Exploring multiple minimal clinically important differences for the esophageal cancer scale QLICP-ES (V2.0) based on anchor-based and distribution-based methods

The minimal clinically important difference (MCID) is an important phrase with big appeal in the interpretation of quality of life (QOL), and a bridge between statistical inference and clinical implication. This study aimed to explore multiple MCIDs of the QLICP-ES(V2.0), an esophageal cancer scale in Quality of Life Instruments system for Cancer Patients. The MCID were established by anchor-based and distribution-based methods. The item Q29 in EORTC QLQ-C30 was selected as an anchor, and Standard A (improves one level after treatment) and standard B (improves at least one level after treatment) were used to define effects in the anchor-based method, whereas methods of effect size (ES), standard error of measurement (SEM), and reliable change index (RCI) were used in the distribution-based method. A total of 232 esophageal cancer patients completed the QLICP-ES (V2.0) assessment. Under criterion A, 102 patients were included and the MCID values for the physical, psychological, social, common symptoms and side effects domains, the general module, the specific module, and the overall score were 15.1, 4.4, 3.1, 6.7, 4.8, 8.5, and 6.0, respectively. Under criterion B, 165 patients were included and the MCID values for the aforementioned domains and overall score were 19.3, 4.2, 4.8, 7.7, 6.5, 9.5, and 7.5, respectively. The MCID values using three approaches of effect size, standard error of measurement and reliable change index were ranged from 1.1 to 13.3. Although the threshold values of anchor-based method and 1.96SEM were recommended to establish the MCID of QLICP-ES(V2.0) in patients with ES, several MCID values have been presented by different methods and standards so that users may select according to different situations.