Derivation of a clinical prediction rule to determine fall risk in community-dwelling individuals with knee osteoarthritis: a cross-sectional study.

Abstract

We aimed to examine the risk factors for falls in individuals with knee osteoarthritis (OA) and derive a clinical prediction rule (CPR) to determine fall risk. Eighty-one individuals with medial compartment knee OA were included. The outcome was whether the participants had a self-reported fall within the past 1year of this study being conducted. The collected data included sex, age, body mass index, Kellgren-Lawrence grade, lesion type (bilateral or unilateral knee OA), pain (rated using the visual analog scale), muscle strength test of the quadriceps femoris, one-leg standing test (OLST), five times sit-to-stand test (FTSST), and 5-m walk test, which were used in binomial logistic regression analysis. The outcome measure of the analysis was whether the study participants belonged to a fall or non-fall group. Receiver operating characteristic (ROC) analysis was performed for the outcome measurements, and the factors were selected by binomial logistic regression analysis. Then, a CPR to determine fall risk was extracted, and its diagnostic characteristics were calculated. Binomial logistic regression analysis showed that the OLST and FTSST were significant. ROC analysis showed that the cut-off values of the OLST and FTSST were 5.3s and 7.9s, respectively. The post-test probability (positive predictive value) increased to 88.2% (positive likelihood ratio = 17.8) when the OLST and FTSST were both positive (the CPR score was 2 points). The CPR obtained from this study would be useful as a screening test to detect the fall risk probability in individuals with knee OA.