Assessing the Viability of the EQ-5D as Part of a Battery of Outcomes in Elderly Total Knee Arthroplasty Patients: A Comparison of Generic, Condition-Specific, and Preference-Based Patient-Reported Outcome Measures

Abstract

Background In the current economic climate it has become increasingly important to evaluate the effectiveness of expensive procedures such as total knee arthroplasties (TKAs). Patient-reported outcome measures, such as generic and condition-specific profile measures, are popular ways of determining outcomes. However, these cannot be used reliably for the purpose of economic evaluations. The EQ-5D, designed for cost-utility analysis, could offer potential advantages to outcome measurement; however, little is known about the viability and performance of this measure in more elderly (_ 75 years) patient cohorts.

Aims The aim of this study was to assess the viability of the EQ-5D for use in the evaluation of TKA and to provide justification for its continued inclusion in the clinical audit of patients undergoing arthroplasty in a large general hospital.

Methods Seven-hundred and seventy-nine consecutive patients undergoing TKA participated in this study. Self-report audit questionnaire booklets were administered at baseline (during a pre-admission clinic appointment) and six months post-operatively (postally). Booklets comprised of the Oxford Knee Scale (OKS;), 12-dimension Short Form (SF-12;), and the EQ-5D ().

Results Two-hundred and forty-six subjects completed both the baseline and follow-up audit surveys. Correlations were strongest between the EQ-5D and OKS instruments (baseline: Tb=-0.58; p0.01; follow-up: Tb=-0.41; p0.01),.. The EQ-5D displayed large effect sizes (d=0.94) and was able to detect clinically important HRQoL improvements (ROC P-value=0.65; CI0.08; p0.001), discriminate those patients who were experiencing poorer pre-operative health (p0.001), and detect those who deteriorated post-operatively (p0.001). Adequate reliability (Cronbach's =0.79) of the EQ-5D instrument was seen post-operatively as well.

Conclusions Results from this study reveal that the EQ-5D instrument shows good responsiveness to health changes in elderly TKA patients post-operatively. Its correlations with the OKS, detection of poorer pre-operative health, and detection of deterioration post-operatively also suggest good validity for the EQ-5D in this cohort. Finally, the relatively large reliability statistic post-operatively supports the use of the EQ-5D in audits. Together, these findings support the continued use of the EQ-5D instrument in future arthroplasty audits and provide evidence that it is viable for the use in cost-utility analysis.