User login
Physician Predictions of Discharge
Hospital discharge planning is a complex process requiring efficient coordination of many different medical and social support services. For this reason, multidisciplinary teams work together to develop individualized discharge plans in an attempt to reduce preventable adverse events related to hospital discharge.[1, 2, 3, 4, 5] Despite these ongoing efforts, optimal discharge strategies have yet to be realized.[1, 4, 5, 6, 7, 8, 9]
One factor that may improve the discharge process is the early identification of patients who are approaching discharge.[10] Multidisciplinary teams cannot fully deploy comprehensive discharge plans until a physician deems a patient to be approaching discharge readiness.[8]
To our knowledge, no studies have examined the performance of physician predictions of upcoming discharge. Instead, prior studies have found that physicians have difficulty predicting the length of stay for patients seen in the emergency room and for elderly patients newly admitted to general medicine floor.[11, 12] The purpose of this study was to evaluate the ability of inpatient general medicine physicians to predict next or same‐day hospital discharges to help inform the timing of discharge planning.
METHODS
We collected daily in‐person predictions from all senior residents and attendings separately on the inpatient general medicine teams (5 resident/attending services and 4 attending‐only services) at a single 950‐bed academic medical center. We asked these physicians to predict whether each patient under their care had a greater than or equal to 80% chance of being discharged the next day, the same day, or neither (ie, no discharge on the next or same day).
Physician predictions of discharge occurred Monday through Friday at 1 of 3 time points: morning (79 am), midday (122 pm), or afternoon (57 pm). Data collection focused on 1 time point per week during 2 different weeks in November 2013 and 1 week in February 2014. Predictions of same‐day discharge could only be made at the morning and midday time points. Each patient could have multiple predictions if they remained hospitalized during subsequent assessments. For each physician making a prediction, we recorded the physician training level (resident or attending).
This protocol was deemed exempt by our university's institutional review board.
Outcomes
We measured the sensitivity (SN), specificity (SP), positive predictive value (PPV), and negative predictive value (NPV) for each type of physician prediction (next day, same day, or no discharge by the end of the next day). We also calculated these measurements for each time point in the time of day subgroup: morning, midday, and afternoon.
Statistical Analyses
Using a normal approximation to the binomial distribution, point estimates and 95% confidence intervals for SN, SP, PPV, and NPV for the group of all patients and for the time of day subgroup are reported. The Cochran‐Armitage trend test was used to examine trends in SN, SP, PPV, and NPV as time to discharge decreased. No adjustments were made for multiple comparisons. A 2‐sided significance level was prespecified at 0.05 for all tests.
For the subset of patients who had discharge predictions made by both a resident and an attending, agreement was examined using the kappa statistic. All analyses were conducted using SAS version 9.3 (SAS Institute, Cary, NC).
RESULTS
A total of 2660 predictions were made by 24 attendings and 15 residents. Nineteen predictions were excluded because of missing prediction type or date of discharge, leaving 2641 predictions for analysis. Table 1 summarizes the total number of predictions within subgroups.
| No. of Predictions | |
|---|---|
| All predictions | 2,641 |
| Day of the week | |
| Monday | 596 |
| Tuesday | 503 |
| Wednesday | 525 |
| Thursday | 551 |
| Friday | 466 |
| Physician training level | |
| Resident | 871 |
| Attending | 1,770 |
| Time of day | |
| Morning (7 am9 am) | 906 |
| Midday (12 pm2 pm) | 832 |
| Afternoon (5 pm7 pm) | 903 |
The overall daily discharge rate in our population was 22.3% (see Supporting Table 1 in the online version of this article for the raw values). The SN and PPV of physician predictions of next‐day discharge were 48% (95% confidence interval [CI]: 43%‐52%) and 51% (95% CI: 46%‐56%), respectively. The SN and PPV for same‐day discharge predictions were 73% (95% CI: 68%‐78%) and 69% (95% CI: 64%‐73%), respectively. The SP for next and same‐day discharge predictions was 90% (95% CI: 89%‐91%) and 95% (95% CI: 94%‐96%), whereas the NPV was 89% (95% CI: 88%‐90%) and 96% (95% CI: 95%‐97%), respectively.
Outcome measures for each prediction type are stratified by time of day and summarized in Table 2. For next‐day discharge predictions, the SN and PPV were lowest in the morning (SN 27%, PPV 33%) and peaked by the afternoon (SN 67%, PPV 69%). Similarly, for same‐day discharges, SN and PPV were highest later in the day (midday SN 88%, PPV 79%). This trend is also demonstrated in the SP and NPV, which increased as time to actual discharge approached, although the trends are not as pronounced as for SN and PPV.
| Validity Measure | Next‐Day Discharge Predictions | Trend P Value | Same‐Day Discharge Predictions | Trend P Value | ||||
|---|---|---|---|---|---|---|---|---|
| Morning | Midday | Afternoon | Morning | Midday | Afternoon | |||
| ||||||||
| Sensitivity | 0.27 (0.210.35) | 0.50 (0.410.59) | 0.67 (0.590.74) | <0.001 | 0.66 (0.590.73) | 0.88 (0.810.93) | <0.001 | |
| Specificity | 0.87 (0.850.90) | 0.90 (0.880.92) | 0.93 (0.910.95) | <0.001 | 0.88 (0.850.90) | 0.95 (0.930.97) | <0.001 | |
| PPV | 0.33 (0.250.41) | 0.48 (0.400.57) | 0.69 (0.610.76) | <0.001 | 0.62 (0.550.68) | 0.79 (0.710.85) | <0.001 | |
| NPV | 0.84 (0.810.87) | 0.91 (0.880.93) | 0.93 (0.910.94) | <0.001 | 0.90 (0.880.92) | 0.98 (0.960.99) | <0.001 | |
The overall agreement between resident and attending predictions was measured and found to have kappa values of 0.51 (P < 0.001) for next‐day predictions and 0.73 (P < 0.001) for same‐day predictions, indicating moderate and substantial agreement, respectively (see Supporting Table 2 in the online version of this article).[13]
DISCUSSION
This is the first study, to our knowledge, to examine the ability of physicians to predict upcoming discharge during the course of routine general medicine inpatient care. We found that although physicians are poor predictors of discharge in the morning prior to the day of expected discharge, their ability to correctly predict inpatient discharges showed continual improvement as the difference between the prediction time and time of actual discharge shortened.
For next‐day predictions, the most accurate time point was the afternoon, when physicians correctly predicted more than two‐thirds of actual next‐day discharges. This finding suggests that physicians can provide meaningful discharge estimates as early as the afternoon prior to expected discharge. This may be an optimal time for physicians to meet with the multidisciplinary discharge teams, as many preparations hinge on timely and accurate predictions of discharge (eg, arranging patient transportation, postdischarge visits by a home health company). Multidisciplinary teams will also be reassured that an afternoon prediction of next‐day discharge would only prematurely activate discharge resources in roughly 3 out of every 10 occurrences. Even in these instances, patients may benefit from the extra time for disease counselling, medication teaching, and arrangement of home services.[4, 5, 6, 7, 8, 9]
This investigation has several limitations. Our study was conducted at a large tertiary care center over brief time periods, with an overall discharge rate of about 1 in 5 patients per day. Thus, the results may not be generalizable to hospitals with different patient populations, volume, or turnover, or when predictions are made at different times throughout the year. Furthermore, we were unable to determine if the outcome measures were affected by prolonged lengths of stay or excessive predictions on relatively few patients. However, we sought to mitigate these constraints by surveying many different respondents with varying experience levels, caring for a heterogeneous patient population at nonconsecutive time points during the year. A review of our hospital's administrative data suggests that the bed occupancy and average length of stay during our surveys were similar with most other time points during the year, and therefore representative of a typical inpatient general medicine service.
Our investigation was a novel investigation into the performance of physician discharge predictions, which are daily predictions made either explicitly or implicitly by physicians caring for patients on a general medicine ward. By utilizing a simple, subjective survey without bulky calculations, this approach closely mirrors real‐world practice patterns, and if further validated, could be easily assimilated into the normal workflow of a wide range of busy clinicians to more effectively activate individualized discharge plans.[1, 2, 3, 4, 5]
Future work could capture additional patient information, such as functional status, diagnosis, and current length of stay, which would allow identification of certain subsets of patients in which physicians are more or less accurate in predicting hospital discharge. Additionally, the outcomes of incorrect predictions, particularly the surprise discharges that left even though they were predicted to stay, could be assessed. If patients were discharged prematurely, this may be reflected by a higher 30‐day readmission rate, lower clinic follow‐up rate, and/or lower patient satisfaction scores.
CONCLUSION
Although physicians are poor predictors of discharge in the morning prior to the day of expected discharge, their ability to correctly predict inpatient discharges steadily improves as the difference between the prediction time and time of actual discharge shortened. It remains to be determined if systematic incorporation of physician discharge predictions into standard workflows will improve the effectiveness of transition of care interventions.
Disclosure: Nothing to report.
- , , , et al. Care transitions project team: Association between quality improvement for care transition in communities and rehospitalizations among Medicare beneficiaries. JAMA. 2013;309(4):381–391.
- , , , , The care transitions intervention: results of a randomized controlled trial. Arch Intern Med. 2006;166:1822–1828.
- , , , et al. Comprehensive discharge planning and home follow‐up of hospitalized elders: a randomized clinical trial. JAMA. 1999;281:613–620.
