Abstract
Background
Academic medical centers are increasingly employing hospitalists to staff teaching wards. Although studies have demonstrated reduced length-of-stay associated with hospitalist care, it is unclear which patients are most likely to benefit.
Objective
To determine whether patients with specific diagnoses or discharge needs account for the association between hospitalist care and reduced length-of-stay.
Design
Hospital admissions were divided into two groups based on type of attending: teaching hospitalist (full-time faculty hospitalist with no outpatient responsibilities), and non-hospitalist (full-time or voluntary faculty contributing one or two months of teaching service per year).
Participants
All patients discharged from an academic teaching service over two years.
Measurements
Data were extracted from Montefiore Medical Center's clinical information system and the social security death registry.
Results
Mean length-of-stay was lower for teaching hospitalists than for non-hospitalists (5.01 days vs. 5.87 days, p < 0.02). The reduction in length-of-stay was greatest for patients requiring close clinical monitoring (patients with congestive heart failure, stroke, asthma, or pneumonia), and for those requiring complex discharge planning. There were no significant differences between the groups in readmission, in-hospital mortality, or 30-day mortality.
Conclusions
Teaching hospitalist care was associated with shorter length-of-stay in patients requiring close clinical monitoring and complex discharge planning, without adversely affecting readmission or mortality rates.
Introduction
Pressure to control costs of inpatient care has led many academic medical centers to employ hospitalists to provide care for medical inpatients and to staff medical teaching rounds.1 Hospitalist care has been associated with shorter length-of-stay2-16 and lower cost3, 4, 7, 9-14, 16, 17 in a variety of settings, including after adjustment for physician experience.18 This reduction in length-of-stay has been demonstrated among patients admitted with pneumonia9, 17 and congestive heart failure.19 However, whether there are other specific patient diagnoses or discharge dispositions that appear to derive particular benefit from hospitalist care has not been firmly established.
The available data on clinical outcomes associated with hospitalist care is less consistent than the data on length-of-stay. One study noted a 33% decrease in in-hospital mortality among patients on the hospitalist service, and this decrease remained significant at 60 days. However, these findings were confounded by non-random assignment to the hospitalist service, and major differences between the treatment groups.11 Another study, in which assignment to the hospitalist service was without preference, found a statistically significant reduction in 30-day mortality in the hospitalist-treated group in the second year of the program,14 but included only two hospitalists. Despite these promising results, randomized studies with long-term follow-up of hospitalist versus non-hospitalist care are lacking.
Our primary objective was to determine which patient groups appear more likely to benefit from hospitalist care by comparing length-of-stay and clinical outcomes (in-hospital mortality, 30-day mortality, and readmission) among hospitalists and non-hospitalists. Our secondary objective was to compare teaching quality between these groups. We compared two models of care: teaching hospitalist (full-time faculty with no outpatient responsibilities), and non-hospitalist (full or part-time faculty with no more than two months per year on the inpatient service). We hypothesized that hospitalist care would be associated with shorter length-of-stay among patients with specific diagnoses, levels of acuity, and discharge needs, without adversely affecting clinical outcomes.
Methods
Study Setting and Data Collection
Montefiore Medical Center's Weiler Hospital is a 381-bed teaching hospital affiliated with Albert Einstein College of Medicine. We extracted data on all discharges from the medical teaching service at Weiler Hospital for two years, from 7/1/02 through 6/30/04. Patients were assigned to a teaching hospitalist or non-hospitalist team, in rotation, at the time of admission by a senior admitting resident. Hospitalist and non-hospitalist teams are identical in every way other than type of attending conducting rounds. Patients assigned to the non-teaching service were not included in this analysis. Although team assignment appears random, we decided not to use this term as we acknowledge that concealed allocation did not occur. Rather, we used a quasi-randomized study design.2 Data on patient age, race-ethnicity, sex, insurance status, history of prior admissions, laboratory values, ICD-9 and DRG codes, length-of-stay, readmission, and in-hospital mortality were extracted from a replicate of Montefiore's clinical information system (Emerging Health Information Technology, Yonkers, NY). The data replicate is merged monthly with the social security death registry, which allowed extraction of 30-day mortality rates after discharge. Montefiore's Institutional Review Board approved the study.
