Validity of Estimating EuroQol (EQ5D) from a Health Assessment Questionnaire (HAQ).
10.4078/jkra.2010.17.3.263
- Author:
Soo Kyung CHO
1
;
Yoon Kyoung SUNG
;
Hyeseon LEE
;
Sang Cheol BAE
Author Information
1. Department of Rheumatology, Hospital for Rheumatic Disease, Hanyang University, College of Medicine, Seoul, Korea. scbae@hanyang.ac.kr
- Publication Type:Original Article
- Keywords:
Health Assessment Questionnaire (HAQ);
EuroQol (EQ5D)
- MeSH:
Arthritis, Rheumatoid;
Cross-Sectional Studies;
Humans;
Linear Models;
Mitoxantrone;
Quality of Life;
Quality-Adjusted Life Years;
Surveys and Questionnaires;
Rheumatic Diseases
- From:The Journal of the Korean Rheumatism Association
2010;17(3):263-271
- CountryRepublic of Korea
- Language:Korean
-
Abstract:
OBJECTIVE: Assessment of health-related quality of life in patients with rheumatoid arthritis (RA) has become important in health research. Health economists have used linear regression equations to mathematically transform changes in HAQ scores into EQ5D data, which can be used to calculate quality adjusted life years (QALYs). We aimed to examine whether a given approach is justified. METHODS: A total of 223 patients with RA were recruited from the Hospital for Rheumatic Diseases at Hanyang University. They completed the HAQ and EQ5D and a correlation analysis was performed between the two instruments. We compared HAQ and EQ5D score changes for patients who completed the EQ5D and HAQ at first and second visits (n=159). Predicted EQ5D was estimated from the HAQ using the calculating method of Bansnack et al. The mean difference between the predicted EQ5D from the HAQ and observed health utility score at the first visit and change during the study were tested by the paired t-test. RESULTS: In the cross-sectional study, EQ5D scores were moderately inversely correlated with HAQ (r=-0.716, p<0.001). However, the predicted EQ5D from the HAQ was significantly different from the observed EQ5D (p=0.001; 95% confidence interval [CI] 0.020~0.079). The change in EQ5D was also inversely correlated with the change in the HAQ (r=-0.615, p<0.001), and change in the predicted EQ5D scores corresponded well with changes in observed health utility scores (p=0.155; 95% CI (-0.0873~0.0140). CONCLUSION: Changes in predicted EQ5D corresponded with observer changes in EQ5D, suggesting that it may be better to use predicted EQ5D form HAQ to identify change in the quality of life.
