and clinical variables.[8, 16, 27, 56] Finally, we lacked readmission data following discharge from rehabilitation; it is possible that the association between FIM score at the time of rehabilitation initiation may have had limited predictive value among patients who successfully completed rehabilitation and were sent home.
CONCLUSION
In conclusion, in this study of patients admitted from acute care hospitals to a single inpatient rehabilitation facility, we observed a strong association between decreased functional status and increased hospital readmission. In particular, medical patients with lower physical functioning exhibited an especially high rate of readmission. Incorporating functional status assessment into routine medical care may help identify patients at higher risk of readmission. Moreover, preventing and treating impaired functional status during inpatient admission, through early activity and mobility, should be evaluated as a way of improving patient outcomes and reducing hospital readmissions.
Disclosures: Erik Hoyer, MD, is supported by the Rehabilitation Medicine Scientist Training Program (RMSTP; 5K12HD001097). The authors report no conflicts of interest.
- , , . Rehospitalizations among patients in the Medicare fee‐for‐service program. N Engl J Med. 2009;360(14):1418–1428.
- . Post‐hospital syndrome—an acquired, transient condition of generalized risk. N Engl J Med. 2013;368(2):100–102.
- , , , et al. Receiving early mobility during an intensive care unit admission is a predictor of improved outcomes in acute respiratory failure. Am J Med Sci. 2011;341(5):373–377.
- , . Association of physical functioning with same‐hospital readmission after stroke. Am J Phys Med Rehabil. 2004;83(6):434–438.
- , , , . Posthospital care transitions: patterns, complications, and risk identification. Health Serv Res. 2004;39(5):1449–1465.
- , , , , , . Risk factors for nonelective hospital readmissions. J Gen Intern Med. 1996;11(12):762–764.
- , , , et al. Risk prediction models for hospital readmission: a systematic review. JAMA. 2011;306(15):1688–1698.
- , , , et al. Hospital readmission in persons with stroke following postacute inpatient rehabilitation. J Gerontol A Biol Sci Med Sci. 2012;67(8):875–881.
- , , , , , . Hospital readmission of persons with hip fracture following medical rehabilitation. Arch Gerontol Geriatr. 2003;36(1):15–22.
- , , , , , . Characteristics of persons rehospitalized after stroke rehabilitation. Arch Phys Med Rehabil. 2001;82(10):1367–1374.
- , , , et al. A classification system for inpatient rehabilitation patients: a review and proposed revisions to the functional independence measure‐function related groups. PB98–105992, September. Washington, DC: US Department of Commerce, National Technical Information Services; 1997.
- , , , , , . A case‐mix classification system for medical rehabilitation. Med Care. 1994;32(4):366–379.
- , , , . The reliability of the functional independence measure: a quantitative review. Arch Phys Med Rehabil. 1996;77(12):1226–1232.
- , , , , , . Four methods for characterizing disability in the formation of function related groups. Arch Phys Med Rehabil. 1994;75(12):1277–1283.
- , , , et al. Association of self‐reported hospital discharge handoffs with 30‐day readmissions. JAMA Intern Med. 2013;173(8):624–629.
- , , , , , . Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951–1958.
- , , , et al. All patient refined diagnosis related groups (APR‐DRGs). Version 15.0. Report No.: 98‐054 Rev. 00. Wallingford, CT: 3M Health Information Systems; 1998.
- The inpatient rehabilitation facility–patient assessment instrument (IRF‐PAI) training manual. 2012. http://www.cms.gov/.
- , , , et al. Relationships between disability measures and nursing effort during medical rehabilitation for patients with traumatic brain and spinal cord injury. Arch Phys Med Rehabil. 1997;78(2):143–149.
- , , , . Interrater reliability of the 7‐level functional independence measure (FIM). Scand J Rehabil Med. 1994;26(3):115–119.
- , , , , , . Comparison of logistic regression and neural networks to predict rehospitalization in patients with stroke. J Clin Epidemiol. 2001;54(11):1159–1165.
- , , . Comparison of the responsiveness of the Barthel Index and the motor component of the Functional Independence Measure in stroke: the impact of using different methods for measuring responsiveness. J Clin Epidemiol. 2002;55(9):922–928.
- , . Prediction of hospital readmission for heart failure: development of a simple risk score based on administrative data. J Am Coll Cardiol. 1999;33(6):1560–1566.
- , , . Are all readmissions bad readmissions? N Engl J Med. 2010;363(3):297–298.
- , . Hospital readmission as an accountability measure. JAMA. 2011;305(5):504–505.
- . Susceptibility to critical illness: reserve, response and therapy. Intensive Care Med. 2000;26(suppl 1):S57–S63.
- , , , , . Predictors of discharge to acute care after inpatient rehabilitation in severely affected stroke patients. Am J Phys Med Rehabil. 2012;91(5):387–392.
- , , , et al. Clinical characteristics and outcomes of sepsis‐related vs non‐sepsis‐related ARDS. Chest. 2010;138(3):559–567.
- , . Long‐term outcomes from sepsis. Curr Infect Dis Rep. 2007;9(5):382–386.
- , . Heart failure performance measures and outcomes: real or illusory gains. JAMA. 2009;302(7):792–794.
- , , , , . Patients' self‐assessed functional status in heart failure by new york heart association class: a prognostic predictor of hospitalizations, quality of life and death. J Card Fail. 2010;16(2):150–156.
- , , , et al. Knee arthroplasty: disabilities in comparison to the general population and to hip arthroplasty using a French national longitudinal survey. PLoS One. 2008;3(7):e2561.
- , , , et al. Gait asymmetry in community‐ambulating stroke survivors. Arch Phys Med Rehabil. 2008;89(2):304–310.
- , , , . Recovery of upper extremity function in stroke patients: The Copenhagen Stroke Study. Arch Phys Med Rehabil. 1994;75(4):394–398.
- , , , , , . The effect of admission physiological variables on 30 day outcome after stroke. J Clin Neurosci. 2005;12(8):905–910.
- , , , et al. Unplanned readmissions after hospital discharge among patients identified as being at high risk for readmission using a validated predictive algorithm. Open Med. 2011;5(2):e104–e111.
- . Can hospitalization‐associated disability be prevented? JAMA. 2011;306(16):1800–1801.
- , , . Hospitalization‐associated disability: “she was probably able to ambulate, but I'm not sure.” JAMA. 2011;306(16):1782–1793.
- , , , , , . Importance of functional measures in predicting mortality among older hospitalized patients. JAMA. 1998;279(15):1187–1193.
- . Mobilizing patients in the intensive care unit: improving neuromuscular weakness and physical function. JAMA. 2008;300(14):1685–1690.
- , , . Technology to enhance physical rehabilitation of critically ill patients. Crit Care Med. 2009;37(10 suppl):S436–S441.
- , , , et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil. 2010;91(4):536–542.
- , , , et al. ICU early physical rehabilitation programs: financial modeling of cost savings. Crit Care Med. 2013;41(3):717–724.
- , , , et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373(9678):1874–1882.
- , , , et al. Early intensive care unit mobility therapy in the treatment of acute respiratory failure. Crit Care Med. 2008;36(8):2238–2243.
- , , , et al. Early activity is feasible and safe in respiratory failure patients. Crit Care Med. 2007;35(1):139–145.
- , . Rehabilitation quality improvement in an intensive care unit setting: implementation of a quality improvement model. Top Stroke Rehabil. 2010;17(4):271–281.
- , , , , . Sustainability and scalability of the hospital elder life program at a community hospital. J Am Geriatr Soc. 2011;59(2):359–365.
- , , , et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med. 1999;340(9):669–676.
- , , , et al. Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med. 2011;364(14):1293–1304.
- , , , et al. Rehabilitation therapy and outcomes in acute respiratory failure: an observational pilot project. J Crit Care. 2010;25(2):254–262.
