Origen: Predicting Adverse Health Outcomes in Nursing Homes: A 9-Year Longitudinal Study and Development of the FRAIL-Minimum Data Set (MDS) Quick Screening Tool – Journal of the American Medical Directors Association
To examine the predictive validity of a quick frailty screening tool, the FRAIL-NH, for adverse health outcomes in nursing home residents, using variables from the Minimum Data Set (MDS). The screening items were compiled from the MDS for potential direct application in long-term care facilities using this health information system.
Longitudinal follow-up study of nursing home residents with annual clinical assessment using the MDS and mortality data between 2005 and 2013.
Six nursing homes operated by a nongovernmental organization in Hong Kong.
Participants included 2380 nursing home residents aged 65 years or older at study baseline.
Frailty assessed using the FRAIL-NH model with items from the MDS. The model covers 8 areas: fatigue, resistance, ambulation, incontinence, polypharmacy, weight loss, nutritional approach, and help with dressing. Adverse health outcomes in subsequent years were measured: incident falls, worsening activities of daily living (ADL) function, hospitalization, and death.
Using a cutoff score of 5 on the FRAIL-NH, the prevalence of frailty was 58.5% in this nursing home sample. Frailty as identified using the FRAIL-NH predicts incident falls, worsening ADL function, hospitalization, and death (hazard ratios [HR] 2.00–3.73). This remained significant after adjusting for sociodemographic and other clinical characteristics. Each level of increase on the FRAIL-NH has strong distinguishing power on the incidence of adverse outcomes. Intermediate frailty status (score 1–4) also significantly predicts adverse health outcomes (HR 1.57–2.06).
The FRAIL-NH is a quick screening tool that can be used to identify frail and prefrail nursing home residents at risk of adverse health outcomes. It can be applied using variables from the MDS, allowing direct adoption in long-term care facilities already using this health information system.