Outcome Variables
The main outcome variables were: length-of-stay, readmission, in-hospital and 30-day mortality, and resident and student teaching evaluations. Length-of-stay was analyzed as a continuous variable and, although not normally distributed, was reported as a mean (as has been done previously).5, 14, 15 Readmission was defined as an admission for any reason, within 30 days of discharge, to Weiler hospital or to the other Montefiore hospital, Moses, and was analyzed as a dichotomous variable. In-hospital and 30-day mortality were analyzed as dichotomous variables.
Independent variables
The major independent variable was the team assignment of each admission (teaching hospitalist vs. non-hospitalist). In addition, we examined patient-level and physician-level covariates for each admission. Patient-level covariates included: demographic characteristics (age, sex, race-ethnicity), insurance status, number of prior admissions within the previous 90 days, albumin value at the time of admission, diagnosis of diabetes (defined as ICD-9 of 250 at discharge), DRG-weight, primary discharge diagnosis (ICD-9 code), and discharge disposition (home with self-care, home with services, or nursing home). Because the source of data was a clinical information system, independent variables used for adjusted analyses could only be those available on a majority of patients. DRG-weight was used as a surrogate for the overall severity of illness, as has been done previously,14 because it was readily available on all patients. For length-of-stay analyses, DRG-weight was categorized into quartiles, which represent levels of patient acuity. Albumin level at the time of admission was also used as a marker for overall disease severity, as has been done in earlier studies,20-22 because it is available on most patients and is thought to be useful in the quantification of risk in research settings.23 ICD-9 codes were grouped into diagnosis clusters using the Healthcare Cost and Utilization Project of the Agency for Healthcare Research and Quality system.24 Some of these diagnoses were further collapsed into clinical categories to make data reporting manageable (e.g. “Neoplasm” contains several AHRQ diagnosis classifications). For length-of-stay analyses, only those diagnoses that were among the top 30 causes for hospitalization in the U.S.A.25 were included, to ensure adequate numbers of patients in each category. The major physician-level covariate of interest was the annual number of months on service. There was no crossover of attendings between groups (i.e. non-hospitalists never became hospitalists, nor vice versa).
Statistical Analysis
First, the two groups (teaching hospitalist vs. non-hospitalist) were compared with respect to demographic and clinical characteristics using t-tests, Mann-Whitney, or chi-square tests. Because DRG-weight is not normally distributed, it is reported both as a mean (using a t-test with unequal variances for comparison between the groups), and as a median (using the Mann-Whitney test for comparison between the groups). Categorical variables are compared using chi-square tests. Second, Somers' D was calculated and used as a rank-based test26 to compare length-of-stay between the groups, with and without standard errors robust to clustering of patients within each physician. The analysis was repeated with and without truncation of length-of-stay to the mean plus three standard deviations to assess the influence of outliers. Then, multivariate linear regression was used to compare length-of-stay between the groups after adjusting for differences in demographic and clinical characteristics. For multivariate analyses the variables for insurance and race-ethnicity were recoded into multiple dummy variables.
For the dichotomous outcomes of readmission and in-hospital and 30-day mortality, rates were initially compared using chi-square tests, and then multivariate logistic regression models were used to assess the impact of hospitalist care after adjusting for differences between the three groups. Variables chosen for multivariate analyses included those that measured a significant difference between the groups (albumin, DRG-weight, insurance); other variables were added, one by one, keeping those with a Wald-statistic p < 0.20. 95% confidence intervals were calculated using standard errors, which are robust to the clustering of patients by physician.
STATA software, version 9.0, (StataCorp, College Station, TX) was used for all statistical analysis and data manipulation.
Results
Patient Population
From 7/1/02 through 6/30/04 there were 9047 discharges from the teaching service at Weiler Hospital. Significant data were missing for ten hospital stays; thus the study sample included 9037. This included 2913 discharges from teaching hospitalist teams (32.2%) and 6124 discharges from non-hospitalist teams (67.8%). The patient populations were similar with respect to age, sex, race-ethnicity, presence of prior admission, and mix of acute diagnoses. The teaching hospitalist group included more patients with Medicaid, a lower mean Albumin, and a lower mean DRG-weight. (Table 1)
Table 1. Demographic and Clinical Characteristics of Patients Assigned To Teaching Hospitalist vs. Non-Hospitalist Services.