- , , . Exercise for acutely hospitalised older medical patients. Cochrane Database Syst Rev. 2007;(1):CD005955.
- , , . Extra physical therapy reduces patient length of stay and improves functional outcomes and quality of life in people with acute or subacute conditions: a systematic review. Arch Phys Med Rehabil. 2011;92(9):1490–1500.
- , , , , , . Fewer emergency readmissions and better quality of life for older adults at risk of hospital readmission: a randomized controlled trial to determine the effectiveness of a 24‐week exercise and telephone follow‐up program. J Am Geriatr Soc. 2009;57(3):395–402.
- , , , , , . Effects of an acute care for elders unit on costs and 30‐day readmissions. JAMA Intern Med. 2013:1–7.
- , , , , . Risks of acute hospital transfer and mortality during stroke rehabilitation. Arch Phys Med Rehabil. 2003;84(5):712–718.
Federally mandated pay‐for‐performance initiatives promote minimizing 30‐day hospital readmissions to improve healthcare quality and reduce costs. Although the reasons for readmissions are multifactorial, many patients are readmitted for a condition other than their initial hospital admitting diagnosis.[1] Impairments in functional status experienced during acute care hospitalization contribute to patients being discharged in a debilitated state and being vulnerable to postdischarge complications and potentially hospital readmission.[2] As such, decreased functional status may be an important and potentially modifiable risk factor for acute care hospital readmission.[3]
Previous studies have suggested that impaired functional status may be an important predictor of rehospitalization.[4, 5, 6, 7] However, inferences from existing studies are limited because they did not consider functional status as their primary focus, they only considered specific patient populations (eg, stroke) or readmissions occurring well beyond the 30‐day period defined by federal pay‐for‐performance standards.[4, 5, 6, 8, 9, 10] Our objective was to evaluate the association between functional status near the time of discharge from acute care hospital and 30‐day readmission for patients admitted to an acute inpatient rehabilitation facility. As a secondary objective, we sought to investigate the relationship between functional status and readmission by diagnostic category (medical, neurologic, or orthopedic).
METHODS
Study Population and Setting
We conducted a single‐center, retrospective study of patients admitted to an inpatient rehabilitation facility at a community hospital between July 1, 2006 and December 31, 2012. This facility provides intensive rehabilitation consisting of 3 hours of therapy per day, skilled nursing care on a 24‐hour basis, and medical care by a physiatrist. We excluded patients who died during inpatient rehabilitation (n=15, 0.2%) and patients not admitted directly from an acute care setting (n=178, 2.0%).
Data Source and Covariates
Data were derived from the Uniform Data System for Medical Rehabilitation (UDSMR), which is an administrative database providing the following data upon admission to an inpatient rehabilitation facility[11, 12, 13]: age, gender, race/ethnicity, marital status, the discharge setting, the admission Functional Independence Measure (FIM) score (details further below), and admission diagnostic category as defined by the primary discharge diagnosis from the acute care hospital and grouped by functional related groups (a case‐mix system for medical rehabilitation).[12, 14] The 3M ClinTrac management software (3M, St. Paul, MN), used for mandatory reporting to the State of Maryland, provided all‐payerrefined diagnosis related group (APRDRG) and severity of illness (SOI) combinations (a tool to group patients into clinically comparable disease and severity‐of‐illness categories expected to use similar resources and experience similar outcomes). The University HealthSystem Consortium (UHC) database provided national readmission rates for all APRDRG‐SOI combinations using a methodology that has been previously described.[15, 16] Expected readmission rates for APRDRG‐SOI combinations served as a patient risk stratification tool based on clinical logic that evaluates age, comorbidities, principal diagnosis during hospitalization, and procedures conducted during hospitalization.[17]
Primary Outcome: Acute Care Readmission
The primary outcome was all‐cause acute care readmission, defined as patient transfer to an acute care hospital during inpatient rehabilitation within 30 days from admission to inpatient rehabilitation. The care model for our inpatient rehabilitation unit is such that when patients become sick or develop a complication, they are admitted directly to a clinical unit (eg, intensive care unit) at the community hospital through a rapid‐response intervention, or the physiatrist arranges with an admitting inpatient attending to accept the patient directly to his or her service.
Primary Exposure: Functional Independence Measure
Functional status was measured using the FIM score.[18] The FIM score is an 18‐item measure of functional status, with each item scored on a scale from 1 to 7 (dependent to independent). Various aspects of motor function and cognitive function are assessed. The FIM has been validated and shown to be reliable and reproducible.[13, 19, 20] By definition for the FIM instrument, admission FIM scores are assessed by trained multidisciplinary personnel first over the 72 hours of the rehabilitation stay, and for this study served as a proxy for patient functional status upon discharge from the acute care setting in our analysis. This 72‐hour time window allows for full assessment by therapists and nurses; however, in clinical practice at the inpatient rehabilitation unit involved in this study, much of the FIM assessment occurs within the first 24 hours of the rehabilitation stay. For our analysis, we divided FIM scores into low, medium, and high functional groups. The thresholds for these groups were based on total FIM score tertiles from a prior study<60, 60 to 76, and >76.[16] As a secondary analysis we created 6 subscales of the overall FIM score based on previous research. These subscales included: transfers (transfer to chair/wheelchair, toilet, and tub/shower), locomotion (walking and stairs), self‐care (eating, grooming, bathing, dressing, and toileting), sphincter control (bladder and bowel management), communication (comprehension and expression), and social cognition (social interaction, problem solving, and memory).[21]
Statistical Analysis
To evaluate differences in patient characteristics by diagnostic category, analysis of variance and 2 tests were used for continuous and dichotomous variables, respectively. Logistic regression was used to evaluate the association between FIM score category and readmission status, adjusting for potentially confounding variables available from the UDSMR and UHC databases. We used interaction terms to test whether the association between the FIM score and readmissions varied significantly across diagnostic categories and by age. As a secondary analysis, we modeled FIM score as a continuous variable. We expressed the odds ratio in this analysis per 10‐point change in FIM, because this represents a clinically relevant change in function.[22] Logistic regression was also used to evaluate the association between FIM subscale scores (transfers, locomotion, self‐care, sphincter control, communication, and social cognition) and readmission status. Statistical significance was defined as a 2‐sided P<0.05. Data were analyzed with R (version 2.15.0;
RESULTS
Readmitted Patients and Diagnostic Categories
A total of 9405 consecutive eligible patients were admitted to the acute inpatient rehabilitation facility between July 1, 2006 and December 31, 2012. A total of 1182 (13%) patients were readmitted back to an acute care hospital from inpatient rehabilitation. Median (interquartile range) time to readmission from acute care hospital discharge was 6 days (310 days), and median length of stay for patients who were discharged to the community from inpatient rehabilitation was 8 days (612 days).