- , , , et al. Hospital‐initiated transitional care interventions as a patient safety strategy: a systematic review. Ann Intern Med. 2013;158:433–440.
- , , , et al. Discharge planning from hospital to home. Cochrane Database Syst Rev. 2013;1:CD000313.
- , , . A prospective study of reasons for prolonged hospitalizations on a general medicine teaching service. J General Intern Med. 2005;20:108–115.
- , , , . The epidemiology of delays in a teaching hospital. Med Care. 1989;27:112–129.
- , , , et al. Development of a checklist of safe discharge practices for hospital patients. J Hosp Med. 2013;8:444–449.
- , , , , Medication reconciliation during transition of care as a patient safety strategy. Ann Intern Med. 2013;158:397–403.
- , , Making effective use of predicted discharge dates to reduce the length of stay in hospital. Nurs Times. 2009;105(15):12–13.
- . Physicians' outcome predictions for elderly patients: Survival, hospital discharge, and length of stay in a department of internal medicine. Scand J Soc Med. 1986;14(3):127–132.
- , , , . Physicians' ability to predict hospital length of stay for patients admitted to the hospital from the emergency department. Emerg Med Int. 2012;2012:824674.
- , . Understanding interobserver agreement: the kappa statistic. Fam Med. 2005;37(5):360–363.
Hospital discharge planning is a complex process requiring efficient coordination of many different medical and social support services. For this reason, multidisciplinary teams work together to develop individualized discharge plans in an attempt to reduce preventable adverse events related to hospital discharge.[1, 2, 3, 4, 5] Despite these ongoing efforts, optimal discharge strategies have yet to be realized.[1, 4, 5, 6, 7, 8, 9]
One factor that may improve the discharge process is the early identification of patients who are approaching discharge.[10] Multidisciplinary teams cannot fully deploy comprehensive discharge plans until a physician deems a patient to be approaching discharge readiness.[8]
To our knowledge, no studies have examined the performance of physician predictions of upcoming discharge. Instead, prior studies have found that physicians have difficulty predicting the length of stay for patients seen in the emergency room and for elderly patients newly admitted to general medicine floor.[11, 12] The purpose of this study was to evaluate the ability of inpatient general medicine physicians to predict next or same‐day hospital discharges to help inform the timing of discharge planning.
METHODS
We collected daily in‐person predictions from all senior residents and attendings separately on the inpatient general medicine teams (5 resident/attending services and 4 attending‐only services) at a single 950‐bed academic medical center. We asked these physicians to predict whether each patient under their care had a greater than or equal to 80% chance of being discharged the next day, the same day, or neither (ie, no discharge on the next or same day).
Physician predictions of discharge occurred Monday through Friday at 1 of 3 time points: morning (79 am), midday (122 pm), or afternoon (57 pm). Data collection focused on 1 time point per week during 2 different weeks in November 2013 and 1 week in February 2014. Predictions of same‐day discharge could only be made at the morning and midday time points. Each patient could have multiple predictions if they remained hospitalized during subsequent assessments. For each physician making a prediction, we recorded the physician training level (resident or attending).
This protocol was deemed exempt by our university's institutional review board.
Outcomes
We measured the sensitivity (SN), specificity (SP), positive predictive value (PPV), and negative predictive value (NPV) for each type of physician prediction (next day, same day, or no discharge by the end of the next day). We also calculated these measurements for each time point in the time of day subgroup: morning, midday, and afternoon.
Statistical Analyses
Using a normal approximation to the binomial distribution, point estimates and 95% confidence intervals for SN, SP, PPV, and NPV for the group of all patients and for the time of day subgroup are reported. The Cochran‐Armitage trend test was used to examine trends in SN, SP, PPV, and NPV as time to discharge decreased. No adjustments were made for multiple comparisons. A 2‐sided significance level was prespecified at 0.05 for all tests.
For the subset of patients who had discharge predictions made by both a resident and an attending, agreement was examined using the kappa statistic. All analyses were conducted using SAS version 9.3 (SAS Institute, Cary, NC).
RESULTS
A total of 2660 predictions were made by 24 attendings and 15 residents. Nineteen predictions were excluded because of missing prediction type or date of discharge, leaving 2641 predictions for analysis. Table 1 summarizes the total number of predictions within subgroups.
| No. of Predictions | |
|---|---|
| All predictions | 2,641 |
| Day of the week | |
| Monday | 596 |
| Tuesday | 503 |
| Wednesday | 525 |
| Thursday | 551 |
| Friday | 466 |
| Physician training level | |
| Resident | 871 |
| Attending | 1,770 |
| Time of day | |
| Morning (7 am9 am) | 906 |
| Midday (12 pm2 pm) | 832 |
| Afternoon (5 pm7 pm) | 903 |
The overall daily discharge rate in our population was 22.3% (see Supporting Table 1 in the online version of this article for the raw values). The SN and PPV of physician predictions of next‐day discharge were 48% (95% confidence interval [CI]: 43%‐52%) and 51% (95% CI: 46%‐56%), respectively. The SN and PPV for same‐day discharge predictions were 73% (95% CI: 68%‐78%) and 69% (95% CI: 64%‐73%), respectively. The SP for next and same‐day discharge predictions was 90% (95% CI: 89%‐91%) and 95% (95% CI: 94%‐96%), whereas the NPV was 89% (95% CI: 88%‐90%) and 96% (95% CI: 95%‐97%), respectively.
Outcome measures for each prediction type are stratified by time of day and summarized in Table 2. For next‐day discharge predictions, the SN and PPV were lowest in the morning (SN 27%, PPV 33%) and peaked by the afternoon (SN 67%, PPV 69%). Similarly, for same‐day discharges, SN and PPV were highest later in the day (midday SN 88%, PPV 79%). This trend is also demonstrated in the SP and NPV, which increased as time to actual discharge approached, although the trends are not as pronounced as for SN and PPV.
| Validity Measure | Next‐Day Discharge Predictions | Trend P Value | Same‐Day Discharge Predictions | Trend P Value | ||||
|---|---|---|---|---|---|---|---|---|
| Morning | Midday | Afternoon | Morning | Midday | Afternoon | |||
| ||||||||
| Sensitivity | 0.27 (0.210.35) | 0.50 (0.410.59) | 0.67 (0.590.74) | <0.001 | 0.66 (0.590.73) | 0.88 (0.810.93) | <0.001 | |
| Specificity | 0.87 (0.850.90) | 0.90 (0.880.92) | 0.93 (0.910.95) | <0.001 | 0.88 (0.850.90) | 0.95 (0.930.97) | <0.001 | |
| PPV | 0.33 (0.250.41) | 0.48 (0.400.57) | 0.69 (0.610.76) | <0.001 | 0.62 (0.550.68) | 0.79 (0.710.85) | <0.001 | |
| NPV | 0.84 (0.810.87) | 0.91 (0.880.93) | 0.93 (0.910.94) | <0.001 | 0.90 (0.880.92) | 0.98 (0.960.99) | <0.001 | |
The overall agreement between resident and attending predictions was measured and found to have kappa values of 0.51 (P < 0.001) for next‐day predictions and 0.73 (P < 0.001) for same‐day predictions, indicating moderate and substantial agreement, respectively (see Supporting Table 2 in the online version of this article).[13]
DISCUSSION
This is the first study, to our knowledge, to examine the ability of physicians to predict upcoming discharge during the course of routine general medicine inpatient care. We found that although physicians are poor predictors of discharge in the morning prior to the day of expected discharge, their ability to correctly predict inpatient discharges showed continual improvement as the difference between the prediction time and time of actual discharge shortened.
For next‐day predictions, the most accurate time point was the afternoon, when physicians correctly predicted more than two‐thirds of actual next‐day discharges. This finding suggests that physicians can provide meaningful discharge estimates as early as the afternoon prior to expected discharge. This may be an optimal time for physicians to meet with the multidisciplinary discharge teams, as many preparations hinge on timely and accurate predictions of discharge (eg, arranging patient transportation, postdischarge visits by a home health company). Multidisciplinary teams will also be reassured that an afternoon prediction of next‐day discharge would only prematurely activate discharge resources in roughly 3 out of every 10 occurrences. Even in these instances, patients may benefit from the extra time for disease counselling, medication teaching, and arrangement of home services.[4, 5, 6, 7, 8, 9]
This investigation has several limitations. Our study was conducted at a large tertiary care center over brief time periods, with an overall discharge rate of about 1 in 5 patients per day. Thus, the results may not be generalizable to hospitals with different patient populations, volume, or turnover, or when predictions are made at different times throughout the year. Furthermore, we were unable to determine if the outcome measures were affected by prolonged lengths of stay or excessive predictions on relatively few patients. However, we sought to mitigate these constraints by surveying many different respondents with varying experience levels, caring for a heterogeneous patient population at nonconsecutive time points during the year. A review of our hospital's administrative data suggests that the bed occupancy and average length of stay during our surveys were similar with most other time points during the year, and therefore representative of a typical inpatient general medicine service.
