A Simulated Randomized Control Trial of Endoscopic Clip Placement to Prevent Delayed Bleeding after Colonic Polypectomy On and Off Anticoagulation

Abstract

Purpose: To determine the cost effectiveness of prophylactic endoscopic clip placement to prevent delayed post-polypectomy bleeding after polypectomy during colonoscopy for patients on and off anti-coagulation and anti-platelet therapy. Methods: A multi-state Markov model was developed using TreeAge software. We simulated three patient cohorts consisting of 10,000 patients undergoing index screening colonoscopy at age 50. The three cohorts consisted of patients: 1) Off anti-coagulation or anti-platelet therapy; 2) On warfarin therapy with a therapeutic INR; 3) On anti-platelet therapy (clopidrogrel). Each patient had a 1 cm pendunculated polyp and patients were randomized to endoscopic clip placement versus no clip on the remnant stalk after polypectomy. The outcome of interest was delayed post-polypectomy bleeding up to 14 days after polypectomy. Post-polypectomy bleeding probabilities were obtained using existing literature pertinent to the study via MEDLINE. Cost was calculated from a third party payer perspective. We conducted a sensitivity analysis, varying the number of clips placed (1-3) and the bleeding probabilities within the range reported in the existing literature. We made the following assumptions in our analysis: 1) Clip placement resulted in a 0% risk of post-polypectomy bleeding; 2) All patients who experienced significant post-polypectomy bleeding were hospitalized; 3) Mortality from post-polypectomy bleeding was 0%. Results: The prophylactic endoscopic clip placement in patients off anticoagulation and anti-platelet therapy was not a cost effective strategy, resulting in an extra $230,000 spent in the clip cohort. However, the placement of an endoscopic clip for patients on warfarin therapy was a cost effective strategy, resulting in a savings of $170,000. The placement of an endoscopic clip was also a cost effective strategy in patients on clopidogrel therapy, resulting in a cost savings of $720,000 in our simulated cohort. The sensitivity analysis showed that placement of more than one clip was not cost effective; however, the placement of a single clip remained cost effective across the range of published delayed post-polypectomy bleeding rates in the anti-coagulation and anti-platelet therapy cohorts. Conclusion: The placement of a prophylactic endoscopic clip appears to be a cost effective strategy to prevent delayed post-polypectomy bleeding in patients on anti-coagulation and anti-platelet therapy based on this simulated randomized control trial. Future prospective studies are needed to better define the role of routine endoscopic clip placement patients undergoing polypectomy.