Hauber B, Johnson FR, Ozdemir S. Conjoint analysis QALYs for acute conditions: an application to vasomotor symptoms. Poster presented at the 9th Annual European Congress of the International Society for Pharmacoeconomics and Outcomes Research; October 31, 2006. Copenhagen, Denmark.

OBJECTIVE: Demonstrate using conjoint analysis to obtain generalized, nonlinear, time-tradeoff estimates of quality-adjusted life-years (QALYs) for clinically relevant durations and severities of acute, non-fatal conditions such as vasomotor symptoms.

METHODS: A self-administered, web-enabled, graded-pairs conjoint-analysis survey elicited women''s preferences for the treatment of vasomotor symptoms. The included treatment attributes are frequency and severity of vasomotor symptoms (daytime hot flashes and nighttime sweats) and duration of these symptoms. Participants also considered vasomotor-symptom, treatment-related side effects. Observed tradeoffs between symptom duration and symptom relief were used to calculate generalized conjoint time equivalents for specified health states without first calculating utilities anchored at 0 and 1.

RESULTS: A total of 523 women with a mean age of 52 years completed the survey. For these women, improvement from severe (severe >6 hot flushes a day and >3 sweats a night) to moderate (moderate 3-6 hot flushes a day and 1-3 sweats a night) vasomotor symptoms over a 7-year treatment period is equivalent to 3.7 years of normal health, while improvement from moderate to mild (mild 1-2 hot flushes a day and no night sweats) symptoms is equivalent to 4.4 years of normal health. QALY benefits of symptom relief are larger for younger women (<=2) than for older women (>52). For example, an improvement from moderate to mild is equivalent to 3.3 years of normal health for younger women, where it is to 6.8 years of normal health for older women.

CONCLUSIONS: Conjoint analysis is a feasible method for estimating QALYs directly for acute health states. This approach avoids ad hoc approaches such as the chaining method. Its advantages over standard-gamble and time-tradeoff methods include avoiding clinically irrelevant tradeoffs involving death or life expectancy, avoiding restrictive assumptions such as linear time preferences, and eliciting preferences in the realistic context of both efficacy and side-effect risks.

Share on: