Inferior vena cava filters and postoperative outcomes in patients undergoing bariatric surgery: a meta-analysis

Abstract

Background: Pulmonary embolism (PE) accounts for almost 40% of perioperative deaths after bariatric surgery. Placement of prophylactic inferior vena cava (IVC) filter before bariatric surgery to improve outcomes has shown varied results. We performed a meta-analysis to evaluate postoperative outcomes associated with the preoperative placement of IVC filters in these patients. Methods: A systematic review was conducted by three investigators independently in PubMed, EMBASE, the Web of Science and Scopus until February 28, 2013. Our search was restricted to studies in adult patients undergoing bariatric surgery with and without IVC filters. Primary outcomes were postoperative deep vein thrombosis (DVT), pulmonary embolism (PE), and postoperative mortality. Meta-analysis used random effects models to account for heterogeneity, and Sidik-Jonkman method to account for scarcity of outcomes and studies. Associations are shown as Relative Risks (RR) and 95% Confidence Intervals (CI). Results: Seven observational studies were identified (n=102,767), with weighted average incidences of DVT (0.9%), PE (1.6%), and mortality (1.0%) for a follow-up ranging from 3 weeks to 3 months. Use of IVC filters was associated with an approximately 3-fold higher risk of DVT and death that was nominally significant for the former outcome, but not the latter (RR 2.81, 95%CI 1.33-5.97, p=0.007; and RR 3.27, 95% CI 0.78-13.64, p=0.1, respectively); there was no difference in the risk of PE (RR 1.02, 95%CI 0.31-3.77, p=0.9). Moderate to high heterogeneity of effects was noted across studies. Conclusions: Placement of IVC filter before bariatric surgery is associated with higher risk of postoperative DVT and mortality. A similar risk of PE in patients with and without IVC filter placement cannot exclude a benefit, given the potential large imbalance in risk at baseline. Randomized trials are needed before IVC placement can be recommended.