Effects of intraoperative colloid administration on outcome in a population-based general surgical cohort: a propensity score analysis

Background. Many studies on colloids have recently been retracted, leaving us with uncertain evidence of their safety. We aimed to analyze whether intraoperative colloid administration is associated with postoperative complications. Methods. The prospectively compiled database of the ARISCAT study of a large, representative cohort of general surgical patients was reanalyzed to compare outcomes according to whether intraoperative colloids were administered or not; a propensity score was used to adjust for potential confounders. The primary outcomes were major postoperative complications. Secondary outcomes were postoperative hospital-free days within 90 days and mortality at 30 and 90 days. In a retrospective survey we asked each center's data collectors to estimate the proportions of the different colloids administered during the study period. Results. Of 2462 patients analyzed, 556 (22.6%) received some type of colloid intraoperatively. The median (25th-75th percentile) of total fluids administered was significantly higher in patients receiving colloids (10.0 [6.9-14.1] mL.kg(-1).h(-1) vs. 8.8 [6.0-12.8] mL.kg(-1).h(-1) for patients not receiving colloids; P<0.01). The median volume of colloids administered was 7.5 (6.3-10.4) mL.kg(-1). An estimated 75.7% of the patients received third-generation hydroxyethyl starches (130/0.4). Significantly associated complications, after propensity score adjustment, were atelectasis, respiratory infection, bronchospasm, arrhythmia, sepsis, paralytic ileum, and hyperglycemia. Patients receiving colloids had 1.9 fewer postoperative hospital-free days (P<0.006). There were no significant differences in 30- and 90-day mortality. Conclusion. Our study suggests an association of intraoperative colloid administration, mainly of 130/0.4 hydroxyethyl starches, with diverse major postoperative complications and longer hospital stay. Controlled studies are urgently needed to assess the safety profile of colloid use in surgical patients.