Association of Nutritional Support With Clinical Outcomes in Malnourished Cancer Patients: A Population-Based Matched Cohort Study

Malnutrition is prevalent in hospitalized cancer patients and has been associated with poor therapy response and unfavorable clinical outcome. While recent studies have shown a survival benefit through nutritional support in a hospitalized malnourished medical population including cancer patients, we aimed to investigate the association of nutritional support with in-hospital mortality and other clinical outcomes in a nationwide inpatient cancer population. In this population-based cohort study, using a large Swiss administrative claims database from April 2013 to December 2018, we created two cohorts of malnourished cancer patients on medical wards. We generated two pairwise cohorts of malnourished patients who received nutritional support by 1:1 propensity-score matching to patients not receiving nutritional support. The primary outcome was all-cause in-hospital mortality. Secondary outcomes were 30-days all-cause hospital readmission and discharge to a post-acute care facility. To account for disease activity, we stratified patients either admitted for cancer as main diagnosis or admitted with cancer as comorbidity. Among 1,851,498 hospitalizations on medical ward, we identified a total of 32,038 malnourished cancer patients. After matching, 11,906 (37%) cases were included in the "cancer main diagnosis cohort" and 5,954 (18.6%) in the "cancer comorbidity cohort." Patients prescribed a nutritional support showed a lower in-hospital mortality in both cohorts as compared to their respective matched controls not receiving nutritional support [cancer main diagnosis cohort: 15.4 vs. 19.4 %, OR 0.76 (95% CI 0.69-0.83); cancer comorbidity cohort: 7.4 vs. 10.2%, OR 0.71 (95% CI 0.59-0.85)]. While we found no difference in 30-days readmission rates, discharge to a post-acute care facility was less frequent in the nutritional support group of both cohorts. In this large cohort study, nutritional support in hospitalized patients with either cancer as main diagnosis or comorbidity was associated with a lower risk of in-hospital mortality and discharge to a post-acute care facility.