A Comparative Effectiveness Trial Between a Post-Acute Care Hospitalist Model and a Community-Based Physician Model of Nursing Home Care

Introduction: To evaluate whether a designated Post-Acute Care Hospitalist (PACH) (an individual physician charged with care of most residents in the nursing home and with set hours to be in the facility each week) could improve some measureable outcomes in the long-term care setting compared with a traditional cadre of community physicians, a comparative trial was designed to measure multiple cost and care variables. Methods: Data were collected in a historical prospective study design for 6 months before the institution of a PACH model in a nursing home in the Baltimore area. Similar data were collected in a similar setting in the same region during the same time frame. During the same 6 months in the following year (ending June 30, 2008) after initiating the PACH program, the same outcome measures, which included demographic information, admissions per census, number of medications per resident, laboratory services and fees per resident day, fallers and falls per resident day, unplanned discharges per resident day, and average pharmacy costs per patient, were collected. Results: Demographic information was similar in all groups. PACH and non-PACH models per resident day were significantly different for laboratory tests (intrafacility pre-PACH $1.93, post-PACH $2.97; P < .005 and interfacility post-PACH $2.97, non-PACH $0.97), fallers (intrafacility pre-PACH 0.006, post-PACH 0.008; P = .01 and interfacility post-PACH 0.008, non-PACH 0.006; P < .005), and falls (intrafacility pre-PACH 0.007, post-PACH 0.01; P < .05 and interfacility post-PACH 0.01, non-PACH 0.008; P < .05). Medication errors after PACH was instituted were 1/6 of pre-PACH rate and 1/60 of the non-PACH facility. Post-PACH pharmacy costs were also better than the non-PACH facility pharmacy costs per resident day by $7.74, but differences for medication errors and pharmacy costs were not statistically significant. Conclusion: Institution of a PACH in a nursing home was associated with a significant increase in laboratory costs and no improvement in fall rates. There was a nonsignificant reduction in medication errors and pharmacy costs. These data support the hypothesis that a PACH model may lead to greater clinician involvement, which may be associated with an increase in clinical testing and a decrease in pharmacy costs and medication errors. If true, the latter would likely far offset any costs associated with additional laboratory testing. These findings warrant further investigation of larger magnitude.