Development and validation of the brain cancer scale among the system of quality of life instruments for cancer patients (QLICP-BN V2.0) based on classical test theory and generalizability theory

Quality of life(QOL) has increasingly gained attention in the medical field. Although specific QOL measurement scales for brain cancer, such as FACT-Br and QLQ-BN20, are available, culturally specific scales reflecting Chinese cultural characteristics are lacking. This study aims to develop and validate the brain cancer scale within the system of QOL instruments for cancer patients QLICP-BN (V2.0), which is suitable for Chinese patients with brain cancer. Using the programmed decision method and instrument development theory, including nominal group and focus group discussions, we developed the QLICP-BN (V2.0) using a modular approach and Chinese cultural characteristics. We used classical test theory (CTT) and generalizability theory (GT) to screen items and evaluate the scale’s reliability, validity, and responsiveness. Specifically, a generalizability study (G-study) in GT identified error sources, and a decision study (D-study) optimized the measurement design accordingly. QLICP-BN (V2.0) was developed with 5 domains: physical, psychological, social, common symptoms and side effects, and the specific module, and a total of 47 items can be considered the optimal number. Regarding reliability, the Cronbach’s α coefficient for the QLICP-BN (V2.0) is 0.919, the test-retest correlation coefficient (r) is 0.805, and the ICC in all domains and on the total scale was higher than 0.80. G-study findings indicate that domain score variation primarily stems from the “participant-item” interaction, and D-study demonstrates that the G-coefficient and ɸ index of the QLICP-BN (V2.0) were 0.926 and 0.899, respectively; the relative and absolute error values were 0.015 and 0.021, respectively, reaching the ideal level. With respect to validity, the scale demonstrates good content validity; Confirmatory factor analysis indicates that the scale dimensions fit well, indicating satisfactory construct validity; And criterion-related validity is acceptable. With respect to responsiveness, the pre-treatment total scale score was 57.321 ± 10.625, while the post-treatment total scale score was 64.125 ± 8.699, with a statistically significant difference ( t = -17.789, P < 0.001), and an effect size SRM = 1.057 representing excellent responsiveness. QLICP-BN (V2.0) has good validity, reliability, and responsiveness for the clinical assessment of QOL in Chinese patients with brain cancer.