Our investigation was a novel investigation into the performance of physician discharge predictions, which are daily predictions made either explicitly or implicitly by physicians caring for patients on a general medicine ward. By utilizing a simple, subjective survey without bulky calculations, this approach closely mirrors real‐world practice patterns, and if further validated, could be easily assimilated into the normal workflow of a wide range of busy clinicians to more effectively activate individualized discharge plans.[1, 2, 3, 4, 5]
Future work could capture additional patient information, such as functional status, diagnosis, and current length of stay, which would allow identification of certain subsets of patients in which physicians are more or less accurate in predicting hospital discharge. Additionally, the outcomes of incorrect predictions, particularly the surprise discharges that left even though they were predicted to stay, could be assessed. If patients were discharged prematurely, this may be reflected by a higher 30‐day readmission rate, lower clinic follow‐up rate, and/or lower patient satisfaction scores.
CONCLUSION
Although physicians are poor predictors of discharge in the morning prior to the day of expected discharge, their ability to correctly predict inpatient discharges steadily improves as the difference between the prediction time and time of actual discharge shortened. It remains to be determined if systematic incorporation of physician discharge predictions into standard workflows will improve the effectiveness of transition of care interventions.
Disclosure: Nothing to report.
Hospital discharge planning is a complex process requiring efficient coordination of many different medical and social support services. For this reason, multidisciplinary teams work together to develop individualized discharge plans in an attempt to reduce preventable adverse events related to hospital discharge.[1, 2, 3, 4, 5] Despite these ongoing efforts, optimal discharge strategies have yet to be realized.[1, 4, 5, 6, 7, 8, 9]
One factor that may improve the discharge process is the early identification of patients who are approaching discharge.[10] Multidisciplinary teams cannot fully deploy comprehensive discharge plans until a physician deems a patient to be approaching discharge readiness.[8]
To our knowledge, no studies have examined the performance of physician predictions of upcoming discharge. Instead, prior studies have found that physicians have difficulty predicting the length of stay for patients seen in the emergency room and for elderly patients newly admitted to general medicine floor.[11, 12] The purpose of this study was to evaluate the ability of inpatient general medicine physicians to predict next or same‐day hospital discharges to help inform the timing of discharge planning.
METHODS
We collected daily in‐person predictions from all senior residents and attendings separately on the inpatient general medicine teams (5 resident/attending services and 4 attending‐only services) at a single 950‐bed academic medical center. We asked these physicians to predict whether each patient under their care had a greater than or equal to 80% chance of being discharged the next day, the same day, or neither (ie, no discharge on the next or same day).
Physician predictions of discharge occurred Monday through Friday at 1 of 3 time points: morning (79 am), midday (122 pm), or afternoon (57 pm). Data collection focused on 1 time point per week during 2 different weeks in November 2013 and 1 week in February 2014. Predictions of same‐day discharge could only be made at the morning and midday time points. Each patient could have multiple predictions if they remained hospitalized during subsequent assessments. For each physician making a prediction, we recorded the physician training level (resident or attending).
This protocol was deemed exempt by our university's institutional review board.
Outcomes
We measured the sensitivity (SN), specificity (SP), positive predictive value (PPV), and negative predictive value (NPV) for each type of physician prediction (next day, same day, or no discharge by the end of the next day). We also calculated these measurements for each time point in the time of day subgroup: morning, midday, and afternoon.
Statistical Analyses
Using a normal approximation to the binomial distribution, point estimates and 95% confidence intervals for SN, SP, PPV, and NPV for the group of all patients and for the time of day subgroup are reported. The Cochran‐Armitage trend test was used to examine trends in SN, SP, PPV, and NPV as time to discharge decreased. No adjustments were made for multiple comparisons. A 2‐sided significance level was prespecified at 0.05 for all tests.
For the subset of patients who had discharge predictions made by both a resident and an attending, agreement was examined using the kappa statistic. All analyses were conducted using SAS version 9.3 (SAS Institute, Cary, NC).
RESULTS
A total of 2660 predictions were made by 24 attendings and 15 residents. Nineteen predictions were excluded because of missing prediction type or date of discharge, leaving 2641 predictions for analysis. Table 1 summarizes the total number of predictions within subgroups.
| No. of Predictions | |
|---|---|
| All predictions | 2,641 |
| Day of the week | |
| Monday | 596 |
| Tuesday | 503 |
| Wednesday | 525 |
| Thursday | 551 |
| Friday | 466 |
| Physician training level | |
| Resident | 871 |
| Attending | 1,770 |
| Time of day | |
| Morning (7 am9 am) | 906 |
| Midday (12 pm2 pm) | 832 |
| Afternoon (5 pm7 pm) | 903 |
The overall daily discharge rate in our population was 22.3% (see Supporting Table 1 in the online version of this article for the raw values). The SN and PPV of physician predictions of next‐day discharge were 48% (95% confidence interval [CI]: 43%‐52%) and 51% (95% CI: 46%‐56%), respectively. The SN and PPV for same‐day discharge predictions were 73% (95% CI: 68%‐78%) and 69% (95% CI: 64%‐73%), respectively. The SP for next and same‐day discharge predictions was 90% (95% CI: 89%‐91%) and 95% (95% CI: 94%‐96%), whereas the NPV was 89% (95% CI: 88%‐90%) and 96% (95% CI: 95%‐97%), respectively.
Outcome measures for each prediction type are stratified by time of day and summarized in Table 2. For next‐day discharge predictions, the SN and PPV were lowest in the morning (SN 27%, PPV 33%) and peaked by the afternoon (SN 67%, PPV 69%). Similarly, for same‐day discharges, SN and PPV were highest later in the day (midday SN 88%, PPV 79%). This trend is also demonstrated in the SP and NPV, which increased as time to actual discharge approached, although the trends are not as pronounced as for SN and PPV.
| Validity Measure | Next‐Day Discharge Predictions | Trend P Value | Same‐Day Discharge Predictions | Trend P Value | ||||
|---|---|---|---|---|---|---|---|---|
| Morning | Midday | Afternoon | Morning | Midday | Afternoon | |||
| ||||||||
| Sensitivity | 0.27 (0.210.35) | 0.50 (0.410.59) | 0.67 (0.590.74) | <0.001 | 0.66 (0.590.73) | 0.88 (0.810.93) | <0.001 | |
| Specificity | 0.87 (0.850.90) | 0.90 (0.880.92) | 0.93 (0.910.95) | <0.001 | 0.88 (0.850.90) | 0.95 (0.930.97) | <0.001 | |
| PPV | 0.33 (0.250.41) | 0.48 (0.400.57) | 0.69 (0.610.76) | <0.001 | 0.62 (0.550.68) | 0.79 (0.710.85) | <0.001 | |
| NPV | 0.84 (0.810.87) | 0.91 (0.880.93) | 0.93 (0.910.94) | <0.001 | 0.90 (0.880.92) | 0.98 (0.960.99) | <0.001 | |
The overall agreement between resident and attending predictions was measured and found to have kappa values of 0.51 (P < 0.001) for next‐day predictions and 0.73 (P < 0.001) for same‐day predictions, indicating moderate and substantial agreement, respectively (see Supporting Table 2 in the online version of this article).[13]
DISCUSSION
This is the first study, to our knowledge, to examine the ability of physicians to predict upcoming discharge during the course of routine general medicine inpatient care. We found that although physicians are poor predictors of discharge in the morning prior to the day of expected discharge, their ability to correctly predict inpatient discharges showed continual improvement as the difference between the prediction time and time of actual discharge shortened.
For next‐day predictions, the most accurate time point was the afternoon, when physicians correctly predicted more than two‐thirds of actual next‐day discharges. This finding suggests that physicians can provide meaningful discharge estimates as early as the afternoon prior to expected discharge. This may be an optimal time for physicians to meet with the multidisciplinary discharge teams, as many preparations hinge on timely and accurate predictions of discharge (eg, arranging patient transportation, postdischarge visits by a home health company). Multidisciplinary teams will also be reassured that an afternoon prediction of next‐day discharge would only prematurely activate discharge resources in roughly 3 out of every 10 occurrences. Even in these instances, patients may benefit from the extra time for disease counselling, medication teaching, and arrangement of home services.[4, 5, 6, 7, 8, 9]
This investigation has several limitations. Our study was conducted at a large tertiary care center over brief time periods, with an overall discharge rate of about 1 in 5 patients per day. Thus, the results may not be generalizable to hospitals with different patient populations, volume, or turnover, or when predictions are made at different times throughout the year. Furthermore, we were unable to determine if the outcome measures were affected by prolonged lengths of stay or excessive predictions on relatively few patients. However, we sought to mitigate these constraints by surveying many different respondents with varying experience levels, caring for a heterogeneous patient population at nonconsecutive time points during the year. A review of our hospital's administrative data suggests that the bed occupancy and average length of stay during our surveys were similar with most other time points during the year, and therefore representative of a typical inpatient general medicine service.
Our investigation was a novel investigation into the performance of physician discharge predictions, which are daily predictions made either explicitly or implicitly by physicians caring for patients on a general medicine ward. By utilizing a simple, subjective survey without bulky calculations, this approach closely mirrors real‐world practice patterns, and if further validated, could be easily assimilated into the normal workflow of a wide range of busy clinicians to more effectively activate individualized discharge plans.[1, 2, 3, 4, 5]
Future work could capture additional patient information, such as functional status, diagnosis, and current length of stay, which would allow identification of certain subsets of patients in which physicians are more or less accurate in predicting hospital discharge. Additionally, the outcomes of incorrect predictions, particularly the surprise discharges that left even though they were predicted to stay, could be assessed. If patients were discharged prematurely, this may be reflected by a higher 30‐day readmission rate, lower clinic follow‐up rate, and/or lower patient satisfaction scores.