Table 1 shows characteristics of all inpatient rehabilitation patients by diagnostic category. For the neurologic category, the most common primary diagnoses were stroke and spinal cord injury; for the medical category, infection, renal failure, congestive heart failure, and chronic obstructive pulmonary disease; and for the orthopedic category, spinal arthrodesis, knee and hip replacements. Mean FIM scores were lowest and highest for patients admitted with a primarily neurologic and orthopedic diagnosis, respectively.
| Characteristic | All Patients, N=9405 | Diagnostic Category | |||
|---|---|---|---|---|---|
| Neurologic, n=3706 | Medical, n=2135 | Orthopedic, n=3564 | P Valueb | ||
| |||||
| Age, y | 67.8 (14.2) | 66.7 (15.3) | 67.0 (14.9) | 69.3 (12.4) | <0.001 |
| Male | 4,068 (43%) | 1,816 (49%) | 1,119 (52%) | 1,133 (32%) | <0.001 |
| Race | <0.001 | ||||
| Caucasian | 6,106 (65%) | 2344 (63%) | 1,320 (62%) | 2,442 (69%) | |
| African American | 2,501 (27%) | 984 (27%) | 658 (31%) | 859 (24%) | |
| Other | 798 (8%) | 378 (10%) | 157 (7%) | 263 (7%) | |
| Married | 4,330 (46%) | 1,683 (45%) | 931 (44%) | 1,716 (48%) | 0.002 |
| APRDRG‐SOI expected readmission rate | 18.0 (7.4) | 20.5 (6.8) | 21.3 (7.5) | 13.5 (5.6) | <0.001 |
| Total admission FIM score | 68.7 (17.2) | 60.4 (18.6) | 69.1 (15.5) | 77.2 (11.7) | <0.001 |
FIM Score Category and Risk of Readmission
Figure 1 shows that patients in the low admission FIM score category had the highest unadjusted rate of readmission for each diagnostic category. In unadjusted analysis, Table 2 shows that younger age, male sex, APDRG‐SOI expected readmission rate, and orthopedic and medical diagnostic categories were associated with readmission. As a continuous variable, FIM scores were linearly associated with readmission (Figure 2), with an unadjusted odds ratio (OR) and 95% confidence interval (CI) of 1.4 (1.4‐1.4, P<0.001) for a 10‐point decrease in FIM. Compared to patients with high admission FIM scores, patients with low and middle FIM scores had higher unadjusted odds of readmission (OR: 4.0; 95% CI: 3.4‐4.7; P<0.001 and OR: 1.8; 95% CI: 1.5‐2.1; P<0.001, respectively). Mean FIM subscale scores for patients readmitted versus not readmitted were transfers (5.3 vs 7.0, P<0.001), locomotion (1.6 vs 2.3, P<0.001), self‐care (17.0 vs 20.8, P<0.001), communication (10.6 vs 11.5, P<0.001), and social cognition (15.1 vs 16.6, P<0.001).
| Bivariable Analysisb | Multivariable Analysisb | |||||
|---|---|---|---|---|---|---|
| Characteristic | All Patients, N=9405 | Readmitted, n=1,182 | OR (95% CI) | P Value | OR (95% CI) | P Value |
| ||||||
| Age, y | 68.0 (14.2) | 66.4 (14.5) | 0.9 (0.91.0) | <0.001 | 0.9 (0.91.0) | <0.001 |
| Male | 3,431 (42%) | 637 (54%) | 1.6 (1.41.8) | <0.001 | 1.3 (1.11.5) | < 0.001 |
| Race | ||||||
| Caucasian | 5,340 (65%) | 766 (65%) | 1.0 | 1.0 | ||
| African American | 2,177 (26%) | 324 (27%) | 1.0 (0.91.2) | 0.60 | 1.0 (0.81.1) | 0.75 |
| Other | 706 (9%) | 92 (8%) | 0.9 (0.71.1) | 0.41 | 0.8 (0.61.0) | 0.12 |
| Married | 3,775 (46%) | 555 (47%) | 1.0 (0.91.2) | 0.50 | 1.0 (0.91.2) | 0.67 |
| Admission diagnosis category | ||||||
| Neurologic | 3,205 (39%) | 501 (42%) | 1.0 | 1.0 | ||
| Medical | 1,726 (21%) | 409 (35%) | 1.5 (1.31.7) | <0.001 | 1.8 (1.62.1) | < 0.001 |
| Orthopedic | 3,292 (40%) | 272 (23%) | 0.5 (0.50.6) | <0.001 | 1.3 (1.11.6) | 0.005 |
| APDRG‐SOI expected readmission rate | 17.4 (7.1%) | 22.2 (8.0%) | 1.1 (1.11.1) | <0.001 | 1.1 (1.01.1) | < 0.001 |
| Total FIM score category | ||||||
| High FIM, >76 points | 3,517 (43%) | 257 (22%) | 1.0 | 1.0 | ||
| Middle FIM, 60points | 2,742 (33%) | 353 (30%) | 1.8 (1.52.1) | <0.001 | 1.5 (1.31.8) | < 0.001 |
| Low FIM, <60 points | 1,964 (24%) | 572 (48%) | 4.0 (3.44.7) | <0.001 | 3.0 (2.53.6) | < 0.001 |
Multivariable and Subset Analyses
Patients with a primary medical diagnosis had higher odds of readmission to the hospital, (OR: 1.8; 95% CI: 1.6‐2.1, P<0.001), relative to patients with a neurologic or orthopedic diagnosis (Table 2). Across all diagnoses, the adjusted odds ratios (95% CIs) for the low and middle versus high FIM score category were 3.0 (2.5‐3.6; P<0.001) and 1.5 (1.3‐1.8; P<0.001) respectively (Table 2). When modeled as a continuous variable, a 10‐point decrease in FIM score was associated with a significantly increased adjusted readmission rate (OR: 1.4; 95% CI: 1.3‐1.4; P<0.001). In adjusted analysis including all subscales of the FIM, only the physical subscales, transfers (P<0.001), locomotion (P=0.002), and self‐care (P<0.001), were significantly associated with readmission. For each diagnostic category, there were similar significant associations between admission FIM score group and readmission status (Table 3). The odds of readmission by FIM score did not differ significantly across the 3 major diagnostic categories (P=0.20 for interaction term), suggesting that the effect of functional status was similar across various types of patients. We also did not observe a statistical interaction between age and FIM score group in predicting readmission (P=0.58). Patients in the lowest FIM group with a medical diagnosis had the highest adjusted readmission rate of 28.7% (Table 3).
| Multivariable Analysisa | Adjusted Readmission Ratesb | |||
|---|---|---|---|---|
| No. | OR (95% CI) | P Value | % (95% CI) | |
| ||||
| Neurologic | ||||
| High FIM (>76 points) | 755 | 1.0 | 7.3 (4.710.0) | |
| Middle FIM (6076 points) | 1,283 | 1.4 (1.02.1) | 0.06 | 9.1 (7.011.1) |
| Low FIM (<60 points) | 1,668 | 3.3 (2.34.7) | <0.001 | 18.7 (16.820.6) |
| Medical | ||||
| High FIM (>76 points) | 807 | 1.0 | 11.2 (8.114.3) | |
| Middle FIM (6076 points) | 766 | 1.8 (1.32.4) | <0.001 | 17.7 (14.520.9) |
| Low FIM (<60 points) | 562 | 3.2 (2.44.3) | <0.001 | 28.7 (25.132.4) |
| Orthopedic | ||||
| High FIM (>76 points) | 2,212 | 1.0 | 6.1 (4.77.6) | |
| Middle FIM (6076 points) | 1,046 | 1.4 (1.11.9) | 0.02 | 8.3 (6.410.1) |
| Low FIM (<60 points) | 306 | 2.2 (1.53.3) | <0.001 | 13.5 (10.416.7) |
DISCUSSION
In this study of 9405 consecutive patients admitted from acute care hospitals to a single inpatient rehabilitation facility, we investigated the association between functional status and readmission to an acute care hospital. We found that low functional status near the time of acute care hospital discharge was strongly associated with higher readmission rates. This relationship was consistently observed across major patient diagnostic categories, with low functioning medical patients having the highest rate of readmission (28.7%). Efforts to maintain or improve functional status during acute care hospitalization may be an important modifiable risk factor for acute care hospital readmission.
