Abstract
Background & aims
Malnutrition is common in hospitalized patients with cancer, yet guideline-recommended nutritional care is implemented inconsistently. We compared nutritional care processes in hospitalized patients with solid tumors on wards with and without a ward-based clinical dietitian. The primary objective was to assess receipt of nutritional intervention among patients meeting a predefined clinical indication. Secondary objectives were discharge nutritional recommendations, NRS-2002 documentation quality, and GLIM-defined malnutrition identification.Methods
Single-center cross-sectional census study at a tertiary oncology hospital (October 2024). Of 1382 hospitalized patients, 986 with solid tumors across 19 wards were included; hematological malignancies and sarcomas, treated almost exclusively on non-dietitian wards were excluded. The primary outcome was receipt of nutritional intervention (dietetic consultation, oral nutritional supplements, enteral or parenteral nutrition) among patients meeting an ESPEN-aligned clinical indication (NRS-2002 ≥ 3, GLIM-defined malnutrition, weight loss >10%, or age-adjusted low BMI). Multivariable logistic regression was adjusted for age, sex, tumor site and disseminated disease with cluster-robust standard errors.Results
Overall, 369 patients met the clinical indication. Patients on dietitian wards were twice as likely to receive nutritional intervention (adjusted rates 91.4% vs 45.2%; adjusted OR 17.32, 95% CI 6.42-46.72), consistently across sensitivity analyses. Discharge nutritional recommendations were also more frequent (78.7% vs 27.8%; adjusted OR 13.48, 95% CI 6.18-29.40). NRS-2002 documentation quality and GLIM-defined malnutrition identification were not independently associated with dietitian presence.Conclusion
Ward-based dietitian presence was associated with a two-fold higher rate of nutritional intervention and more frequent discharge recommendations, but not with screening or diagnostic quality. The single-center, non-randomized design precludes causal inference and limits transportability; multicenter confirmation is needed.