Psychometric Validation of the PROISCD-CG Scale in Patients With Chronic Gastritis: A Structural Equation Modeling Approach

Abstract Background: Chronic gastritis (CG) is a common gastrointestinal condition that adversely affects patients’ physical, psychological, and social functioning. With increasing emphasis on patient-centered care, patient-reported outcomes (PROs) have become essential for evaluating disease burden and treatment effectiveness. Compared with conventional quality-of-life (QOL) instruments, PRO scales offer superior sensitivity, symptom specificity, and clinical interpretability. This study assessed the reliability and validity of the Patient-Reported Clinical Outcome Scale for Chronic Gastritis (PROISCD-CG) using structural equation modeling (SEM). Methods: A total of 174 outpatients and inpatients diagnosed with CG between September and December 2022 were surveyed using the PROISCD-CG scale. Internal consistency, split-half reliability, and criterion validity (with SF-36 as the external criterion) were evaluated. SEM was applied to examine and refine factor structures, assess model fit, and evaluate convergent and discriminant validity. Results: The total scale demonstrated strong reliability (Cronbach’s α = 0.87; split-half = 0.77). Criterion validity against SF-36 was acceptable, with moderate correlations across corresponding domains. The initial SEM model showed suboptimal fit, but modifications—removing low-loading items and adding error covariances—substantially improved fit indices (CFI, IFI, GFI approaching 0.80; RMSEA = 0.082). Convergent validity (AVE) and composite reliability (CR) generally met recommended thresholds, and discriminant validity was acceptable. Conclusions: The PROISCD-CG scale exhibits good psychometric properties and can serve as a reliable and valid tool for assessing patient-centered outcomes in CG. Items with low factor loadings require refinement and further testing in larger cohorts. PRO-based evaluation provides advantages over traditional QOL metrics by more precisely capturing symptom burden, functional impairment, and treatment responsiveness.