Localizing inpatient general medical teams to nursing units has high intuitive validity for improving physician productivity, hospital efficiency, and patient outcomes. Motion or the moving of personnel between tasksso prominent if teams are not localizedis 1 of the 7 wastes in lean thinking.1 In a timemotion study, where hospitalists cared for patients on up to 5 different wards, O'Leary et al2 have reported large parts of hospitalists' workdays spent in indirect patient care (69%), paging (13%), and travel (3%). Localization could increase the amount of time available for direct patient care, decrease time spent for (and interruptions due to) paging, and decrease travel time, all leading to greater productivity.

O'Leary et al3 have also reported the beneficial effects of localization of medical inpatients on communication between nurses and physicians, who could identify each other more often, and reported greater communication (specifically face‐to‐face communication) with each other following localization. This improvement in communication and effective multidisciplinary rounds could lead to safer care4 and better outcomes.

Further investigations about the effect of localization are limited. Roy et al5 have compared the outcomes of patients localized to 2 inpatient pods medically staffed by hospitalists and physician assistants (PAs) to geographically dispersed, but structurally different, house staff teams. They noticed significantly lower costs, slight but nonsignificant increase in length of stay, and no difference in mortality or readmissions, but it is impossible to tease out the affect of localization versus the affect of team composition. In a before‐and‐after study, Findlay et al6 have reported a decrease in mortality and complication rates in clinically homogenous surgical patients (proximal hip fractures) when cared for by junior trainee physicians localized to a unit, but their experience cannot be extrapolated to the much more diverse general medical population.

In our hospital, each general medical team could admit patients dispersed over 14 different units. An internal group, commissioned to evaluate our hospitalist practice, recommended reducing this dispersal to improve physician productivity, hospital efficiency, and outcomes of care. We therefore conducted a project to evaluate the impact of localizing general medical inpatient teams to a single nursing unit.

METHODS

RESULTS

DISCUSSION

We found that general medical patients admitted to H‐PA teams and localized to a single nursing unit had similar risk of 30‐day readmission and charges, but may have had a higher length of stay compared to historical and concurrent controls. The localized teams received far fewer pages, had more patient encounters, generated more RVUs, and walked less during their work day. Taken together, these findings imply that in our study, localization led to greater team productivity and a possible decrease in hospital efficiency, with no significant impact on readmissions or charges incurred.

The higher productivity was likely mediated by the preferential assignments of more patients to the localized teams, and improvements in workflow (such as fewer pages and fewer steps walked), which allowed them to provide more care with the same resources as the control teams. Kocher and Sahni7 recently pointed out that the healthcare sector has experienced no gains in labor productivity in the past 20 years. Our intervention fits their prescription for redesigning healthcare delivery models to achieve higher productivity.

The possibility of a higher LOS associated with localization was a counterintuitive finding, and similar to that reported by Roy et al.5 We propose 3 hypotheses to explain this:

• 1

  Selection bias: Higher workload of the localized teams led to compromised efficiency and a higher length of stay (eg, localized teams had fewer observation admissions, more hospitalizations with an ICU stay, and the AMO was asked to preferentially admit patients to localized teams).

• 2

  Localization provided teams the opportunity to spend more time with their patients (by decreasing nonvalue‐added tasks) and to consequently address more issues before transitioning to outpatient care, or to provide higher quality of care.

• 3

  Gaming: By having a hard cap on total number of occupied beds, we provided a perverse incentive to the localized teams to retain patients longer to keep assigned beds occupied, thereby delaying new admissions to avoid higher workload.

Our study cannot tell us which of these hypotheses represents the dominant phenomenon that led to this surprising finding. Hypothesis 3 is most worrying, and we suggest that others looking to localize their medical teams consider the possibility of unintended perverse incentives.

Differences were more pronounced between the historical control group and the intervention group, as opposed to the intervention group and concurrent controls. This may have occurred if we contaminated the concurrent control by decreasing the number of units they had to go to, by sequestering 1 unit for the intervention team.

Our report has limitations. It is a nonrandomized, quasi‐experimental investigation using a single institution's administrative databases. Our intervention was small in scale (localizing 2 out of 10 general medical teams on 1 out of 14 nursing units). What impact a wider implementation of localization may have on emergency department throughput and hospital occupancy remains to be studied. Nevertheless, our research is the first report, to our knowledge, investigating a wide variety of outcomes of localizing inpatient medical teams, and adds significantly to the limited research on this topic. It also provides significant operational details for other institutions to use when localizing medical teams.

We conclude that our intervention of localization of medical teams to a single nursing unit led to higher productivity and better workflow, but did not impact readmissions or charges incurred. We caution others designing similar localization interventions to protect against possible perverse incentives for inefficient care.

Localizing inpatient general medical teams to nursing units has high intuitive validity for improving physician productivity, hospital efficiency, and patient outcomes. Motion or the moving of personnel between tasksso prominent if teams are not localizedis 1 of the 7 wastes in lean thinking.1 In a timemotion study, where hospitalists cared for patients on up to 5 different wards, O'Leary et al2 have reported large parts of hospitalists' workdays spent in indirect patient care (69%), paging (13%), and travel (3%). Localization could increase the amount of time available for direct patient care, decrease time spent for (and interruptions due to) paging, and decrease travel time, all leading to greater productivity.

O'Leary et al3 have also reported the beneficial effects of localization of medical inpatients on communication between nurses and physicians, who could identify each other more often, and reported greater communication (specifically face‐to‐face communication) with each other following localization. This improvement in communication and effective multidisciplinary rounds could lead to safer care4 and better outcomes.

Further investigations about the effect of localization are limited. Roy et al5 have compared the outcomes of patients localized to 2 inpatient pods medically staffed by hospitalists and physician assistants (PAs) to geographically dispersed, but structurally different, house staff teams. They noticed significantly lower costs, slight but nonsignificant increase in length of stay, and no difference in mortality or readmissions, but it is impossible to tease out the affect of localization versus the affect of team composition. In a before‐and‐after study, Findlay et al6 have reported a decrease in mortality and complication rates in clinically homogenous surgical patients (proximal hip fractures) when cared for by junior trainee physicians localized to a unit, but their experience cannot be extrapolated to the much more diverse general medical population.

In our hospital, each general medical team could admit patients dispersed over 14 different units. An internal group, commissioned to evaluate our hospitalist practice, recommended reducing this dispersal to improve physician productivity, hospital efficiency, and outcomes of care. We therefore conducted a project to evaluate the impact of localizing general medical inpatient teams to a single nursing unit.

METHODS

RESULTS

DISCUSSION

We found that general medical patients admitted to H‐PA teams and localized to a single nursing unit had similar risk of 30‐day readmission and charges, but may have had a higher length of stay compared to historical and concurrent controls. The localized teams received far fewer pages, had more patient encounters, generated more RVUs, and walked less during their work day. Taken together, these findings imply that in our study, localization led to greater team productivity and a possible decrease in hospital efficiency, with no significant impact on readmissions or charges incurred.

The higher productivity was likely mediated by the preferential assignments of more patients to the localized teams, and improvements in workflow (such as fewer pages and fewer steps walked), which allowed them to provide more care with the same resources as the control teams. Kocher and Sahni7 recently pointed out that the healthcare sector has experienced no gains in labor productivity in the past 20 years. Our intervention fits their prescription for redesigning healthcare delivery models to achieve higher productivity.

The possibility of a higher LOS associated with localization was a counterintuitive finding, and similar to that reported by Roy et al.5 We propose 3 hypotheses to explain this:

• 1

  Selection bias: Higher workload of the localized teams led to compromised efficiency and a higher length of stay (eg, localized teams had fewer observation admissions, more hospitalizations with an ICU stay, and the AMO was asked to preferentially admit patients to localized teams).

• 2

  Localization provided teams the opportunity to spend more time with their patients (by decreasing nonvalue‐added tasks) and to consequently address more issues before transitioning to outpatient care, or to provide higher quality of care.

• 3

  Gaming: By having a hard cap on total number of occupied beds, we provided a perverse incentive to the localized teams to retain patients longer to keep assigned beds occupied, thereby delaying new admissions to avoid higher workload.

Our study cannot tell us which of these hypotheses represents the dominant phenomenon that led to this surprising finding. Hypothesis 3 is most worrying, and we suggest that others looking to localize their medical teams consider the possibility of unintended perverse incentives.

Differences were more pronounced between the historical control group and the intervention group, as opposed to the intervention group and concurrent controls. This may have occurred if we contaminated the concurrent control by decreasing the number of units they had to go to, by sequestering 1 unit for the intervention team.

Our report has limitations. It is a nonrandomized, quasi‐experimental investigation using a single institution's administrative databases. Our intervention was small in scale (localizing 2 out of 10 general medical teams on 1 out of 14 nursing units). What impact a wider implementation of localization may have on emergency department throughput and hospital occupancy remains to be studied. Nevertheless, our research is the first report, to our knowledge, investigating a wide variety of outcomes of localizing inpatient medical teams, and adds significantly to the limited research on this topic. It also provides significant operational details for other institutions to use when localizing medical teams.

We conclude that our intervention of localization of medical teams to a single nursing unit led to higher productivity and better workflow, but did not impact readmissions or charges incurred. We caution others designing similar localization interventions to protect against possible perverse incentives for inefficient care.

Localizing inpatient general medical teams to nursing units has high intuitive validity for improving physician productivity, hospital efficiency, and patient outcomes. Motion or the moving of personnel between tasksso prominent if teams are not localizedis 1 of the 7 wastes in lean thinking.1 In a timemotion study, where hospitalists cared for patients on up to 5 different wards, O'Leary et al2 have reported large parts of hospitalists' workdays spent in indirect patient care (69%), paging (13%), and travel (3%). Localization could increase the amount of time available for direct patient care, decrease time spent for (and interruptions due to) paging, and decrease travel time, all leading to greater productivity.

O'Leary et al3 have also reported the beneficial effects of localization of medical inpatients on communication between nurses and physicians, who could identify each other more often, and reported greater communication (specifically face‐to‐face communication) with each other following localization. This improvement in communication and effective multidisciplinary rounds could lead to safer care4 and better outcomes.

Further investigations about the effect of localization are limited. Roy et al5 have compared the outcomes of patients localized to 2 inpatient pods medically staffed by hospitalists and physician assistants (PAs) to geographically dispersed, but structurally different, house staff teams. They noticed significantly lower costs, slight but nonsignificant increase in length of stay, and no difference in mortality or readmissions, but it is impossible to tease out the affect of localization versus the affect of team composition. In a before‐and‐after study, Findlay et al6 have reported a decrease in mortality and complication rates in clinically homogenous surgical patients (proximal hip fractures) when cared for by junior trainee physicians localized to a unit, but their experience cannot be extrapolated to the much more diverse general medical population.

In our hospital, each general medical team could admit patients dispersed over 14 different units. An internal group, commissioned to evaluate our hospitalist practice, recommended reducing this dispersal to improve physician productivity, hospital efficiency, and outcomes of care. We therefore conducted a project to evaluate the impact of localizing general medical inpatient teams to a single nursing unit.

METHODS

RESULTS

DISCUSSION

We found that general medical patients admitted to H‐PA teams and localized to a single nursing unit had similar risk of 30‐day readmission and charges, but may have had a higher length of stay compared to historical and concurrent controls. The localized teams received far fewer pages, had more patient encounters, generated more RVUs, and walked less during their work day. Taken together, these findings imply that in our study, localization led to greater team productivity and a possible decrease in hospital efficiency, with no significant impact on readmissions or charges incurred.

The higher productivity was likely mediated by the preferential assignments of more patients to the localized teams, and improvements in workflow (such as fewer pages and fewer steps walked), which allowed them to provide more care with the same resources as the control teams. Kocher and Sahni7 recently pointed out that the healthcare sector has experienced no gains in labor productivity in the past 20 years. Our intervention fits their prescription for redesigning healthcare delivery models to achieve higher productivity.

The possibility of a higher LOS associated with localization was a counterintuitive finding, and similar to that reported by Roy et al.5 We propose 3 hypotheses to explain this:

• 1

  Selection bias: Higher workload of the localized teams led to compromised efficiency and a higher length of stay (eg, localized teams had fewer observation admissions, more hospitalizations with an ICU stay, and the AMO was asked to preferentially admit patients to localized teams).

• 2

  Localization provided teams the opportunity to spend more time with their patients (by decreasing nonvalue‐added tasks) and to consequently address more issues before transitioning to outpatient care, or to provide higher quality of care.

• 3

  Gaming: By having a hard cap on total number of occupied beds, we provided a perverse incentive to the localized teams to retain patients longer to keep assigned beds occupied, thereby delaying new admissions to avoid higher workload.

Our study cannot tell us which of these hypotheses represents the dominant phenomenon that led to this surprising finding. Hypothesis 3 is most worrying, and we suggest that others looking to localize their medical teams consider the possibility of unintended perverse incentives.

Differences were more pronounced between the historical control group and the intervention group, as opposed to the intervention group and concurrent controls. This may have occurred if we contaminated the concurrent control by decreasing the number of units they had to go to, by sequestering 1 unit for the intervention team.

Our report has limitations. It is a nonrandomized, quasi‐experimental investigation using a single institution's administrative databases. Our intervention was small in scale (localizing 2 out of 10 general medical teams on 1 out of 14 nursing units). What impact a wider implementation of localization may have on emergency department throughput and hospital occupancy remains to be studied. Nevertheless, our research is the first report, to our knowledge, investigating a wide variety of outcomes of localizing inpatient medical teams, and adds significantly to the limited research on this topic. It also provides significant operational details for other institutions to use when localizing medical teams.

We conclude that our intervention of localization of medical teams to a single nursing unit led to higher productivity and better workflow, but did not impact readmissions or charges incurred. We caution others designing similar localization interventions to protect against possible perverse incentives for inefficient care.

Many patients are admitted and readmitted to acute care hospitals with alcohol‐related diagnoses, including alcohol withdrawal syndrome (AWS), and experience significant morbidity and mortality. In patients with septic shock or at risk for acute respiratory distress syndrome (ARDS), chronic alcohol abuse is associated with increased ARDS and severity of multiple organ dysfunction.1 Among intensive care unit (ICU) patients, those with alcohol dependence have higher morbidity, including septic shock, and higher hospital mortality.2 Patients who experience AWS as a result of alcohol dependence may experience life‐threatening complications, such as seizures and delirium tremens.3,4 In‐hospital mortality from AWS is historically high,5 but with benzodiazepines used in a symptom‐driven manner to treat the complications of alcohol use, hospital mortality rates are more recently reported at 2.4%.6

As inpatient outcomes1,2,7 and hospital mortality810 are negatively affected by alcohol abuse, the post‐hospital course of these patients is also of interest. Specifically, patients are often admitted and readmitted with alcohol‐related diagnoses or AWS to acute care hospitals, but relatively little quantitative data exist on readmission factors in this population.11 Patients readmitted to detoxification units or alcohol and substance abuse units have been studied, and factors associated with readmission include psychiatric disorder,1217 female gender,14,15 and delay in rehabilitation aftercare.18

These results cannot be generalized to patients with AWS who are admitted and readmitted to acute‐care hospitals. First, patients hospitalized for alcohol withdrawal symptoms are often medically ill with more severe symptoms, and more frequent coexisting medical and psychiatric illnesses, that complicate the withdrawal syndrome. Detoxification units and substance abuse units require patients to be medically stable before admission, because they do not have the ability to provide a high level of supervision and treatment. Second, much of what we know regarding risk factors for readmission to detoxification centers and substance abuse units comes from studies of administrative data of the Veterans Health Administration,12,13 Medicare Provider Analysis and Review file,16 privately owned outpatient substance abuse centers,14 and publicly funded detoxification centers.18 These results may be difficult to generalize to other patient populations. Accordingly, the objective of this study was to identify demographic and clinical factors associated with multiple admissions to a general medicine service for treatment of AWS over a 3‐year period. Characterization of these high‐risk patients and their hospital course may help focus intervention and reduce these revolving door admissions.

METHODS

The Mayo Clinic Institutional Review Board deemed the study exempt.

Patient Selection

The study was conducted at an 1157‐bed academic tertiary referral hospital, located in the Midwest, that has approximately 15,000 inpatient admissions to general medicine services annually and serves as the main referral center for the region. Patients included in this study were adults admitted to general medicine services and treated with symptom‐triggered Clinical Institute Withdrawal AssessmentAlcohol Revised (CIWA‐Ar) protocol19 between January 1, 2006 and December 31, 2008. Patients were identified using the Mayo Clinic Quality Improvement Resource Center database, as done in a previous CIWA‐Ar study.20 Patients were excluded if the primary diagnosis was a nonalcohol‐related diagnosis (Figure 1).

Figure 1

Patients were placed in 1 of 2 groups based on number of admissions during the study period, either a single‐admission group or a multiple‐admissions group. While most readmission studies use a 30‐day mark from discharge, we used 3 years to better capture relapse and recidivism in this patient population. The 2 groups were then compared retrospectively. To insure that a single admission was not arbitrarily created by the December 2008 cutoff, we reviewed the single‐admission group for additional admissions through June of 2009. If a patient did have a subsequent admission, then the patient was moved to the multiple‐admissions group.

RESULTS

The CIWA‐Ar protocol was ordered on 1199 admissions during the study period. Of these, 411 (34.3%) admissions were excluded because AWS was not the primary diagnosis, leaving 788 (65.7%) admissions for 322 patients, which formed the study population. Of the 322 patients, 180 (56%) had a single admission and 142 (44%) had multiple admissions.

Univariate analyses of demographic and clinical variables are shown in Tables 1 and 2, respectively. Patients with multiple admissions were more likely divorced (P = 0.028), have a high school education or less (P = 0.002), have a higher CCI score (P < 0.0001), a higher CIWA‐Ar score (max) (P < 0.0001), a higher ALT level (P = 0.050), more psychiatric comorbidity (P < 0.026), and a positive urine drug screen (P < 0.001). Adverse events were not significantly different between the 2 groups (Table 2).

Multivariate logistic model analysis was performed using the variables age, male gender, divorced marital status, high school education or less, CIWA‐Ar score (max), CCI score, psychiatric comorbidity, and positive urine drug screen. With a stepwise elimination process, the final model showed that multiple admissions were associated with high school education or less (P = 0.0071), higher CCI score (P = 0.0010), higher CIWA‐Ar score (max) (P < 0.0001), a positive urine drug screen (P = 0.0002), and psychiatric comorbidity (P = 0.0303) (Table 3).

DISCUSSION

We provide important information regarding identification of individuals at high risk for multiple admissions to general medicine services for treatment of AWS. This study found that patients with multiple admissions for AWS had more medical comorbidity. They had more cases of diabetes mellitus, cardiovascular disease, and cerebrovascular disease, and their CCI scores were higher. They also had higher CIWA‐Ar (max) scores, as well as higher CIWA‐Ar component scores, indicating a more severe withdrawal.

Further, psychiatric comorbidity was also associated with multiple admissions. Consistent with the high prevalence in alcoholic patients, psychiatric comorbidity was common in both patients with a single admission and multiple admissions. We also found that a positive urine drug screen was associated with multiple admissions. Interestingly, few patients in each group had a diagnosis recorded in the medical record of an additional substance abuse disorder, yet 14% of patients with a single admission and 29% of patients with multiple admissions had a positive urine drug screen for a non‐alcohol substance. Psychiatric comorbidity, including additional substance abuse, is a well‐established risk factor for readmission to detoxification centers.1215, 17,22,23 Also, people with either an alcohol or non‐alcohol drug addiction, are known to be 7 times more likely to have another addiction than the rest of the population.24 This study suggests clinicians may underrecognize additional substance abuse disorders which are common in this patient population.