Previous studies have suggested that functional status may serve as an indicator of physiological reserve, and therefore vulnerability to medical complications and readmission.[6, 16, 23, 24, 25] Physiologic reserve refers to a person's ability to endure acute illness and is influenced by a number of factors, such as the adequacy of oxygen delivery to tissues, cardiovascular health, immune state, and nutritional status.[26] We found that motor subscales of the FIM score (transfers, locomotion, and self‐care), but not the other subscales, were independently associated with readmissions, which may suggest that lower motor scores are a stronger marker of physiologic reserve.[10, 16, 27] Although not our primary focus, we did note in our multivariable models that after adjusting for functional status, patients in a medical diagnostic category had higher readmission rates compared to patients with a primary neurologic or orthopedic diagnosis, but the impact of FIM score was consistent across all these diagnostic categories. We speculate that medical conditions that result in hospitalization, such as sepsis or acute kidney failure, may be more likely to result in multiorgan dysfunction that may impair physiological reserve and increase susceptibility to medical complications.[28, 29, 30, 31] In comparison, acute neurologic and orthopedic diagnoses, such as stroke or hip arthroplasty, directly impair gross motor function,[32, 33, 34, 35] with relative sparing of overall physiologic reserve.
The association between low functional status and readmissions is supported by previous studies across multiple hospital settings.[4, 5, 7, 8, 9, 27, 36] Despite this finding, routine inpatient medical practice may not fully address functional impairments. For instance, systematic measurement and documentation of functional status on admission and during hospitalization are not routine and may be a barrier to identifying medical patients at high risk for readmission.[37, 38, 39] Moreover, without recognition of functional impairment and its implications, current clinical practice may suboptimally prevent and treat physical impairments during inpatient care. However, such barriers can be surmounted. For example, in the medical intensive care unit setting, there is growing recognition that proactive and aggressive management of hospital‐acquired functional impairments through early rehabilitation is safe and feasible, improving patient outcomes while reducing hospital costs and readmissions.[3, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51] Moreover, 2 recent meta‐analyses have shown that physical therapy hospital‐based exercise programs can improve length of stay, overall hospital costs, and rates of discharge to home.[52, 53] Finally, a randomized trial has demonstrated that an individualized exercise regimen started in the acute hospital setting with long‐term telephone follow‐up can significantly reduce emergency hospital readmissions and improve quality of life in older adults.[54] Therefore, decreased functional status likely represents a modifiable risk factor for hospital readmission, and further research is necessary to more systematically identify low‐functioning patients and implement early mobility and activity programs to reduce hospital‐acquired functional impairment.[2, 49, 55]
Our analysis has potential limitations. First, this was an observational study and we are unable to demonstrate a direct cause‐and‐effect relationship between functional status and readmission. However, our results are consistent with prior literature in this field. Second, our cohort only included patients who were discharged from an acute hospital to a rehabilitation facility, which may limit its generalizability. However, we included a large patient sample size with a broad range of admission FIM scores, and our findings are consistent with other studies conducted in different clinical settings. Third, although 1 of our goals was to evaluate how readmission rates differed by diagnostic category, it is possible that individual diagnoses within each category may have different risks for readmission, and future larger studies could evaluate more detailed diagnostic grouping approaches. Fourth, we also recognize that although FIM score assessment has been validated, admission assessment occurs over a 72‐hour time period, during which patients' function could potentially change a clinically meaningful degree. Fifth, there may be residual confounding because of limitations in available data within our administrative dataset; however, we did account for severity of illness using a standardized measure, and prior research has demonstrated that the relationship between functional status and readmissions may be minimally confounded by demographic and clinical variables.[8, 16, 27, 56] Finally, we lacked readmission data following discharge from rehabilitation; it is possible that the association between FIM score at the time of rehabilitation initiation may have had limited predictive value among patients who successfully completed rehabilitation and were sent home.
CONCLUSION
In conclusion, in this study of patients admitted from acute care hospitals to a single inpatient rehabilitation facility, we observed a strong association between decreased functional status and increased hospital readmission. In particular, medical patients with lower physical functioning exhibited an especially high rate of readmission. Incorporating functional status assessment into routine medical care may help identify patients at higher risk of readmission. Moreover, preventing and treating impaired functional status during inpatient admission, through early activity and mobility, should be evaluated as a way of improving patient outcomes and reducing hospital readmissions.
Disclosures: Erik Hoyer, MD, is supported by the Rehabilitation Medicine Scientist Training Program (RMSTP; 5K12HD001097). The authors report no conflicts of interest.
Federally mandated pay‐for‐performance initiatives promote minimizing 30‐day hospital readmissions to improve healthcare quality and reduce costs. Although the reasons for readmissions are multifactorial, many patients are readmitted for a condition other than their initial hospital admitting diagnosis.[1] Impairments in functional status experienced during acute care hospitalization contribute to patients being discharged in a debilitated state and being vulnerable to postdischarge complications and potentially hospital readmission.[2] As such, decreased functional status may be an important and potentially modifiable risk factor for acute care hospital readmission.[3]
Previous studies have suggested that impaired functional status may be an important predictor of rehospitalization.[4, 5, 6, 7] However, inferences from existing studies are limited because they did not consider functional status as their primary focus, they only considered specific patient populations (eg, stroke) or readmissions occurring well beyond the 30‐day period defined by federal pay‐for‐performance standards.[4, 5, 6, 8, 9, 10] Our objective was to evaluate the association between functional status near the time of discharge from acute care hospital and 30‐day readmission for patients admitted to an acute inpatient rehabilitation facility. As a secondary objective, we sought to investigate the relationship between functional status and readmission by diagnostic category (medical, neurologic, or orthopedic).
METHODS
Study Population and Setting
We conducted a single‐center, retrospective study of patients admitted to an inpatient rehabilitation facility at a community hospital between July 1, 2006 and December 31, 2012. This facility provides intensive rehabilitation consisting of 3 hours of therapy per day, skilled nursing care on a 24‐hour basis, and medical care by a physiatrist. We excluded patients who died during inpatient rehabilitation (n=15, 0.2%) and patients not admitted directly from an acute care setting (n=178, 2.0%).
Data Source and Covariates
Data were derived from the Uniform Data System for Medical Rehabilitation (UDSMR), which is an administrative database providing the following data upon admission to an inpatient rehabilitation facility[11, 12, 13]: age, gender, race/ethnicity, marital status, the discharge setting, the admission Functional Independence Measure (FIM) score (details further below), and admission diagnostic category as defined by the primary discharge diagnosis from the acute care hospital and grouped by functional related groups (a case‐mix system for medical rehabilitation).[12, 14] The 3M ClinTrac management software (3M, St. Paul, MN), used for mandatory reporting to the State of Maryland, provided all‐payerrefined diagnosis related group (APRDRG) and severity of illness (SOI) combinations (a tool to group patients into clinically comparable disease and severity‐of‐illness categories expected to use similar resources and experience similar outcomes). The University HealthSystem Consortium (UHC) database provided national readmission rates for all APRDRG‐SOI combinations using a methodology that has been previously described.[15, 16] Expected readmission rates for APRDRG‐SOI combinations served as a patient risk stratification tool based on clinical logic that evaluates age, comorbidities, principal diagnosis during hospitalization, and procedures conducted during hospitalization.[17]
Primary Outcome: Acute Care Readmission
The primary outcome was all‐cause acute care readmission, defined as patient transfer to an acute care hospital during inpatient rehabilitation within 30 days from admission to inpatient rehabilitation. The care model for our inpatient rehabilitation unit is such that when patients become sick or develop a complication, they are admitted directly to a clinical unit (eg, intensive care unit) at the community hospital through a rapid‐response intervention, or the physiatrist arranges with an admitting inpatient attending to accept the patient directly to his or her service.
