Development and validation of a symptom assessment scale for gastric cancer patients undergoing neoadjuvant therapy: A COSMIN-guided study.

PURPOSE: To develop and validate a symptom assessment scale for gastric cancer patients undergoing neoadjuvant therapy based on the COSMIN framework and the Symptom Experience Model.

METHODS

An exploratory mixed methods design was used. Initial items were constructed through literature review, theoretical framework analysis, and qualitative interviews (n = 20), refined by Delphi consultation (n = 14) and psychometrically tested in 384 patients from three tertiary hospitals. Item analysis, content validity, structural validity, reliability, and known-group validity were evaluated. Structural validity was examined using exploratory factor analysis (EFA), parallel analysis, exploratory graph analysis (EGA), and confirmatory factor analysis (CFA).

RESULTS

The final instrument consisted of 25 symptom items and 6 symptom-interference items. Item analysis resulted in the removal of one item with inadequate result. Content validity was satisfactory, with item-level content validity indices ranging from 0.78 to 1.00 and a scale-level content validity index of 0.94. EFA, parallel analysis, and EGA collectively supported a four-factor structure comprising emotional-psychological symptoms, treatment-related systemic symptoms, gastrointestinal symptoms, and systemic/metabolic symptoms. CFA confirmed acceptable model fit (χ2/df = 2.119, RMSEA = 0.070, CFI = 0.911, TLI = 0.901, IFI = 0.912). Discriminant validity was supported (HTMT< 0.85). The overall scale demonstrated excellent internal consistency (Cronbach's α = 0.926; McDonald's ω = 0.929) and acceptable split-half reliability (0.746). Significant differences were observed between clinically distinct patient groups, supporting known-group validity.

CONCLUSION

The developed symptom assessment scale demonstrates satisfactory validity and reliability, and may serve as a valuable instrument for symptom assessment and supportive care in this population.