A comparison of the EuroQol and the Health Utilities Index in patients treated for congenital anomalies

It seems generally believed that the HUI3 is a more responsive utility measure than the EQ-5D because of the crude level structure of the EQ-5D compared to the HUI3. As empirical evidence to support this hypothesis is lacking, we undertook a study to compare the construct validity of the utility indices of the EQ-5Dindex and the HUI3index in 135 patients treated for congenital anorectal malformation and 57 patients treated for congenital diaphragmatic hernia. Discriminant validity was tested by the ability of the HUI3index and EQ-5Dindex to distinguish clinically relevant subgroups in the patient populations. Convergent validity was tested using Pearson correlations of the HUI3index and the EQ-5Dindex with the symptom scores. In general the index scores were in line with expectations: the higher the level of symptomaticity, the lower the indices. The HUI3 classified patients in more health states than the EQ-5D and classified a smaller percentage in the best health state. Nevertheless, the EQ-5Dindex distinguished more clinically relevant subgroups than the HUI3index. The two indices were correlated weakly to moderately with the symptom scores, which was expected given the relatively good quality of life of the patients. Despite the higher number of health states of the HUI, the EQ-5Dindex discriminated more clinically relevant subgroups than the HUI3index. This means that discriminative power of the utility indices is not determined merely by the number of health states of the underlying classification system. Differences in the description of the health states might explain this finding.