**Background:** Achieving adequate nutrition in hospitalised oncology patients is essential for optimal health outcomes, yet patients often fail to meet nutritional requirements due to disease and treatment side effects. The relationship between nutritional intake (energy and protein) and clinical outcomes such as length of stay (LOS) and 30-day readmission rates in oncology populations is not well understood. This study aimed to investigate these interrelationships in a cohort of hospitalised adult oncology patients at a specialist tertiary cancer centre.
**Methods:** This retrospective cross-sectional study was conducted at the Peter MacCallum Cancer Centre (117 beds) in Melbourne, Australia. Nutritional intake data were obtained from patients admitted between 25 May and 25 July 2022. Inclusion criteria: adult patients (≥18 years) admitted and discharged during the study period, consuming regular ward meals, clear fluid, liquid, or pureed foods. Exclusion criteria: deceased during study period, nil by mouth, receiving only enteral or parenteral nutrition, or receiving end-of-life care. Of 465 participants with food intake data, 331 (71.2%) met inclusion criteria. Estimated energy requirements were calculated as 125 kJ/kg body weight/day and protein requirements as 1.2 g/kg body weight/day per ESPEN guidelines. Nutritional intake was estimated using the Mobile Intake® system, a validated 5-point visual scale completed by trained healthcare professionals. Due to COVID-19-related staffing shortages, dinner data were incomplete; a scaling factor was derived from a subgroup with complete dinner data and applied to estimate mean daily intakes. LOS and 30-day readmission data were obtained from electronic medical records. Statistical analyses included independent t-tests, Mann–Whitney tests, chi-squared tests, Kruskal–Wallis H tests, and multivariable linear and binary logistic regression.
**Key Results:** At admission, 28.3% of participants were identified as at risk of malnutrition (MST ≥ 2). Mean daily intake was 75.22 ± 34.33 kJ/kg/day energy and 1.00 ± 0.54 g/kg/day protein. Only 9.1% (n=30) met estimated daily energy requirements, and 25.4% (n=84) met estimated daily protein requirements. Participants at risk of malnutrition consumed significantly less energy (−898.89 kJ, p < 0.001) and protein (−10.37 g, p = 0.015) compared with those not at risk. Median LOS was 5 days (IQR: 3–10 days). No statistically significant relationship was found between nutritional intake per kilogram body weight per day and LOS (energy: p = 0.286; protein: p = 0.651). However, participants at risk of malnutrition had a longer LOS by 1.33 days (p = 0.008). The multiple regression model explained 10.2% of the variance in LOS (adjusted R² = 0.102, p < 0.001), with inpatient diet type being the largest contributor (beta: 0.273, 95% CI: 0.162–0.384, p < 0.001). The all-cause 30-day hospital readmission rate was 20.2% (n=67). No association was found between readmissions and mean daily nutritional intake. Age (r = −0.133, p = 0.015), presence of metastases (r = 0.125, p = 0.02), and longer LOS (1.34 days, r = 0.145, p = 0.02) were associated with readmissions. Sarcoma (43.5%), gynaecological (36.8%), and lung (40.0%) cancers had the highest readmission rates. The logistic regression model explained 23.9% of the variability in readmissions (Nagelkerke R² = 0.239).
**Clinical Implications:** This study found no direct relationship between nutritional intake and LOS or 30-day readmission rates in hospitalised oncology patients, consistent with several previous mixed-diagnostic cohort studies. However, malnutrition risk at admission was associated with both lower intake and longer LOS, suggesting that early identification and management of malnutrition may be more impactful than focusing on intake alone. The high rates of failure to meet estimated energy (90.9%) and protein (74.6%) requirements highlight the need for improved nutritional support strategies. The variation in readmission rates by cancer type (particularly sarcoma, gynaecological, and lung cancers) suggests that diagnosis-specific approaches to nutritional care and discharge planning may be warranted. Limitations include the short study duration, missing dinner data (modelled via scaling factor), lack of data on between-meal snacks and oral nutritional supplements, and inability to distinguish planned from unplanned readmissions. Future research should focus on cancer diagnoses with increased malnutrition risk and include unplanned readmissions and emergency department presentations.