reports

Event analysis of all-cause and ambulatory care sensitive hospitalization of long-stay nursing home residents

Baltimore, MD, USA: Centers for Medicare and Medicaid Services.

Schneider Institutes for Health Policy, The Heller School for Social Policy and Management, Brandeis University, Waltham, MA, USA

The free-standing nursing home is the setting for ambulatory medical care for Medicare beneficiaries who are long-stay residents. About three-quarters of these long-stay beneficiaries have one or more chronic conditions. To gain insight into factors associated with more effective care for chronic illness, the study used event (survival) analysis to analyze the relationship between nursing home characteristics and the risk of all-cause hospitalization and hospitalization for ambulatory care sensitive (ACS) conditions for long-stay nursing home residents with and without specific chronic illness diagnoses.

       Study Design: Time to event for two types of hospitalization was modeled as a function of resident, nursing home, market area and state policy factors. ACS hospitalization was defined according to an expanded definition adapted to the nursing home setting. The event analysis approach enabled appropriate treatment of censoring by death, discharge and observation time, including left-censoring for long-stay residents in hospital at the start of observation. The multivariate models estimated the impact of nursing home characteristics, including staffing, on length of time to hospitalization while adjusting for individual beneficiary characteristics, most importantly health and disability status.

  Population Studied: Medicare Aged beneficiaries with and without identified chronic illness who were long-stay (> 3 months) residents of freestanding nursing homes. Using MDS data, 74,279 beneficiaries were identified as long-stay residents as of 1/1/2000 in the Medicare 5% sample. 54,250 lived in free-standing nursing homes that could be matched from OSCAR files. 75.2% of long-stay beneficiaries had a chronic illness diagnosis.

Principal Findings: Nursing home and area characteristics (ownership, staffing by type, any special care unit, nurse aide training in the home, Medicaid rate, state bed hold policy, hospital beds per elder) were significant predictors of time to all-cause and ACS hospitalization for long-stay residents even after accounting for individual health and other characteristics. Greater registered nurse (RN) hours per resident day were protective against risk of adverse event; each additional hour of RN staff time increased time to hospitalization by a factor of 1.24 for all-cause hospitalization and by 1.29 for ACS hospitalization. Nursing assistant (NA) hours per resident day were not significantly associated with time to either type of hospitalization. Nonprofit ownership was protective against hospitalization, lengthening time to event. Time to both types of hospitalization was greater if the state paid higher Medicaid rates or did not have a policy of paying for beds while Medicaid residents are hospitalized.

        Conclusions: The analysis uses a novel approach (event analysis) that corroborates findings of previous research concerning the impact of nursing home resources (RN staffing, Medicaid rate), orientation (ownership, proportion Medicare, special care units) and environment (bed hold policy, hospital bed availability) on hospitalization of long-stay residents. A definition of ACS hospitalization for nursing home residents is proposed that combines features of previous definitions, but findings are not highly sensitive to the definition employed.

Implications for Policy, Delivery or Practice: The nursing home is the setting for chronic illness care for the Medicare beneficiaries who live there. Nursing home RN staffing and services may be significantly associated with the risk of adverse events for long-stay residents, suggesting that care for beneficiaries with chronic illness may be improved by increasing certain nursing home care resources.