Lung Cancer 2022 March 29 [Link]
Michael Koller, Jammbe Z Musoro, Krzysztof Tomaszewski, Corneel Coens, Madeleine T King, Mirjam A G Sprangers, Mogens Groenvold, Kim Cocks, Galina Velikova, Hans-Henning Flechtner, Andrew Bottomley
Objectives: A minimally important difference (MID) is the smallest difference in quality of life (QoL) perceived as relevant by patients or clinicians. MIDs aid interpretation of QOL data in research and clinical practice. We aimed to determine MIDs for the EORTC QLQ-C30 for patients with lung cancer or malignant pleural mesothelioma.
Materials and methods: Data were drawn from two EORTC-sponsored randomized clinical trials (RCTs): a three-arm RCT of two cisplatin-based treatments and paclitaxel plus gemcitabine in advanced non-small-cell lung cancer, and an RCT comparing cisplatin with or without raltitrexed in patients with malignant pleural mesothelioma. MIDs for interpreting within-group change and between-group differences in change over time were computed using anchor-based approaches, for improvements and deteriorations separately. Distribution-based approaches provided corroborative evidence.
Results: The combined data from the trials comprised 730 patients. Available data allowed us to determine 8/14 anchor-based MIDs for EORTC scales for improvements, and 9/14 MIDs for deterioration. Furthermore, we provided distribution-based estimates for all 14 QLQ-C30 scales. Most MIDs for improvements ranged between 5 and 10, for both within-group and between-group differences. Outliers were appetite loss and constipation, with MIDs up to 15 score points. MIDs were slightly larger for within-group deterioration, ranging from -5 to – 15, with the largest for Nausea/vomiting (-1 to 4) and Appetite loss (-1 to 5). MIDs for between-group differences in deterioration ranged from – 4 (Physical, Role, and Social functioning, and Global quality of life) to -9 (Nausea/vomiting, Appetite loss and Constipation).
Conclusions: MIDs vary over scales and for between- versus within-group comparisons; this must be taken into account when interpreting changes. Nevertheless, the majority of MIDs range between 5 and 10 score points, in line with previously used thresholds for QLQ-C30. These findings and those from other tumor-specific MID analyses will inform a planned consensus process identifying commonalities and differences across tumor sites.