Improving the Hearing Status Discrimination of the Health Utilities Index, Mark 3: Design of the Hearing Status Classification System.

OBJECTIVE

Redesign the health status classification system of the Health Utilities Index, Mark 3 (HUI-3) "hearing" attribute to improve its discrimination of hearing-impaired health states.

STUDY

DESIGN

Domain and item selection from a previously generated item set guided by a domain and item importance survey, structural independence, and cognitive interviews with patients.

SETTING

Tertiary referral center.

PARTICIPANTS

Patients with a range of hearing loss severities, etiologies, and treatment experiences participated in the domain and item importance survey (n = 108) and hour-long cognitive interviews (n = 10).

MAIN OUTCOME AND MEASURES: Subattributes and levels for the novel Hearing attribute. Domain and item importance was scored on a seven-point Likert scale (1, not at all important; 7, extremely important).

RESULTS

Mean domain importance was highest for "speech in noise" (6.21; 95% confidence interval, 5.98-6.43) and lowest for "benefits of hearing loss" (3.46; 95% confidence interval, 3.03-3.89). Domains with moderate or greater ( r ≥ 0.5) domain importance Pearson correlation or construct overlap that interfered with structural independence were combined into single subattributes. Iterative adjustments to instructions, items, and phrasing were guided by cognitive interviews to derive the final instrument with seven subattributes: speech, environmental sounds, localization, listening effort, tinnitus, music, and assistive devices. The novel hearing attribute defines 25,920 unique hearing states.

CONCLUSION

The novel HUI-hearing is part of a comprehensive health utility instrument designed for individuals with hearing loss. Pending derivation of a hearing single attribute utility function and evaluation of measurement properties, our innovative approach could be used to improve health utility measurement for impairments described by any of the other HUI-3 attributes.