Abstract / Summary
Malnutrition is common among patients undergoing chemotherapy after gastrointestinal cancer surgery and results in reduced treatment tolerance, impaired recovery and diminished quality of life. Existing nutritional assessment tools often address isolated aspects and lack a cohesive framework that incorporates chemotherapy-related side effects and regional dietary and cultural differences, thereby limiting their clinical effectiveness in personalized care. This study aims to develop and validate a clinically applicable and culturally tailored Nutritional Literacy Scale for Chinese patients receiving postoperative chemotherapy for gastrointestinal cancers and assess its psychometric properties. A two-phase mixed-methods design was adopted for scale development and multicenter cross-sectional validation. Phase I involved a preliminary scale construction through a systematic literature review and consultation with experts, utilizing Nutbeam’s health literacy framework and the knowledge-attitude-practice model. Phase II involved recruiting 404 Chinese patients with gastrointestinal cancers undergoing chemotherapy via purposive sampling. Eligible participants were randomly divided into exploratory ( N = 202) and confirmatory ( N = 202) subsamples, with 202 participants in each subsample. The derivation cohort underwent item analysis and exploratory factor analysis (EFA) to refine the items. The validation cohort underwent confirmatory factor analysis (CFA), validity tests (content validity, construct validity, criterion-related validity), reliability assessment (Cronbach’s α coefficient, split-half reliability, test-retest reliability), known-groups validity and clinical cutoff determination. The finalized Nutritional Literacy Scale comprised 33 items across four dimensions: Cognitive, Functional, Interactive and Critical literacy. The scale demonstrated high content validity (I-CVI = 0.83-1.00; S-CVI/Ave = 0.970). The EFA identified a four-factor structure explaining 61.80% of the variance. The CFA confirmed a good model fit (χ²/df = 1.221, RMSEA = 0.033, CFI = 0.971, TLI = 0.969, SRMR = 0.044, with WLSMV estimation employed), with factor loadings ranging from 0.642 to 0.804. The scale exhibited excellent reliability, as indicated by Cronbach’s α (0.960), split-half reliability (0.959) and test-retest reliability (0.926) .Furthermore, the scale demonstrated good known-groups validity and a clinical cutoff score of 91.5 was established via ROC analysis (AUC = 0.977), providing an effective tool for identifying “inadequate nutritional literacy”. The scale demonstrated satisfactory reliability and validity, a stable factor structure, and clinical applicability. It can be used as a preliminary, context-specific assessment tool for evaluating nutritional literacy among patients receiving postoperative chemotherapy for gastrointestinal cancers in Chinese tertiary hospitals. By incorporating chemotherapy-specific content and local adaptation, the scale may support high-risk patient identification and individualized nutritional interventions. Experts participated in expert meetings and Delphi consultations to provide professional insights and guide questionnaire development. Patients who underwent postoperative chemotherapy for gastrointestinal cancers participated in completing surveys to provide data on factors influencing nutritional literacy, aligning with the research aims.