| All (n = 9037) |
TH (n = 2913) |
NH (n =6124) |
p value* | |
|---|---|---|---|---|
| Demographics | ||||
| Age- mean ± SD | 64.9 ± 17.8 | 65.3 ± 17.6 | 64.7 ± 17.9 | 0.21 |
| Female- no.(%) | 5254 (58.1) | 1659 (57.0) | 3595 (58.7) | 0.12 |
| Service- no.(%) | 6650 (73.6) | 2167 (74.4) | 4483 (73.2) | 0.23 |
| Prior Admit no. (%) | 2390 (26.4) | 797 (27.4) | 1593 (26.0) | 0.18 |
| Albumin mean ± SD | 3.80 ± 0.70 | 3.77 ± 0.68 | 3.82 ± 0.66 | 0.002 |
| DRG wt- mean ± SD | 1.24 ± 1.56 | 1.20 ± 1.33 | 1.27 ± 1.66 | 0.03 |
| DRG wt- median | 0.95 | 0.95 | 0.96 | 0.90 |
| Insurance No. (%) | 0.04 | |||
| Self-Pay | 78 (0.9) | 19 (0.7) | 59 (1.0) | |
| Medicaid | 2964 (32.8) | 1008 (34.6) | 1956 (31.9) | |
| Medicare | 3984 (44.1) | 1244 (42.7) | 2740 (44.7) | |
| Commercial | 2008 (22.2) | 640 (22.0) | 1368 (22.3) | |
| Race-Ethnicity No. (%) | 0.44 | |||
| White | 2527 (28.0) | 786 (27.0) | 1741 (28.4) | |
| Black | 2617 (29.0) | 847 (29.1) | 1770 (28.9) | |
| Hispanic | 2958 (32.7) | 964 (33.1) | 1994 (32.6) | |
| Unknown/Other | 935 (10.3) | 316 (10.8) | 619 (10.1) | |
| Diagnosis† No. (%) | ||||
| Congestive Heart Failure | 817 (9.0) | 262 (9.0) | 555 (9.1) | 0.92 |
| Chest Pain | 679 (7.5) | 221 (7.6) | 458 (7.5) | 0.86 |
| Acute Myocardial Infarction | 497 (5.5) | 167 (5.7) | 330 (5.4) | 0.50 |
| Pneumonia | 618 (6.8) | 180 (6.2) | 438 (7.2) | 0.09 |
| Asthma/COPD | 523 (5.8) | 168 (5.8) | 355 (5.8) | 0.96 |
| Coronary Disease w/o Infarction | 406 (4.5) | 131 (4.5) | 275 (4.5) | 0.99 |
| Diabetes Decompensated‡ | 333 (3.7) | 96 (3.3) | 237 (3.9) | 0.18 |
| Arrhythmia | 294 (3.3) | 85 (2.9) | 209 (3.4) | 0.22 |
| Sepsis | 257 (2.8) | 79 (2.7) | 178 (2.9) | 0.60 |
| Gastrointestinal Bleeding | 247 (2.7) | 79 (2.7) | 168 (2.7) | 0.93 |
| Syncope | 232 (2.6) | 80 (2.7) | 152 (2.5) | 0.46 |
| Urinary Tract Infection | 228 (2.5) | 82 (2.8) | 146 (2.4) | 0.22 |
| Fluid/Electrolyte Disorder | 209 (2.3) | 76 (2.6) | 133 (2.2) | 0.20 |
| Neoplasm | 200 (2.2) | 69 (2.4) | 131 (2.1) | 0.49 |
| Enteritis/Diarrhea | 191 (2.1) | 55 (1.9) | 136 (2.2) | 0.30 |
| Renal Failure | 190 (2.1) | 66 (2.3) | 124 (2.0) | 0.46 |
| Acid/Peptic Disease | 177 (2.0) | 71 (2.4) | 106 (1.7) | 0.02 |
| Cellulitis | 175 (1.9) | 62 (2.1) | 113 (1.8) | 0.36 |
| Complication of Surgery/Procedure | 162 (1.8) | 57 (2.0) | 105 (1.7) | 0.42 |
| Hepatitis/Cirrhosis | 153 (1.7) | 53 (1.8) | 100 (1.6) | 0.52 |
| Acute Cerebro-vascular Accident | 129 (1.4) | 37 (1.3) | 92 (1.5) | 0.39 |
| Thromboembolic Disease | 120 (1.3) | 43 (1.5) | 77 (1.3) | 0.40 |
| HIV | 119 (1.3) | 40 (1.4) | 79 (1.3) | 0.75 |
| Anemia | 101 (1.1) | 31 (1.1) | 70 (1.1) | 0.74 |
| Pancreatitis | 72 (0.8) | 29 (1.0) | 43 (0.7) | 0.14 |
| Hemoglobin SS Disease | 38 (0.4) | 13 (0.5) | 25 (0.4) | 0.79 |
| Other Diagnosis | 1870 (20.7) | 581 (19.9) | 1289 (21.0) | 0.23 |
TH= Teaching Hospitalist; NH= Non-Hospitalist
Continuous variables reported as mean ± Standard deviation and groups compared using t-test with equal or unequal variance, as appropriate.
