Direct Mapping Between Oswestry Disability Index Scores and EQ-5D-3L Utilities: The Need for Utility Scores to...

Area of Science:

• Health Economics
• Rehabilitation Medicine
• Biostatistics

Background:

• Disease-specific questionnaires (e.g., Oswestry Disability Index, ODI) are common but not suitable for cost-utility analysis.
• Preference-based generic questionnaires (e.g., EuroQoL 5-dimension 3-level, EQ-5D-3L) are essential for cost-utility analysis but less frequently used in clinical practice.
• A mapping algorithm is needed to convert ODI scores to EQ-5D-3L utilities.

Purpose of the Study:

• To develop a direct mapping algorithm from Oswestry Disability Index (ODI) scores to EuroQoL five-dimension-three level (EQ-5D-3L) utilities.
• To validate the algorithm in Spanish patients diagnosed with low back pain.

Main Methods:

• Retrospective cross-sectional study with 296 Spanish low back pain patients (592 questionnaire pairs).
• Evaluated ordinary least squares, Tobit, and adjusted limited dependent variable mixture models.
• Used cross-validation to select the best model based on goodness-of-fit and accuracy.

Main Results:

• Excellent correlation observed between total ODI scores and EQ-5D-3L utilities (Spearman's rho = -0.910).
• The selected ordinary least squares model, using ODI total scores (including cubic terms), accurately predicted EQ-5D-3L utilities.
• The final model explained 90.56% of the variance with a root mean squared error of 0.1125.

Conclusions:

• Accurate prediction of EQ-5D-3L utilities from ODI total scores is feasible for low back pain patients.
• The developed algorithm facilitates the integration of clinical data (ODI) into economic evaluations (EQ-5D-3L utilities).