Margaret Tsien, MD, FAAP

Clinical question: Can one predict whether nonoperative management of spinal epidural abscesses will fail?

Background: Even though spinal epidural abscesses have a low incidence and nonspecific presentation, a delay in treatment can lead to significant morbidity. Previously, operative management was the preferred treatment; however, improvements in imaging and timing of diagnosis have led to an increased interest in nonoperative management. Few studies have identified possible predictors of failure for nonoperative management, and no algorithm exists for weighing the different possible predictors with the outcome of nonoperative management failure.

Study design: Retrospective cohort study.

Setting: A Massachusetts hospital system with two tertiary academic medical centers and three regional community hospitals.

Synopsis: The study evaluated 1,053 patients admitted with a spinal epidural abscess during 1993-2016. Of these, 432 patients were managed nonoperatively, and 367 were included in the analysis. Failure of nonoperative management occurred in 99 patients (27%). These patients were compared with 266 patients with successful nonoperative management with more than 60 days of follow-up. Six independent factors were associated with failure of nonoperative management including motor deficit at presentation (odds ratio, 7.85), pathological or compression fractures (OR, 6.12), active malignancy (OR, 3.32), diabetes (OR, 2.92), sensory changes at presentation (3.48), and location of the abscess dorsal to the thecal sac (OR, 0.29). Subsequently, a clinical algorithm was created to predict the likelihood of failure of nonoperative management.

Because of its retrospective design, the study was unable to assess the efficacy of surgery versus nonoperative management.
 

Bottom line: Specific measures of general health, neurologic status at presentation, and anatomical data of a patient with a spinal epidural abscess have led to the development of a clinical algorithm to determine the risk of failure in nonoperative management of spinal epidural abscesses.

Citation: Shah AA et al. Nonoperative management of spinal epidural abscess: Development of a predictive algorithm for failure. J Bone Joint Surg Am. 2018;100(7):546-55.

Dr. Tsien is a hospitalist in the division of hospital medicine in the department of medicine at Loyola University Chicago, Maywood, Ill.

Clinical question: Can one predict whether nonoperative management of spinal epidural abscesses will fail?

Background: Even though spinal epidural abscesses have a low incidence and nonspecific presentation, a delay in treatment can lead to significant morbidity. Previously, operative management was the preferred treatment; however, improvements in imaging and timing of diagnosis have led to an increased interest in nonoperative management. Few studies have identified possible predictors of failure for nonoperative management, and no algorithm exists for weighing the different possible predictors with the outcome of nonoperative management failure.

Study design: Retrospective cohort study.

Setting: A Massachusetts hospital system with two tertiary academic medical centers and three regional community hospitals.

Synopsis: The study evaluated 1,053 patients admitted with a spinal epidural abscess during 1993-2016. Of these, 432 patients were managed nonoperatively, and 367 were included in the analysis. Failure of nonoperative management occurred in 99 patients (27%). These patients were compared with 266 patients with successful nonoperative management with more than 60 days of follow-up. Six independent factors were associated with failure of nonoperative management including motor deficit at presentation (odds ratio, 7.85), pathological or compression fractures (OR, 6.12), active malignancy (OR, 3.32), diabetes (OR, 2.92), sensory changes at presentation (3.48), and location of the abscess dorsal to the thecal sac (OR, 0.29). Subsequently, a clinical algorithm was created to predict the likelihood of failure of nonoperative management.

Because of its retrospective design, the study was unable to assess the efficacy of surgery versus nonoperative management.
 

Bottom line: Specific measures of general health, neurologic status at presentation, and anatomical data of a patient with a spinal epidural abscess have led to the development of a clinical algorithm to determine the risk of failure in nonoperative management of spinal epidural abscesses.

Citation: Shah AA et al. Nonoperative management of spinal epidural abscess: Development of a predictive algorithm for failure. J Bone Joint Surg Am. 2018;100(7):546-55.

Dr. Tsien is a hospitalist in the division of hospital medicine in the department of medicine at Loyola University Chicago, Maywood, Ill.

Clinical question: Can one predict whether nonoperative management of spinal epidural abscesses will fail?

Background: Even though spinal epidural abscesses have a low incidence and nonspecific presentation, a delay in treatment can lead to significant morbidity. Previously, operative management was the preferred treatment; however, improvements in imaging and timing of diagnosis have led to an increased interest in nonoperative management. Few studies have identified possible predictors of failure for nonoperative management, and no algorithm exists for weighing the different possible predictors with the outcome of nonoperative management failure.

