Serum albumin and age are predictors of length of hospital stay in surgical patients

The purpose of our project was to identify inexpensive nutritional indices that predict the length of hospital stay in surgical patients. Demographic, anthropometric, dietary, and biochemical measurements were taken on 110 patients upon their admission to general surgical wards at Security Forces Hospital in Riyadh, Saudi Arabia. Patients were divided into two groups 1) digestive cases (except biliary tract, DC) and 2) non-digestive cases (NDC). Length of hospital stay was calculated from discharge information. Statistical analyses were performed using student's t-test, Chi-square, and multiple linear regression. Patients' mean age was 38.7 +/- 16.2 years (mean +/- SD) and there was no difference in age between the two groups. Patients in the DC group had a shorter length of stay (5.1 +/- 3.0 days) than patients in the NDC group (7.5 +/- 5.5 days; P < 0.01) Age and serum albumin were the two variables that best predicted length of hospital stay. Thirty-six percent of DC patients and 31% of NDC patients had serum albumins < 35 g/L. DC patients with low serum albumins (< 35 g/L) stayed in the hospital three days longer than patients with normal serum albumins (> 35 g/L) (P < 0.01) and in the NDC group, patients with low serum albumins stayed in the hospital five days longer than patients with normal serum albumins (P < 0.01). We conclude that regression equations which include age and serum albumin can be used to predict the length of hospital stay for surgical patients. Copyright (C) 1996 Elsevier Science Inc.