Using the effect size to model change in preference values from descriptive health status

Area of Science:

• Health Economics
• Psychometrics
• Biostatistics

Background:

• Translating group-level health status changes into preference values is crucial for economic analyses.
• Existing methods may not fully capture the nuances of preference shifts.
• Effect size (ES) offers a standardized metric for group-level health improvement.

Purpose of the Study:

• To develop and describe a modeling approach for translating group-level health status changes into preference value changes.
• To utilize effect size (ES) as a summary measure for group-level improvement.
• To inform economic analyses by predicting preference value changes from existing data.

Main Methods:

• The study employed a pilot modeling approach using effect sizes (ES), defined as standardized mean differences.
• Vignettes with varying severity decrements on the SF-12 mental health scale were valued by general practitioners (n=42) using rating scale (RS) and time trade-off (TTO) methods.
• Translation factors were developed to convert ES differences into predicted changes in preference values for five mental disorders.

Main Results:

• An effect size difference in health status corresponded to an average preference value difference of 0.171–0.204 using the RS method.
• Using the time trade-off (TTO) method, the average preference value difference was 0.104–0.158.
• These findings establish a quantitative link between effect size and changes in preference values.

Conclusions:

• The observed relationship between effect size and preference value changes may be specific to the measures used.
• Further research with diverse raters and preference measures is recommended.
• This approach holds potential for expanding the evidence base for modeling preference changes in economic evaluations.