Background: The rate of unscheduled readmissions is an important quality indicator with financial implications for hospitals.
Objective: To determine if resident-staffed services have more favorable outcomes compared to hospitalist services, predictors of readmissions were determined within an academic hospital.
Methods: From November 1, 2006 to April 30, 2007, 5943 admissions were assigned to a resident-staffed teaching service (n = 2244) or to a hospitalist-based service (n = 3699). Data on age, race, sex, insurance status, case mix index (CMI), length of stay (LOS), and unscheduled hospital readmission within 30 days were analyzed.
Results: Patients admitted to the hospitalist service were older and more likely to be female compared with those admitted to the teaching service. There were no significant differences in the health care insurance carrier of the patients admitted to the two services. The rate of unscheduled hospital readmissions within 30 days was significantly higher in the teaching service compared to the hospitalist service (14.1% vs 10.4%; P < 0.001). LOS was significantly higher (4.95 +/- 7.77 vs 4.14 +/- 5.95; P < 0.001), and the CMI was significantly lower (1.04 +/- 0.94 vs 1.14 +/- 1.02; P < 0.001) in the hospitalist service compared to the resident-staffed service. Probability of readmission was significantly reduced with increasing LOS and discharge home with self care.
Conclusions: The modestly increased unscheduled readmission rate to the resident-staffed service compared to the hospitalist service may be related to lower LOS. Increased CMI of patients in the resident service may have contributed to the increased rate of readmissions.