Primary Exposure: Functional Independence Measure
Functional status was measured using the FIM score.[18] The FIM score is an 18‐item measure of functional status, with each item scored on a scale from 1 to 7 (dependent to independent). Various aspects of motor function and cognitive function are assessed. The FIM has been validated and shown to be reliable and reproducible.[13, 19, 20] By definition for the FIM instrument, admission FIM scores are assessed by trained multidisciplinary personnel first over the 72 hours of the rehabilitation stay, and for this study served as a proxy for patient functional status upon discharge from the acute care setting in our analysis. This 72‐hour time window allows for full assessment by therapists and nurses; however, in clinical practice at the inpatient rehabilitation unit involved in this study, much of the FIM assessment occurs within the first 24 hours of the rehabilitation stay. For our analysis, we divided FIM scores into low, medium, and high functional groups. The thresholds for these groups were based on total FIM score tertiles from a prior study<60, 60 to 76, and >76.[16] As a secondary analysis we created 6 subscales of the overall FIM score based on previous research. These subscales included: transfers (transfer to chair/wheelchair, toilet, and tub/shower), locomotion (walking and stairs), self‐care (eating, grooming, bathing, dressing, and toileting), sphincter control (bladder and bowel management), communication (comprehension and expression), and social cognition (social interaction, problem solving, and memory).[21]
Statistical Analysis
To evaluate differences in patient characteristics by diagnostic category, analysis of variance and 2 tests were used for continuous and dichotomous variables, respectively. Logistic regression was used to evaluate the association between FIM score category and readmission status, adjusting for potentially confounding variables available from the UDSMR and UHC databases. We used interaction terms to test whether the association between the FIM score and readmissions varied significantly across diagnostic categories and by age. As a secondary analysis, we modeled FIM score as a continuous variable. We expressed the odds ratio in this analysis per 10‐point change in FIM, because this represents a clinically relevant change in function.[22] Logistic regression was also used to evaluate the association between FIM subscale scores (transfers, locomotion, self‐care, sphincter control, communication, and social cognition) and readmission status. Statistical significance was defined as a 2‐sided P<0.05. Data were analyzed with R (version 2.15.0;
RESULTS
Readmitted Patients and Diagnostic Categories
A total of 9405 consecutive eligible patients were admitted to the acute inpatient rehabilitation facility between July 1, 2006 and December 31, 2012. A total of 1182 (13%) patients were readmitted back to an acute care hospital from inpatient rehabilitation. Median (interquartile range) time to readmission from acute care hospital discharge was 6 days (310 days), and median length of stay for patients who were discharged to the community from inpatient rehabilitation was 8 days (612 days).
Table 1 shows characteristics of all inpatient rehabilitation patients by diagnostic category. For the neurologic category, the most common primary diagnoses were stroke and spinal cord injury; for the medical category, infection, renal failure, congestive heart failure, and chronic obstructive pulmonary disease; and for the orthopedic category, spinal arthrodesis, knee and hip replacements. Mean FIM scores were lowest and highest for patients admitted with a primarily neurologic and orthopedic diagnosis, respectively.
| Characteristic | All Patients, N=9405 | Diagnostic Category | |||
|---|---|---|---|---|---|
| Neurologic, n=3706 | Medical, n=2135 | Orthopedic, n=3564 | P Valueb | ||
| |||||
| Age, y | 67.8 (14.2) | 66.7 (15.3) | 67.0 (14.9) | 69.3 (12.4) | <0.001 |
| Male | 4,068 (43%) | 1,816 (49%) | 1,119 (52%) | 1,133 (32%) | <0.001 |
| Race | <0.001 | ||||
| Caucasian | 6,106 (65%) | 2344 (63%) | 1,320 (62%) | 2,442 (69%) | |
| African American | 2,501 (27%) | 984 (27%) | 658 (31%) | 859 (24%) | |
| Other | 798 (8%) | 378 (10%) | 157 (7%) | 263 (7%) | |
| Married | 4,330 (46%) | 1,683 (45%) | 931 (44%) | 1,716 (48%) | 0.002 |
| APRDRG‐SOI expected readmission rate | 18.0 (7.4) | 20.5 (6.8) | 21.3 (7.5) | 13.5 (5.6) | <0.001 |
| Total admission FIM score | 68.7 (17.2) | 60.4 (18.6) | 69.1 (15.5) | 77.2 (11.7) | <0.001 |
FIM Score Category and Risk of Readmission
Figure 1 shows that patients in the low admission FIM score category had the highest unadjusted rate of readmission for each diagnostic category. In unadjusted analysis, Table 2 shows that younger age, male sex, APDRG‐SOI expected readmission rate, and orthopedic and medical diagnostic categories were associated with readmission. As a continuous variable, FIM scores were linearly associated with readmission (Figure 2), with an unadjusted odds ratio (OR) and 95% confidence interval (CI) of 1.4 (1.4‐1.4, P<0.001) for a 10‐point decrease in FIM. Compared to patients with high admission FIM scores, patients with low and middle FIM scores had higher unadjusted odds of readmission (OR: 4.0; 95% CI: 3.4‐4.7; P<0.001 and OR: 1.8; 95% CI: 1.5‐2.1; P<0.001, respectively). Mean FIM subscale scores for patients readmitted versus not readmitted were transfers (5.3 vs 7.0, P<0.001), locomotion (1.6 vs 2.3, P<0.001), self‐care (17.0 vs 20.8, P<0.001), communication (10.6 vs 11.5, P<0.001), and social cognition (15.1 vs 16.6, P<0.001).
| Bivariable Analysisb | Multivariable Analysisb | |||||
|---|---|---|---|---|---|---|
| Characteristic | All Patients, N=9405 | Readmitted, n=1,182 | OR (95% CI) | P Value | OR (95% CI) | P Value |
| ||||||
| Age, y | 68.0 (14.2) | 66.4 (14.5) | 0.9 (0.91.0) | <0.001 | 0.9 (0.91.0) | <0.001 |
| Male | 3,431 (42%) | 637 (54%) | 1.6 (1.41.8) | <0.001 | 1.3 (1.11.5) | < 0.001 |
| Race | ||||||
| Caucasian | 5,340 (65%) | 766 (65%) | 1.0 | 1.0 | ||
| African American | 2,177 (26%) | 324 (27%) | 1.0 (0.91.2) | 0.60 | 1.0 (0.81.1) | 0.75 |
| Other | 706 (9%) | 92 (8%) | 0.9 (0.71.1) | 0.41 | 0.8 (0.61.0) | 0.12 |
| Married | 3,775 (46%) | 555 (47%) | 1.0 (0.91.2) | 0.50 | 1.0 (0.91.2) | 0.67 |
| Admission diagnosis category | ||||||
| Neurologic | 3,205 (39%) | 501 (42%) | 1.0 | 1.0 | ||
| Medical | 1,726 (21%) | 409 (35%) | 1.5 (1.31.7) | <0.001 | 1.8 (1.62.1) | < 0.001 |
| Orthopedic | 3,292 (40%) | 272 (23%) | 0.5 (0.50.6) | <0.001 | 1.3 (1.11.6) | 0.005 |
| APDRG‐SOI expected readmission rate | 17.4 (7.1%) | 22.2 (8.0%) | 1.1 (1.11.1) | <0.001 | 1.1 (1.01.1) | < 0.001 |
| Total FIM score category | ||||||
| High FIM, >76 points | 3,517 (43%) | 257 (22%) | 1.0 | 1.0 | ||
| Middle FIM, 60points | 2,742 (33%) | 353 (30%) | 1.8 (1.52.1) | <0.001 | 1.5 (1.31.8) | < 0.001 |
| Low FIM, <60 points | 1,964 (24%) | 572 (48%) | 4.0 (3.44.7) | <0.001 | 3.0 (2.53.6) | < 0.001 |
Multivariable and Subset Analyses
Patients with a primary medical diagnosis had higher odds of readmission to the hospital, (OR: 1.8; 95% CI: 1.6‐2.1, P<0.001), relative to patients with a neurologic or orthopedic diagnosis (Table 2). Across all diagnoses, the adjusted odds ratios (95% CIs) for the low and middle versus high FIM score category were 3.0 (2.5‐3.6; P<0.001) and 1.5 (1.3‐1.8; P<0.001) respectively (Table 2). When modeled as a continuous variable, a 10‐point decrease in FIM score was associated with a significantly increased adjusted readmission rate (OR: 1.4; 95% CI: 1.3‐1.4; P<0.001). In adjusted analysis including all subscales of the FIM, only the physical subscales, transfers (P<0.001), locomotion (P=0.002), and self‐care (P<0.001), were significantly associated with readmission. For each diagnostic category, there were similar significant associations between admission FIM score group and readmission status (Table 3). The odds of readmission by FIM score did not differ significantly across the 3 major diagnostic categories (P=0.20 for interaction term), suggesting that the effect of functional status was similar across various types of patients. We also did not observe a statistical interaction between age and FIM score group in predicting readmission (P=0.58). Patients in the lowest FIM group with a medical diagnosis had the highest adjusted readmission rate of 28.7% (Table 3).
