Why EQ-5D and the Oxford Hip and Knee scores do not measure the same things

Abstract

Introduction: The outcome of treatment is always paramount for patients and healthcare professionals. Patient-reported outcome measures have been developed to measure outcomes.Since 2009, all patients in England having hip and knee replacement surgery have been asked to complete the generic EuroQol EQ-5D-3L and EQ visual analogue scale (EQ-VAS) and the condition-specific Oxford Hip Score or Oxford Knee Score for hips and knees, respectively.

Methods: EQ-5D-3L has five dimensions with three options each. Each combination has been scaled relative to the best conceivable health state (value 1.0) and the state of dead (value 0) to produce a relative severity score (EQ-Index) with a range from -0.594 to 1.0. This can be used to calculate quality-adjusted life-years. The EQ-VAS is a visual analogue scale from 0 (dead) to 100 (best conceivable health state).The Oxford Hip and Knee scores are similar to each other. They have 12 questions with five options each, scored 0-4. These scores are added, giving a scale with range 0 (no problems) to 48 (extreme problems on all questions).Using over 40 000 records for patients undergoing hip and knee replacements from the National Health Service patient-reported outcome measure database, we compare EQ-5D-3L with the Oxford Hip and Knee scores. To aid comparison, each score was transformed arithmetically to a common 0 (floor) to 100 (ceiling) scale.

Results: EQ-Index, EQ-VAS and the Oxford scores give very different results in terms of change, effect size and correlation.

Discussion: More research is needed, but some speculative ideas are put forward, which could explain these findings.

Keywords: Health Impact Assessment; Patient Reported Outcome Measures; Quality improvement; Quality improvement methodologies; Quality measurement.