Abstract
PURPOSE: To derive preference weights in Trinidad and Tobago for Quality of Well-being Scale (QWB) health states in order to calculate QWB scores that can be compared to scores calculated from US-derived preference weights. The comparison was to determine whether the QWB scores from these different preference weights would lead to similar conclusions. METHODS: We conducted in-person household interviews to elicit preferences for 65 health states using a probability sample of 235 adults from Port of Spain, Chaguanas and San Fernando, Trinidad and Tobago. A regression model with correction for within-person clustering of observations was used to obtain preference weights based on case judgments on a 0 (dead) to 10 ("perfect health") scale. The independent variables were the components of the QWB entered as indicator (0, 1) variables. RESULTS: One hundred and nineteen (51%) respondents provided ratings. The respondents that provided ratings were demographically no different from those that did not. The QWB response patterns were very similar using Trinidad and US weights. The mean (SD) QWB score was 0.750 (0.130) for female respondents and 0.784 (0.125) for male respondents using Trinidad coefficients (t2, 233=-2.05, P=0.04) and 0.747 (0.131) for female respondents and 0.783 (0.126) for male respondents using US weights (t2, 233=-2.17, P=0.03). CONCLUSIONS: Overall, we found the US and Trinidad and Tobago weights were highly similar and that the choice of either set of weights would lead to similar conclusions.