Outcomes of Patients Treated by Hospitalists, General Internists, and Family Physicians

Background:

The hospitalist model has been rapidly adopted across the United States; however, supporting evidence is derived from a small number of studies examining the practice of few physicians.

Methods:

We conducted a retrospective cohort study of 76,296 patients aged 18 and older who were hospitalized at 1 of 45 U.S. hospitals that used an expanded set of physician specialty codes which included an option to categorize attending physicians as hospitalists. We examined the records of patients discharged between September 1, 2002, and June 30, 2005, with a principal diagnosis of pneumonia, heart failure, chest pain, ischemic stroke, urinary tract infection, acute exacerbation of chronic obstructive pulmonary disease, or acute myocardial infarction. We limited our analysis to patients cared for by a hospitalist, general internist, or family physician. We developed a series of multivariable models to assess the independent effect of physician specialty on length of stay, cost, inpatient mortality, and readmission while adjusting for patient age, sex, ethnicity, insurance, principal diagnosis, comorbidities, hospital characteristics (ie size, teaching status, location, urban/rural), and physician's annual case volume. Generalized estimating equation (GEE) models were used to account for the clustering of patients within physicians and physicians within hospitals. An interaction term of physician specialty with principal diagnosis was included in all models. Analyses of length of stay and costs were restricted to cases with values within 3 standard deviations of the mean. Logit link GEE models were used to assess the binary outcomes of in‐hospital mortality and 14‐day readmission.

Results:

On average, the 971 physicians categorized as family physicians cared for 20 patients with the 7 selected diagnoses each year, the 993 general internists cared for 30, and the 284 hospitalists cared for 75. When compared to patients cared for by general internists, those cared for by hospitalists had a shorter length of stay (adjusted difference ‐0.6 days, 95% CI: ‐0.7, ‐0.5 days) and lower costs (adjusted difference ‐$417, 95% CI: ‐$704, ‐$131), but similar inpatient mortality (OR = 0.95, 95% CI: 0.85, 1.05) and 14‐day readmission risk (OR = 0.98, 95% CI: 0.91, 1.05). When compared with those cared for by family physicians, length of stay of patients cared for by hospitalists was shorter (adjusted difference ‐0.4 days, 95% CI: ‐0.6, ‐0.3 days), whereas costs (adjusted difference $8, 95% CI ‐$296, $311), mortality (OR = 0.95, 95% CI 0.83, 1.07) , and readmissions (OR = 0.95, 95% CI: 0.87, 1.04) were similar. Differences in annualized case volume for the 7 diagnoses did not explain the observed variation in outcomes between physician groups.

Conclusions:

For patients with common inpatient diagnoses, the hospitalist model reduces length of stay and costs without adversely effecting mortality or readmission.

Author Disclosure:

P. K Lindenauer, None; M. B. Rothberg, None; P. S. Pekow, None; C. Kenwood, None; E. M. Benjamin, None; A. D. Auerbach, None.