Breaking the cycle: Identifying and addressing risk factors for 30-day hospital readmission among people experiencing homelessness

This 30-minute oral presentation will highlight findings from a recent study that used Medicaid claims data merged with HCH electronic medical record data to investigate the primary risk factors for 30-day inpatient hospital readmissions among a sample of individuals who were homeless in Boston. Specific patient-level and admission-level risk factors that were examined included, but were not limited to, patient race, sex, preferred language, age, care team at Boston Health Care for the Homeless Program (BHCHP), admission timing (month, season, day of week, weekend vs. weekday), admitting diagnosis, length of stay, medical vs. surgical service during inpatient stay, admitting source, discharge location (e.g., AMA, discharge to street or shelter), housing status, admitting hospital, substance-related disorder, alcohol-related disorder, mental illness, visit with BHCHP within 30 days of discharge, number of prescribed medications at discharge, and more. The presentation will address how results from the study, which analyzed over 1,300 inpatient hospital admissions over a two-year period, will be used to inform the development of innovative and targeted readmission reduction interventions for the homeless population. This is critical because people experiencing homelessness have been shown to be high utilizers of costly emergency and inpatient services, and HCH grantees are continually seeking new ways to improve care transitions and health outcomes for patients who are exiting the hospital without a stable home in which to recover. While there has been increasing pressure by the Centers for Medicare and Medicaid Services (CMS) to address specific quality measures aimed at reducing hospital readmissions, few studies have examined readmission rates and specific risk factors for 30-day readmission among people who are experiencing homelessness. This is particularly important in the era of accountable care, as safety net providers, including HCH grantees and their partners, increasingly take on financial risk for their patients' total quality and cost of care, including inpatient hospital costs and quality measures related to readmissions. Preliminary results from this study suggest that having a Health Care for the Homeless behavioral health visit within 30 days of an initial or "index" hospital admission significantly reduced the risk of a readmission to the hospital within that 30-day period. Presenters will discuss the implications of such findings, particularly with respect to how a homeless health center can collaborate with area hospitals to use these findings, "working together for quality," while also highlighting ways that other HCH grantees may replicate this study with their own data. No prior knowledge is required of attendees, though familiarity with research methods and terminology (e.g., statistical significance, odds ratios) will be helpful.