Related Experiment Video
Updated: Sep 13, 2026
Fu's Subcutaneous Needling for Knee Osteoarthritis Pain
Published on: March 24, 2023
Comparative, validity and responsiveness of the HOOS-PS and KOOS-PS to the WOMAC physical function subscale in total
A M Davis1, A V Perruccio, M Canizares
1Division of Health Care and Outcomes Research, Toronto Western Research Institute, Toronto, Ontario, Canada. adavis@uhnresearch.ca
Objective:
To evaluate the internal consistency of the Hip disability and Osteoarthritis Outcome Score-Physical Function Short-form (HOOS-PS) and the Knee injury and Osteoarthritis Outcome Score-Physical Function Short-form (KOOS-PS) in total hip replacement (THR) and total knee (TKR) replacement. Construct validity and responsiveness were compared to the Western Ontario McMaster Universities' Osteoarthritis Index (WOMAC) Likert 3.0 physical function (PF) subscale and the PF excluding the items in the short measures (PF-exclusions).
Methods:
Participants completed the full HOOS or KOOS, measures of fatigue, anxiety, depression and the Chronic Pain Grade (CPG) pre-surgery and the HOOS or KOOS 6 months post-surgery. Internal consistency for the HOOS-PS and KOOS-PS was calculated using Cronbach's alpha. For construct validity, it was hypothesized that correlations between the HOOS-PS or KOOS-PS and PF and PF-exclusions with fatigue, CPG, anxiety and depression and HOOS/KOOS pain scales would differ by magnitudes of <0.1. Standardized response means (SRMs) were calculated for the HOOS-PS, KOOS-PS, PF and PF-exclusions and hypothesized to be >1.
Results:
The THR group (n=201) had a mean age of 62.3 years; 53.2% were female. The TKR group (n=248) had a mean age of 64.5 years; 63.7% were female. Cronbach's alpha was 0.79 and 0.89 for the HOOS-PS and KOOS-PS, respectively, confirming that the measures represented a homogeneous construct. The correlation of the HOOS-PS to the PF and PF-exclusions was 0.90 and 0.86, respectively; r=0.90 (PF) and r=0.85 (PF-exclusions) for the KOOS-PS. The results supported the construct validity hypotheses. For THR, the SRM was 1.5, 1.7 and 1.7 for the HOOS-PS, PF and PF-exclusions; for TKR, the SRM was 1.4, 1.5 and 1.7, respectively.
Conclusions:
The short HOOS-PS and KOOS-PS represent homogenous short measures of PF with similar construct validity and responsiveness to the 17-item PF. The HOOS-PS and KOOS-PS are parsimonious, valid and responsive for evaluating PF in THR and TKR.
