Background: Many studies on colloids have recently been retracted, leaving us with uncertain evidence of their safety.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.