CONCLUSION
Although physicians are poor predictors of discharge in the morning prior to the day of expected discharge, their ability to correctly predict inpatient discharges steadily improves as the difference between the prediction time and time of actual discharge shortened. It remains to be determined if systematic incorporation of physician discharge predictions into standard workflows will improve the effectiveness of transition of care interventions.
Disclosure: Nothing to report.
- , , , et al. Care transitions project team: Association between quality improvement for care transition in communities and rehospitalizations among Medicare beneficiaries. JAMA. 2013;309(4):381–391.
- , , , , The care transitions intervention: results of a randomized controlled trial. Arch Intern Med. 2006;166:1822–1828.
- , , , et al. Comprehensive discharge planning and home follow‐up of hospitalized elders: a randomized clinical trial. JAMA. 1999;281:613–620.
- , , , et al. Hospital‐initiated transitional care interventions as a patient safety strategy: a systematic review. Ann Intern Med. 2013;158:433–440.
- , , , et al. Discharge planning from hospital to home. Cochrane Database Syst Rev. 2013;1:CD000313.
- , , . A prospective study of reasons for prolonged hospitalizations on a general medicine teaching service. J General Intern Med. 2005;20:108–115.
- , , , . The epidemiology of delays in a teaching hospital. Med Care. 1989;27:112–129.
- , , , et al. Development of a checklist of safe discharge practices for hospital patients. J Hosp Med. 2013;8:444–449.
- , , , , Medication reconciliation during transition of care as a patient safety strategy. Ann Intern Med. 2013;158:397–403.
- , , Making effective use of predicted discharge dates to reduce the length of stay in hospital. Nurs Times. 2009;105(15):12–13.
- . Physicians' outcome predictions for elderly patients: Survival, hospital discharge, and length of stay in a department of internal medicine. Scand J Soc Med. 1986;14(3):127–132.
- , , , . Physicians' ability to predict hospital length of stay for patients admitted to the hospital from the emergency department. Emerg Med Int. 2012;2012:824674.
- , . Understanding interobserver agreement: the kappa statistic. Fam Med. 2005;37(5):360–363.
- , , , et al. Care transitions project team: Association between quality improvement for care transition in communities and rehospitalizations among Medicare beneficiaries. JAMA. 2013;309(4):381–391.
- , , , , The care transitions intervention: results of a randomized controlled trial. Arch Intern Med. 2006;166:1822–1828.
- , , , et al. Comprehensive discharge planning and home follow‐up of hospitalized elders: a randomized clinical trial. JAMA. 1999;281:613–620.
- , , , et al. Hospital‐initiated transitional care interventions as a patient safety strategy: a systematic review. Ann Intern Med. 2013;158:433–440.
- , , , et al. Discharge planning from hospital to home. Cochrane Database Syst Rev. 2013;1:CD000313.
- , , . A prospective study of reasons for prolonged hospitalizations on a general medicine teaching service. J General Intern Med. 2005;20:108–115.
- , , , . The epidemiology of delays in a teaching hospital. Med Care. 1989;27:112–129.
- , , , et al. Development of a checklist of safe discharge practices for hospital patients. J Hosp Med. 2013;8:444–449.
- , , , , Medication reconciliation during transition of care as a patient safety strategy. Ann Intern Med. 2013;158:397–403.
- , , Making effective use of predicted discharge dates to reduce the length of stay in hospital. Nurs Times. 2009;105(15):12–13.
- . Physicians' outcome predictions for elderly patients: Survival, hospital discharge, and length of stay in a department of internal medicine. Scand J Soc Med. 1986;14(3):127–132.
- , , , . Physicians' ability to predict hospital length of stay for patients admitted to the hospital from the emergency department. Emerg Med Int. 2012;2012:824674.
- , . Understanding interobserver agreement: the kappa statistic. Fam Med. 2005;37(5):360–363.
Left Atrial Appendage Closure Noninferior to Warfain for Cardioembolic Event Prophylaxis in Nonvalvular Afibrillation
Clinical question: Is mechanical, left atrial appendage (LAA) closure as effective as warfarin therapy in preventing cardioembolic events in patients with nonvalvular atrial fibrillation (Afib)?
Background: Anticoagulation with warfarin has long been the standard therapy for prevention of thromboembolic complications of nonvalvular Afib; however, its use is limited by the need for monitoring and lifelong adherence, as well as its many dietary and medication interactions. Prior studies investigating the efficacy of a deployable device intended to close the LAA have shown noninferiority of the device when compared with standard warfarin anticoagulation. This study evaluated LAA closure device efficacy after a 3.8-year interval.
Study design: Randomized, unblinded controlled trial.
Setting: Fifty-nine centers in the U.S. and Europe.
Synopsis: Authors randomized 707 participants 18 years or older with nonvalvular Afib and CHADS2 score ≥1 in a 2:1 fashion to the intervention and warfarin therapy groups. The primary outcome was a composite endpoint including stroke, systemic embolism, and cardiovascular or unexplained death. The event rate in the device group was 2.3 per 100 patient-years, compared with 3.8 in the warfarin group. Rate ratio was 0.60, meeting noninferiority criteria. Primary safety events were not statistically different.
Although the authors concluded that LAA device closure was noninferior to warfarin therapy, it should be noted that there was a high dropout rate, especially in the warfarin group, motivated either by a desire to try a novel oral anticoagulant or the perception that warfarin therapy was not beneficial. It should also be noted that device placement involved not only a percutaneous procedure, but also 45 days of aspirin and warfarin therapy initially to promote endothelization, followed by six months of clopidogrel.
Bottom line: Percutaneous device closure of the LAA appears to be noninferior to warfarin therapy in the prevention of cardioembolic events over a period of several years, and might be superior.
Clinical question: Is mechanical, left atrial appendage (LAA) closure as effective as warfarin therapy in preventing cardioembolic events in patients with nonvalvular atrial fibrillation (Afib)?
Background: Anticoagulation with warfarin has long been the standard therapy for prevention of thromboembolic complications of nonvalvular Afib; however, its use is limited by the need for monitoring and lifelong adherence, as well as its many dietary and medication interactions. Prior studies investigating the efficacy of a deployable device intended to close the LAA have shown noninferiority of the device when compared with standard warfarin anticoagulation. This study evaluated LAA closure device efficacy after a 3.8-year interval.
Study design: Randomized, unblinded controlled trial.
Setting: Fifty-nine centers in the U.S. and Europe.
Synopsis: Authors randomized 707 participants 18 years or older with nonvalvular Afib and CHADS2 score ≥1 in a 2:1 fashion to the intervention and warfarin therapy groups. The primary outcome was a composite endpoint including stroke, systemic embolism, and cardiovascular or unexplained death. The event rate in the device group was 2.3 per 100 patient-years, compared with 3.8 in the warfarin group. Rate ratio was 0.60, meeting noninferiority criteria. Primary safety events were not statistically different.
Although the authors concluded that LAA device closure was noninferior to warfarin therapy, it should be noted that there was a high dropout rate, especially in the warfarin group, motivated either by a desire to try a novel oral anticoagulant or the perception that warfarin therapy was not beneficial. It should also be noted that device placement involved not only a percutaneous procedure, but also 45 days of aspirin and warfarin therapy initially to promote endothelization, followed by six months of clopidogrel.
Bottom line: Percutaneous device closure of the LAA appears to be noninferior to warfarin therapy in the prevention of cardioembolic events over a period of several years, and might be superior.
Clinical question: Is mechanical, left atrial appendage (LAA) closure as effective as warfarin therapy in preventing cardioembolic events in patients with nonvalvular atrial fibrillation (Afib)?
Background: Anticoagulation with warfarin has long been the standard therapy for prevention of thromboembolic complications of nonvalvular Afib; however, its use is limited by the need for monitoring and lifelong adherence, as well as its many dietary and medication interactions. Prior studies investigating the efficacy of a deployable device intended to close the LAA have shown noninferiority of the device when compared with standard warfarin anticoagulation. This study evaluated LAA closure device efficacy after a 3.8-year interval.
Study design: Randomized, unblinded controlled trial.
Setting: Fifty-nine centers in the U.S. and Europe.
Synopsis: Authors randomized 707 participants 18 years or older with nonvalvular Afib and CHADS2 score ≥1 in a 2:1 fashion to the intervention and warfarin therapy groups. The primary outcome was a composite endpoint including stroke, systemic embolism, and cardiovascular or unexplained death. The event rate in the device group was 2.3 per 100 patient-years, compared with 3.8 in the warfarin group. Rate ratio was 0.60, meeting noninferiority criteria. Primary safety events were not statistically different.
Although the authors concluded that LAA device closure was noninferior to warfarin therapy, it should be noted that there was a high dropout rate, especially in the warfarin group, motivated either by a desire to try a novel oral anticoagulant or the perception that warfarin therapy was not beneficial. It should also be noted that device placement involved not only a percutaneous procedure, but also 45 days of aspirin and warfarin therapy initially to promote endothelization, followed by six months of clopidogrel.
Bottom line: Percutaneous device closure of the LAA appears to be noninferior to warfarin therapy in the prevention of cardioembolic events over a period of several years, and might be superior.
Tramadol Associated with Increased Rate of Hypoglycemia
Clinical question: Does tramadol increase rates of hospitalization from hypoglycemia compared to other opioids?
