Mapping Health State Utility from Disease-Specific Measures in Spinal Muscular Atrophy and Paroxysmal Nocturnal Hemoglobinuria.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND: An economic evaluation is widely used to facilitate decision making regarding drug reimbursement in many healthcare systems. However, the absence of preference-based measurement in clinical trials has hindered the health economic evaluation of drugs for rare diseases. OBJECTIVE: This study aims to develop mapping algorithms that convert disease-specific scales-Spinal Muscular Atrophy Independence Scale (SMAIS) for spinal muscular atrophy and Functional Assessment of Cancer Therapy-Anemia (FACT-An) for paroxysmal nocturnal hemoglobinuria-into five-level EQ-5D (EQ-5D-5L) and SF-6D version 2 (SF-6Dv2) utility values, thereby enabling the economic evaluation of related drugs. METHODS: Data were collected from two online surveys conducted in China. Both direct and indirect mapping methods were explored, including ordinary least squares regression, Tobit regression model, censored least absolute deviation, generalized linear model, beta mixture regression, adjusted limited dependent variable mixture model, ordinal logistic regression (OLOGIT), and multinomial logistic regression (MLOGIT). Model performance was assessed by mean absolute error (MAE), root mean squared error (RMSE), and adjusted R-square (adjusted R2). The optimal model was selected based on the lowest average ranking value, derived from the MAE and RMSE through five-fold cross-validation. RESULTS: A total of 192 patients with spinal muscular atrophy and 306 patients with paroxysmal nocturnal hemoglobinuria were included in the analysis. For spinal muscular atrophy, the MLOGIT, which included SMAIS total score and sex as predictors, demonstrated the best performance, with the lowest MAE and RMSE (EQ-5D-5L: MAE: 0.1471; RMSE: 0.1839; adjusted R2: 0.5932; SF-6Dv2: MAE: 0.1208; RMSE: 0.1563; adjusted R2: 0.4323) after five-fold cross-validation. For paroxysmal nocturnal hemoglobinuria, the OLOGIT model using the FACT-An dimension score performed best (EQ-5D-5L: MAE: 0.1068; RMSE: 0.1431; adjusted R2: 0.5394; SF-6Dv2: MAE: 0.0877; RMSE: 0.1162; adjusted R2: 0.6754). CONCLUSIONS: These newly developed mapping algorithms enable the estimation of EQ-5D-5L and SF-6Dv2 utilities in the absence of a preference-based measurement, thus supporting health economic evaluations of therapies for spinal muscular atrophy and paroxysmal nocturnal hemoglobinuria.