In contrast to studies of patients readmitted to detoxification units and substance abuse units,14,15,18,22 we found level of education, specifically a high school education or less, to be associated with multiple admissions. In a study of alcoholics, Greenfield and colleagues found that lower education in alcoholics predicts shorter time to relapse.25 A lack of education may result in inadequate healthcare literacy. Poor health behavior choices that follow may lead to relapse and subsequent admissions for AWS. With respect to other demographic variables, patients in our study population were predominantly men, which is not surprising. Gender differences in alcoholism are well established, with alcohol abuse and dependence more prevalent in men.26 We did not find gender associated with multiple admissions.

Our findings have management and treatment implications. First, providers who care for patients with AWS should not simply focus on treating withdrawal signs and symptoms, but also screen for and address other medical issues, which may not be apparent or seem stable at first glance. While a comorbid medical condition may not be the primary reason for hospital admission, comorbid medical conditions are known to be a source of psychological distress27 and have a negative effect on abstinence. Second, all patients should be screened for additional substance abuse. Initial laboratory testing should include a urine drug screen. Third, before discharging the patient, providers should establish primary care follow‐up for continued surveillance of medical issues. There is evidence that primary care services are predictive of better substance abuse treatment outcomes for those with medical problems.28,29

Finally, inpatient psychiatric consultation, upon admission, is essential for several reasons. First, the psychiatric team can help with initial management and treatment of the alcohol withdrawal regardless of stage and severity, obtain a more comprehensive psychiatric history, and assess for the presence of psychiatric comorbidities that may contribute to, aggravate, or complicate the clinical picture. The team can also address other substance abuse issues when detected by drug screen or clinical history. The psychiatric team, along with chemical‐dependency counselors and social workers, can provide valuable input regarding chemical‐dependency resources available on discharge and help instruct the patient in healthy behaviors. Because healthcare illiteracy may be an issue in this patient population, these instructions should be tailored to the patient's educational level. Prior to discharge, the psychiatry team, social workers, or chemical‐dependency counselors can also assist with, or arrange, rehabilitation aftercare for patients. Recent work shows that patients were less likely to be readmitted to crisis detoxification if they entered rehabilitation care within 3 days of discharge.18

Our study has significant limitations. This study was performed with data at a single academic medical center with an ethnically homogeneous patient population, limiting the external validity of its results. Because this is a retrospective study, data analyses are limited by the quality and accuracy of data in the electronic medical record. Also, our follow‐up period may not have been long enough to detect additional admissions, and we did not screen for patient admissions prior to the study period. By limiting data collection to admissions for AWS to general medical services, we may have missed cases of AWS when admitted for other reasons or to subspecialty services, and we may have missed severe cases requiring admission to an intensive care unit. While we believe we were able to capture most admissions, we may underreport this number since we cannot account for those events that may have occurred at other facilities and locations. Lastly, without a control group, this study is limited in its ability to show an association between any variable and readmission.

In our study, 142 patients accounted for 608 admissions during the 3‐year study period, which speaks to the high recidivism rates for patients with AWS. This disease is associated with high morbidity, high medical costs, and high utilization of healthcare. Our study provides insight regarding identification of patients at high risk of multiple admissions with respect to demographic (lower level of education) and clinical characteristics (worse withdrawal severity, more medical and psychiatric comorbidity, and polysubstance abuse). We believe collaboration between social services, chemical‐dependency counselors, psychiatry, and medicine is necessary to effectively treat this population of patients and assist with the crucial transition to the outpatient setting. Future studies should include key social factors, such as health literacy in the readmission risk assessment, as well as primary care follow‐up and rehabilitation aftercare.

Many patients are admitted and readmitted to acute care hospitals with alcohol‐related diagnoses, including alcohol withdrawal syndrome (AWS), and experience significant morbidity and mortality. In patients with septic shock or at risk for acute respiratory distress syndrome (ARDS), chronic alcohol abuse is associated with increased ARDS and severity of multiple organ dysfunction.1 Among intensive care unit (ICU) patients, those with alcohol dependence have higher morbidity, including septic shock, and higher hospital mortality.2 Patients who experience AWS as a result of alcohol dependence may experience life‐threatening complications, such as seizures and delirium tremens.3,4 In‐hospital mortality from AWS is historically high,5 but with benzodiazepines used in a symptom‐driven manner to treat the complications of alcohol use, hospital mortality rates are more recently reported at 2.4%.6

As inpatient outcomes1,2,7 and hospital mortality810 are negatively affected by alcohol abuse, the post‐hospital course of these patients is also of interest. Specifically, patients are often admitted and readmitted with alcohol‐related diagnoses or AWS to acute care hospitals, but relatively little quantitative data exist on readmission factors in this population.11 Patients readmitted to detoxification units or alcohol and substance abuse units have been studied, and factors associated with readmission include psychiatric disorder,1217 female gender,14,15 and delay in rehabilitation aftercare.18

These results cannot be generalized to patients with AWS who are admitted and readmitted to acute‐care hospitals. First, patients hospitalized for alcohol withdrawal symptoms are often medically ill with more severe symptoms, and more frequent coexisting medical and psychiatric illnesses, that complicate the withdrawal syndrome. Detoxification units and substance abuse units require patients to be medically stable before admission, because they do not have the ability to provide a high level of supervision and treatment. Second, much of what we know regarding risk factors for readmission to detoxification centers and substance abuse units comes from studies of administrative data of the Veterans Health Administration,12,13 Medicare Provider Analysis and Review file,16 privately owned outpatient substance abuse centers,14 and publicly funded detoxification centers.18 These results may be difficult to generalize to other patient populations. Accordingly, the objective of this study was to identify demographic and clinical factors associated with multiple admissions to a general medicine service for treatment of AWS over a 3‐year period. Characterization of these high‐risk patients and their hospital course may help focus intervention and reduce these revolving door admissions.

METHODS

The Mayo Clinic Institutional Review Board deemed the study exempt.

Patient Selection

The study was conducted at an 1157‐bed academic tertiary referral hospital, located in the Midwest, that has approximately 15,000 inpatient admissions to general medicine services annually and serves as the main referral center for the region. Patients included in this study were adults admitted to general medicine services and treated with symptom‐triggered Clinical Institute Withdrawal AssessmentAlcohol Revised (CIWA‐Ar) protocol19 between January 1, 2006 and December 31, 2008. Patients were identified using the Mayo Clinic Quality Improvement Resource Center database, as done in a previous CIWA‐Ar study.20 Patients were excluded if the primary diagnosis was a nonalcohol‐related diagnosis (Figure 1).

Figure 1

Patients were placed in 1 of 2 groups based on number of admissions during the study period, either a single‐admission group or a multiple‐admissions group. While most readmission studies use a 30‐day mark from discharge, we used 3 years to better capture relapse and recidivism in this patient population. The 2 groups were then compared retrospectively. To insure that a single admission was not arbitrarily created by the December 2008 cutoff, we reviewed the single‐admission group for additional admissions through June of 2009. If a patient did have a subsequent admission, then the patient was moved to the multiple‐admissions group.

RESULTS

The CIWA‐Ar protocol was ordered on 1199 admissions during the study period. Of these, 411 (34.3%) admissions were excluded because AWS was not the primary diagnosis, leaving 788 (65.7%) admissions for 322 patients, which formed the study population. Of the 322 patients, 180 (56%) had a single admission and 142 (44%) had multiple admissions.

Univariate analyses of demographic and clinical variables are shown in Tables 1 and 2, respectively. Patients with multiple admissions were more likely divorced (P = 0.028), have a high school education or less (P = 0.002), have a higher CCI score (P < 0.0001), a higher CIWA‐Ar score (max) (P < 0.0001), a higher ALT level (P = 0.050), more psychiatric comorbidity (P < 0.026), and a positive urine drug screen (P < 0.001). Adverse events were not significantly different between the 2 groups (Table 2).

Multivariate logistic model analysis was performed using the variables age, male gender, divorced marital status, high school education or less, CIWA‐Ar score (max), CCI score, psychiatric comorbidity, and positive urine drug screen. With a stepwise elimination process, the final model showed that multiple admissions were associated with high school education or less (P = 0.0071), higher CCI score (P = 0.0010), higher CIWA‐Ar score (max) (P < 0.0001), a positive urine drug screen (P = 0.0002), and psychiatric comorbidity (P = 0.0303) (Table 3).

DISCUSSION

We provide important information regarding identification of individuals at high risk for multiple admissions to general medicine services for treatment of AWS. This study found that patients with multiple admissions for AWS had more medical comorbidity. They had more cases of diabetes mellitus, cardiovascular disease, and cerebrovascular disease, and their CCI scores were higher. They also had higher CIWA‐Ar (max) scores, as well as higher CIWA‐Ar component scores, indicating a more severe withdrawal.

Further, psychiatric comorbidity was also associated with multiple admissions. Consistent with the high prevalence in alcoholic patients, psychiatric comorbidity was common in both patients with a single admission and multiple admissions. We also found that a positive urine drug screen was associated with multiple admissions. Interestingly, few patients in each group had a diagnosis recorded in the medical record of an additional substance abuse disorder, yet 14% of patients with a single admission and 29% of patients with multiple admissions had a positive urine drug screen for a non‐alcohol substance. Psychiatric comorbidity, including additional substance abuse, is a well‐established risk factor for readmission to detoxification centers.1215, 17,22,23 Also, people with either an alcohol or non‐alcohol drug addiction, are known to be 7 times more likely to have another addiction than the rest of the population.24 This study suggests clinicians may underrecognize additional substance abuse disorders which are common in this patient population.

