Background & aims: Reducing length of stay (LOS) is a priority for hospitals but patients' decline in nutritional status may have a negative impact. The aims of the study were to assess the change in nutritional status during hospitalization and determine if its decline is associated with prolonged LOS.
Methods: This is a prospective cohort study conducted in 18 Canadian hospitals. Subjective global assessment (SGA) and weight measurements were performed at admission and discharge. Patient information was collected at admission and extracted from the chart during hospitalization. Association between LOS and changes in SGA or weight loss ≥5% was tested using multivariate Cox PH approach. Results are expressed as hazard ratios (HR) and their 95% CI.
Results: 409 patients (53% male) with a LOS >7 days were analyzed. Patients' median (q1,q3) age was 68 years (58,79) and LOS was 11 days (8,17). At admission, 49% of patients were well nourished (SGA A), 37% were moderately malnourished (SGA B) and 14% were severely malnourished (SGA C). From admission to discharge, 34% remained well-nourished, 29% remained malnourished (SGA B or C), 20% deteriorated and 17% improved. Of the 409 patients, 373 had weight measurements at admission and discharge: 92 (25%) had ≥5% weight loss. Multivariate models showed that after adjusting for covariates, decline in nutritional status from SGA A to B/C or SGA B to C (HR: 0.62, CI: (0.44, 0.87); HR: 0.35, CI: (0.20, 0.62) respectively) and weight loss ≥5% (HR: 0.52; CI: 0.40, 0.69) were significantly associated with longer LOS.
Conclusion: In-hospital decline in nutritional status as assessed by SGA or weight loss ≥5% is associated with prolonged LOS independently of factors reflecting demographics, living accommodations and disease severity. This suggests a role for nutrition care in reducing LOS.
Keywords: Hospital malnutrition; Length of stay; Nutritional decline; Nutritional status.
Copyright © 2015 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.