Background: As tramadol use has increased in the general population, there have been multiple reports of hypoglycemia after initiation of the painkiller, including in patients with no other known risk factors, such as diabetes mellitus.
Study design: Case control study.
Setting: United Kingdom.
Synopsis: Using the United Kingdom’s Clinical Practice Research Datalink, a cohort of 334,034 patients was identified, including 1,105 hospitalized for hypoglycemia. To compare incidence of hypoglycemia in patients taking tramadol versus nontramadol opioids, patients newly treated with tramadol for noncancer pain were compared with those treated with codeine.
Use of tramadol was associated with increase in hospitalization for treatment of hypoglycemia compared with codeine. Specifically, tramadol use had an odds ratio (OR) of 1.52 (95% confidence interval, 1.09-2.10). The risk of hypoglycemia was higher in the first 30 days of use, with an OR of 2.61 (95% confidence interval, 1.61-4.23).
Since tramadol prescribing has increased over the past 10 years, clinicians should be mindful of the potential association between tramadol and severe hypoglycemia requiring hospitalization. Although the details of the pathophysiology leading to this outcome remain unclear, evidence of a causal relationship is mounting. The association with hypoglycemia was seen particularly in the first 30 days of therapy. The incidence of less severe hypoglycemia not requiring hospitalization remains unknown.
Bottom line: Tramadol use is associated with increased rates of hypoglycemia requiring hospitalization.
Clinical question: Does tramadol increase rates of hospitalization from hypoglycemia compared to other opioids?
Background: As tramadol use has increased in the general population, there have been multiple reports of hypoglycemia after initiation of the painkiller, including in patients with no other known risk factors, such as diabetes mellitus.
Study design: Case control study.
Setting: United Kingdom.
Synopsis: Using the United Kingdom’s Clinical Practice Research Datalink, a cohort of 334,034 patients was identified, including 1,105 hospitalized for hypoglycemia. To compare incidence of hypoglycemia in patients taking tramadol versus nontramadol opioids, patients newly treated with tramadol for noncancer pain were compared with those treated with codeine.
Use of tramadol was associated with increase in hospitalization for treatment of hypoglycemia compared with codeine. Specifically, tramadol use had an odds ratio (OR) of 1.52 (95% confidence interval, 1.09-2.10). The risk of hypoglycemia was higher in the first 30 days of use, with an OR of 2.61 (95% confidence interval, 1.61-4.23).
Since tramadol prescribing has increased over the past 10 years, clinicians should be mindful of the potential association between tramadol and severe hypoglycemia requiring hospitalization. Although the details of the pathophysiology leading to this outcome remain unclear, evidence of a causal relationship is mounting. The association with hypoglycemia was seen particularly in the first 30 days of therapy. The incidence of less severe hypoglycemia not requiring hospitalization remains unknown.
Bottom line: Tramadol use is associated with increased rates of hypoglycemia requiring hospitalization.
Clinical question: Does tramadol increase rates of hospitalization from hypoglycemia compared to other opioids?
Background: As tramadol use has increased in the general population, there have been multiple reports of hypoglycemia after initiation of the painkiller, including in patients with no other known risk factors, such as diabetes mellitus.
Study design: Case control study.
Setting: United Kingdom.
Synopsis: Using the United Kingdom’s Clinical Practice Research Datalink, a cohort of 334,034 patients was identified, including 1,105 hospitalized for hypoglycemia. To compare incidence of hypoglycemia in patients taking tramadol versus nontramadol opioids, patients newly treated with tramadol for noncancer pain were compared with those treated with codeine.
Use of tramadol was associated with increase in hospitalization for treatment of hypoglycemia compared with codeine. Specifically, tramadol use had an odds ratio (OR) of 1.52 (95% confidence interval, 1.09-2.10). The risk of hypoglycemia was higher in the first 30 days of use, with an OR of 2.61 (95% confidence interval, 1.61-4.23).
Since tramadol prescribing has increased over the past 10 years, clinicians should be mindful of the potential association between tramadol and severe hypoglycemia requiring hospitalization. Although the details of the pathophysiology leading to this outcome remain unclear, evidence of a causal relationship is mounting. The association with hypoglycemia was seen particularly in the first 30 days of therapy. The incidence of less severe hypoglycemia not requiring hospitalization remains unknown.
Bottom line: Tramadol use is associated with increased rates of hypoglycemia requiring hospitalization.
Longer Surgeries Associated with Increased VTE Risk
Clinical question: Does duration of surgical procedure influence venous thromboembolism (VTE) risk?
Background: The relationship between surgical procedure length and VTE risk has not been vigorously assessed, although it has been postulated that longer procedures are associated with increased VTE risk. Improved understanding of this relationship may be beneficial to surgeons deciding on VTE prophylaxis strategies or determining whether to perform coupled procedures.
Study design: Retrospective cohort study.
Setting: Data collected from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP).
Synopsis: Study authors divided 1,432,855 surgical cases during which general anesthesia was administered for a specified duration into five quintiles based on length of operative time, defined as the period during which a patient was under general anesthesia. The primary outcome was the development of a VTE within 30 days of the procedure, defined as deep venous thrombosis (DVT), pulmonary embolism (PE), or both. Logistic regression analyses were performed to assess the relationship between procedure length and VTE occurrence.
The middle quintile of procedures carried a VTE rate of 0.86%. There was a significant association between procedure duration and VTE risk when the first and second quintiles, and fourth and fifth quintiles, were compared to the middle quintile. The association was present across all surgical subspecialties.
Bottom line: Longer duration of surgical procedures is associated with increased VTE risk.
Clinical question: Does duration of surgical procedure influence venous thromboembolism (VTE) risk?
Background: The relationship between surgical procedure length and VTE risk has not been vigorously assessed, although it has been postulated that longer procedures are associated with increased VTE risk. Improved understanding of this relationship may be beneficial to surgeons deciding on VTE prophylaxis strategies or determining whether to perform coupled procedures.
Study design: Retrospective cohort study.
Setting: Data collected from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP).
Synopsis: Study authors divided 1,432,855 surgical cases during which general anesthesia was administered for a specified duration into five quintiles based on length of operative time, defined as the period during which a patient was under general anesthesia. The primary outcome was the development of a VTE within 30 days of the procedure, defined as deep venous thrombosis (DVT), pulmonary embolism (PE), or both. Logistic regression analyses were performed to assess the relationship between procedure length and VTE occurrence.
The middle quintile of procedures carried a VTE rate of 0.86%. There was a significant association between procedure duration and VTE risk when the first and second quintiles, and fourth and fifth quintiles, were compared to the middle quintile. The association was present across all surgical subspecialties.
Bottom line: Longer duration of surgical procedures is associated with increased VTE risk.
Clinical question: Does duration of surgical procedure influence venous thromboembolism (VTE) risk?
Background: The relationship between surgical procedure length and VTE risk has not been vigorously assessed, although it has been postulated that longer procedures are associated with increased VTE risk. Improved understanding of this relationship may be beneficial to surgeons deciding on VTE prophylaxis strategies or determining whether to perform coupled procedures.
Study design: Retrospective cohort study.
Setting: Data collected from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP).
Synopsis: Study authors divided 1,432,855 surgical cases during which general anesthesia was administered for a specified duration into five quintiles based on length of operative time, defined as the period during which a patient was under general anesthesia. The primary outcome was the development of a VTE within 30 days of the procedure, defined as deep venous thrombosis (DVT), pulmonary embolism (PE), or both. Logistic regression analyses were performed to assess the relationship between procedure length and VTE occurrence.
The middle quintile of procedures carried a VTE rate of 0.86%. There was a significant association between procedure duration and VTE risk when the first and second quintiles, and fourth and fifth quintiles, were compared to the middle quintile. The association was present across all surgical subspecialties.
Bottom line: Longer duration of surgical procedures is associated with increased VTE risk.
Mortality, Readmission Rates Unchanged by Duty Hour Reforms
Clinical question: Did the 2011 Accreditation Council for Graduate Medical Education (ACGME) duty hour reforms change mortality rates or readmission rates at teaching hospitals?
Background: The 2011 ACGME duty hour reforms maintained the 80-hour weekly work limit for medical residents, decreased the number of continuous hours to 16 hours from 30 hours for interns, and decreased the number of continuous hours for residents to 24 hours, with an additional four hours allowed for transitions of care. These changes have raised concerns about increased handoffs and potential changes in patient safety.
Study design: Observational study of Medicare admissions before and after duty hour reforms.
Setting: Short-term, acute-care hospitals.
Synopsis: Investigators compared 4,325,854 inpatient Medicare admissions from the two years prior to duty hour reforms with 2,058,419 admissions the year after the reforms. For each time period, the 30-day mortality and 30-day readmission rates were assessed; outcomes from more intensive teaching hospitals were compared with the outcomes from less intensive teaching hospitals. Teaching intensity was assessed according to the resident-to-bed ratio, a measure that has been used in prior research.
No significant differences were found in the primary outcomes of 30-day all-location mortality or 30-day all-cause readmissions.
When looking at specific diagnoses, only stroke was found to have a higher risk of readmission in the post-reform period (OR 1.06, 95% CI 1.01-1.13).
Although 2011 duty hour reforms represented a large, national structural change in resident education, no significant positive or negative effect was found on these important patient safety measures, consistent with what has been found in prior reviews.