| Multivariable Analysisa | Adjusted Readmission Ratesb | |||
|---|---|---|---|---|
| No. | OR (95% CI) | P Value | % (95% CI) | |
| ||||
| Neurologic | ||||
| High FIM (>76 points) | 755 | 1.0 | 7.3 (4.710.0) | |
| Middle FIM (6076 points) | 1,283 | 1.4 (1.02.1) | 0.06 | 9.1 (7.011.1) |
| Low FIM (<60 points) | 1,668 | 3.3 (2.34.7) | <0.001 | 18.7 (16.820.6) |
| Medical | ||||
| High FIM (>76 points) | 807 | 1.0 | 11.2 (8.114.3) | |
| Middle FIM (6076 points) | 766 | 1.8 (1.32.4) | <0.001 | 17.7 (14.520.9) |
| Low FIM (<60 points) | 562 | 3.2 (2.44.3) | <0.001 | 28.7 (25.132.4) |
| Orthopedic | ||||
| High FIM (>76 points) | 2,212 | 1.0 | 6.1 (4.77.6) | |
| Middle FIM (6076 points) | 1,046 | 1.4 (1.11.9) | 0.02 | 8.3 (6.410.1) |
| Low FIM (<60 points) | 306 | 2.2 (1.53.3) | <0.001 | 13.5 (10.416.7) |
DISCUSSION
In this study of 9405 consecutive patients admitted from acute care hospitals to a single inpatient rehabilitation facility, we investigated the association between functional status and readmission to an acute care hospital. We found that low functional status near the time of acute care hospital discharge was strongly associated with higher readmission rates. This relationship was consistently observed across major patient diagnostic categories, with low functioning medical patients having the highest rate of readmission (28.7%). Efforts to maintain or improve functional status during acute care hospitalization may be an important modifiable risk factor for acute care hospital readmission.
Previous studies have suggested that functional status may serve as an indicator of physiological reserve, and therefore vulnerability to medical complications and readmission.[6, 16, 23, 24, 25] Physiologic reserve refers to a person's ability to endure acute illness and is influenced by a number of factors, such as the adequacy of oxygen delivery to tissues, cardiovascular health, immune state, and nutritional status.[26] We found that motor subscales of the FIM score (transfers, locomotion, and self‐care), but not the other subscales, were independently associated with readmissions, which may suggest that lower motor scores are a stronger marker of physiologic reserve.[10, 16, 27] Although not our primary focus, we did note in our multivariable models that after adjusting for functional status, patients in a medical diagnostic category had higher readmission rates compared to patients with a primary neurologic or orthopedic diagnosis, but the impact of FIM score was consistent across all these diagnostic categories. We speculate that medical conditions that result in hospitalization, such as sepsis or acute kidney failure, may be more likely to result in multiorgan dysfunction that may impair physiological reserve and increase susceptibility to medical complications.[28, 29, 30, 31] In comparison, acute neurologic and orthopedic diagnoses, such as stroke or hip arthroplasty, directly impair gross motor function,[32, 33, 34, 35] with relative sparing of overall physiologic reserve.
The association between low functional status and readmissions is supported by previous studies across multiple hospital settings.[4, 5, 7, 8, 9, 27, 36] Despite this finding, routine inpatient medical practice may not fully address functional impairments. For instance, systematic measurement and documentation of functional status on admission and during hospitalization are not routine and may be a barrier to identifying medical patients at high risk for readmission.[37, 38, 39] Moreover, without recognition of functional impairment and its implications, current clinical practice may suboptimally prevent and treat physical impairments during inpatient care. However, such barriers can be surmounted. For example, in the medical intensive care unit setting, there is growing recognition that proactive and aggressive management of hospital‐acquired functional impairments through early rehabilitation is safe and feasible, improving patient outcomes while reducing hospital costs and readmissions.[3, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51] Moreover, 2 recent meta‐analyses have shown that physical therapy hospital‐based exercise programs can improve length of stay, overall hospital costs, and rates of discharge to home.[52, 53] Finally, a randomized trial has demonstrated that an individualized exercise regimen started in the acute hospital setting with long‐term telephone follow‐up can significantly reduce emergency hospital readmissions and improve quality of life in older adults.[54] Therefore, decreased functional status likely represents a modifiable risk factor for hospital readmission, and further research is necessary to more systematically identify low‐functioning patients and implement early mobility and activity programs to reduce hospital‐acquired functional impairment.[2, 49, 55]
Our analysis has potential limitations. First, this was an observational study and we are unable to demonstrate a direct cause‐and‐effect relationship between functional status and readmission. However, our results are consistent with prior literature in this field. Second, our cohort only included patients who were discharged from an acute hospital to a rehabilitation facility, which may limit its generalizability. However, we included a large patient sample size with a broad range of admission FIM scores, and our findings are consistent with other studies conducted in different clinical settings. Third, although 1 of our goals was to evaluate how readmission rates differed by diagnostic category, it is possible that individual diagnoses within each category may have different risks for readmission, and future larger studies could evaluate more detailed diagnostic grouping approaches. Fourth, we also recognize that although FIM score assessment has been validated, admission assessment occurs over a 72‐hour time period, during which patients' function could potentially change a clinically meaningful degree. Fifth, there may be residual confounding because of limitations in available data within our administrative dataset; however, we did account for severity of illness using a standardized measure, and prior research has demonstrated that the relationship between functional status and readmissions may be minimally confounded by demographic and clinical variables.[8, 16, 27, 56] Finally, we lacked readmission data following discharge from rehabilitation; it is possible that the association between FIM score at the time of rehabilitation initiation may have had limited predictive value among patients who successfully completed rehabilitation and were sent home.
CONCLUSION
In conclusion, in this study of patients admitted from acute care hospitals to a single inpatient rehabilitation facility, we observed a strong association between decreased functional status and increased hospital readmission. In particular, medical patients with lower physical functioning exhibited an especially high rate of readmission. Incorporating functional status assessment into routine medical care may help identify patients at higher risk of readmission. Moreover, preventing and treating impaired functional status during inpatient admission, through early activity and mobility, should be evaluated as a way of improving patient outcomes and reducing hospital readmissions.
Disclosures: Erik Hoyer, MD, is supported by the Rehabilitation Medicine Scientist Training Program (RMSTP; 5K12HD001097). The authors report no conflicts of interest.
- , , . Rehospitalizations among patients in the Medicare fee‐for‐service program. N Engl J Med. 2009;360(14):1418–1428.
- . Post‐hospital syndrome—an acquired, transient condition of generalized risk. N Engl J Med. 2013;368(2):100–102.
- , , , et al. Receiving early mobility during an intensive care unit admission is a predictor of improved outcomes in acute respiratory failure. Am J Med Sci. 2011;341(5):373–377.
- , . Association of physical functioning with same‐hospital readmission after stroke. Am J Phys Med Rehabil. 2004;83(6):434–438.
