Abstract
PURPOSE: To evaluate demographic, clinical, and laboratory variables, and their associations with in-hospital mortality, among elderly internal medicine patients with nasogastric tube (NGT) feeding. PATIENTS AND METHODS: Demographic, clinical, and laboratory data were collected retrospectively for 129 patients aged ≥80 years who initiated NGT feeding during their hospitalization in internal medicine wards. The data were compared between survivors and non-survivors. Multivariate logistic regressions were performed to identify the variables most significantly associated with in-hospital mortality. RESULTS: The in-hospital mortality rate was 60.5%. Compared to survivors, non-survivors more often presented with pressure sores (P=0.005) and lymphopenia (P<0.001), were more often treated with invasive mechanical ventilation (P<0.001), and less often underwent geriatric assessment (P<0.001). Non-survivors demonstrated higher mean levels of C-reactive protein, and lower mean values of serum cholesterol, triglycerides, total protein, and albumin (P<0.001 for all comparisons). On multivariate analysis, the following variables were most significantly associated with in-hospital mortality in the entire cohort: the presence of pressure sores (odds ratio [OR], 4.34; 95% confidence interval [CI], 1.68-11.48; P=0.003) and lymphopenia (OR, 4.09; 95% CI, 1.51-11.08; P=0.006), and serum cholesterol (OR, 0.98; 95% CI, 0.96-0.99; P=0.003). CONCLUSION: Among elderly acutely ill patients who initiated NGT feeding during hospitalization, in-hospital mortality was extremely high. The factors most strongly associated with in-hospital mortality were the presence of pressure sores and lymphopenia, and lower serum cholesterol levels. These findings may provide useful prognostic information for decision-making regarding initiation of NGT feeding in elderly hospitalized patients.