In contrast to studies of patients readmitted to detoxification units and substance abuse units,14,15,18,22 we found level of education, specifically a high school education or less, to be associated with multiple admissions. In a study of alcoholics, Greenfield and colleagues found that lower education in alcoholics predicts shorter time to relapse.25 A lack of education may result in inadequate healthcare literacy. Poor health behavior choices that follow may lead to relapse and subsequent admissions for AWS. With respect to other demographic variables, patients in our study population were predominantly men, which is not surprising. Gender differences in alcoholism are well established, with alcohol abuse and dependence more prevalent in men.26 We did not find gender associated with multiple admissions.

Our findings have management and treatment implications. First, providers who care for patients with AWS should not simply focus on treating withdrawal signs and symptoms, but also screen for and address other medical issues, which may not be apparent or seem stable at first glance. While a comorbid medical condition may not be the primary reason for hospital admission, comorbid medical conditions are known to be a source of psychological distress27 and have a negative effect on abstinence. Second, all patients should be screened for additional substance abuse. Initial laboratory testing should include a urine drug screen. Third, before discharging the patient, providers should establish primary care follow‐up for continued surveillance of medical issues. There is evidence that primary care services are predictive of better substance abuse treatment outcomes for those with medical problems.28,29

Finally, inpatient psychiatric consultation, upon admission, is essential for several reasons. First, the psychiatric team can help with initial management and treatment of the alcohol withdrawal regardless of stage and severity, obtain a more comprehensive psychiatric history, and assess for the presence of psychiatric comorbidities that may contribute to, aggravate, or complicate the clinical picture. The team can also address other substance abuse issues when detected by drug screen or clinical history. The psychiatric team, along with chemical‐dependency counselors and social workers, can provide valuable input regarding chemical‐dependency resources available on discharge and help instruct the patient in healthy behaviors. Because healthcare illiteracy may be an issue in this patient population, these instructions should be tailored to the patient's educational level. Prior to discharge, the psychiatry team, social workers, or chemical‐dependency counselors can also assist with, or arrange, rehabilitation aftercare for patients. Recent work shows that patients were less likely to be readmitted to crisis detoxification if they entered rehabilitation care within 3 days of discharge.18

Our study has significant limitations. This study was performed with data at a single academic medical center with an ethnically homogeneous patient population, limiting the external validity of its results. Because this is a retrospective study, data analyses are limited by the quality and accuracy of data in the electronic medical record. Also, our follow‐up period may not have been long enough to detect additional admissions, and we did not screen for patient admissions prior to the study period. By limiting data collection to admissions for AWS to general medical services, we may have missed cases of AWS when admitted for other reasons or to subspecialty services, and we may have missed severe cases requiring admission to an intensive care unit. While we believe we were able to capture most admissions, we may underreport this number since we cannot account for those events that may have occurred at other facilities and locations. Lastly, without a control group, this study is limited in its ability to show an association between any variable and readmission.

In our study, 142 patients accounted for 608 admissions during the 3‐year study period, which speaks to the high recidivism rates for patients with AWS. This disease is associated with high morbidity, high medical costs, and high utilization of healthcare. Our study provides insight regarding identification of patients at high risk of multiple admissions with respect to demographic (lower level of education) and clinical characteristics (worse withdrawal severity, more medical and psychiatric comorbidity, and polysubstance abuse). We believe collaboration between social services, chemical‐dependency counselors, psychiatry, and medicine is necessary to effectively treat this population of patients and assist with the crucial transition to the outpatient setting. Future studies should include key social factors, such as health literacy in the readmission risk assessment, as well as primary care follow‐up and rehabilitation aftercare.

Many patients are admitted and readmitted to acute care hospitals with alcohol‐related diagnoses, including alcohol withdrawal syndrome (AWS), and experience significant morbidity and mortality. In patients with septic shock or at risk for acute respiratory distress syndrome (ARDS), chronic alcohol abuse is associated with increased ARDS and severity of multiple organ dysfunction.1 Among intensive care unit (ICU) patients, those with alcohol dependence have higher morbidity, including septic shock, and higher hospital mortality.2 Patients who experience AWS as a result of alcohol dependence may experience life‐threatening complications, such as seizures and delirium tremens.3,4 In‐hospital mortality from AWS is historically high,5 but with benzodiazepines used in a symptom‐driven manner to treat the complications of alcohol use, hospital mortality rates are more recently reported at 2.4%.6

As inpatient outcomes1,2,7 and hospital mortality810 are negatively affected by alcohol abuse, the post‐hospital course of these patients is also of interest. Specifically, patients are often admitted and readmitted with alcohol‐related diagnoses or AWS to acute care hospitals, but relatively little quantitative data exist on readmission factors in this population.11 Patients readmitted to detoxification units or alcohol and substance abuse units have been studied, and factors associated with readmission include psychiatric disorder,1217 female gender,14,15 and delay in rehabilitation aftercare.18

These results cannot be generalized to patients with AWS who are admitted and readmitted to acute‐care hospitals. First, patients hospitalized for alcohol withdrawal symptoms are often medically ill with more severe symptoms, and more frequent coexisting medical and psychiatric illnesses, that complicate the withdrawal syndrome. Detoxification units and substance abuse units require patients to be medically stable before admission, because they do not have the ability to provide a high level of supervision and treatment. Second, much of what we know regarding risk factors for readmission to detoxification centers and substance abuse units comes from studies of administrative data of the Veterans Health Administration,12,13 Medicare Provider Analysis and Review file,16 privately owned outpatient substance abuse centers,14 and publicly funded detoxification centers.18 These results may be difficult to generalize to other patient populations. Accordingly, the objective of this study was to identify demographic and clinical factors associated with multiple admissions to a general medicine service for treatment of AWS over a 3‐year period. Characterization of these high‐risk patients and their hospital course may help focus intervention and reduce these revolving door admissions.

METHODS

The Mayo Clinic Institutional Review Board deemed the study exempt.

Patient Selection

The study was conducted at an 1157‐bed academic tertiary referral hospital, located in the Midwest, that has approximately 15,000 inpatient admissions to general medicine services annually and serves as the main referral center for the region. Patients included in this study were adults admitted to general medicine services and treated with symptom‐triggered Clinical Institute Withdrawal AssessmentAlcohol Revised (CIWA‐Ar) protocol19 between January 1, 2006 and December 31, 2008. Patients were identified using the Mayo Clinic Quality Improvement Resource Center database, as done in a previous CIWA‐Ar study.20 Patients were excluded if the primary diagnosis was a nonalcohol‐related diagnosis (Figure 1).

Figure 1

Patients were placed in 1 of 2 groups based on number of admissions during the study period, either a single‐admission group or a multiple‐admissions group. While most readmission studies use a 30‐day mark from discharge, we used 3 years to better capture relapse and recidivism in this patient population. The 2 groups were then compared retrospectively. To insure that a single admission was not arbitrarily created by the December 2008 cutoff, we reviewed the single‐admission group for additional admissions through June of 2009. If a patient did have a subsequent admission, then the patient was moved to the multiple‐admissions group.

RESULTS

The CIWA‐Ar protocol was ordered on 1199 admissions during the study period. Of these, 411 (34.3%) admissions were excluded because AWS was not the primary diagnosis, leaving 788 (65.7%) admissions for 322 patients, which formed the study population. Of the 322 patients, 180 (56%) had a single admission and 142 (44%) had multiple admissions.

Univariate analyses of demographic and clinical variables are shown in Tables 1 and 2, respectively. Patients with multiple admissions were more likely divorced (P = 0.028), have a high school education or less (P = 0.002), have a higher CCI score (P < 0.0001), a higher CIWA‐Ar score (max) (P < 0.0001), a higher ALT level (P = 0.050), more psychiatric comorbidity (P < 0.026), and a positive urine drug screen (P < 0.001). Adverse events were not significantly different between the 2 groups (Table 2).

Multivariate logistic model analysis was performed using the variables age, male gender, divorced marital status, high school education or less, CIWA‐Ar score (max), CCI score, psychiatric comorbidity, and positive urine drug screen. With a stepwise elimination process, the final model showed that multiple admissions were associated with high school education or less (P = 0.0071), higher CCI score (P = 0.0010), higher CIWA‐Ar score (max) (P < 0.0001), a positive urine drug screen (P = 0.0002), and psychiatric comorbidity (P = 0.0303) (Table 3).

DISCUSSION

We provide important information regarding identification of individuals at high risk for multiple admissions to general medicine services for treatment of AWS. This study found that patients with multiple admissions for AWS had more medical comorbidity. They had more cases of diabetes mellitus, cardiovascular disease, and cerebrovascular disease, and their CCI scores were higher. They also had higher CIWA‐Ar (max) scores, as well as higher CIWA‐Ar component scores, indicating a more severe withdrawal.

Further, psychiatric comorbidity was also associated with multiple admissions. Consistent with the high prevalence in alcoholic patients, psychiatric comorbidity was common in both patients with a single admission and multiple admissions. We also found that a positive urine drug screen was associated with multiple admissions. Interestingly, few patients in each group had a diagnosis recorded in the medical record of an additional substance abuse disorder, yet 14% of patients with a single admission and 29% of patients with multiple admissions had a positive urine drug screen for a non‐alcohol substance. Psychiatric comorbidity, including additional substance abuse, is a well‐established risk factor for readmission to detoxification centers.1215, 17,22,23 Also, people with either an alcohol or non‐alcohol drug addiction, are known to be 7 times more likely to have another addiction than the rest of the population.24 This study suggests clinicians may underrecognize additional substance abuse disorders which are common in this patient population.

