Background:
Perioperative warfarin management is a clinical challenge as it requires careful assessment of both thrombosis and bleeding risk associated with the planned procedure. In our experience, many clinicians favor bridging over not bridging and often use full dose unfractionated or low molecular weight heparin even when a prophylactic dose would be a reasonable alternative. To improve and standardize the complex decision-making process for anticoagulation management for patients undergoing elective surgery we introduced an evidenced-based perioperative bridging anticoagulation protocol in 2017 based on the 9th edition of American College of Chest Physicians guidelines. The aim of the study was to investigate the impact of the perioperative bridging anticoagulation protocol on postoperative bleeding and thrombotic events in patients undergoing elective surgery at our institution. We hypothesized that adherence to the proposed protocol reduces the incidence of postoperative bleeding complications without an increase in thrombotic risk.
Methods:
We performed a retrospective chart review for all surgical inpatients who were on long-term anticoagulation with warfarin and who required interruption of warfarin around surgery at Dartmouth-Hitchcock Medical Center from June 2016, when the protocol was introduced, to June 2019. Data were extracted from the electronic medical record for each study subject and included demographics, clinical and radiologic data, indication for long-term anticoagulation (atrial fibrillation, venous thromboembolism, or mechanical heart valve), risk categories for thrombotic events, type of surgery, anticoagulation management before and after surgery including type and dose of short-acting anticoagulation used for bridging, and bleeding and thrombotic complications within 30 days of surgery. We evaluated adherence to the bridging protocol by individual providers and divided subjects into two groups for comparison: 1) subjects who received protocol-directed care and 2) subjects who received non-protocol care. We compared the risk of postoperative bleeding and thrombotic events in the two groups using the R3.5.1. Concordance of bridge protocol and crude associations between discrete variables were assessed by McNemar's and two-tailed Fisher's exact test, respectively. Multivariate logistic regression was used to estimate covariate-adjusted association groups and the outcomes of interest. P < 0.05 was used as the criterion for statistical significance.
Results:
194 subjects met entry criteria and were included in the study; 114 subjects were managed according to the protocol (58.8%; 95% CI = 51.5%-65.8%, McNemar's P = 0.034). Clinical characteristics were similar in the protocol-adherent and nonadherent groups. Most patients (50%) were on long-term warfarin for stroke prevention with atrial fibrillation. The incidence of bleeding and thrombotic complications for the entire population was 7.8% and 4.7 %, respectively. Two-thirds of subjects who experienced either bleeding (66.7% vs 33.3%, 95% CI= 0.91-11.93, P = 0.055, two tailed Fischer's exact test) or thrombotic (66.7% vs 33.3%, 95% CI= 0.62-19.19, P = 0.16, two tailed Fischer's exact test) complications were in the protocol-nonadherent group. Controlling for age and sex using multivariate logistic regression, subjects whose anticoagulation was not managed postoperatively in adherence to the protocol were over three times as likely to experience a bleeding complication (adjusted odds ratio=3.36, 95% CI= 1.13-11.40, P = 0.036) and a thrombotic complication (adjusted odds ratio=3.54, 95% CI=0.88-17.77, P=0.088) when compared to patients whose postoperative anticoagulation management was protocol-driven.
Discussion:
Overall adherence to an evidence-based anticoagulation bridging protocol by providers was only 59%. When adhered to, protocol-based practice helped to reduce bleeding complications by 60% and was associated with a trend toward a lower risk for thrombosis. Our study results emphasize the value of an evidence-based perioperative anticoagulation bridging protocol and the importance of protocol adherence. To improve adherence, we have now embedded the protocol as an order-set in our electronic medical record system and have provided additional education to providers within our institution with a plan to assess improvement in adherence subsequently.
Disclosures
No relevant conflicts of interest to declare.
Outpatient Management of the Chronically Anticoagulated Patient for Elective Surgery
