Background: Venous thromboembolism is a dreaded complication following microsurgical breast reconstruction. Although the high-risk nature of the procedure is well known, a thorough analysis of modifiable risk factors has not been performed. The purpose of this study was to analyze the association of such factors with the postoperative occurrence of venous thromboembolism longitudinally.
Methods: Using the Truven MarketScan Database, a retrospective cohort study of women who underwent microsurgical breast reconstruction from 2007 to 2015 and who developed postoperative venous thromboembolism within 90 days of reconstruction was performed. Predictor variables included age, timing of reconstruction, body mass index, history of radiation therapy, history of venous thromboembolism, Elixhauser Comorbidity Index, and length of stay. Univariate analyses were performed, in addition to logistic and zero-inflated Poisson regressions, to evaluate predictors of venous thromboembolism and changes in venous thromboembolism over the study period, respectively.
Results: Twelve thousand seven hundred seventy-eight women were identified, of which 167 (1.3 percent) developed venous thromboembolism. The majority of venous thromboembolisms (67.1 percent) occurred following discharge, with no significant change from 2007 to 2015. Significant predictors of venous thromboembolism included Elixhauser score (p < 0.01), history of venous thromboembolism (p < 0.03), and length of stay (p < 0.001). Compared to patients who developed a venous thromboembolism during the inpatient stay, patients who developed a postdischarge venous thromboembolism had a lower mean Elixhauser score (p < 0.001).
Conclusions: Postoperative venous thromboembolism continues to be an inadequately addressed problem, as evidenced by a stable incidence over the study period. Identification of modifiable risk factors, such as length of stay, provides potential avenues for intervention. As the majority of venous thromboembolisms occur following discharge, future studies are warranted to investigate the role for an intervention in this period.
Clinical question/level of evidence: Risk, III.