Objectives: A simple method of numerical scoring of value-sets emerging from EuroQol five-dimensional questionnaires is proposed satisfying desired properties including meaningful comparison of patients and group of patients across time and space.
Methods: Data driven weights are assigned to different levels of different dimensions and continuous scores (Y)corresponding to a health state are obtained as weighted sum. Y-scores are further transformed by linear transformations to P-scores ∈ [1,100] following Normal distribution. Illustration of the proposed method of scoring health states and its properties are given with a hypothetical data
Results: The method helps in assessment of health profile, progress/deterioration with time, classifying and ranking of a group of patients, meaningful comparisons and application of statistical techniques in parametric set up along with computation of treatment effectiveness, reliability avoiding unidimensionality assumption. The method may work best for sample of patients suffering from same disease
Conclusions: The approach makes no assumption about distributions of dimension/ level scores.
The proposed simple measures will help the researchers and practitioners to make meaningful analysis and draw meaningful conclusions. The proposed scoring method is applicable for EQ-5D-5L, EQ-5D-3L and EQ-5D-Y. Future studies suggested on psychometric properties of the proposed scoring system.
Satyendra Nath Chakrabartty