"Nutritional" markers as prognostic indicators of postoperative sepsis in cancer patients.

A study was undertaken to identify the nutritional parameters associated with a high risk of postoperative sepsis. The nutritional status of 162 cancer patients subjected to clean or clean-contaminated elective surgery was preoperatively evaluated according to the following parameters: percentage weight loss, arm circumference, triceps skinfold, arm muscle circumference, creatinine-height index, total serum protein, serum albumin, total iron-binding capacity, cholinesterase, peripheral lymphocytes, complement C3-C4 components, and skin tests. Patients were followed postoperatively according to a precise protocol to classify them as infected or noninfected. Postoperative sepsis was present in 40 patients who had significantly different mean values for four nutritional parameters from those of 114 patients with no complications, ie, total serum protein, 6.60 vs 6.99 g/dl, p = 0.008; serum albumin, 3.39 vs 3.66 g/dl, p = 0.001; total iron-binding capacity 301.32 vs 337.17 mmg/dl, p = 0.006; and cholinesterase, 2389.77 vs 2770.10 mU/ml, p = 0.005. Moreover, the relative risk and the attributable risk for these variables were evaluated and the significance was tested by the chi 2 test. By using multiple logistic analysis it appeared that only total serum protein and total iron-binding capacity gave an independent contribution to the risk of postoperative sepsis, while serum albumin disappeared and cholinesterase became non significant when the contribution of the first two variables was accounted for. It was also possible to identify, in a small number of patients, combinations of two variables that were associated with a very high risk of postoperative sepsis.(ABSTRACT TRUNCATED AT 250 WORDS)