Identification and Validation of Risk Factors for Postoperative Infectious Complications Following Hepatectomy

Abstract

SummaryPostoperative infectious complications (PICs) are associated with significant morbidity after abdominal surgery. Using multivariate analysis of data from a prospective database, our study focused on the risk factors for PICs and the prevention of these complications after hepatectomy, with the goal of improving outcomes and reducing the length of hospital stays. BackgroundPICs following surgery are associated with significant morbidity, increase the length of hospital stays, and have a negative impact on long-term oncological outcome. The aim of this study was to determine the risk factors for PICs following partial hepatectomy and to validate these results with an external database. MethodsBetween January 2006 and December 2009, 555 patients underwent elective partial hepatectomy. We prospectively collected and retrospectively analyzed demographic data, pathological variables, associated pathological conditions, and preoperative, intraoperative, and postoperative variables. The dependent variables studied were the occurrence of PICs, defined as development of one or more of the following conditions: pneumonia, sepsis, Central line-associated bloodstream infection, urinary tract infection, wound infection, and infected intra-abdominal fluid collection. PICs were devised in medical (PIMCs) and surgical (PISCs) complications. The incidence of PICs and validation of the predictive score were determined using an external prospective database of 342 patients. ResultsThe multivariate analysis identified three independent risk factors for PICs: the presence of a nasogastric tube (OR = 1.8), blood transfusion (OR = 1.9), and diabetes (OR = 2.4). The multivariate analysis identified only one independent risk factor for PISCs: an associated portal venous resection (OR = 5.5). The multivariate analysis identified four independent risk factors for PIMCs: presence of a biliary drainage (OR = 1.9), blood transfusion (OR = 2.1), diabetes (OR = 2.9), and presence of atrial fibrillation (OR = 3.6). According to the three predictive factors, the observed rates of PICs ranged from 18.8 % to 77.8 % and ranged from 24.2 % to 100 % in the external database. Predicted and observed risks of PICs were not statistically different. ConclusionsThe correction of modifiable risk factors among the identified factors could reduce the incidence of PICs and, as a consequence, improve patient outcomes and reduce the length of hospital stays.