In contrast to studies of patients readmitted to detoxification units and substance abuse units,14,15,18,22 we found level of education, specifically a high school education or less, to be associated with multiple admissions. In a study of alcoholics, Greenfield and colleagues found that lower education in alcoholics predicts shorter time to relapse.25 A lack of education may result in inadequate healthcare literacy. Poor health behavior choices that follow may lead to relapse and subsequent admissions for AWS. With respect to other demographic variables, patients in our study population were predominantly men, which is not surprising. Gender differences in alcoholism are well established, with alcohol abuse and dependence more prevalent in men.26 We did not find gender associated with multiple admissions.

Our findings have management and treatment implications. First, providers who care for patients with AWS should not simply focus on treating withdrawal signs and symptoms, but also screen for and address other medical issues, which may not be apparent or seem stable at first glance. While a comorbid medical condition may not be the primary reason for hospital admission, comorbid medical conditions are known to be a source of psychological distress27 and have a negative effect on abstinence. Second, all patients should be screened for additional substance abuse. Initial laboratory testing should include a urine drug screen. Third, before discharging the patient, providers should establish primary care follow‐up for continued surveillance of medical issues. There is evidence that primary care services are predictive of better substance abuse treatment outcomes for those with medical problems.28,29

Finally, inpatient psychiatric consultation, upon admission, is essential for several reasons. First, the psychiatric team can help with initial management and treatment of the alcohol withdrawal regardless of stage and severity, obtain a more comprehensive psychiatric history, and assess for the presence of psychiatric comorbidities that may contribute to, aggravate, or complicate the clinical picture. The team can also address other substance abuse issues when detected by drug screen or clinical history. The psychiatric team, along with chemical‐dependency counselors and social workers, can provide valuable input regarding chemical‐dependency resources available on discharge and help instruct the patient in healthy behaviors. Because healthcare illiteracy may be an issue in this patient population, these instructions should be tailored to the patient's educational level. Prior to discharge, the psychiatry team, social workers, or chemical‐dependency counselors can also assist with, or arrange, rehabilitation aftercare for patients. Recent work shows that patients were less likely to be readmitted to crisis detoxification if they entered rehabilitation care within 3 days of discharge.18

Our study has significant limitations. This study was performed with data at a single academic medical center with an ethnically homogeneous patient population, limiting the external validity of its results. Because this is a retrospective study, data analyses are limited by the quality and accuracy of data in the electronic medical record. Also, our follow‐up period may not have been long enough to detect additional admissions, and we did not screen for patient admissions prior to the study period. By limiting data collection to admissions for AWS to general medical services, we may have missed cases of AWS when admitted for other reasons or to subspecialty services, and we may have missed severe cases requiring admission to an intensive care unit. While we believe we were able to capture most admissions, we may underreport this number since we cannot account for those events that may have occurred at other facilities and locations. Lastly, without a control group, this study is limited in its ability to show an association between any variable and readmission.

In our study, 142 patients accounted for 608 admissions during the 3‐year study period, which speaks to the high recidivism rates for patients with AWS. This disease is associated with high morbidity, high medical costs, and high utilization of healthcare. Our study provides insight regarding identification of patients at high risk of multiple admissions with respect to demographic (lower level of education) and clinical characteristics (worse withdrawal severity, more medical and psychiatric comorbidity, and polysubstance abuse). We believe collaboration between social services, chemical‐dependency counselors, psychiatry, and medicine is necessary to effectively treat this population of patients and assist with the crucial transition to the outpatient setting. Future studies should include key social factors, such as health literacy in the readmission risk assessment, as well as primary care follow‐up and rehabilitation aftercare.

Hospitalists are physicians whose primary focus is the general medical care of hospitalized patients. Hospitalists are uniquely positioned to implement strategies to improve patient flow and efficiency.1 With emergency department (ED) diversion reaching rates upward of 70%, lack of access to inpatient beds leads to delayed care with worsened outcomes.25

To improve access to hospital beds, hospitals may increase capacity by either adding beds or by more efficiently using existing beds. Operations management principles have been applied to healthcare to ensure efficient use of beds. These include: reducing variability of scheduled admissions, remeasuring length of stay (LOS) and bed demand after implementing strategies to reduce practice variation, and employing queuing theory to generate predictions of optimal beds needed.6 The Joint Commission implemented a leadership standard (LD 04.03.11) that hospitals develop and implement plans to identify and mitigate impediments to efficient patient flow through the hospital.

To improve access, hospital leaders expect hospitalists to staff in inpatient medicine programs, surgical comanagement, short stay and chest pain units, and active bed management.7 In the following review, we define hospitalists' roles in the aforementioned programs and their effect on patient flow. We also touch on preoperative clinics, palliative care, geographic rounding, and flexible staffing models.

ACUTE INPATIENT CARE

Hospitalists are one of the fastest growing physician groups in the United States.810 Hospitalists improve efficiency and quality of care across a variety of demographic, geographic, and healthcare settings.11, 12 A 2002 retrospective cohort study in a community‐based urban teaching hospital showed that hospitalists decreased LOS by 0.61 days and lowered risk for death in the hospital (adjusted relative hazard, 0.71; 95% confidence interval [CI], 0.540.93).13 A 2004 prospective quasi‐experimental observational study done at an academic teaching hospital showed an adjusted LOS that was 16.2% lower, and adjusted cost 9.7% lower, for patients on the hospitalists' service.14 In 2007, Lindenauer and colleagues found that a national sample of hospitalists decreased LOS by 0.4 days and lowered cost by $286 per patient.15 The findings of these individual studies were supported in a 2009 systematic review of 33 studies by Peterson which showed that hospitalists decrease LOS.16 In a recent study, Kuo and Goodwin showed that while hospitalists decrease LOS and cost, the patients they care for have higher Medicare costs after discharge by $322 per patient, and are more likely to be readmitted (odds ratio, 1.08; CI, 1.041.14).17

The hospitalist model of care continues to grow, and hospitalists will soon number as many as 30,000.18 For acute medical inpatients, the evidence suggests that hospitalists improve patient flow by decreasing LOS while improving other aspects of quality of care. However, Kuo and Goodwin's findings suggest that the transition of care from inpatient to outpatient settings still requires attention.17

SURGICAL COMANAGEMENT

The Society of Hospital Medicine (SHM) core competencies include perioperative medicine.19, 20 In the 2006 SHM national survey, 85% of hospital medicine groups indicated that they participated in surgical comanagement.21

Hospitalists have improved patient flow and outcomes for orthopedic patients. Hospitalist management of hip fracture patients decreases time to surgery and LOS compared to standard care.2224 Phy and colleagues studied 466 patients for 2 years after the inception of hospital medicine comanagement of surgical patients, and found that care by hospitalists decreased LOS by 2.2 days.22 In a retrospective study of 118 patients, Roy and colleagues found that hospitalist‐managed patients had shorter time to consultation and surgery, decreased LOS, and lower costs.23 In a retrospective cohort study, Batsis looked at mortality in 466 patients with hip fracture, and found no difference between hospitalist management and standard care.24 In patients undergoing elective hip and knee arthroplasty, Huddleston and colleagues reported that patients managed by hospitalists had fewer complications and shorter LOS. The nurses and orthopedic surgeons preferred the hospitalistorthopedist comanagement model.25

The benefits of hospitalist comanagement are not limited to adult patients undergoing orthopedic surgery. For high‐risk patients undergoing lower extremity reconstruction surgery, Pinzur and colleagues noted that LOS was shorter for a cohort of patients managed by hospitalists than for a group of historical controls not treated by hospitalists.26 Simon and colleagues studied comanagement for pediatric spinal fusion patients, and found a decrease in LOS from 6.5 to 4.8 days.27

Several factors should be considered in developing and implementing a successful comanagement program. Since comanagement duties may fall upon hospitalists in order to protect surgeons' time,28 hospital medicine groups should ensure adequate staffing prior to taking on additional services. Clear guidelines to delineate roles and responsibilities of the comanaging groups also need to be developed and implemented.29, 30

Comanaging may also involve additional training. Hospitalists who manage neurologic, neurosurgical, trauma, and psychiatric patients report being undertrained for such conditions.31, 32 Hospital medicine groups need to ensure training needs are met and supported. Given the successes of comanagement and the increasing complexity of surgical patients,33 this practice will likely expand to a greater variety of non‐medical patients.

SHORT STAY UNITS

In 2003, short stay units (SSU) were present in approximately 20% of US hospitals, with 11% of hospitals planning on opening one in the next year.34 SSU are designed to manage acute, self‐limited medical conditions that require brief staysusually less than 72 hours. Approximately 80% of SSU patients are discharged home, avoiding hospitalization.35 Historically, SSU have been under the domain of the ED; however, there is an emerging role for hospitalist‐run SSU.36

Despite demand for SSU, little research has been performed on hospitalist‐led SSU. In 2000, Abenhaim and colleagues showed that a hospitalist‐run SSU at a university‐affiliated teaching hospital had a shorter LOS and lower rates of complications and readmissions when compared to medicine teaching services.37 In 2008, Northwestern Memorial Hospital opened a 30‐bed hospitalist‐run SSU; for those patients, LOS decreased by 2 days.38 In 2010, Leykum and colleagues showed that a hospitalist‐run observation unit can decrease LOS from 2.4 days to 2.2 days.39 Careful selection of SSU patients is needed to obtain these results. Lucas and colleagues found that whether or not SSU patients required assistance of specialists was the strongest predictor of unsuccessful stays (>72 hours or inpatient conversion) in SSU.36

Whether SSU are run by hospital medicine or emergency medicine is decided at an institutional level. Location of SSU in a specifically designated area is crucial, as it allows physicians to round efficiently on patients and to work with staff trained in observation services. Development of admission criteria that include specific diagnoses which match hospitalists' scope of practice is also important (Table 1).32

The protocol‐based and diagnosis‐specific nature of SSU may enhance quality of care through standardization. Future research may delineate the utility of SSU.

