FRAIL-NH Predicts Outcomes in Long Term Care.

Abstract

To investigate the predictive validity of the short, simple FRAIL-NH frailty screening tool in the long term care population and to then compare the predictive validity with the frailty index (FI) for 6-month adverse health outcomes. Retrospective study using the Minimum Data Set (MDS) 3.0 and chart review from June-December 2014. Two Long Term Care Facilities in Saint Louis, MO. 270 patients ages ≥ 65 years old residing in long term care. Frailty was measured using the FRAIL-NH and Frailty Index (FI) criteria. Adverse outcomes measured at 6-month follow-up included falls, hospitalizations, and hospice enrollment/mortality. Based on screening tool used frailty prevalence was 48.7% for FRAIL-NH and 30.3% for FI. The FRAIL-NH pre-frail (Adjusted Odds Ratio [AOR]=2.62; 95% Confidence Interval [CI]=1.25-5.54; p=0.11) classification was associated with 6 month risk of falling and mortality/hospice enrollment was associated with the frail classification, AOR=3.96 (1.44-10.87, p=0.007). Combining the pre-frail and frail categories both measures predicted 6 month mortality with the FRAIL-NH being the strongest predictor (AOR=3.36; 95%CI=1.26-8.98; p=0.016) and the FI was a more modest predictor with an AOR of 2.28; 95%CI=1.01-5.15; p=0.047. When directly comparing the FRAIL-NH to the FI, the FRAIL-NH pre-frail were at increased risk of falling, AOR=2.42 (1.11-5.92, p=0.027) and the FRAIL-NH frail were at increased risk of hospice enrollment/death, OR=3.25 (1.04- 10.86) p=0.044. In comparison to the FI, the FRAIL-NH preformed just as well at screening for frailty and was a slightly better predictor of adverse outcomes. The FRAIL-NH is a brief, easy-to-administer frailty screening tool appropriate for long term care patients and predicts increased risk of falls in the pre-frail and mortality/hospice enrollment in the frail.