Bottom line: The 2011 ACGME duty hour reforms showed no significant changes in mortality or readmissions when comparing hospitals with intensive teaching roles to those with fewer trainees.
Clinical question: Did the 2011 Accreditation Council for Graduate Medical Education (ACGME) duty hour reforms change mortality rates or readmission rates at teaching hospitals?
Background: The 2011 ACGME duty hour reforms maintained the 80-hour weekly work limit for medical residents, decreased the number of continuous hours to 16 hours from 30 hours for interns, and decreased the number of continuous hours for residents to 24 hours, with an additional four hours allowed for transitions of care. These changes have raised concerns about increased handoffs and potential changes in patient safety.
Study design: Observational study of Medicare admissions before and after duty hour reforms.
Setting: Short-term, acute-care hospitals.
Synopsis: Investigators compared 4,325,854 inpatient Medicare admissions from the two years prior to duty hour reforms with 2,058,419 admissions the year after the reforms. For each time period, the 30-day mortality and 30-day readmission rates were assessed; outcomes from more intensive teaching hospitals were compared with the outcomes from less intensive teaching hospitals. Teaching intensity was assessed according to the resident-to-bed ratio, a measure that has been used in prior research.
No significant differences were found in the primary outcomes of 30-day all-location mortality or 30-day all-cause readmissions.
When looking at specific diagnoses, only stroke was found to have a higher risk of readmission in the post-reform period (OR 1.06, 95% CI 1.01-1.13).
Although 2011 duty hour reforms represented a large, national structural change in resident education, no significant positive or negative effect was found on these important patient safety measures, consistent with what has been found in prior reviews.
Bottom line: The 2011 ACGME duty hour reforms showed no significant changes in mortality or readmissions when comparing hospitals with intensive teaching roles to those with fewer trainees.
Clinical question: Did the 2011 Accreditation Council for Graduate Medical Education (ACGME) duty hour reforms change mortality rates or readmission rates at teaching hospitals?
Background: The 2011 ACGME duty hour reforms maintained the 80-hour weekly work limit for medical residents, decreased the number of continuous hours to 16 hours from 30 hours for interns, and decreased the number of continuous hours for residents to 24 hours, with an additional four hours allowed for transitions of care. These changes have raised concerns about increased handoffs and potential changes in patient safety.
Study design: Observational study of Medicare admissions before and after duty hour reforms.
Setting: Short-term, acute-care hospitals.
Synopsis: Investigators compared 4,325,854 inpatient Medicare admissions from the two years prior to duty hour reforms with 2,058,419 admissions the year after the reforms. For each time period, the 30-day mortality and 30-day readmission rates were assessed; outcomes from more intensive teaching hospitals were compared with the outcomes from less intensive teaching hospitals. Teaching intensity was assessed according to the resident-to-bed ratio, a measure that has been used in prior research.
No significant differences were found in the primary outcomes of 30-day all-location mortality or 30-day all-cause readmissions.
When looking at specific diagnoses, only stroke was found to have a higher risk of readmission in the post-reform period (OR 1.06, 95% CI 1.01-1.13).
Although 2011 duty hour reforms represented a large, national structural change in resident education, no significant positive or negative effect was found on these important patient safety measures, consistent with what has been found in prior reviews.
Bottom line: The 2011 ACGME duty hour reforms showed no significant changes in mortality or readmissions when comparing hospitals with intensive teaching roles to those with fewer trainees.
Emergency Department Utilization May Be Lower for Attending-Only Physician Visits versus Supervised Visits
Clinical question: Does supervised learning in the ED lead to higher rates of resource utilization?
Background: Care at academic medical centers might be more expensive than nonteaching hospitals because of the increased testing and resource utilization that occurs among learners. Although there is a growing emphasis in training programs on cost-conscious care, little data has looked at resource use as an outcome.
Study design: Cross-sectional study of the National Hospital Ambulatory Medical Care Survey in 2010.
Setting: Probability sample of American EDs and ED visits.
Synopsis: Using the 2010 National Hospital Ambulatory Medical Care Survey ED sub-file, a probability sample of 29,182 ED visits was obtained—25,808 attending-only visits and 3,374 supervised visits.
Supervised visits were more likely to lead to hospital admissions (21% versus 14%), advanced imaging (28% vs. 21%), and a longer median ED stay, but not with more blood testing than attending-only ED visits. EDs were placed into three categories: “nonteaching”; “minor teaching,” where trainees are involved in fewer than 50% of visits; and “major teaching,” where trainees are involved in more than 50% of visits. Study results showed no increase in resource utilization in major teaching EDs, except for an increase in ED length of stay.
Although there was an attempt to adjust for biased selection and complexity, there is a risk that biased selection of “teaching cases” in minor teaching EDs could explain some of the higher resource use for these cases. This study does not imply causation; however, it suggests that further studies might be warranted to evaluate the relationship between learners and resource utilization.
Bottom line: Supervised visits were associated with increased hospital admissions, advanced imaging, and longer ED length of stay (LOS), but other than LOS, this relationship did not persist in major teaching EDs.
Clinical question: Does supervised learning in the ED lead to higher rates of resource utilization?
Background: Care at academic medical centers might be more expensive than nonteaching hospitals because of the increased testing and resource utilization that occurs among learners. Although there is a growing emphasis in training programs on cost-conscious care, little data has looked at resource use as an outcome.
Study design: Cross-sectional study of the National Hospital Ambulatory Medical Care Survey in 2010.
Setting: Probability sample of American EDs and ED visits.
Synopsis: Using the 2010 National Hospital Ambulatory Medical Care Survey ED sub-file, a probability sample of 29,182 ED visits was obtained—25,808 attending-only visits and 3,374 supervised visits.
Supervised visits were more likely to lead to hospital admissions (21% versus 14%), advanced imaging (28% vs. 21%), and a longer median ED stay, but not with more blood testing than attending-only ED visits. EDs were placed into three categories: “nonteaching”; “minor teaching,” where trainees are involved in fewer than 50% of visits; and “major teaching,” where trainees are involved in more than 50% of visits. Study results showed no increase in resource utilization in major teaching EDs, except for an increase in ED length of stay.
Although there was an attempt to adjust for biased selection and complexity, there is a risk that biased selection of “teaching cases” in minor teaching EDs could explain some of the higher resource use for these cases. This study does not imply causation; however, it suggests that further studies might be warranted to evaluate the relationship between learners and resource utilization.
Bottom line: Supervised visits were associated with increased hospital admissions, advanced imaging, and longer ED length of stay (LOS), but other than LOS, this relationship did not persist in major teaching EDs.
Clinical question: Does supervised learning in the ED lead to higher rates of resource utilization?
Background: Care at academic medical centers might be more expensive than nonteaching hospitals because of the increased testing and resource utilization that occurs among learners. Although there is a growing emphasis in training programs on cost-conscious care, little data has looked at resource use as an outcome.
Study design: Cross-sectional study of the National Hospital Ambulatory Medical Care Survey in 2010.
Setting: Probability sample of American EDs and ED visits.
Synopsis: Using the 2010 National Hospital Ambulatory Medical Care Survey ED sub-file, a probability sample of 29,182 ED visits was obtained—25,808 attending-only visits and 3,374 supervised visits.
Supervised visits were more likely to lead to hospital admissions (21% versus 14%), advanced imaging (28% vs. 21%), and a longer median ED stay, but not with more blood testing than attending-only ED visits. EDs were placed into three categories: “nonteaching”; “minor teaching,” where trainees are involved in fewer than 50% of visits; and “major teaching,” where trainees are involved in more than 50% of visits. Study results showed no increase in resource utilization in major teaching EDs, except for an increase in ED length of stay.
Although there was an attempt to adjust for biased selection and complexity, there is a risk that biased selection of “teaching cases” in minor teaching EDs could explain some of the higher resource use for these cases. This study does not imply causation; however, it suggests that further studies might be warranted to evaluate the relationship between learners and resource utilization.
Bottom line: Supervised visits were associated with increased hospital admissions, advanced imaging, and longer ED length of stay (LOS), but other than LOS, this relationship did not persist in major teaching EDs.
Physician Spending Habits During Residency Training Can Persist for Years
Clinical question: For primary care physicians (PCPs), does residency training area affect the pattern of physician spending after training is complete?
Background: Regional and system-level variations in the intensity of medical services provided are common in the U.S. Residency training practice patterns could explain these variations. This study examines the relationship between spending patterns in the region of residency training and individual physician practice spending patterns after training.
Study design: Secondary, multilevel, multivariable analysis of 2011 Medicare claims data.
Setting: Random, nationally representative sample of family and internal medicine physicians completing residency between 1992 and 2010, with Medicare patient panels of 40 or more patients.
Synopsis: Investigators randomly selected 2,851 PCPs who completed residency training from 1992-2010, providing care to 491,948 Medicare beneficiaries. Practice locations and residency training were matched with the Dartmouth Atlas Hospital Referral Region (HRR) files. Training and practice HRRs were categorized into low-, average-, and high-spending groups.
Physicians practicing in high-spending regions who trained in high-spending regions spent $1,926 more per Medicare beneficiary than those trained in low-spending regions. In average-spending regions, physicians who trained in high-spending regions spent an average of $897 higher than those who trained in low-spending regions. No differences were found in low-spending regions.
