MO588: Is The ‘Short-Form 12 Health Survey’ A Valid Assessment of Physical Function in People Living with Chronic Kidney Disease?

Abstract

Abstract BACKGROUND AND AIMS Compromised physical function is well-recognised in chronic kidney disease (CKD) and is associated with reduced quality of life (QoL) and poor outcome. CKD is often considered to be the model of ‘accelerated aging,’ thus quick and reliable measurements of physical function are essential as part of healthy aging in the CKD population. However, monitoring physical function via objective assessments can be impractical in clinical and some research settings. The ‘Short-Form 12 Health Survey’ (SF-12) is a commonly used self-report instrument to assess the quality of life, and it has also been used to assess physical health and function through its physical component subscale (PCS), and the physical function domain (PF). However, the validity of the SF-12 against objective physical function tests is unknown. In this analysis, we aimed to validate the SF-12 against multiple objective physical function measures. METHOD Forty-six non-dialysis CKD non-dialysis patients (21 men, mean age: 66.8 (±11.7) years, eGFR 42.8 (±19.6) mL/min/1.73 m2) completed the SF-12 and performed objective physical function tests [short physical performance battery (SPPB), sit to stand 60 (STS60), sit to stand 5 (STS5), timed up and go (TUAG), gait speed and handgrip strength (HGS)]. Based on these tests, patients were categorised into having ‘low’ and ‘normal’ physical function via previously established reference points. To establish construct validity, partial correlation adjusted for age, eGFR and sex were conducted. The predictive ability of the SF-12 was assessed by receiver operating curves (ROC) and the area under curve (AUC), as well as sensitivity (%) and specificity (%). RESULTS The PF domain and PCS subscale had moderate to strong correlation with the SPPB (r = 0.455, P = 0.010 and r = 0.432, P = 0.012), and the STS60 (r = 0.659, P = 0.000 and r = 0.729, P < 0.000) tests and a moderate negative correlation with the STS5 (r= −0.431, P = 0.022 and r = −0.471, P = 0.009). PCS was also associated with gait speed (r = 0.369, P = 0.41). The ability of the PF domain and PCS subscale to predict physical function was ‘Fair’ (P = 0.03) and ‘Good’ (P = 0.01) based on the STS60 test but not the other tests (Table 1). A cut off score, for the PF domain, using Youden's J statistics was determined at a 0.343 with a score of 75 (sensitivity: 76% and specificity: 58%). A cut off score for the PCS component, was determined at Youden's J of 0.400 with score of ≤41.04 (sensitivity: 65% and specificity: 75%). CONCLUSION These data demonstrate that the SF-12 PF domain and PCS component have adequate construct validity when compared with lower body functional tests such as the STS60, STS5 or the SPPB. The PF domain demonstrated a fair accuracy while the PCS component showed good accuracy to predict ‘low’ and ‘normal’ functional status based on the STS60 test. These data also confirm previously assumed cut off point of 75 scores on the SF-12 PF domain, and it also introduces a cut-off point of ≤41.04 scores for the SF-12 PCS subscale to predict low physical function in this population. Thus, the SF-12 may be an adequate tool to measure perceived functional status in nondialysis CKD patients. Therefore, it could be a useful instrument to measure physical function as well as QoL as core components of healthy aging in this population.