CHEST PAIN UNITS

In the United States, in 2004, approximately 6 million patients present annually to EDs with chest pain.40 Cost of care of patients unnecessarily admitted to coronary care units has been estimated to be nearly $3 billion annually.41 Still, as many as 3% of patients with acute myocardial infarction are discharged home.42 Chest pain units (CPU) were developed to facilitate evaluation of patients with chest pain, at low risk for acute coronary syndrome, without requiring inpatient admission. A number of studies have suggested that admission to a CPU is a safe and cost‐effective alternative to hospital admission.4348

CPU have traditionally been staffed by ED physicians and/or cardiologists. In a prepost study, Krantz and colleagues found that a CPU model, incorporating hospitalists at an academic public safety‐net hospital, decreased ED LOS with no difference in 30‐day cardiac event rate.49 Myers and colleagues created a hospitalist‐directed nonteaching service in an academic medical center to admit low‐risk chest pain patients. Patients admitted to the hospitalist service had a statistically significant lower median LOS (23 hours vs 33 hours) and approximately half the median hospital charges than those admitted to teaching services.50 At the same academic medical center, Bayley and colleagues showed that 91% of patients admitted for chest pain waited more than 3 hours for a bed. This adversely affected ED revenue by tying up beds, resulting in an estimated annual loss of $168,300 of hospital revenue. Creation of a hospitalist‐managed service for low‐acuity chest pain patients reduced hospital LOS by 7 hours.51 Somekh and colleagues demonstrated that a protocol‐driven, cardiologist‐run CPU results in a decreased LOS and readmission rate compared to usual care.52 In a non‐peer reviewed case study, Cox Health opened an 8‐bed, hospitalist‐led CPU in 2003. They decreased LOS from 72 to 18 hours, while increasing revenue by $2.5 million a year.53 These studies suggest that hospitalist‐run CPU can decrease LOS, increase revenue, and relieve ED overcrowding.

Development of a successful CPU depends upon clear inclusion/exclusion criteria; close collaboration among ED physicians, hospitalists, and cardiologists; the development of evidence‐based protocols, and the availability of stress testing.

ACTIVE BED MANAGEMENT

As of 2007, 90% of EDs were crowded beyond their capacity.2 ED crowding leads to ambulance diversion,54 which can delay care and increase mortality rates.55 One of the main causes of ED crowding is the boarding of admitted patients.56 Boarded, admitted patients have been shown to have decreased quality of care and patient satisfaction.35

Active bed management (ABM) by hospitalists can decrease ED diversion. Howell and colleagues instituted ABM where hospitalists, as active bed managers, facilitate placement of patients to their inpatient destinations to assist ED flow.57 This 24‐hour, hospitalist‐led, active bed management service decreased both ED LOS and ambulance diversion. The bed manager collaborated real‐time with medicine and ED attending physicians, nursing supervisors, and charge nurses to change patient care status, and assign and facilitate transfer of patients to appropriate units. These hospitalist bed managers were also empowered to activate additional resources when pre‐diversion rounds identified resource limitations and impending ED divert. They found overall ED LOS for admitted patients decreased by 98 minutes, while LOS for non‐admitted patients stayed the same. AMB decreased diversion due to critically ill and telemetry patients by 28% (786 hours), and diversion due to lower acuity patients by 6% (182 hours). This intervention proved cost‐effective. Three full‐time equivalent (FTE) hospitalists' salaries staff 1 active bed manager working 24/7. Nearly 1000 hours of diversion were avoided at an annual savings of $1086 per hour of diversion decreased.

ABM is a new frontier for hospitals in general, and hospitalists in particular. Chadaga and colleagues found that a hospital medicine‐ED team participating in active bed management, while caring for admitted patients boarded in the ED, can decrease ED diversion and improve patient flow. The percentage of patients transferred to a medicine floor and discharged within 8 hours was reduced by 67% (P 0.01), while the number of discharges from the ED of admitted medicine patients increased by 61% (P 0.001).58

To decrease initial investment, components of ABM (ED triage, bed assignment, discharge facilitation) can be instituted in parts. Hospital medicine groups with limited resources may only provide a triage service by phone for difficult ED cases. Bedside evaluations and collaboration with nursing staff to improve bed placement may be a next step, with floor and/or intensive care unit (ICU) rounds to facilitate early discharges as a final component.

OTHER AREAS

CONCLUSION

While many hospital medicine groups were started to provide acute inpatient medical care, most have found that their value to hospitals reaches beyond bedside care. With an epidemic of ED diversion and lack of access to hospital beds and services, optimizing throughput has become imperative for hospital systems. While hospital access can be improved with addition of new beds, improving throughput by decreasing LOS maximizes utilization of existing resources.

We have reviewed how hospitalists improve patient flow in acute inpatient medicine, surgical comanagement, short stay units, chest pain units, and active bed management. In each instance, the literature supports measures for decreasing LOS while maintaining or improving quality of care. Hinami and colleagues showed physician satisfaction with hospitalist‐provided patient care.75 Most studies have been limited by tracking upstream effects of improved efficiency. As there is now some evidence that decreasing LOS may increase readmissions,17 future studies should incorporate this metric into their outcomes. The effect of formal operations management principles on patient flow and bed efficiency is not well known and should be further examined.

In addition, we have touched on other areas (perioperative clinics, palliative care, geographic rounding, and flexible staffing models) where hospitalists may impact patient throughput. These areas represent excellent opportunities for future research.

Hospitalist participation in many of these areas is in its infancy. Hospital medicine programs interested in expanding their services, beyond acute inpatient care, have the opportunity to develop standards and continue research on the effect of hospital medicine‐led services on patient care and flow.

Hospitalists are physicians whose primary focus is the general medical care of hospitalized patients. Hospitalists are uniquely positioned to implement strategies to improve patient flow and efficiency.1 With emergency department (ED) diversion reaching rates upward of 70%, lack of access to inpatient beds leads to delayed care with worsened outcomes.25

To improve access to hospital beds, hospitals may increase capacity by either adding beds or by more efficiently using existing beds. Operations management principles have been applied to healthcare to ensure efficient use of beds. These include: reducing variability of scheduled admissions, remeasuring length of stay (LOS) and bed demand after implementing strategies to reduce practice variation, and employing queuing theory to generate predictions of optimal beds needed.6 The Joint Commission implemented a leadership standard (LD 04.03.11) that hospitals develop and implement plans to identify and mitigate impediments to efficient patient flow through the hospital.

To improve access, hospital leaders expect hospitalists to staff in inpatient medicine programs, surgical comanagement, short stay and chest pain units, and active bed management.7 In the following review, we define hospitalists' roles in the aforementioned programs and their effect on patient flow. We also touch on preoperative clinics, palliative care, geographic rounding, and flexible staffing models.

ACUTE INPATIENT CARE

Hospitalists are one of the fastest growing physician groups in the United States.810 Hospitalists improve efficiency and quality of care across a variety of demographic, geographic, and healthcare settings.11, 12 A 2002 retrospective cohort study in a community‐based urban teaching hospital showed that hospitalists decreased LOS by 0.61 days and lowered risk for death in the hospital (adjusted relative hazard, 0.71; 95% confidence interval [CI], 0.540.93).13 A 2004 prospective quasi‐experimental observational study done at an academic teaching hospital showed an adjusted LOS that was 16.2% lower, and adjusted cost 9.7% lower, for patients on the hospitalists' service.14 In 2007, Lindenauer and colleagues found that a national sample of hospitalists decreased LOS by 0.4 days and lowered cost by $286 per patient.15 The findings of these individual studies were supported in a 2009 systematic review of 33 studies by Peterson which showed that hospitalists decrease LOS.16 In a recent study, Kuo and Goodwin showed that while hospitalists decrease LOS and cost, the patients they care for have higher Medicare costs after discharge by $322 per patient, and are more likely to be readmitted (odds ratio, 1.08; CI, 1.041.14).17

The hospitalist model of care continues to grow, and hospitalists will soon number as many as 30,000.18 For acute medical inpatients, the evidence suggests that hospitalists improve patient flow by decreasing LOS while improving other aspects of quality of care. However, Kuo and Goodwin's findings suggest that the transition of care from inpatient to outpatient settings still requires attention.17

SURGICAL COMANAGEMENT

The Society of Hospital Medicine (SHM) core competencies include perioperative medicine.19, 20 In the 2006 SHM national survey, 85% of hospital medicine groups indicated that they participated in surgical comanagement.21

Hospitalists have improved patient flow and outcomes for orthopedic patients. Hospitalist management of hip fracture patients decreases time to surgery and LOS compared to standard care.2224 Phy and colleagues studied 466 patients for 2 years after the inception of hospital medicine comanagement of surgical patients, and found that care by hospitalists decreased LOS by 2.2 days.22 In a retrospective study of 118 patients, Roy and colleagues found that hospitalist‐managed patients had shorter time to consultation and surgery, decreased LOS, and lower costs.23 In a retrospective cohort study, Batsis looked at mortality in 466 patients with hip fracture, and found no difference between hospitalist management and standard care.24 In patients undergoing elective hip and knee arthroplasty, Huddleston and colleagues reported that patients managed by hospitalists had fewer complications and shorter LOS. The nurses and orthopedic surgeons preferred the hospitalistorthopedist comanagement model.25

The benefits of hospitalist comanagement are not limited to adult patients undergoing orthopedic surgery. For high‐risk patients undergoing lower extremity reconstruction surgery, Pinzur and colleagues noted that LOS was shorter for a cohort of patients managed by hospitalists than for a group of historical controls not treated by hospitalists.26 Simon and colleagues studied comanagement for pediatric spinal fusion patients, and found a decrease in LOS from 6.5 to 4.8 days.27

Several factors should be considered in developing and implementing a successful comanagement program. Since comanagement duties may fall upon hospitalists in order to protect surgeons' time,28 hospital medicine groups should ensure adequate staffing prior to taking on additional services. Clear guidelines to delineate roles and responsibilities of the comanaging groups also need to be developed and implemented.29, 30

Comanaging may also involve additional training. Hospitalists who manage neurologic, neurosurgical, trauma, and psychiatric patients report being undertrained for such conditions.31, 32 Hospital medicine groups need to ensure training needs are met and supported. Given the successes of comanagement and the increasing complexity of surgical patients,33 this practice will likely expand to a greater variety of non‐medical patients.

