Revisiting minimally important changes for the Oxford Hip and Knee scores

TL;DR

This study shows that the Oxford Hip and Knee scores have baseline dependency, meaning a single measure of meaningful change may not accurately reflect patient outcomes.

Contribution

The study quantifies baseline dependency in the Oxford Hip and Knee scores using a large dataset and multiple categorization methods.

Findings

The minimally important change (MIC) for the Oxford Hip and Knee scores varies depending on baseline scores.

Baseline dependency was observed regardless of the method used to categorize change scores.

Using a single MIC may not capture the full range of meaningful patient change.

Abstract

A number of measures have been proposed to evaluate meaningful within-person change in the Oxford Hip and Oxford Knee scores (OHS and OKS), however there is evidence of lower baseline scores being associated with higher change scores, that is, these instruments potentially demonstrate baseline dependency. The study aim was to identify and quantify the impact of baseline dependency for the OHS and OKS.

The data were collated from the National Health Service in England including the OHS, OKS, EQ-5D-3L and a global transition item (GTI). Change scores, including the minimally important change (MIC) were derived and categorised by the GTI and baseline scores for the OHS and OKS. Baseline dependency was evaluated using different baseline categories (OHS/OKS, EQ-5D quartiles and split-item method).

A total of 387,524 records were extracted. Although the overall MIC were in-line with previous research, the results showed these measures varied by pre-operative scores. Baseline dependency was present irrespective of the method employed to categorise change scores.

The MICs for both the OHS and OKS show distinct baseline dependencies. The use of a single MIC for either instrument is unlikely to capture the full range of meaningful change experienced by individual patients and therefore has implications for the interpretation of interventional outcomes with these instruments. A multifaceted approach involving multiple sources of patient-relevant measures is recommended to provide more robust measures for the evaluation of patient outcomes and healthcare services.

The online version contains supplementary material available at 10.1186/s41687-026-01024-1.

Figures