Study design: Retrospective cohort study.

Setting: A Massachusetts hospital system with two tertiary academic medical centers and three regional community hospitals.

Synopsis: The study evaluated 1,053 patients admitted with a spinal epidural abscess during 1993-2016. Of these, 432 patients were managed nonoperatively, and 367 were included in the analysis. Failure of nonoperative management occurred in 99 patients (27%). These patients were compared with 266 patients with successful nonoperative management with more than 60 days of follow-up. Six independent factors were associated with failure of nonoperative management including motor deficit at presentation (odds ratio, 7.85), pathological or compression fractures (OR, 6.12), active malignancy (OR, 3.32), diabetes (OR, 2.92), sensory changes at presentation (3.48), and location of the abscess dorsal to the thecal sac (OR, 0.29). Subsequently, a clinical algorithm was created to predict the likelihood of failure of nonoperative management.

Because of its retrospective design, the study was unable to assess the efficacy of surgery versus nonoperative management.
 

Bottom line: Specific measures of general health, neurologic status at presentation, and anatomical data of a patient with a spinal epidural abscess have led to the development of a clinical algorithm to determine the risk of failure in nonoperative management of spinal epidural abscesses.

Citation: Shah AA et al. Nonoperative management of spinal epidural abscess: Development of a predictive algorithm for failure. J Bone Joint Surg Am. 2018;100(7):546-55.

Dr. Tsien is a hospitalist in the division of hospital medicine in the department of medicine at Loyola University Chicago, Maywood, Ill.

Clinical question: Does patient experience correlate with specific hospital care coordination and transition strategies, and if so, which strategies most strongly correlate with higher patient experience scores?

Background: Patient experience is an increasingly important measure in value-based payment programs. However, progress has been slow in improving patient experience, and little empirical data exist regarding which strategies are effective. Care transitions are critical times during a hospitalization, with many hospitals already implementing measures to improve the discharge process and prevent readmission of patients. It is not known whether those measures also influence patient experience scores, and if they do improve scores, which measures are most effective at doing so.

Study design: An analytic observational survey design.

Setting: Hospitals eligible for the Hospital Readmissions Reduction Program (HRRP) between June 2013 and December 2014.

Synopsis: A survey was developed and given to chief medical officers at 1,600 hospitals between June 2013 and December 2014; the survey assessed care coordination strategies employed by these institutions. 992 hospitals (62% response rate) were subsequently categorized as “low-strategy,” “mid-strategy,” or “high-strategy” hospitals. Patient satisfaction scores from the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey in 2014 were correlated to the number of strategies and the specific strategies each hospital employed. In general, the higher-strategy hospitals had significantly higher HCAHPS survey scores than did low-strategy hospitals (+2.23 points; P less than .001). Specifically, creating and sharing a discharge summary prior to discharge (+1.43 points; P less than .001), using a discharge planner (+1.71 points; P less than .001), and calling patients 48 hours post discharge (+1.64 points; P less than .001) all resulted in overall higher hospital ratings by patients.
 

One limitation of this study is that no causal inference can be made between the specific strategies associated with higher HCAHPS scores and care coordination strategies.

Bottom line: Hospital-led care transition strategies with direct patient interactions led to higher patient satisfaction scores.

Citation: Figueroa JF et al. Hospital-level care coordination strategies associated with better patient experience. BMJ Qual Saf. 2018 Apr 4. doi: 10.1136/bmjqs-2017-007597.

Dr. Tsien is a hospitalist in the division of hospital medicine in the department of medicine at Loyola University Chicago, Maywood, Ill.

Clinical question: Does patient experience correlate with specific hospital care coordination and transition strategies, and if so, which strategies most strongly correlate with higher patient experience scores?

Background: Patient experience is an increasingly important measure in value-based payment programs. However, progress has been slow in improving patient experience, and little empirical data exist regarding which strategies are effective. Care transitions are critical times during a hospitalization, with many hospitals already implementing measures to improve the discharge process and prevent readmission of patients. It is not known whether those measures also influence patient experience scores, and if they do improve scores, which measures are most effective at doing so.

Study design: An analytic observational survey design.

Setting: Hospitals eligible for the Hospital Readmissions Reduction Program (HRRP) between June 2013 and December 2014.