Categorical variables reported as number (percent) and compared using χ2. Medians of Age and DRG-Weight compared using Mann-Whitney.
For zero-sum variables race-ethnicity and insurance χ2 reported for whole group and by category.
Diagnoses grouped by ICD-9 code according to clinical classifications by the Agency for Healthcare Research and Quality, and collapsed into clinical categories.
Includes Diabetic-ketoacidosis, hyper-osmolar non-ketotic, and hyperglycemia
Length-of-stay
The mean length-of-stay was 5.01 ± 6.12 days in the teaching hospitalist group vs. 5.87 ± 7.51 days in the non-hospitalist group (p = 0.02; Table 2). The reduction in length-of-stay was still large and significant after adjustment for potential confounders, (unadjusted reduction = 0.86 days, p = 0.004; adjusted reduction = 0.78 days, p < 0.001, adjusted for race-ethnicity, insurance, DRG-weight, albumin, and prior admissions). There was a statistically significant negative linear association between length-of-stay (in days) and number of months the attending was on service per year, with each extra month on service associated with a 0.19 day reduction in length-of-stay (β = -0.19 days per month on service, p < 0.002).
Table 2. Length-of-Stay Among Patients Assigned to Teaching Hospitalist vs. Non-Hospitalist Teams.
| All | TH | NH | ||
|---|---|---|---|---|
| All Admissions | (n=9037) | (n=2913) | (n=6124) | p value* |
| Mean Length-of-stay | 5.59 ± 7.10 | 5.01 ± 6.12 | 5.87 ± 7.51 | 0.02 |
| Mean Truncated Length-of-stay† | 5.31 ± 5.24 | 4.82 ± 4.80 | 5.54 ± 5.42 | 0.02 |
| First Admission | (n = 6883) | (n = 2150) | (n = 4733) | |
| Mean Length-of-stay | 5.46 ± 7.10 | 4.96 ± 5.93 | 5.69 ± 7.56 | 0.04 |
| Mean Truncated Length-of-stay† | 5.17 ± 5.09 | 4.79 ± 4.76 | 5.34 ± 5.23 | 0.04 |
TH= Teaching Hospitalist; NH= Non-Hospitalist; LOS = Length-of-stay
Rank-based testing using Somers' D, with standard errors robust to clusters of patients by physician
LOS truncated at mean + 3 standard deviations
To assess the influence of outliers on length-of-stay, the analysis was repeated using the variable length-of-stay truncated at the mean plus three standard deviations. In truncated analysis, teaching hospitalist care was associated with reduced length-of-stay when compared with non-hospitalist care (4.82 days vs. 5.54 days, p = 0.02).
To assess the influence of patients returning with multiple admissions, the analysis was performed using only the first admission for each patient during the time period (n = 6883). This first-admission analysis also yielded significant reductions in length-of-stay for teaching hospitalists vs. non-hospitalists (4.96 days vs. 5.69 days, p = 0.04).
Length-of-stay Among Patients with Specific Diagnoses and Discharge Dispositions
To determine whether reduction in length-of-stay is more pronounced among patients with specific diagnoses, level of acuity, or discharge needs, length-of-stay data were analyzed after stratifying by clinical conditions, DRG-weight, and discharge disposition (Table 3). Reductions in length-of-stay associated with hospitalist care were most pronounced in patients with diagnoses of cerebrovascular accident (CVA or stroke), sepsis, asthma/COPD, or urinary tract infection, without appreciable differences in length-of-stay among patients with diagnoses of acute MI, coronary disease without MI, arrhythmia, or gastrointestinal bleeding. Lengths-of-stay were marginally shorter in patients with pneumonia (p = 0.13) and congestive heart failure (p = 0.06). The mean length-of-stay was low for patients admitted with chest pain in both groups, but slightly lower in the teaching hospitalist group (p = 0.04). Regarding overall acuity, differences between hospitalists and non-hospitalists were significant in patients with lower overall acuity (DRG-weight quartiles 1-3; mean difference = 0.49 days; p = 0.02), but widened and remained significant in patients with the highest overall acuity (4th quartile of DRG-weight; mean difference = 1.95 days; p <0.001). Regarding discharge disposition, significant differences between groups were smaller in patients who were discharged home with self-care, widened when home services were required, and were largest when transfer to a nursing facility at the time of discharge was required.
