Objective Although goal-directed fluid therapy has been proven to be effective in reducing the incidence of postoperative complications, the underlying mechanisms remain unknown. The aim of this study was to examine the mediating role of intraoperative hemodynamic lability in the association between goal-directed fluid therapy (GDFT) and the incidence of postoperative complications. We further tested the role of this mediation effect using mean arterial pressure, a hemodynamic indicator.
Methods This is a secondary analysis of a completed nonrandomized controlled study purposed to investigate the effect of GDFT on the incidence of postoperative complications in patients undergoing posterior spine arthrodesis. We used a simple mediation model to test whether the mediation effect of average real variability (ARV) between the association of GDFT and postoperative complication exists. Mediation analysis was conducted using the mediation package in R (version 3.1.2), based on 5,000 bootstrapped samples, adjusting for covariates.
Results Among the 300 patients in the study, 40% (120/300) developed postoperative complications within 30 days. GDFT was associated with fewer 30-day postoperative complications after adjustment for confounders (odds ratio: 0.460, 95% CI: 0.278, 0.761; P = 0.003). The total effect of GDFT on postoperative complications was -0.18 (95% CI: -0.28, -0.07; P < 0.01). The average causal mediation effect was -0.08 (95% CI: -0.15, -0.04; P < 0.01). The average direct effect was -0.09 (95% CI: -0.20, 0.03; P = 0.17). The proportion mediated was 49.9% (95% CI; 18.3%, 140.0%).
Conclusions The intraoperative blood pressure lability mediated the relationship between GDFT and the incidence of postoperative complications. Future research is needed to clarify whether actively reducing intraoperative blood pressure lability can prevent postoperative complications.

Key words: goal-directed fluid therapy, mediation analysis, postoperative complications, hemodynamic stability