This association varied significantly according to years in practice. For physicians in the first seven years of practice, patient expenditures in the highest-spending training HRR were 29% greater than those in the lowest-spending training HRR. After 16 years of practice, this variation disappeared.
Although this study does not establish causality, there may be opportunities to control spending with focused interventions in residency.
Bottom line: Spending patterns vary even within HRRs; however, this study’s findings suggest that physicians’ practice patterns are developed in residency training and that training in high-spending regions likely leads to increased expenditures. Focusing on cost-conscious care during residency training could be a significant option for curtailing healthcare costs in the future.
Clinical question: For primary care physicians (PCPs), does residency training area affect the pattern of physician spending after training is complete?
Background: Regional and system-level variations in the intensity of medical services provided are common in the U.S. Residency training practice patterns could explain these variations. This study examines the relationship between spending patterns in the region of residency training and individual physician practice spending patterns after training.
Study design: Secondary, multilevel, multivariable analysis of 2011 Medicare claims data.
Setting: Random, nationally representative sample of family and internal medicine physicians completing residency between 1992 and 2010, with Medicare patient panels of 40 or more patients.
Synopsis: Investigators randomly selected 2,851 PCPs who completed residency training from 1992-2010, providing care to 491,948 Medicare beneficiaries. Practice locations and residency training were matched with the Dartmouth Atlas Hospital Referral Region (HRR) files. Training and practice HRRs were categorized into low-, average-, and high-spending groups.
Physicians practicing in high-spending regions who trained in high-spending regions spent $1,926 more per Medicare beneficiary than those trained in low-spending regions. In average-spending regions, physicians who trained in high-spending regions spent an average of $897 higher than those who trained in low-spending regions. No differences were found in low-spending regions.
This association varied significantly according to years in practice. For physicians in the first seven years of practice, patient expenditures in the highest-spending training HRR were 29% greater than those in the lowest-spending training HRR. After 16 years of practice, this variation disappeared.
Although this study does not establish causality, there may be opportunities to control spending with focused interventions in residency.
Bottom line: Spending patterns vary even within HRRs; however, this study’s findings suggest that physicians’ practice patterns are developed in residency training and that training in high-spending regions likely leads to increased expenditures. Focusing on cost-conscious care during residency training could be a significant option for curtailing healthcare costs in the future.
Clinical question: For primary care physicians (PCPs), does residency training area affect the pattern of physician spending after training is complete?
Background: Regional and system-level variations in the intensity of medical services provided are common in the U.S. Residency training practice patterns could explain these variations. This study examines the relationship between spending patterns in the region of residency training and individual physician practice spending patterns after training.
Study design: Secondary, multilevel, multivariable analysis of 2011 Medicare claims data.
Setting: Random, nationally representative sample of family and internal medicine physicians completing residency between 1992 and 2010, with Medicare patient panels of 40 or more patients.
Synopsis: Investigators randomly selected 2,851 PCPs who completed residency training from 1992-2010, providing care to 491,948 Medicare beneficiaries. Practice locations and residency training were matched with the Dartmouth Atlas Hospital Referral Region (HRR) files. Training and practice HRRs were categorized into low-, average-, and high-spending groups.
Physicians practicing in high-spending regions who trained in high-spending regions spent $1,926 more per Medicare beneficiary than those trained in low-spending regions. In average-spending regions, physicians who trained in high-spending regions spent an average of $897 higher than those who trained in low-spending regions. No differences were found in low-spending regions.
This association varied significantly according to years in practice. For physicians in the first seven years of practice, patient expenditures in the highest-spending training HRR were 29% greater than those in the lowest-spending training HRR. After 16 years of practice, this variation disappeared.
Although this study does not establish causality, there may be opportunities to control spending with focused interventions in residency.
Bottom line: Spending patterns vary even within HRRs; however, this study’s findings suggest that physicians’ practice patterns are developed in residency training and that training in high-spending regions likely leads to increased expenditures. Focusing on cost-conscious care during residency training could be a significant option for curtailing healthcare costs in the future.
Malpractice Reform Does Not Change Physician Practice Patterns
Clinical question: Do malpractice reform policies shift physician practice patterns toward lower utilization of healthcare resources?
Background: Physician-reported fears of lawsuits lead to defensive medicine practices, which contribute to high healthcare costs. It is unclear whether malpractice reform legislation reduces these costly physician practice patterns. The ED is a high-risk environment that may promote defensive medicine practices and is the focus of recent malpractice reform legislation in Texas, Georgia, and South Carolina.
Study design: Case-control.
Setting: EDs in Texas, Georgia, South Carolina, and adjacent states.
Synopsis: Using a 5% random sample of Medicare claims from 1997-2011, the investigators evaluated the impact of recent malpractice reform legislation on intensity of practice by ED physicians, as defined by rates of use of advanced imaging (computed tomography [CT] or magnetic resonance imaging [MRI]), hospital admission, and average charges. ED claims from the three reform states (Texas, Ga., and S.C.) were compared to neighboring (control) states.
Adjusted analysis of 3,868,110 ED visits from 1,166 eligible hospitals demonstrated no significant reductions in CT/MRI utilization, rates of hospital admission, or (in two of the three reform states) average per-visit ED charges attributable to policy reforms.
Bottom line: Broadly protective malpractice reform had minimal impact on emergency physicians’ intensity of practice, as measured by rates of advanced imaging use, hospital admission, and average charges. Such “pro-physician” legal reforms may be inadequate in isolation to significantly reduce costs.
Clinical question: Do malpractice reform policies shift physician practice patterns toward lower utilization of healthcare resources?
Background: Physician-reported fears of lawsuits lead to defensive medicine practices, which contribute to high healthcare costs. It is unclear whether malpractice reform legislation reduces these costly physician practice patterns. The ED is a high-risk environment that may promote defensive medicine practices and is the focus of recent malpractice reform legislation in Texas, Georgia, and South Carolina.
Study design: Case-control.
Setting: EDs in Texas, Georgia, South Carolina, and adjacent states.
Synopsis: Using a 5% random sample of Medicare claims from 1997-2011, the investigators evaluated the impact of recent malpractice reform legislation on intensity of practice by ED physicians, as defined by rates of use of advanced imaging (computed tomography [CT] or magnetic resonance imaging [MRI]), hospital admission, and average charges. ED claims from the three reform states (Texas, Ga., and S.C.) were compared to neighboring (control) states.
Adjusted analysis of 3,868,110 ED visits from 1,166 eligible hospitals demonstrated no significant reductions in CT/MRI utilization, rates of hospital admission, or (in two of the three reform states) average per-visit ED charges attributable to policy reforms.
Bottom line: Broadly protective malpractice reform had minimal impact on emergency physicians’ intensity of practice, as measured by rates of advanced imaging use, hospital admission, and average charges. Such “pro-physician” legal reforms may be inadequate in isolation to significantly reduce costs.
Clinical question: Do malpractice reform policies shift physician practice patterns toward lower utilization of healthcare resources?
Background: Physician-reported fears of lawsuits lead to defensive medicine practices, which contribute to high healthcare costs. It is unclear whether malpractice reform legislation reduces these costly physician practice patterns. The ED is a high-risk environment that may promote defensive medicine practices and is the focus of recent malpractice reform legislation in Texas, Georgia, and South Carolina.
Study design: Case-control.
Setting: EDs in Texas, Georgia, South Carolina, and adjacent states.
Synopsis: Using a 5% random sample of Medicare claims from 1997-2011, the investigators evaluated the impact of recent malpractice reform legislation on intensity of practice by ED physicians, as defined by rates of use of advanced imaging (computed tomography [CT] or magnetic resonance imaging [MRI]), hospital admission, and average charges. ED claims from the three reform states (Texas, Ga., and S.C.) were compared to neighboring (control) states.
Adjusted analysis of 3,868,110 ED visits from 1,166 eligible hospitals demonstrated no significant reductions in CT/MRI utilization, rates of hospital admission, or (in two of the three reform states) average per-visit ED charges attributable to policy reforms.
Bottom line: Broadly protective malpractice reform had minimal impact on emergency physicians’ intensity of practice, as measured by rates of advanced imaging use, hospital admission, and average charges. Such “pro-physician” legal reforms may be inadequate in isolation to significantly reduce costs.
Updated Guidelines for Management of Non-ST-Elevation Acute Coronary Syndrome
Clinical question: What is the recommended approach for management of non-ST-elevation acute coronary syndrome (NSTE-ACS)?
Background: This is the first comprehensive update from the American Heart Association/American College of Cardiology (AHA/ACC) on NSTE-ACS since 2007 and follows a focused update published in 2012.
Synopsis: This guideline provides recommendations for acute and long-term care of patients with NSTE-ACS.
Cardiac-specific troponin assays (troponin I or T) are the mainstay for ACS diagnosis. When contemporary troponin assays are used for diagnosis, other biomarkers (CK-MB, myoglobin) are not useful.
Initial hospital care for all patients with NSTE-ACS should include early initiation of beta-blockers (within the first 24 hours), high-intensity statin therapy, P2Y12 inhibitor (clopidogrel or ticagrelor) plus aspirin, and parenteral anticoagulation.
An early invasive strategy (diagnostic angiography within 24 hours with intent to perform revascularization based on coronary anatomy) is preferred to an ischemia-guided strategy, particularly in high-risk NSTE-ACS patients (Global Registry of Acute Coronary Events [GRACE] score >140).