Synopsis: A survey was developed and given to chief medical officers at 1,600 hospitals between June 2013 and December 2014; the survey assessed care coordination strategies employed by these institutions. 992 hospitals (62% response rate) were subsequently categorized as “low-strategy,” “mid-strategy,” or “high-strategy” hospitals. Patient satisfaction scores from the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey in 2014 were correlated to the number of strategies and the specific strategies each hospital employed. In general, the higher-strategy hospitals had significantly higher HCAHPS survey scores than did low-strategy hospitals (+2.23 points; P less than .001). Specifically, creating and sharing a discharge summary prior to discharge (+1.43 points; P less than .001), using a discharge planner (+1.71 points; P less than .001), and calling patients 48 hours post discharge (+1.64 points; P less than .001) all resulted in overall higher hospital ratings by patients.
 

One limitation of this study is that no causal inference can be made between the specific strategies associated with higher HCAHPS scores and care coordination strategies.

Bottom line: Hospital-led care transition strategies with direct patient interactions led to higher patient satisfaction scores.

Citation: Figueroa JF et al. Hospital-level care coordination strategies associated with better patient experience. BMJ Qual Saf. 2018 Apr 4. doi: 10.1136/bmjqs-2017-007597.

Dr. Tsien is a hospitalist in the division of hospital medicine in the department of medicine at Loyola University Chicago, Maywood, Ill.

Clinical question: Does patient experience correlate with specific hospital care coordination and transition strategies, and if so, which strategies most strongly correlate with higher patient experience scores?

Background: Patient experience is an increasingly important measure in value-based payment programs. However, progress has been slow in improving patient experience, and little empirical data exist regarding which strategies are effective. Care transitions are critical times during a hospitalization, with many hospitals already implementing measures to improve the discharge process and prevent readmission of patients. It is not known whether those measures also influence patient experience scores, and if they do improve scores, which measures are most effective at doing so.

Study design: An analytic observational survey design.

Setting: Hospitals eligible for the Hospital Readmissions Reduction Program (HRRP) between June 2013 and December 2014.

Synopsis: A survey was developed and given to chief medical officers at 1,600 hospitals between June 2013 and December 2014; the survey assessed care coordination strategies employed by these institutions. 992 hospitals (62% response rate) were subsequently categorized as “low-strategy,” “mid-strategy,” or “high-strategy” hospitals. Patient satisfaction scores from the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey in 2014 were correlated to the number of strategies and the specific strategies each hospital employed. In general, the higher-strategy hospitals had significantly higher HCAHPS survey scores than did low-strategy hospitals (+2.23 points; P less than .001). Specifically, creating and sharing a discharge summary prior to discharge (+1.43 points; P less than .001), using a discharge planner (+1.71 points; P less than .001), and calling patients 48 hours post discharge (+1.64 points; P less than .001) all resulted in overall higher hospital ratings by patients.
 

One limitation of this study is that no causal inference can be made between the specific strategies associated with higher HCAHPS scores and care coordination strategies.

Bottom line: Hospital-led care transition strategies with direct patient interactions led to higher patient satisfaction scores.

Citation: Figueroa JF et al. Hospital-level care coordination strategies associated with better patient experience. BMJ Qual Saf. 2018 Apr 4. doi: 10.1136/bmjqs-2017-007597.

Dr. Tsien is a hospitalist in the division of hospital medicine in the department of medicine at Loyola University Chicago, Maywood, Ill.

CLINICAL QUESTION: Is there variation in patient-to-intensivist ratios (PIR) across ICUs, and does that ratio affect hospital mortality?

BACKGROUND: Most studies show that intensivists improve ICU patient outcomes. With increasing ICU patients but stable intensivist staffing, patient-to-intensivist ratios are increasing. It is unclear if that rising ratio is adversely affecting patient mortality.

STUDY DESIGN: Multicenter retrospective cohort analysis.

SETTING: ICUs in the United Kingdom from 2010 to 2013.

The median PIR was 8.5 but varied substantially – PIRs were often larger. The association between PIR and mortality was U shaped. There was a decrease in mortality as the PIR reached 7.5, after which the mortality increased again. The higher mortality with very low PIRs could reflect a volume-outcome relationship. Less patients could mean less experience, different levels of ancillary staff, and so on.

This study did not take into account the possible differences in the multidisciplinary makeup of the ICU teams that would affect the intensivist’s level of responsibility.

BOTTOM LINE: There seems to be an optimal PIR for mortality, though that optimal number would likely depend on the ancillary staff, level of trainees, and patient acuity.

CITATIONS: Gershengorn HB, Harrison DA, Garland A, et al. “Association of Intensive Care Unit Patient-to-Intensivist Ratios With Hospital Mortality.” JAMA Intern Med. 2017 Mar 1;177(3):388-96.

