Data aggregation for research, clinical care, and quality can be facilitated by linking scores on patient-reported outcome measures. For a study, researchers sought to correlate Hip Disability and Osteoarthritis Outcome Score-Physical Function Short Form (HOOS-PS) with Patient-reported Outcomes Measurement Information System Physical Function (PROMIS PF).
From 2017 to 2020, a retrospective analysis was undertaken on patients with hip osteoarthritis who received routine clinical treatment from an orthopedic surgeon. The study comprised 3,382 individuals who completed 7,369 pairs of HOOS-PS and PROMIS PF measurements at a single nonsurgical, preoperative, or postoperative time point. In the linkage analysis sample, they included one randomly selected time point of scores for each patient. The linkage accuracy was examined using four methodologies, including equipercentile and item response theory-based approaches.
The correlation between PROMIS PF and HOOS-PS scores was strong (r=-0.827 for raw HOOS-PS scores and r=0.820 for summary HOOS-PS scores). The assumptions for the equipercentile and item response theory linking techniques were satisfied. They chose the Stocking-Lord technique based on item response theory as the best crosswalk and calculated item parameters for the HOOS-PS items using the PROMIS measure. A sensitivity study revealed that the crosswalk estimates were generally robust in nonsurgical, preoperative, and postoperative patients.
These crosswalks may be used to translate HOOS-PS and PROMIS PF metric scores at the group level, which is useful for data aggregation. Individual patient-level data conversion is not suggested due to the higher chance of mistakes.