Table 3. Length-of-stay in Days Among Patients with Specific Diagnoses, Levels of Overall Acuity, and Discharge Dispositions Assigned to Teaching Hospitalist vs. Non-Hospitalist Services.
| Group (No.) | TH | NH | Difference (days) | p value* |
|---|---|---|---|---|
| Diagnosis | ||||
| Acute Cerebro-vascular Accident (129) | 8.59 ± 7.46 | 12.45 ± 8.74 | 3.86 | 0.001 |
| Sepsis (257) | 8.63 ± 8.25 | 12.33 ± 16.58 | 3.70 | 0.03 |
| Pneumonia (618) | 6.97 ± 8.40 | 8.35 ± 10.05 | 1.38 | 0.13 |
| Congestive Heart Failure (817) | 4.63 ± 3.72 | 5.81 ± 5.63 | 1.18 | 0.06 |
| Asthma/COPD (523) | 3.53 ± 2.69 | 4.46 ± 4.29 | 0.93 | 0.009 |
| Urinary Tract Infection (228) | 4.48 ± 4.61 | 5.31 ± 6.21 | 0.83 | 0.05 |
| Diabetes Decompensated† (333) | 4.50 ± 6.12 | 5.03 ± 6.47 | 0.53 | 0.31 |
| Cellulitis (175) | 5.15 ± 7.66 | 5.66 ± 5.06 | 0.51 | 0.17 |
| Fluid/Electrolyte disorder (209) | 4.26 ± 2.55 | 4.62 ± 4.03 | 0.36 | 0.54 |
| Acute Myocardial Infarction (497) | 7.11 ± 6.22 | 7.44 ± 6.48 | 0.33 | 0.57 |
| Gastrointestinal Bleeding (247) | 5.25 ± 4.68 | 5.44 ± 5.99 | 0.19 | 0.73 |
| Chest Pain (679) | 1.76 ± 1.50 | 1.85 ± 1.36 | 0.09 | 0.04 |
| Coronary Disease w/o Infarction (406) | 3.56 ± 3.27 | 3.62 ± 3.36 | 0.06 | 0.43 |
| Arrhythmia (294) | 4.27 ± 4.73 | 4.03 ± 3.39 | -0.24 | 0.35 |
| Quartiles of DRG-weight | ||||
| 1st Quartile (2309) | 2.22 ± 2.00 | 2.68 ± 2.51 | 0.46 | < 0.001 |
| 2nd Quartile (2210) | 4.14 ± 3.72 | 4.58 ± 3.94 | 0.44 | 0.07 |
| 3rd Quartile (2259) | 4.96 ± 4.22 | 5.59 ± 5.03 | 0.63 | 0.12 |
| 4th Quartile (2259) | 8.74 ± 9.59 | 10.69 ± 11.97 | 1.95 | < 0.001 |
| Disposition | ||||
| Home-Self (5175) | 3.40 ± 3.13 | 3.80 ± 3.59 | 0.40 | 0.02 |
| Home Care (1601) | 5.24 ± 4.42 | 6.25 ± 5.18 | 1.01 | 0.005 |
| SNF (1402) | 8.21 ± 9.56 | 10.22 ± 10.74 | 2.01 | 0.005 |
TH= Teaching Hospitalist; NH= Non-Hospitalist
Variables reported as mean ± standard deviation
Rank-based testing using Somers' D, with standard errors robust to clusters of patients by physician
Diagnoses grouped by ICD-9 code according to clinical classifications by the Agency for Healthcare Research and Quality, and collapsed into clinical categories.
Includes Diabetic-ketoacidosis, hyper-osmolar non-ketotic, and hyperglycemia
Dispositions: Home-self= discharged home without services; Home Care= discharged home with arranged visiting nurse;
SNF= at discharge, transferred to a skilled nursing facility
Readmission and Mortality Rates
After adjustment for demographic and clinical characteristics, there were no differences between the groups in readmission, in-hospital mortality, or 30-day mortality (Table 4). In first-admission only analysis, after adjustment, there were no differences between the two groups in readmission (OR = 0.92; 95% CI: 0.80 – 1.06), in-hospital mortality (OR = 0.95; 95% CI: 0.72 – 1.27), or 30-day mortality (OR = 0.99; 95% CI: 0.82 – 1.18).
