UNPACKING TRANSITIONAL CARE: FACILITATORS, BARRIERS, AND QUALITY IMPACTS OF A MULTIDISCIPLINARY PROGRAM TO REDUCE 30-DAY READMISSIONS

Several provisions in the Affordable Care Act incentivize healthcare systems to move from traditional fee–for-service towards payments rewarding high-quality care. Notably, the Hospital Readmission Reduction Program penalizes hospitals with excess readmissions. This dissertation sought to evaluate the facilitators, barriers and quality impacts of a computer- based transitional care checklist implemented at a large academic medical center. The evaluation is composed of two parts: 1) a process evaluation to understand program implementation and 2) an impact evaluation to understand program effectiveness. For the process evaluation, we conducted 25 semi-structured interviews with transitional care providers. The impact evaluation was a retrospective cohort analysis of discharges from UNC Hospitals. Patients were eligible for the study if they were discharged between July 2014 and September 2015 and had a 20-30% risk of readmission. We looked at how the overall checklist and the different provider components of the checklist affected unplanned all-cause 30-day readmissions. Our sample included 10,083 eligible discharges. The overall readmission rate was 25.9%. While the entire sample was qualified for the program, out of five possible components, on average patients had less than two complete. Among provider components, completion rates ranged from 17.5% for discharging physicians to 46.7% for case managers. We found a not statistically significant but modest protective effect of the checklist on readmissions (OR = 0.92; 95% CI, 0.81-1.04). In addition, most provider components were protective against readmissions, but confidence intervals included one. Odds ratio ranged from 0.87 for the discharging physician component (95% CI=0.75-1.00) to 0.95 for the case management component (95% CI= 0.83-1.09). We predicted having all the components would reduce readmissions approximately 5-8 percentage points if implemented completely. This study describes the potential of hospitals in real-world settings to reduce readmission using a pre-discharge checklist. Overall, the checklist had a promising but limited impact on readmissions. This is at least partially related to challenges with implementation. Healthcare systems could improve the implementation of similar programs by utilizing mid-level managers and champions to: communicate how the program aligns with existing priorities, clarify team roles and interdependencies, and ensure tools support new workflows.Doctor of Philosoph