Comparing statistically reliable and patient-perceived improvement following hip arthroplasty: Rasch-based and anchor-based approaches

Background Interpreting meaningful change in patient-reported outcome measures (PROMs) is essential for assessing hip replacement outcomes. The Oxford Hip Score (OHS) is widely used, yet statistically detectable change may not align with patient-perceived improvement. Objectives To compare Rasch-based classifications of OHS change, following Hobart’s effect size methodology, with anchor-based interpretations derived from patient Global Impression of Change (GIC). Methods OHS data from a large hip replacement cohort ( N = 184,509) were analysed using two complementary approaches. Rasch analysis transformed ordinal scores into interval-level person measures, and effect sizes with confidence intervals classified patients as significantly improved, non-significantly improved, unchanged, or worse. MIC was estimated using a five-point patient GIC. Median pre-operative, post-operative, and change scores were summarised per category. Results Both methods demonstrated graded outcome patterns. Concordance was strongest for large improvements, with statistically significant change closely matching reports of being “much better.” Divergence occurred for modest improvement and worsening: patients often reported little or no benefit despite measurable OHS improvement. Conclusions Rasch-based and anchor-based approaches capture overlapping but distinct constructs. Integrating both provides a nuanced interpretation of PROM change after hip replacement.