Variations in hospitalization rates among nursing home residents: The role of facility and market attributes

Purpose: This study examined the contribution of facility-level and area market-level attributes to variations in hospitalization rates among nursing home residents. Design and Methods: Three years (1991-1994) of state quarterly Medicaid case-mix reimbursement data from 527 nursing homes (NH) in Massachusetts were linked with Medicare Provider Analysis and Review hospital claims and nursing facility attribute data to produce a longitudinal, analytical file containing 72,319 person-quarter observations. Logistic regression models were used to estimate the influence of facility-level and market-level factors on hospital use, after controlling for individual-level resident attributes, including: NH diagnoses, resident-level quality of care indicators, and diagnostic cost grouping classification from previous hospital stays. Results: Multivariate findings suggest that resident heterogeneity alone does not account for the wide variations in hospitalization rates across nursing homes. Instead, facility characteristics such as profit status, nurse staffing patterns, NH size, chain affiliation, and percentage of Medicaid and Medicare reimbursed days significantly influence NH residents' risk of hospitalization. Broader area market factors also appear to contribute to variations in hospitalization rates. Implications: Variations in hospitalization rates may reflect underutilization, as well as overutilization. Continued efforts toward identifying-medically necessary hospitalizations are needed.