Deriving a mapping algorithm for converting SF-36 scores to EQ-5D utility score in a Korean population

Area of Science:

• Health Economics
• Psychometrics
• Biostatistics

Background:

• Limited research exists on mapping algorithms between the EQ-5D and SF-36 in Korea.
• Developing such algorithms is crucial for consistent health outcome assessment.

Purpose of the Study:

• To derive and validate predictive models for converting the SF-36 health profile to the EQ-5D index using Korean data.
• To establish reliable methods for health status measurement and comparison.

Main Methods:

• Individual data from 2211 participants across three studies were used, split into derivation, internal, and external validation sets.
• Ordinary Least-Squares (OLS) regression, two-part modeling, and multinomial logistic modeling were employed.
• SF-36 scale and summary scores served as independent variables to predict the EQ-5D index and its dimensions.

Main Results:

• Three distinct scoring algorithms were evaluated, showing comparable performance based on Mean Absolute Error (MAE) and R2 values.
• The OLS model, incorporating specific SF-36 dimensions, was identified as the optimal algorithm.
• Consistent MAE values were observed across derivation (0.087-0.109) and external validation (0.082-0.097) sets.

Conclusions:

• The study successfully provides mapping algorithms to estimate the EQ-5D index from the SF-36 profile in Korea.
• These algorithms demonstrate significant explanatory power and low prediction errors, facilitating cross-walk analyses.
• The findings support the use of these models for robust health outcome evaluations in the Korean population.