Dr. Tsien is assistant professor in the division of hospital medicine, Loyola University Chicago, Maywood, Ill.

CLINICAL QUESTION: Is there variation in patient-to-intensivist ratios (PIR) across ICUs, and does that ratio affect hospital mortality?

BACKGROUND: Most studies show that intensivists improve ICU patient outcomes. With increasing ICU patients but stable intensivist staffing, patient-to-intensivist ratios are increasing. It is unclear if that rising ratio is adversely affecting patient mortality.

STUDY DESIGN: Multicenter retrospective cohort analysis.

SETTING: ICUs in the United Kingdom from 2010 to 2013.

The median PIR was 8.5 but varied substantially – PIRs were often larger. The association between PIR and mortality was U shaped. There was a decrease in mortality as the PIR reached 7.5, after which the mortality increased again. The higher mortality with very low PIRs could reflect a volume-outcome relationship. Less patients could mean less experience, different levels of ancillary staff, and so on.

This study did not take into account the possible differences in the multidisciplinary makeup of the ICU teams that would affect the intensivist’s level of responsibility.

BOTTOM LINE: There seems to be an optimal PIR for mortality, though that optimal number would likely depend on the ancillary staff, level of trainees, and patient acuity.

CITATIONS: Gershengorn HB, Harrison DA, Garland A, et al. “Association of Intensive Care Unit Patient-to-Intensivist Ratios With Hospital Mortality.” JAMA Intern Med. 2017 Mar 1;177(3):388-96.

Dr. Tsien is assistant professor in the division of hospital medicine, Loyola University Chicago, Maywood, Ill.

CLINICAL QUESTION: Is there variation in patient-to-intensivist ratios (PIR) across ICUs, and does that ratio affect hospital mortality?

BACKGROUND: Most studies show that intensivists improve ICU patient outcomes. With increasing ICU patients but stable intensivist staffing, patient-to-intensivist ratios are increasing. It is unclear if that rising ratio is adversely affecting patient mortality.

STUDY DESIGN: Multicenter retrospective cohort analysis.

SETTING: ICUs in the United Kingdom from 2010 to 2013.

The median PIR was 8.5 but varied substantially – PIRs were often larger. The association between PIR and mortality was U shaped. There was a decrease in mortality as the PIR reached 7.5, after which the mortality increased again. The higher mortality with very low PIRs could reflect a volume-outcome relationship. Less patients could mean less experience, different levels of ancillary staff, and so on.

This study did not take into account the possible differences in the multidisciplinary makeup of the ICU teams that would affect the intensivist’s level of responsibility.

BOTTOM LINE: There seems to be an optimal PIR for mortality, though that optimal number would likely depend on the ancillary staff, level of trainees, and patient acuity.

CITATIONS: Gershengorn HB, Harrison DA, Garland A, et al. “Association of Intensive Care Unit Patient-to-Intensivist Ratios With Hospital Mortality.” JAMA Intern Med. 2017 Mar 1;177(3):388-96.

Dr. Tsien is assistant professor in the division of hospital medicine, Loyola University Chicago, Maywood, Ill.

CLINICAL QUESTION: What are the accuracy and clinical utility of risk scoring systems in the assessment of patients with upper gastrointestinal (GI) bleeding?

BACKGROUND: There are several pre- and postendoscopy risk scores to predict clinically relevant outcomes such as transfusion, mortality, endoscopy treatment, surgery, and length of hospital stay for upper GI bleeding. The accuracy and applicability of these risk scores has not been well established.

STUDY DESIGN: International multicenter prospective study.

SETTING: Six large hospitals in Europe, North America, Asia, and Oceania.

SYNOPSIS: This is a prospective study comparing three pre-endoscopy scoring systems (Rockall, AIMS65, Glasgow Blatchford) and two postendoscopy scoring systems (PNED, full Rockall) in 3,012 patients with upper GI bleeding over six hospitals. It examined clinically relevant outcomes: intervention (transfusion, endoscopy, interventional radiology, surgery), mortality, rebleeding, and length of hospital stay.

The Glasgow Blatchford risk score was the most accurate at predicting the need for hospitalization and death across all hospitals, compared with the other scoring systems. It was determined that a Glasgow Blatchford score of less than 1 is an optimal threshold for outpatient management, with a 98.6% sensitivity in identifying those who would not require intervention or die. The utility of these scores to direct management in high-risk patients is limited and needs further studies. No scoring system predicted rebleeding or length of hospital stay.

A weakness of the study is that patients who bled while already inpatients were excluded.

BOTTOM LINE: The Glasgow Blatchford risk score can help direct care of very low risk patients (score, less than 1) with upper GI bleeding toward outpatient management.

CITATIONS: Stanley AJ, Laine L, Dalton HR, et al. “Comparison of risk scoring systems for patients presenting with upper gastrointestinal bleeding: International multicenter prospective study.” BMJ. 2017 Jan 4;356:i6432.
 

Dr. Tsien is assistant professor in the division of hospital medicine, Loyola University Chicago, Maywood, Ill.

CLINICAL QUESTION: What are the accuracy and clinical utility of risk scoring systems in the assessment of patients with upper gastrointestinal (GI) bleeding?

BACKGROUND: There are several pre- and postendoscopy risk scores to predict clinically relevant outcomes such as transfusion, mortality, endoscopy treatment, surgery, and length of hospital stay for upper GI bleeding. The accuracy and applicability of these risk scores has not been well established.

STUDY DESIGN: International multicenter prospective study.

SETTING: Six large hospitals in Europe, North America, Asia, and Oceania.

SYNOPSIS: This is a prospective study comparing three pre-endoscopy scoring systems (Rockall, AIMS65, Glasgow Blatchford) and two postendoscopy scoring systems (PNED, full Rockall) in 3,012 patients with upper GI bleeding over six hospitals. It examined clinically relevant outcomes: intervention (transfusion, endoscopy, interventional radiology, surgery), mortality, rebleeding, and length of hospital stay.

The Glasgow Blatchford risk score was the most accurate at predicting the need for hospitalization and death across all hospitals, compared with the other scoring systems. It was determined that a Glasgow Blatchford score of less than 1 is an optimal threshold for outpatient management, with a 98.6% sensitivity in identifying those who would not require intervention or die. The utility of these scores to direct management in high-risk patients is limited and needs further studies. No scoring system predicted rebleeding or length of hospital stay.

A weakness of the study is that patients who bled while already inpatients were excluded.

BOTTOM LINE: The Glasgow Blatchford risk score can help direct care of very low risk patients (score, less than 1) with upper GI bleeding toward outpatient management.

CITATIONS: Stanley AJ, Laine L, Dalton HR, et al. “Comparison of risk scoring systems for patients presenting with upper gastrointestinal bleeding: International multicenter prospective study.” BMJ. 2017 Jan 4;356:i6432.
 

Dr. Tsien is assistant professor in the division of hospital medicine, Loyola University Chicago, Maywood, Ill.

CLINICAL QUESTION: What are the accuracy and clinical utility of risk scoring systems in the assessment of patients with upper gastrointestinal (GI) bleeding?

BACKGROUND: There are several pre- and postendoscopy risk scores to predict clinically relevant outcomes such as transfusion, mortality, endoscopy treatment, surgery, and length of hospital stay for upper GI bleeding. The accuracy and applicability of these risk scores has not been well established.

STUDY DESIGN: International multicenter prospective study.

SETTING: Six large hospitals in Europe, North America, Asia, and Oceania.

SYNOPSIS: This is a prospective study comparing three pre-endoscopy scoring systems (Rockall, AIMS65, Glasgow Blatchford) and two postendoscopy scoring systems (PNED, full Rockall) in 3,012 patients with upper GI bleeding over six hospitals. It examined clinically relevant outcomes: intervention (transfusion, endoscopy, interventional radiology, surgery), mortality, rebleeding, and length of hospital stay.

The Glasgow Blatchford risk score was the most accurate at predicting the need for hospitalization and death across all hospitals, compared with the other scoring systems. It was determined that a Glasgow Blatchford score of less than 1 is an optimal threshold for outpatient management, with a 98.6% sensitivity in identifying those who would not require intervention or die. The utility of these scores to direct management in high-risk patients is limited and needs further studies. No scoring system predicted rebleeding or length of hospital stay.

A weakness of the study is that patients who bled while already inpatients were excluded.

BOTTOM LINE: The Glasgow Blatchford risk score can help direct care of very low risk patients (score, less than 1) with upper GI bleeding toward outpatient management.

CITATIONS: Stanley AJ, Laine L, Dalton HR, et al. “Comparison of risk scoring systems for patients presenting with upper gastrointestinal bleeding: International multicenter prospective study.” BMJ. 2017 Jan 4;356:i6432.
 

Dr. Tsien is assistant professor in the division of hospital medicine, Loyola University Chicago, Maywood, Ill.