Do Hospitalist Teaching Services Provide Cost-Effective Care?
Abstract & Commentary
Increasing evidence suggests that a hospitalist model is an effective means to provide care for acutely ill patients. However, prior studies have not evaluated how the involvement of house staff affects efficiency of this model. Over a 12-month interval, Hackner and associates prospectively identified all Medicaid patients ³ 18 years of age who were cared for by a hospitalist teaching service (n = 477) or private physicians (n = 1160) at 1 medical center. Patients were excluded if their attending physician was a pediatrician, psychiatrist, or if they were admitted for a major surgical procedure.
The hospitalist service was comprised of 4 teams. Each team consisted of 1 attending physician and 4 second- or third-year residents. Each team admitted patients every 4 nights and rounded on a daily basis with the hospitalist attending. Hospitalist physicians spent ³ 50% of their time in direct inpatient clinical activities, were paid a flat fee for their services, and did not receive a bonus based on cost or service reductions. Nonhospitalist patients were cared for by a variety of private providers, most of whom spent < 25% of their time in direct inpatient clinical activities.
Overall median length of stay (LOS) was 4 days for the private patients vs. 3 days for hospitalist cases (P < 0.0001). Median total costs per case were $4853 for private patients vs. $4002 for hospitalist cases (P < 0.0001). Private patients were older (56 vs. 53 years; P < 0.0001), more likely to be female (P = 0.04), and African-American (P < 0.004). However, neither African-American ethnicity nor gender had a significant correlation with LOS. Severity of illness accounted for differences in LOS within the population as a whole, but it did not account for the difference in LOS between private and hospitalist patients.
To explore the relationship between age and LOS, patients were stratified by age. Only patients ³ 65 years of age showed statistically significant reductions in both LOS (P < 0.0001) and total cost (P = 0.002). There were no significant differences between group differences in ICU costs, mortality, interfacility transfers, or 30-day hospital readmission rates. However, there were differences in some cost centers. Median pharmacy costs per case were higher for private patients (15.2% of total costs) than for hospitalist cases (12.7%) (P = 0.03). Imaging costs were also higher for private patients (12.7% of total costs) vs. hospitalist cases (11.3%) (P = 0.009). Subspecialty consultation rates were 37.6% for private patients and 16.6% for hospitalist cases (P < 0.0001). With increasing age, consultation rates increased for private patients from 29.5% to 42.9% without a significant increase for hospitalist cases (Hackner D, et al. The value of a hospitalist service. Efficient care for the aging population? Chest. 2001;119:580-589).
Comment by Leslie A. Hoffman, PhD, RN
Findings of this study suggested that a hospitalist with an active teaching role can provide care that results in lower LOS, costs, and consultation rates compared to private control subjects with a similar insurance payor (Medicaid). Demographic variables did not appear to account for the observed differences. When costs of major service and ancillary therapies were examined, no between group differences were found for laboratory services, blood bank services, or pathology. However, pharmacy and imaging costs were lower for hospitalist cases. In addition, there were more subspecialty consultations in private patients.
This difference was particularly pronounced for patients ³ 65 years of age. As age increased to older than 65 years, there were significant increases in total costs, LOS, and consultations. However, this increase was much smaller for hospitalist cases. In fact, the lower subspecialty consultation rate among older patients appeared to be a key efficiency factor for the hospitalist. Patients older than 65 years typically qualify for Medicare, in addition to Medicaid. It is possible that this change in payor status prompted a different pattern in regard to consultations, test ordering, and medication choices that increased costs. Conversely, lack of access to an alternate payor may have restricted consultations, tests, and medication choices for younger patients.
The effect of payor source on practice patterns is a highly controversial, but important area, which needs further study. Regardless, findings of this study suggest that hospitalist services can provide cost-effective care for older Americans, a group that represents a rapidly increasing population. Further, the study suggests that academic (teaching) interests do not negatively affect the efficiency of care. Rather, they have a positive effect.