SHORT STAY UNITS

In 2003, short stay units (SSU) were present in approximately 20% of US hospitals, with 11% of hospitals planning on opening one in the next year.34 SSU are designed to manage acute, self‐limited medical conditions that require brief staysusually less than 72 hours. Approximately 80% of SSU patients are discharged home, avoiding hospitalization.35 Historically, SSU have been under the domain of the ED; however, there is an emerging role for hospitalist‐run SSU.36

Despite demand for SSU, little research has been performed on hospitalist‐led SSU. In 2000, Abenhaim and colleagues showed that a hospitalist‐run SSU at a university‐affiliated teaching hospital had a shorter LOS and lower rates of complications and readmissions when compared to medicine teaching services.37 In 2008, Northwestern Memorial Hospital opened a 30‐bed hospitalist‐run SSU; for those patients, LOS decreased by 2 days.38 In 2010, Leykum and colleagues showed that a hospitalist‐run observation unit can decrease LOS from 2.4 days to 2.2 days.39 Careful selection of SSU patients is needed to obtain these results. Lucas and colleagues found that whether or not SSU patients required assistance of specialists was the strongest predictor of unsuccessful stays (>72 hours or inpatient conversion) in SSU.36

Whether SSU are run by hospital medicine or emergency medicine is decided at an institutional level. Location of SSU in a specifically designated area is crucial, as it allows physicians to round efficiently on patients and to work with staff trained in observation services. Development of admission criteria that include specific diagnoses which match hospitalists' scope of practice is also important (Table 1).32

The protocol‐based and diagnosis‐specific nature of SSU may enhance quality of care through standardization. Future research may delineate the utility of SSU.

CHEST PAIN UNITS

In the United States, in 2004, approximately 6 million patients present annually to EDs with chest pain.40 Cost of care of patients unnecessarily admitted to coronary care units has been estimated to be nearly $3 billion annually.41 Still, as many as 3% of patients with acute myocardial infarction are discharged home.42 Chest pain units (CPU) were developed to facilitate evaluation of patients with chest pain, at low risk for acute coronary syndrome, without requiring inpatient admission. A number of studies have suggested that admission to a CPU is a safe and cost‐effective alternative to hospital admission.4348

CPU have traditionally been staffed by ED physicians and/or cardiologists. In a prepost study, Krantz and colleagues found that a CPU model, incorporating hospitalists at an academic public safety‐net hospital, decreased ED LOS with no difference in 30‐day cardiac event rate.49 Myers and colleagues created a hospitalist‐directed nonteaching service in an academic medical center to admit low‐risk chest pain patients. Patients admitted to the hospitalist service had a statistically significant lower median LOS (23 hours vs 33 hours) and approximately half the median hospital charges than those admitted to teaching services.50 At the same academic medical center, Bayley and colleagues showed that 91% of patients admitted for chest pain waited more than 3 hours for a bed. This adversely affected ED revenue by tying up beds, resulting in an estimated annual loss of $168,300 of hospital revenue. Creation of a hospitalist‐managed service for low‐acuity chest pain patients reduced hospital LOS by 7 hours.51 Somekh and colleagues demonstrated that a protocol‐driven, cardiologist‐run CPU results in a decreased LOS and readmission rate compared to usual care.52 In a non‐peer reviewed case study, Cox Health opened an 8‐bed, hospitalist‐led CPU in 2003. They decreased LOS from 72 to 18 hours, while increasing revenue by $2.5 million a year.53 These studies suggest that hospitalist‐run CPU can decrease LOS, increase revenue, and relieve ED overcrowding.

Development of a successful CPU depends upon clear inclusion/exclusion criteria; close collaboration among ED physicians, hospitalists, and cardiologists; the development of evidence‐based protocols, and the availability of stress testing.

ACTIVE BED MANAGEMENT

As of 2007, 90% of EDs were crowded beyond their capacity.2 ED crowding leads to ambulance diversion,54 which can delay care and increase mortality rates.55 One of the main causes of ED crowding is the boarding of admitted patients.56 Boarded, admitted patients have been shown to have decreased quality of care and patient satisfaction.35

Active bed management (ABM) by hospitalists can decrease ED diversion. Howell and colleagues instituted ABM where hospitalists, as active bed managers, facilitate placement of patients to their inpatient destinations to assist ED flow.57 This 24‐hour, hospitalist‐led, active bed management service decreased both ED LOS and ambulance diversion. The bed manager collaborated real‐time with medicine and ED attending physicians, nursing supervisors, and charge nurses to change patient care status, and assign and facilitate transfer of patients to appropriate units. These hospitalist bed managers were also empowered to activate additional resources when pre‐diversion rounds identified resource limitations and impending ED divert. They found overall ED LOS for admitted patients decreased by 98 minutes, while LOS for non‐admitted patients stayed the same. AMB decreased diversion due to critically ill and telemetry patients by 28% (786 hours), and diversion due to lower acuity patients by 6% (182 hours). This intervention proved cost‐effective. Three full‐time equivalent (FTE) hospitalists' salaries staff 1 active bed manager working 24/7. Nearly 1000 hours of diversion were avoided at an annual savings of $1086 per hour of diversion decreased.

ABM is a new frontier for hospitals in general, and hospitalists in particular. Chadaga and colleagues found that a hospital medicine‐ED team participating in active bed management, while caring for admitted patients boarded in the ED, can decrease ED diversion and improve patient flow. The percentage of patients transferred to a medicine floor and discharged within 8 hours was reduced by 67% (P 0.01), while the number of discharges from the ED of admitted medicine patients increased by 61% (P 0.001).58

To decrease initial investment, components of ABM (ED triage, bed assignment, discharge facilitation) can be instituted in parts. Hospital medicine groups with limited resources may only provide a triage service by phone for difficult ED cases. Bedside evaluations and collaboration with nursing staff to improve bed placement may be a next step, with floor and/or intensive care unit (ICU) rounds to facilitate early discharges as a final component.

OTHER AREAS

CONCLUSION

While many hospital medicine groups were started to provide acute inpatient medical care, most have found that their value to hospitals reaches beyond bedside care. With an epidemic of ED diversion and lack of access to hospital beds and services, optimizing throughput has become imperative for hospital systems. While hospital access can be improved with addition of new beds, improving throughput by decreasing LOS maximizes utilization of existing resources.

We have reviewed how hospitalists improve patient flow in acute inpatient medicine, surgical comanagement, short stay units, chest pain units, and active bed management. In each instance, the literature supports measures for decreasing LOS while maintaining or improving quality of care. Hinami and colleagues showed physician satisfaction with hospitalist‐provided patient care.75 Most studies have been limited by tracking upstream effects of improved efficiency. As there is now some evidence that decreasing LOS may increase readmissions,17 future studies should incorporate this metric into their outcomes. The effect of formal operations management principles on patient flow and bed efficiency is not well known and should be further examined.

In addition, we have touched on other areas (perioperative clinics, palliative care, geographic rounding, and flexible staffing models) where hospitalists may impact patient throughput. These areas represent excellent opportunities for future research.

Hospitalist participation in many of these areas is in its infancy. Hospital medicine programs interested in expanding their services, beyond acute inpatient care, have the opportunity to develop standards and continue research on the effect of hospital medicine‐led services on patient care and flow.

Hospitalists are physicians whose primary focus is the general medical care of hospitalized patients. Hospitalists are uniquely positioned to implement strategies to improve patient flow and efficiency.1 With emergency department (ED) diversion reaching rates upward of 70%, lack of access to inpatient beds leads to delayed care with worsened outcomes.25

To improve access to hospital beds, hospitals may increase capacity by either adding beds or by more efficiently using existing beds. Operations management principles have been applied to healthcare to ensure efficient use of beds. These include: reducing variability of scheduled admissions, remeasuring length of stay (LOS) and bed demand after implementing strategies to reduce practice variation, and employing queuing theory to generate predictions of optimal beds needed.6 The Joint Commission implemented a leadership standard (LD 04.03.11) that hospitals develop and implement plans to identify and mitigate impediments to efficient patient flow through the hospital.

To improve access, hospital leaders expect hospitalists to staff in inpatient medicine programs, surgical comanagement, short stay and chest pain units, and active bed management.7 In the following review, we define hospitalists' roles in the aforementioned programs and their effect on patient flow. We also touch on preoperative clinics, palliative care, geographic rounding, and flexible staffing models.