Table 4. Clinical Outcomes Among Patients Assigned to Teaching Hospitalist and Non-Hospitalist Services.
| TH | NH | |
|---|---|---|
| 30-Day Readmission* | 0.97 (0.88 - 1.08) |
reference |
| In-Hosp Mortality† | 1.03 (0.83 - 1.28) |
reference |
| 30-Day Mortality‡ | 1.05 (0.88 - 1.26) |
reference |
TH= Teaching Hospitalist; NH= Non-Hospitalist
Odds Ratios (95% confidence intervals) for events if assigned to a hospitalist team using non-hospitalist teams as reference
95% confidence intervals reported cluster robust for clusters of patients by physician
Adjusted for age, sex, insurance, diabetes, albumin, DRG-weight, number of admissions in prior 90 days
Adjusted for age, sex, race-ethnicity, insurance, diabetes, albumin, DRG-weight
Adjusted for age, sex, insurance, albumin, DRG-Weight, number of admissions in prior 90-days
30-day mortality taken from social security death registry using social security number matching.
Discussion
Our experience with a teaching hospitalist service confirmed several now-familiar findings of hospitalist care, including lower lengths-of-stay for patients on teaching hospitalist teams. We further showed that teaching hospitalist care appears to have a stronger association with length-of-stay in patients requiring specialized services upon discharge, such as homecare services or transfer to a nursing facility. We were able to measure discharge planning needs using a separate discharge disposition code, assigned independent of billing at the time of discharge. We believe the greater reduction in length-of-stay associated with complex discharge planning reflects hospitalist skills in working with ancillary staff, such as social workers or discharge planners.
In addition, we found that hospitalist care had the strongest association with length-of-stay in patients with the highest acuity, as measured by the DRG-weight. This important finding might help guide the use of hospitalist services, given that small numbers of patients at the highest acuity levels account for disproportionate numbers of total hospital days.27
Hospitalist care had the strongest association with length-of-stay in patients with specific diagnoses, including cerebrovascular accidents (strokes), congestive heart failure, pneumonia, sepsis, urinary tract infection (UTI) and asthma/COPD. These findings may reflect hospitalists' ability to closely monitor patients with dynamic conditions, such as congestive heart failure, pneumonia, sepsis, UTI, or asthma/COPD. The close monitoring and continuous presence offered by hospitalists may allow for earlier discharge, because hospitalists are more likely to detect clinical improvement in real time, and to make appropriate adjustments in treatment regimens. For stroke patients in particular, lower lengths-of-stay associated with hospitalist care may also reflect more effective attention to complex discharge needs, such as rehabilitation services.
When we analyzed length-of-stay by the number of months the attending spent on service per year, we showed a statistically significant negative linear trend towards lower lengths-of-stay with increasing numbers of inpatient months. This suggests that extra inpatient experience is associated with even larger reductions in length-of-stay, and supports earlier findings that showed that reductions in length-of-stay with hospitalist care were greater,14 or became significant,11 in the second year of hospitalist programs.
In contrast to other analyses, the patients in the two groups compared in this analysis were largely similar, but some differences persisted. Slightly “sicker” patients may have been assigned to teaching hospitalist teams because at any given time, due to lower length-of-stay, the existing census was likely to be lower on a teaching hospitalist team.
Despite the strengths of our study, it has some limitations. Similar to other studies of hospitalist care, the groups of attendings were small (five teaching hospitalists and 54 non-hospitalists), and thus an individual outlier may have had a significant effect on the outcomes. Given the large number of diagnoses examined, the results of length-of-stay in specific diagnoses needs to be interpreted with caution. An additional limitation is that we had limited power to detect small differences in mortality rates. While it is reassuring that we did not find a statistically significant difference in readmission or mortality rates associated with hospitalist care, it would be erroneous to infer that there is no difference. Given our sample size and the event rates in the non-hospitalist group, requiring a power of 80% we have a minimal detectable odds ratio of 1.18 for readmission, 1.38 for in-hospital events, and 1.24 for 30-day mortality. The strongest claim that can be inferred from this analysis is that if there is a difference in these outcomes with hospitalist care, it is likely less than the above. Smaller differences in mortality rates would only be detectable with a larger sample. Finally, although our data were collected in 2002-04, they reflect a practice in common use today, namely, hospitalists being used for inpatient teaching service in academic medical centers.
