Abstracts

Rationale: Estimate minimally important change (MIC) to facilitate analysis and interpretation of QOLIE-31 scores. Methods: Data from three placebo-controlled double-blind phase II/III trials of adjunctive lacosamide in partial onset seizures (POS) were pooled. Patients completed the QOLIE-31 at baseline and at endpoint (after 16-18 weeks of treatment or early discontinuation, n=1046), and a 7-point Patient Global Impression of Change (PGIC) scale at endpoint (Phase III studies only, n=805). Distribution-based MIC values were calculated based on a 0.3 Effect Size (ES), 0.3 Standardized Response Mean (SRM), one Standard Error of Measurement (SEM), and one Reliable Change Index (RCI). Anchor-based MIC values were calculated as the mean QOLIE-31 score change for different levels of PGIC answers. Receiver Operating Characteristic (ROC) curves were also used. Results: Distribution-based methods yielded QOLIE-31 MIC values of 4.73 (ES), 3.90 (SRM), 5.23 (SEM) and 7.39 (RCI) for the Total score, on a range of 0 to 100. Larger values were found for the Social Functioning, Seizure Worry and Medication Effects subscales, especially with SEM and RCI calculations. For anchor-based methods, combining the minimally and much improved and worsened categories of the PGIC resulted in similar MIC as the previous methods, with 5.19 for the Total score. The ROC method yielded an optimal threshold of 3.88. Conclusions: The various methods used produced generally comparable QOLIE-31 MIC values (approximately 5 out of 100 for the Total score). These MIC values were considerably lower than those reported in the literature (10-11 points). A number of aspects including variation in population and study design most likely contributed to those differences. These results suggest that smaller MIC values should be considered when analyzing and interpreting QOLIE-31 score changes in refractory patients with POS.