ACUTE INPATIENT CARE

Hospitalists are one of the fastest growing physician groups in the United States.810 Hospitalists improve efficiency and quality of care across a variety of demographic, geographic, and healthcare settings.11, 12 A 2002 retrospective cohort study in a community‐based urban teaching hospital showed that hospitalists decreased LOS by 0.61 days and lowered risk for death in the hospital (adjusted relative hazard, 0.71; 95% confidence interval [CI], 0.540.93).13 A 2004 prospective quasi‐experimental observational study done at an academic teaching hospital showed an adjusted LOS that was 16.2% lower, and adjusted cost 9.7% lower, for patients on the hospitalists' service.14 In 2007, Lindenauer and colleagues found that a national sample of hospitalists decreased LOS by 0.4 days and lowered cost by $286 per patient.15 The findings of these individual studies were supported in a 2009 systematic review of 33 studies by Peterson which showed that hospitalists decrease LOS.16 In a recent study, Kuo and Goodwin showed that while hospitalists decrease LOS and cost, the patients they care for have higher Medicare costs after discharge by $322 per patient, and are more likely to be readmitted (odds ratio, 1.08; CI, 1.041.14).17

The hospitalist model of care continues to grow, and hospitalists will soon number as many as 30,000.18 For acute medical inpatients, the evidence suggests that hospitalists improve patient flow by decreasing LOS while improving other aspects of quality of care. However, Kuo and Goodwin's findings suggest that the transition of care from inpatient to outpatient settings still requires attention.17

SURGICAL COMANAGEMENT

The Society of Hospital Medicine (SHM) core competencies include perioperative medicine.19, 20 In the 2006 SHM national survey, 85% of hospital medicine groups indicated that they participated in surgical comanagement.21

Hospitalists have improved patient flow and outcomes for orthopedic patients. Hospitalist management of hip fracture patients decreases time to surgery and LOS compared to standard care.2224 Phy and colleagues studied 466 patients for 2 years after the inception of hospital medicine comanagement of surgical patients, and found that care by hospitalists decreased LOS by 2.2 days.22 In a retrospective study of 118 patients, Roy and colleagues found that hospitalist‐managed patients had shorter time to consultation and surgery, decreased LOS, and lower costs.23 In a retrospective cohort study, Batsis looked at mortality in 466 patients with hip fracture, and found no difference between hospitalist management and standard care.24 In patients undergoing elective hip and knee arthroplasty, Huddleston and colleagues reported that patients managed by hospitalists had fewer complications and shorter LOS. The nurses and orthopedic surgeons preferred the hospitalistorthopedist comanagement model.25

The benefits of hospitalist comanagement are not limited to adult patients undergoing orthopedic surgery. For high‐risk patients undergoing lower extremity reconstruction surgery, Pinzur and colleagues noted that LOS was shorter for a cohort of patients managed by hospitalists than for a group of historical controls not treated by hospitalists.26 Simon and colleagues studied comanagement for pediatric spinal fusion patients, and found a decrease in LOS from 6.5 to 4.8 days.27

Several factors should be considered in developing and implementing a successful comanagement program. Since comanagement duties may fall upon hospitalists in order to protect surgeons' time,28 hospital medicine groups should ensure adequate staffing prior to taking on additional services. Clear guidelines to delineate roles and responsibilities of the comanaging groups also need to be developed and implemented.29, 30

Comanaging may also involve additional training. Hospitalists who manage neurologic, neurosurgical, trauma, and psychiatric patients report being undertrained for such conditions.31, 32 Hospital medicine groups need to ensure training needs are met and supported. Given the successes of comanagement and the increasing complexity of surgical patients,33 this practice will likely expand to a greater variety of non‐medical patients.

SHORT STAY UNITS

In 2003, short stay units (SSU) were present in approximately 20% of US hospitals, with 11% of hospitals planning on opening one in the next year.34 SSU are designed to manage acute, self‐limited medical conditions that require brief staysusually less than 72 hours. Approximately 80% of SSU patients are discharged home, avoiding hospitalization.35 Historically, SSU have been under the domain of the ED; however, there is an emerging role for hospitalist‐run SSU.36

Despite demand for SSU, little research has been performed on hospitalist‐led SSU. In 2000, Abenhaim and colleagues showed that a hospitalist‐run SSU at a university‐affiliated teaching hospital had a shorter LOS and lower rates of complications and readmissions when compared to medicine teaching services.37 In 2008, Northwestern Memorial Hospital opened a 30‐bed hospitalist‐run SSU; for those patients, LOS decreased by 2 days.38 In 2010, Leykum and colleagues showed that a hospitalist‐run observation unit can decrease LOS from 2.4 days to 2.2 days.39 Careful selection of SSU patients is needed to obtain these results. Lucas and colleagues found that whether or not SSU patients required assistance of specialists was the strongest predictor of unsuccessful stays (>72 hours or inpatient conversion) in SSU.36

Whether SSU are run by hospital medicine or emergency medicine is decided at an institutional level. Location of SSU in a specifically designated area is crucial, as it allows physicians to round efficiently on patients and to work with staff trained in observation services. Development of admission criteria that include specific diagnoses which match hospitalists' scope of practice is also important (Table 1).32

The protocol‐based and diagnosis‐specific nature of SSU may enhance quality of care through standardization. Future research may delineate the utility of SSU.

CHEST PAIN UNITS

In the United States, in 2004, approximately 6 million patients present annually to EDs with chest pain.40 Cost of care of patients unnecessarily admitted to coronary care units has been estimated to be nearly $3 billion annually.41 Still, as many as 3% of patients with acute myocardial infarction are discharged home.42 Chest pain units (CPU) were developed to facilitate evaluation of patients with chest pain, at low risk for acute coronary syndrome, without requiring inpatient admission. A number of studies have suggested that admission to a CPU is a safe and cost‐effective alternative to hospital admission.4348

CPU have traditionally been staffed by ED physicians and/or cardiologists. In a prepost study, Krantz and colleagues found that a CPU model, incorporating hospitalists at an academic public safety‐net hospital, decreased ED LOS with no difference in 30‐day cardiac event rate.49 Myers and colleagues created a hospitalist‐directed nonteaching service in an academic medical center to admit low‐risk chest pain patients. Patients admitted to the hospitalist service had a statistically significant lower median LOS (23 hours vs 33 hours) and approximately half the median hospital charges than those admitted to teaching services.50 At the same academic medical center, Bayley and colleagues showed that 91% of patients admitted for chest pain waited more than 3 hours for a bed. This adversely affected ED revenue by tying up beds, resulting in an estimated annual loss of $168,300 of hospital revenue. Creation of a hospitalist‐managed service for low‐acuity chest pain patients reduced hospital LOS by 7 hours.51 Somekh and colleagues demonstrated that a protocol‐driven, cardiologist‐run CPU results in a decreased LOS and readmission rate compared to usual care.52 In a non‐peer reviewed case study, Cox Health opened an 8‐bed, hospitalist‐led CPU in 2003. They decreased LOS from 72 to 18 hours, while increasing revenue by $2.5 million a year.53 These studies suggest that hospitalist‐run CPU can decrease LOS, increase revenue, and relieve ED overcrowding.

Development of a successful CPU depends upon clear inclusion/exclusion criteria; close collaboration among ED physicians, hospitalists, and cardiologists; the development of evidence‐based protocols, and the availability of stress testing.

ACTIVE BED MANAGEMENT

As of 2007, 90% of EDs were crowded beyond their capacity.2 ED crowding leads to ambulance diversion,54 which can delay care and increase mortality rates.55 One of the main causes of ED crowding is the boarding of admitted patients.56 Boarded, admitted patients have been shown to have decreased quality of care and patient satisfaction.35

Active bed management (ABM) by hospitalists can decrease ED diversion. Howell and colleagues instituted ABM where hospitalists, as active bed managers, facilitate placement of patients to their inpatient destinations to assist ED flow.57 This 24‐hour, hospitalist‐led, active bed management service decreased both ED LOS and ambulance diversion. The bed manager collaborated real‐time with medicine and ED attending physicians, nursing supervisors, and charge nurses to change patient care status, and assign and facilitate transfer of patients to appropriate units. These hospitalist bed managers were also empowered to activate additional resources when pre‐diversion rounds identified resource limitations and impending ED divert. They found overall ED LOS for admitted patients decreased by 98 minutes, while LOS for non‐admitted patients stayed the same. AMB decreased diversion due to critically ill and telemetry patients by 28% (786 hours), and diversion due to lower acuity patients by 6% (182 hours). This intervention proved cost‐effective. Three full‐time equivalent (FTE) hospitalists' salaries staff 1 active bed manager working 24/7. Nearly 1000 hours of diversion were avoided at an annual savings of $1086 per hour of diversion decreased.

ABM is a new frontier for hospitals in general, and hospitalists in particular. Chadaga and colleagues found that a hospital medicine‐ED team participating in active bed management, while caring for admitted patients boarded in the ED, can decrease ED diversion and improve patient flow. The percentage of patients transferred to a medicine floor and discharged within 8 hours was reduced by 67% (P 0.01), while the number of discharges from the ED of admitted medicine patients increased by 61% (P 0.001).58

To decrease initial investment, components of ABM (ED triage, bed assignment, discharge facilitation) can be instituted in parts. Hospital medicine groups with limited resources may only provide a triage service by phone for difficult ED cases. Bedside evaluations and collaboration with nursing staff to improve bed placement may be a next step, with floor and/or intensive care unit (ICU) rounds to facilitate early discharges as a final component.

OTHER AREAS

CONCLUSION

While many hospital medicine groups were started to provide acute inpatient medical care, most have found that their value to hospitals reaches beyond bedside care. With an epidemic of ED diversion and lack of access to hospital beds and services, optimizing throughput has become imperative for hospital systems. While hospital access can be improved with addition of new beds, improving throughput by decreasing LOS maximizes utilization of existing resources.

We have reviewed how hospitalists improve patient flow in acute inpatient medicine, surgical comanagement, short stay units, chest pain units, and active bed management. In each instance, the literature supports measures for decreasing LOS while maintaining or improving quality of care. Hinami and colleagues showed physician satisfaction with hospitalist‐provided patient care.75 Most studies have been limited by tracking upstream effects of improved efficiency. As there is now some evidence that decreasing LOS may increase readmissions,17 future studies should incorporate this metric into their outcomes. The effect of formal operations management principles on patient flow and bed efficiency is not well known and should be further examined.

In addition, we have touched on other areas (perioperative clinics, palliative care, geographic rounding, and flexible staffing models) where hospitalists may impact patient throughput. These areas represent excellent opportunities for future research.

Hospitalist participation in many of these areas is in its infancy. Hospital medicine programs interested in expanding their services, beyond acute inpatient care, have the opportunity to develop standards and continue research on the effect of hospital medicine‐led services on patient care and flow.