We conclude that teaching hospitalist care was associated with reduced length-of-stay on an academic teaching service, but not with increased readmission or mortality rates. There were several patient groups identified in whom this association was particularly strong: patients who require close clinical monitoring, patients in whom real-time adjustment of therapy is critical, patients with high overall acuity, and patients in whom complex discharge planning is necessary.
Acknowledgments
We would like to thank Lori Ashton for programming support.
Financial support: The Institute for Medical Effectiveness Research, a joint project of the Albert Einstein College of Medicine and the North Shore-LIJ Health System, and the Clinical Investigation Core of the Center for AIDS Research at the Albert Einstein College of Medicine and Montefiore Medical Center, funded by the National Institutes of Health (NIH P30 AI51519)
Footnotes
Portions of this research were presented at the 29th Annual Meeting of the Society for General Internal Medicine, April 2006, Los Angeles, California.
Reference List
- 1.Pham HH, Devers KJ, Kuo S, Berenson R. Health care market trends and the evolution of hospitalist use and roles. J Gen Intern Med. 2005 February;20(2):101–7. doi: 10.1111/j.1525-1497.2005.40184.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wachter RM, Goldman L. The hospitalist movement 5 years later. JAMA. 2002 January 23;287(4):487–94. doi: 10.1001/jama.287.4.487. [DOI] [PubMed] [Google Scholar]
- 3.Wachter RM, Katz P, Showstack J, Bindman AB, Goldman L. Reorganizing an academic medical service: impact on cost, quality, patient satisfaction, and education. JAMA. 1998 May 20;279(19):1560–5. doi: 10.1001/jama.279.19.1560. [DOI] [PubMed] [Google Scholar]
- 4.Diamond HS, Goldberg E, Janosky JE. The effect of full-time faculty hospitalists on the efficiency of care at a community teaching hospital. Ann Intern Med. 1998 August 1;129(3):197–203. doi: 10.7326/0003-4819-129-3-199808010-00006. [DOI] [PubMed] [Google Scholar]
- 5.Craig DE, Hartka L, Likosky WH, Caplan WM, Litsky P, Smithey J. Implementation of a hospitalist system in a large health maintenance organization: the Kaiser Permanente experience. Ann Intern Med. 1999 February 16;130(4 Pt 2):355–9. doi: 10.7326/0003-4819-130-4-199902161-00005. [DOI] [PubMed] [Google Scholar]
- 6.Freese RB. The Park Nicollet experience in establishing a hospitalist system. Ann Intern Med. 1999 February 16;130(4 Pt 2):350–4. doi: 10.7326/0003-4819-130-4-199902161-00004. [DOI] [PubMed] [Google Scholar]
- 7.Halpert AP, Pearson SD, LeWine HE, Mckean SC. The impact of an inpatient physician program on quality, utilization, and satisfaction. Am J Manag Care. 2000 May;6(5):549–55. [PubMed] [Google Scholar]
- 8.Molinari C, Short R. Effects of an HMO hospitalist program on inpatient utilization. Am J Manag Care. 2001 November;7(11):1051–7. [PubMed] [Google Scholar]
- 9.Rifkin WD, Conner D, Silver A, Eichorn A. Comparison of processes and outcomes of pneumonia care between hospitalists and community-based primary care physicians. Mayo Clin Proc. 2002 October;77(10):1053–8. doi: 10.4065/77.10.1053. [DOI] [PubMed] [Google Scholar]
- 10.Davis KM, Koch KE, Harvey JK, Wilson R, Englert J, Gerard PD. Effects of hospitalists on cost, outcomes, and patient satisfaction in a rural health system. Am J Med. 2000 June 1;108(8):621–6. doi: 10.1016/s0002-9343(00)00362-4. [DOI] [PubMed] [Google Scholar]
- 11.Auerbach AD, Wachter RM, Katz P, Showstack J, Baron RB, Goldman L. Implementation of a voluntary hospitalist service at a community teaching hospital: improved clinical efficiency and patient outcomes. Ann Intern Med. 2002 December 3;137(11):859–65. doi: 10.7326/0003-4819-137-11-200212030-00006. [DOI] [PubMed] [Google Scholar]
