Objective: To investigate the measurement properties of a hearing item ('bolt-on') added to the EQ-5D-5L descriptive system.
Methods: Cross-sectional data on the EQ-5D-5L descriptive system with the addition of a hearing bolt-on item, namely, 'I have no/slight/moderate/severe/extreme problems hearing', was collected through face-to-face interviews with 418 participants of a community-based hearing loss screening program in Singapore. The distribution of responses to the EQ-5D and the hearing items were compared in terms of the proportion of 'no problems' and the Shannon index (H'). The measurement properties of the hearing item were assessed by testing its correlation with the hearing thresholds in dBHL; and with the standard EQ-5D items. The properties were also evaluated by comparing the proportion of 'full health' in participants with different hearing severities between standard and hearing bolt-on EQ-5D-5L; and the ability in identifying the participants between the standard EQ-5D-5L index score and the level sum score (LSS) of the hearing bolt-on system.
Results: Compared to the standard EQ-5D items, the hearing item has a lower proportion of 'no problems' (80.1% vs. 80.9-97.9%); more evenly-distributed responses (H' value: 0.92 vs. 0.18-0.87); and much stronger correlation with the hearing thresholds (|r|: 0.322 and 0.325 vs. 0.008-0.139). It generally has a weak correlation with EQ-5D items (|r|: 0.055-0.160); and reduces the proportion of 'full health' for the participants with different hearing functions with the percentage reduction varying from 6.6 to 22.7%. The LSS is more discriminative than the standard EQ-5D-5L index score as well (F-statistic: 12.71 and 13.93 vs.4.06 and 4.70).
Conclusions: A hearing bolt-on to EQ-5D-5L can validly measure hearing severities and is likely to increase the sensitivity of the resultant preference-based EQ-5D index to hearing loss.
Keywords: Bolt-on; EQ-5D; Hearing; Measurement properties; Singapore.
© 2022. Crown.