Comparison of Outcomes between Hospitalists and General Internists in a Tertiary Academic Medical Center

Background:

A recent study published in the New England Journal of Medicine showed that hospitalists only modestly reduced length of stay and observed costs. This is a finding different from that observed in prior studies. Our study aim was to compare length of stay (LOS), readmission rate, mortality rate, and observed costs between hospitalist service and general internal medicine service.

Methods:

All admissions to the hospitalist service (HS) and general internal medicine service (GIMS) from October 1, 2006, to September 30, 2007, were reviewed. The GIMS has 4 teaching teams with 1 attending, 1 senior resident, and 2 interns in each team. The HS has 2 teams composed of 4 attendings. Each HS team works every other week. Admissions are randomly assigned through a nursing‐run bed management team in a 2:1 ratio.

Results:

A total of 7933 admissions were reviewed, and the HS provided care for 32.3% of patients. Information on 817 admissions was incomplete and was removed from the final analysis. Significant differences were noted among patients cared for by the HS versus the GIMS in mean age (54.7 ± 17.6 vs. 57.7 ± 17.9 years, P < .01), race (nonwhites 39.9% vs. 36.8%, P < .01), severity of illness (minor, 19.0% vs. 14.2%; moderate, 38% vs. 34.5%; major, 32.3% vs. 34.6%; extreme, 10.7% vs. 16.7%; P < .01). Overall LOS was significantly lesser for the HS than for the GIMS (5.3 ± 10.3 vs. 6.5 ± 9.7 days, P < .001). When analyzed based on the severity of illness, LOS was shorter for the HS than for the GIMS across most levels of severity (extreme [15.4 ± 27.4 vs. 16.6 ± 18.3 days, P = .451], major [6.0 ± 5.2 vs. 6.4 ± 6.2 days, P = .146], moderate [3.2 ± 2.4 vs. 3.5 ± 2.8 days, P = .01], and minor (2.5 ± 1.9 vs. 2.4 ± 1.8 days, P = .261]). Mean observed costs were significantly lower for the HS than for the GIMS in categories of extreme illness ($35,544 ± $66,468 vs. $47,069 ± $61,915, P < .01) and moderate illness ($6553 ± $4588 vs. $7233 ± $5121, P < .001), but was not significantly different for major illness ($12,007 ± $10,699 vs. $12,817 ± $11,821, P = .107) and minor illness ($5318 ± $3490 vs. $5135 ± $2979, P = .349). Readmission rate at 30 days was slightly higher for the HS than for the GIMS (7.96% vs. 7.53%, P = .04). Compared to the GIMS, the HS had a lower early death rate (0.2% vs. 0.7%, P = .006) and a lower mortality index (0.29 vs. 0.42), but these differences did not persist after adjusting for severity of illness.

Conclusions:

In our study, hospitalists reduced the length of stay compared with general internists, even after adjusting for severity of illness, without compromising quality of care. Hospitalists reduced overall observed costs in certain subgroups of patients. Identifying special areas of high performance may further improve quality of care outcomes by hospitalists.

Author Disclosure:

V. Jeevanantham, Wake Forest University, faculty; T. Reavis, Wake Forest University, staff; P. Agborbesong, Wake Forest University, faculty.