Thirty-Day Hospital Readmission Among Homeless Individuals With Medicaid in Massachusetts

Background: National efforts are underway to reduce hospital readmissions. Few studies have used administrative data to provide a global view of readmission among people experiencing homelessness, who often utilize multiple hospital systems. Objective: To examine the 30-day hospital readmission rate and factors associated with readmission following discharge among homeless Medicaid members in Massachusetts. Methods: We analyzed medical record and Medicaid administrative data for 1269 hospitalizations between 2013 and 2014 for 458 unique patients attributed to Boston Health Care for the Homeless Program. Generalized Estimating Equations were used to investigate factors associated with readmission. Results: Of all hospitalizations, 27% resulted in readmission, more than double the average national Medicaid readmission rate. Leaving against medical advice was associated with increased readmission, while having a Health Care for the Homeless primary care practitioner was associated with reduced readmission. Among the most frequently admitted individuals, being discharged to medical respite care was associated with reduced readmission. Conclusions: To break the readmission cycle, health care providers serving homeless individuals could focus on assuring access to medical respite care and extending outreach efforts that increase primary care engagement. This may be especially important for accountable care systems, as safety net providers increasingly assume financial risk for patients' total cost and quality of care.