Predictive Validity of the Short Physical Performance Battery for Frail States in Community-Dwelling Older Adults

Abstract ID
4125
Authors' names
Grace Chia1, Natalie Teo1, Joanne Kua1
Author's provenances
1Department of Geriatric Medicine, Tan Tock Seng Hospital
Abstract category
Abstract sub-category

Abstract

INTRODUCTION

Frailty is prevalent in Singapore's rapidly ageing population. The Clinical Frailty Scale (CFS) is a widely adopted frailty assessment tool chosen for its ease of use and predictive accuracy. However, CFS grading requires clinical judgment, introducing subjectivity and potential inter-rater variability. The Short Physical Performance Battery (SPPB) is an objective measure of physical function that can predict disability and mortality in older adults, with declining scores suggesting increased frailty. This study aimed to determine the correlation of SPPB to CFS in identifying frail states among community-dwelling older adults.

METHODS

CFS scores were determined by an independent assessor, and individuals were classified into non-frail (CFS 1-3), pre-frail (CFS 4), and frail (CFS 5-6) groups. SPPB scores were measured for all participants. Receiver operating characteristic (ROC) analyses were performed to evaluate the discriminative ability of SPPB for frail states.

RESULTS

The sample included 95 non-frail, 34 pre-frail, and 30 frail individuals (mean ages 76, 77, and 80 years respectively). Median SPPB scores were 11, 10, and 9.5 for non-frail, pre-frail, and frail groups respectively. For pre-frailty, an SPPB cutoff of ≤10 achieved an AUC of 0.72 (95% CI: 0.63-0.82), with 79% sensitivity and 67% specificity. For frailty, an SPPB cutoff of ≤8 resulted in an AUC of 0.69 (95% CI: 0.59-0.80), with 37% sensitivity and 90% specificity.

CONCLUSIONS

SPPB scores correlate well with CFS-defined frailty states despite fundamental differences in assessment approaches. While SPPB showed acceptable diagnostic performance overall, its low sensitivity for frailty suggests it may miss individuals whose deficits are primarily non-physical. However, SPPB's objectivity and reproducibility makes it a reliable alternative for community-based frailty screening, particularly where inter-rater variability is a concern. Our findings support using SPPB as a practical screening tool to identify those who may benefit from early intervention.

Comments

Thank you for a very clinically relevant and practical study. Given the relatively low sensitivity of SPPB for frailty at the ≤8 cut-off, will you recommend combining SPPB with other brief markers, such as cognition, nutrition, mood or social vulnerability, to improve case detection?

Submitted by kai.ping.sze@n… on

Permalink

Based on your SD results,  a SPPB score of 10 could be non-frail, pre-frail or frail. How should a clinician differentiate the status?

Submitted by ellen.freiberg… on

Permalink