- 12.Hackner D, Tu G, Braunstein GD, Ault M, Weingarten S, Mohsenifar Z. The value of a hospitalist service: efficient care for the aging population? Chest. 2001 February;119(2):580–9. doi: 10.1378/chest.119.2.580. [DOI] [PubMed] [Google Scholar]
- 13.Palmer HC, Armistead NS, Elnicki DM, et al. The effect of a hospitalist service with nurse discharge planner on patient care in an academic teaching hospital. Am J Med. 2001 December 1;111(8):627–32. doi: 10.1016/s0002-9343(01)00976-7. [DOI] [PubMed] [Google Scholar]
- 14.Meltzer D, Manning WG, Morrison J, et al. Effects of physician experience on costs and outcomes on an academic general medicine service: results of a trial of hospitalists. Ann Intern Med. 2002 December 3;137(11):866–74. doi: 10.7326/0003-4819-137-11-200212030-00007. [DOI] [PubMed] [Google Scholar]
- 15.Bellet PS, Whitaker RC. Evaluation of a pediatric hospitalist service: impact on length of stay and hospital charges. Pediatrics. 2000 March;105(3 Pt 1):478–84. doi: 10.1542/peds.105.3.478. [DOI] [PubMed] [Google Scholar]
- 16.Landrigan CP, Srivastava R, Muret-Wagstaff S, et al. Impact of a health maintenance organization hospitalist system in academic pediatrics. Pediatrics. 2002 October;110(4):720–8. doi: 10.1542/peds.110.4.720. [DOI] [PubMed] [Google Scholar]
- 17.Stein MD, Hanson S, Tammaro D, Hanna L, Most AS. Economic effects of community versus hospital-based faculty pneumonia care. J Gen Intern Med. 1998 November;13(11):774–7. doi: 10.1046/j.1525-1497.1998.00231.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rifkin WD, Holmboe E, Scherer H, Sierra H. Comparison of hospitalists and nonhospitalists in inpatient length of stay adjusting for patient and physician characteristics. J Gen Intern Med. 2004 November;19(11):1127–32. doi: 10.1111/j.1525-1497.2004.1930415.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lindenauer PK, Chehabeddine R, Pekow P, Fitzgerald J, Benjamin EM. Quality of care for patients hospitalized with heart failure: assessing the impact of hospitalists. Arch Intern Med. 2002 June 10;162(11):1251–6. doi: 10.1001/archinte.162.11.1251. [DOI] [PubMed] [Google Scholar]
- 20.Hollenbeck BK, Miller DC, Taub DA, et al. The effects of adjusting for case mix on mortality and length of stay following radical cystectomy. J Urol. 2006 October;176(4 Pt 1):1363–8. doi: 10.1016/j.juro.2006.06.015. [DOI] [PubMed] [Google Scholar]
- 21.Wannamethee SG, Shaper AG, Lennon L, Whincup PH. Height loss in older men: associations with total mortality and incidence of cardiovascular disease. Arch Intern Med. 2006 December 11;166(22):2546–52. doi: 10.1001/archinte.166.22.2546. [DOI] [PubMed] [Google Scholar]
- 22.Maraldi C, Volpato S, Kritchevsky SB, et al. Impact of inflammation on the relationship among alcohol consumption, mortality, and cardiac events: the health, aging, and body composition study. Arch Intern Med. 2006 July 24;166(14):1490–7. doi: 10.1001/archinte.166.14.1490. [DOI] [PubMed] [Google Scholar]
- 23.Goldwasser P, Feldman J. Association of serum albumin and mortality risk. J Clin Epidemiol. 1997 June;50(6):693–703. doi: 10.1016/s0895-4356(97)00015-2. [DOI] [PubMed] [Google Scholar]
- 24.Clinical Classifications Software (CCS) 2005 [computer program] 2007. [Google Scholar]
- 25.Healthcare Cost and Utilization Project (HCUP) website Of the Agency for Healthcare Quality and Research (AHRQ) 2007. Ref Type: Generic. [Google Scholar]
- 26.Newson R. Confidence intervals for rank statistics: Somers' D and extensions. The STATA Journal. 2006;6(3):309–34. [Google Scholar]
- 27.Cots F, Elvira D, Castells X, Saez M. Relevance of outlier cases in case mix systems and evaluation of trimming methods. Health Care Manag Sci. 2003 February;6(1):27–35. doi: 10.1023/a:1021908220013. [DOI] [PubMed] [Google Scholar]