Comparing Measurement Methods in Health State Evaluation - Case of the Burden of Tinnitus

Abstract

The objective of this study is to compare four methods for valuing health states in their ability to reflect the burden of Tinnitus. Valuations arr elicited from 210 patients using frequently applied measurement methods: Time Tradeoff (TTO), Standard Gamble (SG) and two different versions of Visual Analog Scales (VAS). Correlation, factor and regression analysis are performed to investigate whether these valuations corresponded to impairments of Tinnitus. The standardised form of the VAS, which is assumed to correct for reference point biases, perform best. Its correlation with symptoms is highest compared to other methods. Factor analysis supports this finding. In addition, least-square regression analysis show that standardised VAS explain more of overall variance than other instruments. However, a definite judgement on the best measurement method cannot be made since the analysis was restricted to a single health state. Further research, including a broader range of health conditions, is required.