“Ischemia-guided” strategy replaces the term “conservative management strategy,” and its focus on aggressive medical therapy is an option in selected low-risk patient populations (e.g. thrombolysis in myocardial infarction [TIMI] risk score 0 or 1, GRACE score <109, low-risk troponin-negative females). Patients managed with an ischemia-guided strategy should undergo pre-discharge noninvasive stress testing for further risk stratification.
Regardless of angiography strategy (invasive vs. ischemia-guided), post-discharge dual antiplatelet therapy (clopidogrel or ticagrelor) is recommended for up to 12 months in all patients with NSTE-ACS. Prasugrel is an appropriate P2Y12 inhibitor option for patients following percutaneous coronary intervention with stent placement.
All patients with NSTE-ACS should be referred to an outpatient comprehensive cardiovascular rehabilitation program.
Clinical question: What is the recommended approach for management of non-ST-elevation acute coronary syndrome (NSTE-ACS)?
Background: This is the first comprehensive update from the American Heart Association/American College of Cardiology (AHA/ACC) on NSTE-ACS since 2007 and follows a focused update published in 2012.
Synopsis: This guideline provides recommendations for acute and long-term care of patients with NSTE-ACS.
Cardiac-specific troponin assays (troponin I or T) are the mainstay for ACS diagnosis. When contemporary troponin assays are used for diagnosis, other biomarkers (CK-MB, myoglobin) are not useful.
Initial hospital care for all patients with NSTE-ACS should include early initiation of beta-blockers (within the first 24 hours), high-intensity statin therapy, P2Y12 inhibitor (clopidogrel or ticagrelor) plus aspirin, and parenteral anticoagulation.
An early invasive strategy (diagnostic angiography within 24 hours with intent to perform revascularization based on coronary anatomy) is preferred to an ischemia-guided strategy, particularly in high-risk NSTE-ACS patients (Global Registry of Acute Coronary Events [GRACE] score >140).
“Ischemia-guided” strategy replaces the term “conservative management strategy,” and its focus on aggressive medical therapy is an option in selected low-risk patient populations (e.g. thrombolysis in myocardial infarction [TIMI] risk score 0 or 1, GRACE score <109, low-risk troponin-negative females). Patients managed with an ischemia-guided strategy should undergo pre-discharge noninvasive stress testing for further risk stratification.
Regardless of angiography strategy (invasive vs. ischemia-guided), post-discharge dual antiplatelet therapy (clopidogrel or ticagrelor) is recommended for up to 12 months in all patients with NSTE-ACS. Prasugrel is an appropriate P2Y12 inhibitor option for patients following percutaneous coronary intervention with stent placement.
All patients with NSTE-ACS should be referred to an outpatient comprehensive cardiovascular rehabilitation program.
Clinical question: What is the recommended approach for management of non-ST-elevation acute coronary syndrome (NSTE-ACS)?
Background: This is the first comprehensive update from the American Heart Association/American College of Cardiology (AHA/ACC) on NSTE-ACS since 2007 and follows a focused update published in 2012.
Synopsis: This guideline provides recommendations for acute and long-term care of patients with NSTE-ACS.
Cardiac-specific troponin assays (troponin I or T) are the mainstay for ACS diagnosis. When contemporary troponin assays are used for diagnosis, other biomarkers (CK-MB, myoglobin) are not useful.
Initial hospital care for all patients with NSTE-ACS should include early initiation of beta-blockers (within the first 24 hours), high-intensity statin therapy, P2Y12 inhibitor (clopidogrel or ticagrelor) plus aspirin, and parenteral anticoagulation.
An early invasive strategy (diagnostic angiography within 24 hours with intent to perform revascularization based on coronary anatomy) is preferred to an ischemia-guided strategy, particularly in high-risk NSTE-ACS patients (Global Registry of Acute Coronary Events [GRACE] score >140).
“Ischemia-guided” strategy replaces the term “conservative management strategy,” and its focus on aggressive medical therapy is an option in selected low-risk patient populations (e.g. thrombolysis in myocardial infarction [TIMI] risk score 0 or 1, GRACE score <109, low-risk troponin-negative females). Patients managed with an ischemia-guided strategy should undergo pre-discharge noninvasive stress testing for further risk stratification.
Regardless of angiography strategy (invasive vs. ischemia-guided), post-discharge dual antiplatelet therapy (clopidogrel or ticagrelor) is recommended for up to 12 months in all patients with NSTE-ACS. Prasugrel is an appropriate P2Y12 inhibitor option for patients following percutaneous coronary intervention with stent placement.
All patients with NSTE-ACS should be referred to an outpatient comprehensive cardiovascular rehabilitation program.
New Guidelines for Platelet Transfusions in Adults
Clinical question: What is the recommended approach to platelet transfusion in several common clinical scenarios?
Background: The AABB (formerly American Association of Blood Banks) developed these guidelines from a recent systematic review on platelet transfusion.
Synopsis: One strong recommendation was made based on moderate-quality evidence. Four weak or uncertain recommendations were made based on low- or very low-quality evidence.
For hospitalized patients with therapy-induced hypoproliferative thrombocytopenia, transfusion of up to a single unit of platelets is recommended for a platelet count of 10x109 cells/L or less to reduce the risk of spontaneous bleeding (strong recommendation, moderate-quality evidence).
For patients undergoing elective central venous catheter placement, platelet transfusion is recommended for a platelet count of less than 20x109 cells/L (weak recommendation, low-quality evidence).
For patients undergoing elective diagnostic lumbar puncture, platelet transfusion is recommended for a platelet count of less than 50x109 cells/L (weak recommendation, very low-quality evidence).
For patients undergoing major elective non-neuraxial surgery, platelet transfusion is recommended for a platelet count of less than 50x109 cells/L (weak recommendation, very low-quality evidence).
For patients undergoing cardiopulmonary bypass surgery, it is recommended that surgeons not perform routine transfusion of platelets in non-thrombocytopenic patients. For patients who have peri-operative bleeding with thrombocytopenia and/or evidence of platelet dysfunction, platelet transfusion is recommended (weak recommendation, very low-quality evidence).
There is insufficient evidence to recommend for or against platelet transfusion in patients with intracranial hemorrhage who are taking antiplatelet medications (uncertain recommendation, very low-quality evidence).
Clinical question: What is the recommended approach to platelet transfusion in several common clinical scenarios?
Background: The AABB (formerly American Association of Blood Banks) developed these guidelines from a recent systematic review on platelet transfusion.
Synopsis: One strong recommendation was made based on moderate-quality evidence. Four weak or uncertain recommendations were made based on low- or very low-quality evidence.
For hospitalized patients with therapy-induced hypoproliferative thrombocytopenia, transfusion of up to a single unit of platelets is recommended for a platelet count of 10x109 cells/L or less to reduce the risk of spontaneous bleeding (strong recommendation, moderate-quality evidence).
For patients undergoing elective central venous catheter placement, platelet transfusion is recommended for a platelet count of less than 20x109 cells/L (weak recommendation, low-quality evidence).
For patients undergoing elective diagnostic lumbar puncture, platelet transfusion is recommended for a platelet count of less than 50x109 cells/L (weak recommendation, very low-quality evidence).
For patients undergoing major elective non-neuraxial surgery, platelet transfusion is recommended for a platelet count of less than 50x109 cells/L (weak recommendation, very low-quality evidence).
For patients undergoing cardiopulmonary bypass surgery, it is recommended that surgeons not perform routine transfusion of platelets in non-thrombocytopenic patients. For patients who have peri-operative bleeding with thrombocytopenia and/or evidence of platelet dysfunction, platelet transfusion is recommended (weak recommendation, very low-quality evidence).
There is insufficient evidence to recommend for or against platelet transfusion in patients with intracranial hemorrhage who are taking antiplatelet medications (uncertain recommendation, very low-quality evidence).
Clinical question: What is the recommended approach to platelet transfusion in several common clinical scenarios?
Background: The AABB (formerly American Association of Blood Banks) developed these guidelines from a recent systematic review on platelet transfusion.
Synopsis: One strong recommendation was made based on moderate-quality evidence. Four weak or uncertain recommendations were made based on low- or very low-quality evidence.
For hospitalized patients with therapy-induced hypoproliferative thrombocytopenia, transfusion of up to a single unit of platelets is recommended for a platelet count of 10x109 cells/L or less to reduce the risk of spontaneous bleeding (strong recommendation, moderate-quality evidence).
For patients undergoing elective central venous catheter placement, platelet transfusion is recommended for a platelet count of less than 20x109 cells/L (weak recommendation, low-quality evidence).
For patients undergoing elective diagnostic lumbar puncture, platelet transfusion is recommended for a platelet count of less than 50x109 cells/L (weak recommendation, very low-quality evidence).
For patients undergoing major elective non-neuraxial surgery, platelet transfusion is recommended for a platelet count of less than 50x109 cells/L (weak recommendation, very low-quality evidence).
For patients undergoing cardiopulmonary bypass surgery, it is recommended that surgeons not perform routine transfusion of platelets in non-thrombocytopenic patients. For patients who have peri-operative bleeding with thrombocytopenia and/or evidence of platelet dysfunction, platelet transfusion is recommended (weak recommendation, very low-quality evidence).
There is insufficient evidence to recommend for or against platelet transfusion in patients with intracranial hemorrhage who are taking antiplatelet medications (uncertain recommendation, very low-quality evidence).