Floor and ceiling effects in the OHS: an analysis of the NHS PROMs data set

Abstract

Objectives: The objective was to examine whether the Oxford Hip Score (OHS) demonstrated a floor or a ceiling effect when used to measure the outcome of hip replacement surgery in a large national cohort. Setting: Secondary database analysis of a national audit conducted in England and Wales on patient undergoing hip and knee arthroplasty in a secondary care setting. Participants: 93 253 primary arthroplasty patients completed preoperative OHS questionnaires and 69 361 completed 6-month postoperative OHS questionnaires. The population had a mean age of 67.78 (range 14–100, SD 11.3) and 59% were female. Primary Secondary Outcome Measures: Primary outcome measure was the Oxford Hip Score (OHS). Secondary outcome measures were the OHS-FCS and OHS-PCS. Floor and ceiling effects were considered present if >15% of patients achieved the worst score/floor effect (0/48) or best/ceiling effect (48/48) score. Results: Preoperatively, 0% of patients achieved the best score (48) and 0.1% achieved the worst score (0). Postoperatively, 0.1% patients achieved the worst score, but the percentage achieving the best score increased to 11.6%. Subgroup analyses demonstrated that patients between 50 and 59 years of age had the highest postoperative best score, at 15.3%. The highest postoperative OHS worst score percentage was in a group of patients who had a preoperative OHS above 41/48 at 28%. Furthermore, 22.6% of patients achieved the best postoperative OHS-PCS and 19.9% best postoperative OHS-FCS. Conclusions: Based on NHS PROMS data the overall OHS does not exhibit a ceiling or floor effect and should continue to be used as a valid measure of patient-reported outcomes for patients undergoing total hip arthroplasty. However, subscale analysis does indicate some limitations in the OHS-PCS and OHS-FCS. Trial registration number: NDORMS. Introducing standardised and evidence-based thresholds for hip and knee replacement surgery. The Arthroplasty Candidacy Help Engine (ACHE tool). HTA Project 11/63/01.