Propensity score analysis confirms the independent effect of parenteral nutrition on the risk of central venous catheter-related bloodstream infection in oncological patients

Background & aims: Parenteral nutrition is known as a high-risk factor for central venous catheter-related bloodstream infection (CVC-RBSI) in cancer patients. Owing to ethical and technical problems, the studies in the literature have nonrandomized designs and are therefore often confounded by biases. We performed a propensity score analysis to estimate the effect of parenteral nutrition on CVC-RBSI in digestive cancer patients who underwent chemotherapy. Methods: Data were collected prospectively. A logistic regression model was used to calculate a propensity score, which was the probability of receiving parenteral nutrition. Kaplan-Meier survival and Cox regression model were used to estimate the effect of the parenteral nutrition on CVC-RBSI after adjustment for the propensity score. Results: Before the propensity score analysis, the differences between patients with (n = 113) and without (n = 312) parenteral nutrition were identified including: male gender, body weight, weight loss, performance status, location of primary cancer, FOLFIRI, and previous long-term corticotherapy. After propensity score stratification, all of the covariates were balanced within each stratum. After adjustment, patients with parenteral nutrition were at a higher risk for CVC-RBSI. Conclusion: By using the propensity score analysis, this study confirmed that parenteral nutrition was an independent risk factor for CVC-RBSI in digestive cancer patients. (c) 2012 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.