Background & aims: Patients admitted to intensive care units (ICUs) often need enteral nutrition (EN) support. For patients with type 2 diabetes (T2D), standard EN formulas may not provide ideal nutrients. The purpose was to investigate whether use of a diabetes-specific formula (DSF) could provide clinical and health economic benefits (compared to standard formulas) in critically ill patients with T2D.
Methods: This study was a retrospective analysis of medical records and expenditure data covering a 5-year period (2009-2013) from the hospitalization database of the National Taiwan University Hospital. Records of ICU patients who had T2D and were receiving enteral feeding with either the DSF or non-diabetes-specific formula (non-DSF) for at least 5 days were included in the analysis. Mortality, ICU length of stay (LOS), diabetes-related medications, and total costs of care (including all costs covered by the National Health Insurance and private expenses) were considered as the primary outcomes.
Results: A total of 158 patient records were analyzed in the DSF group and 794 in the non-DSF group. The baseline demographics including age, gender, weight, body mass index (BMI), and comorbidity patterns were mostly comparable between the groups. Compared to those receiving non-DSF, patients with T2D receiving DSF were found to have significantly decreased mortality (5.1% vs. 12.3%, P = 0.0118) and reduced need for insulin prescription (29.1% vs. 38.4%, P = 0.0269). ICU LOS was shorter for DSF patients, but no statistical difference was found (13.0 days vs. 15.1 days, P = 0.1843). However, significantly lower total ICU costs were reported for DSF patients (6700 USD vs. 9200 USD, P < 0.0001).
Conclusions: The use of DSF in ICU patients with T2D is correlated with significant reduction in mortality and improved health economic outcomes.
Keywords: Critical illness; Diabetes; Diabetes-specific formulas; Enteral nutrition; Health economics.
Copyright © 2016 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.