- , , , . Posthospital care transitions: patterns, complications, and risk identification. Health Serv Res. 2004;39(5):1449–1465.
- , , , , , . Risk factors for nonelective hospital readmissions. J Gen Intern Med. 1996;11(12):762–764.
- , , , et al. Risk prediction models for hospital readmission: a systematic review. JAMA. 2011;306(15):1688–1698.
- , , , et al. Hospital readmission in persons with stroke following postacute inpatient rehabilitation. J Gerontol A Biol Sci Med Sci. 2012;67(8):875–881.
- , , , , , . Hospital readmission of persons with hip fracture following medical rehabilitation. Arch Gerontol Geriatr. 2003;36(1):15–22.
- , , , , , . Characteristics of persons rehospitalized after stroke rehabilitation. Arch Phys Med Rehabil. 2001;82(10):1367–1374.
- , , , et al. A classification system for inpatient rehabilitation patients: a review and proposed revisions to the functional independence measure‐function related groups. PB98–105992, September. Washington, DC: US Department of Commerce, National Technical Information Services; 1997.
- , , , , , . A case‐mix classification system for medical rehabilitation. Med Care. 1994;32(4):366–379.
- , , , . The reliability of the functional independence measure: a quantitative review. Arch Phys Med Rehabil. 1996;77(12):1226–1232.
- , , , , , . Four methods for characterizing disability in the formation of function related groups. Arch Phys Med Rehabil. 1994;75(12):1277–1283.
- , , , et al. Association of self‐reported hospital discharge handoffs with 30‐day readmissions. JAMA Intern Med. 2013;173(8):624–629.
- , , , , , . Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951–1958.
- , , , et al. All patient refined diagnosis related groups (APR‐DRGs). Version 15.0. Report No.: 98‐054 Rev. 00. Wallingford, CT: 3M Health Information Systems; 1998.
- The inpatient rehabilitation facility–patient assessment instrument (IRF‐PAI) training manual. 2012. http://www.cms.gov/.
- , , , et al. Relationships between disability measures and nursing effort during medical rehabilitation for patients with traumatic brain and spinal cord injury. Arch Phys Med Rehabil. 1997;78(2):143–149.
- , , , . Interrater reliability of the 7‐level functional independence measure (FIM). Scand J Rehabil Med. 1994;26(3):115–119.
- , , , , , . Comparison of logistic regression and neural networks to predict rehospitalization in patients with stroke. J Clin Epidemiol. 2001;54(11):1159–1165.
- , , . Comparison of the responsiveness of the Barthel Index and the motor component of the Functional Independence Measure in stroke: the impact of using different methods for measuring responsiveness. J Clin Epidemiol. 2002;55(9):922–928.
- , . Prediction of hospital readmission for heart failure: development of a simple risk score based on administrative data. J Am Coll Cardiol. 1999;33(6):1560–1566.
- , , . Are all readmissions bad readmissions? N Engl J Med. 2010;363(3):297–298.
- , . Hospital readmission as an accountability measure. JAMA. 2011;305(5):504–505.
- . Susceptibility to critical illness: reserve, response and therapy. Intensive Care Med. 2000;26(suppl 1):S57–S63.
- , , , , . Predictors of discharge to acute care after inpatient rehabilitation in severely affected stroke patients. Am J Phys Med Rehabil. 2012;91(5):387–392.
- , , , et al. Clinical characteristics and outcomes of sepsis‐related vs non‐sepsis‐related ARDS. Chest. 2010;138(3):559–567.
- , . Long‐term outcomes from sepsis. Curr Infect Dis Rep. 2007;9(5):382–386.
- , . Heart failure performance measures and outcomes: real or illusory gains. JAMA. 2009;302(7):792–794.
- , , , , . Patients' self‐assessed functional status in heart failure by new york heart association class: a prognostic predictor of hospitalizations, quality of life and death. J Card Fail. 2010;16(2):150–156.
- , , , et al. Knee arthroplasty: disabilities in comparison to the general population and to hip arthroplasty using a French national longitudinal survey. PLoS One. 2008;3(7):e2561.
- , , , et al. Gait asymmetry in community‐ambulating stroke survivors. Arch Phys Med Rehabil. 2008;89(2):304–310.
- , , , . Recovery of upper extremity function in stroke patients: The Copenhagen Stroke Study. Arch Phys Med Rehabil. 1994;75(4):394–398.
- , , , , , . The effect of admission physiological variables on 30 day outcome after stroke. J Clin Neurosci. 2005;12(8):905–910.
- , , , et al. Unplanned readmissions after hospital discharge among patients identified as being at high risk for readmission using a validated predictive algorithm. Open Med. 2011;5(2):e104–e111.
- . Can hospitalization‐associated disability be prevented? JAMA. 2011;306(16):1800–1801.
- , , . Hospitalization‐associated disability: “she was probably able to ambulate, but I'm not sure.” JAMA. 2011;306(16):1782–1793.
- , , , , , . Importance of functional measures in predicting mortality among older hospitalized patients. JAMA. 1998;279(15):1187–1193.
- . Mobilizing patients in the intensive care unit: improving neuromuscular weakness and physical function. JAMA. 2008;300(14):1685–1690.
- , , . Technology to enhance physical rehabilitation of critically ill patients. Crit Care Med. 2009;37(10 suppl):S436–S441.
- , , , et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil. 2010;91(4):536–542.
- , , , et al. ICU early physical rehabilitation programs: financial modeling of cost savings. Crit Care Med. 2013;41(3):717–724.
- , , , et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373(9678):1874–1882.
- , , , et al. Early intensive care unit mobility therapy in the treatment of acute respiratory failure. Crit Care Med. 2008;36(8):2238–2243.
- , , , et al. Early activity is feasible and safe in respiratory failure patients. Crit Care Med. 2007;35(1):139–145.
- , . Rehabilitation quality improvement in an intensive care unit setting: implementation of a quality improvement model. Top Stroke Rehabil. 2010;17(4):271–281.
- , , , , . Sustainability and scalability of the hospital elder life program at a community hospital. J Am Geriatr Soc. 2011;59(2):359–365.
- , , , et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med. 1999;340(9):669–676.
- , , , et al. Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med. 2011;364(14):1293–1304.
- , , , et al. Rehabilitation therapy and outcomes in acute respiratory failure: an observational pilot project. J Crit Care. 2010;25(2):254–262.
- , , . Exercise for acutely hospitalised older medical patients. Cochrane Database Syst Rev. 2007;(1):CD005955.
- , , . Extra physical therapy reduces patient length of stay and improves functional outcomes and quality of life in people with acute or subacute conditions: a systematic review. Arch Phys Med Rehabil. 2011;92(9):1490–1500.
- , , , , , . Fewer emergency readmissions and better quality of life for older adults at risk of hospital readmission: a randomized controlled trial to determine the effectiveness of a 24‐week exercise and telephone follow‐up program. J Am Geriatr Soc. 2009;57(3):395–402.
- , , , , , . Effects of an acute care for elders unit on costs and 30‐day readmissions. JAMA Intern Med. 2013:1–7.
- , , , , . Risks of acute hospital transfer and mortality during stroke rehabilitation. Arch Phys Med Rehabil. 2003;84(5):712–718.
- , , . Rehospitalizations among patients in the Medicare fee‐for‐service program. N Engl J Med. 2009;360(14):1418–1428.
- . Post‐hospital syndrome—an acquired, transient condition of generalized risk. N Engl J Med. 2013;368(2):100–102.
- , , , et al. Receiving early mobility during an intensive care unit admission is a predictor of improved outcomes in acute respiratory failure. Am J Med Sci. 2011;341(5):373–377.
- , . Association of physical functioning with same‐hospital readmission after stroke. Am J Phys Med Rehabil. 2004;83(6):434–438.
