Risk Factors for Early Hospital Readmission in Low‐Income Elderly Adults

Objectives To identify risk factors for early hospital readmission in low‐income community‐dwelling older adults. Design Prospective cohort study. Setting University‐affiliated urban safety‐net healthcare system in I ndianapolis, I ndiana. Participants Community‐dwelling adults aged 65 and older with annual income less than 200% of the federal poverty level and enrolled in the G eriatric R esources for A ssessment and C are of E lders ( GRACE ) randomized controlled trial (N = 951). Measurements Participant health and functional status at baseline and 6, 12, 18, and 24 months. Early readmission was defined as a repeat hospitalization occurring within 30 days of a prior hospital discharge. Candidate risk factors included sociodemographic characteristics, health and functional status, prior care, lifestyle, and satisfaction with care. Results Of 457 index admissions in 328 participants, 85 (19%) were followed by an early readmission. The independent risk factors for early readmission identified according to regression analysis were living alone (odds ratio ( OR ) = 1.71, 95% confidence interval ( CI ) = 1.02–2.87), fair or poor satisfaction with primary care physician ( OR  = 2.12, 95% CI  = 1.01–4.46), not having M edicaid ( OR  = 1.80, 95% CI  = 1.05–3.11), receiving a new assistive device in the past 6 months ( OR  = 2.26, 95% CI  = 1.26–4.05), and staying in a nursing home in the past 6 months ( OR  = 5.08, 95% CI  = 1.56–16.53). Age, race, sex, education, and chronic diseases were not associated with early readmission. Conclusion A broad range of nonmedical risk factors played a greater role than previously recognized in early hospital readmission of low‐income seniors.