- , , , . Posthospital care transitions: patterns, complications, and risk identification. Health Serv Res. 2004;39(5):1449–1465.
- , , , , , . Risk factors for nonelective hospital readmissions. J Gen Intern Med. 1996;11(12):762–764.
- , , , et al. Risk prediction models for hospital readmission: a systematic review. JAMA. 2011;306(15):1688–1698.
- , , , et al. Hospital readmission in persons with stroke following postacute inpatient rehabilitation. J Gerontol A Biol Sci Med Sci. 2012;67(8):875–881.
- , , , , , . Hospital readmission of persons with hip fracture following medical rehabilitation. Arch Gerontol Geriatr. 2003;36(1):15–22.
- , , , , , . Characteristics of persons rehospitalized after stroke rehabilitation. Arch Phys Med Rehabil. 2001;82(10):1367–1374.
- , , , et al. A classification system for inpatient rehabilitation patients: a review and proposed revisions to the functional independence measure‐function related groups. PB98–105992, September. Washington, DC: US Department of Commerce, National Technical Information Services; 1997.
- , , , , , . A case‐mix classification system for medical rehabilitation. Med Care. 1994;32(4):366–379.
- , , , . The reliability of the functional independence measure: a quantitative review. Arch Phys Med Rehabil. 1996;77(12):1226–1232.
- , , , , , . Four methods for characterizing disability in the formation of function related groups. Arch Phys Med Rehabil. 1994;75(12):1277–1283.
- , , , et al. Association of self‐reported hospital discharge handoffs with 30‐day readmissions. JAMA Intern Med. 2013;173(8):624–629.
- , , , , , . Functional status impairment is associated with unplanned readmissions. Arch Phys Med Rehabil. 2013;94(10):1951–1958.
- , , , et al. All patient refined diagnosis related groups (APR‐DRGs). Version 15.0. Report No.: 98‐054 Rev. 00. Wallingford, CT: 3M Health Information Systems; 1998.
- The inpatient rehabilitation facility–patient assessment instrument (IRF‐PAI) training manual. 2012. http://www.cms.gov/.
- , , , et al. Relationships between disability measures and nursing effort during medical rehabilitation for patients with traumatic brain and spinal cord injury. Arch Phys Med Rehabil. 1997;78(2):143–149.
- , , , . Interrater reliability of the 7‐level functional independence measure (FIM). Scand J Rehabil Med. 1994;26(3):115–119.
- , , , , , . Comparison of logistic regression and neural networks to predict rehospitalization in patients with stroke. J Clin Epidemiol. 2001;54(11):1159–1165.
- , , . Comparison of the responsiveness of the Barthel Index and the motor component of the Functional Independence Measure in stroke: the impact of using different methods for measuring responsiveness. J Clin Epidemiol. 2002;55(9):922–928.
- , . Prediction of hospital readmission for heart failure: development of a simple risk score based on administrative data. J Am Coll Cardiol. 1999;33(6):1560–1566.
- , , . Are all readmissions bad readmissions? N Engl J Med. 2010;363(3):297–298.
- , . Hospital readmission as an accountability measure. JAMA. 2011;305(5):504–505.
- . Susceptibility to critical illness: reserve, response and therapy. Intensive Care Med. 2000;26(suppl 1):S57–S63.
- , , , , . Predictors of discharge to acute care after inpatient rehabilitation in severely affected stroke patients. Am J Phys Med Rehabil. 2012;91(5):387–392.
- , , , et al. Clinical characteristics and outcomes of sepsis‐related vs non‐sepsis‐related ARDS. Chest. 2010;138(3):559–567.
- , . Long‐term outcomes from sepsis. Curr Infect Dis Rep. 2007;9(5):382–386.
- , . Heart failure performance measures and outcomes: real or illusory gains. JAMA. 2009;302(7):792–794.
- , , , , . Patients' self‐assessed functional status in heart failure by new york heart association class: a prognostic predictor of hospitalizations, quality of life and death. J Card Fail. 2010;16(2):150–156.
- , , , et al. Knee arthroplasty: disabilities in comparison to the general population and to hip arthroplasty using a French national longitudinal survey. PLoS One. 2008;3(7):e2561.
- , , , et al. Gait asymmetry in community‐ambulating stroke survivors. Arch Phys Med Rehabil. 2008;89(2):304–310.
- , , , . Recovery of upper extremity function in stroke patients: The Copenhagen Stroke Study. Arch Phys Med Rehabil. 1994;75(4):394–398.
- , , , , , . The effect of admission physiological variables on 30 day outcome after stroke. J Clin Neurosci. 2005;12(8):905–910.
- , , , et al. Unplanned readmissions after hospital discharge among patients identified as being at high risk for readmission using a validated predictive algorithm. Open Med. 2011;5(2):e104–e111.
- . Can hospitalization‐associated disability be prevented? JAMA. 2011;306(16):1800–1801.
- , , . Hospitalization‐associated disability: “she was probably able to ambulate, but I'm not sure.” JAMA. 2011;306(16):1782–1793.
- , , , , , . Importance of functional measures in predicting mortality among older hospitalized patients. JAMA. 1998;279(15):1187–1193.
- . Mobilizing patients in the intensive care unit: improving neuromuscular weakness and physical function. JAMA. 2008;300(14):1685–1690.
- , , . Technology to enhance physical rehabilitation of critically ill patients. Crit Care Med. 2009;37(10 suppl):S436–S441.
- , , , et al. Early physical medicine and rehabilitation for patients with acute respiratory failure: a quality improvement project. Arch Phys Med Rehabil. 2010;91(4):536–542.
- , , , et al. ICU early physical rehabilitation programs: financial modeling of cost savings. Crit Care Med. 2013;41(3):717–724.
- , , , et al. Early physical and occupational therapy in mechanically ventilated, critically ill patients: a randomised controlled trial. Lancet. 2009;373(9678):1874–1882.
- , , , et al. Early intensive care unit mobility therapy in the treatment of acute respiratory failure. Crit Care Med. 2008;36(8):2238–2243.
- , , , et al. Early activity is feasible and safe in respiratory failure patients. Crit Care Med. 2007;35(1):139–145.
- , . Rehabilitation quality improvement in an intensive care unit setting: implementation of a quality improvement model. Top Stroke Rehabil. 2010;17(4):271–281.
- , , , , . Sustainability and scalability of the hospital elder life program at a community hospital. J Am Geriatr Soc. 2011;59(2):359–365.
- , , , et al. A multicomponent intervention to prevent delirium in hospitalized older patients. N Engl J Med. 1999;340(9):669–676.
- , , , et al. Functional disability 5 years after acute respiratory distress syndrome. N Engl J Med. 2011;364(14):1293–1304.
- , , , et al. Rehabilitation therapy and outcomes in acute respiratory failure: an observational pilot project. J Crit Care. 2010;25(2):254–262.
- , , . Exercise for acutely hospitalised older medical patients. Cochrane Database Syst Rev. 2007;(1):CD005955.
- , , . Extra physical therapy reduces patient length of stay and improves functional outcomes and quality of life in people with acute or subacute conditions: a systematic review. Arch Phys Med Rehabil. 2011;92(9):1490–1500.
- , , , , , . Fewer emergency readmissions and better quality of life for older adults at risk of hospital readmission: a randomized controlled trial to determine the effectiveness of a 24‐week exercise and telephone follow‐up program. J Am Geriatr Soc. 2009;57(3):395–402.
- , , , , , . Effects of an acute care for elders unit on costs and 30‐day readmissions. JAMA Intern Med. 2013:1–7.
- , , , , . Risks of acute hospital transfer and mortality during stroke rehabilitation. Arch Phys Med Rehabil. 2003;84(5):712–718.
© 2014 Society of Hospital Medicine
