Association between peripartum hysterectomy and venous thromboembolism.

Abstract

BACKGROUND\nThe risk of venous thromboembolism (VTE) postpartum is modified by mode of delivery, with the risk of VTE being higher after cesarean compared to vaginal delivery. The risk of VTE following peripartum hysterectomy is largely unknown.\n\n\nOBJECTIVE\nThe objective of this study was to compare the incidence and risk of VTE among women who had and did not have a peripartum hysterectomy. We also sought to compare the risk of VTE after hysterectomy with other patient, pregnancy, and delivery risk factors known to be associated with VTE.\n\n\nSTUDY DESIGN\nThis is a cross-sectional study of women with delivery encounters identified in the Nationwide Readmission Database from October 2015 to December 2017. Delivery encounters and all variables of interest were identified using ICD-10 diagnosis and procedure codes. The incidence of VTE during the delivery encounter and rehospitalizations within 6 weeks of discharge was compared among women who had and did not have a peripartum hysterectomy. Multivariable logistic regressions were used to estimate associations between VTE after hysterectomy, adjusted for the following characteristics known to be associated with VTE risk: maternal age, payer at time of delivery, obesity, hypertension, diabetes, tobacco use, multifetal gestation, peripartum infection, and peripartum hemorrhage. Similarly, VTE risk by mode of delivery, including hysterectomy, were compared. Diagnoses that may have been indications for peripartum hysterectomy were identified among patients who underwent hysterectomy and compared between those who did and did not have VTE. Analyses used survey weights to obtain population estimates.\n\n\nRESULTS\nOf the 4,419,037 women with deliveries, 5,098 (11.5 per 10,000 deliveries) underwent hysterectomy. 110 patients (215.8 per 10,000 deliveries) were diagnosed with VTE after hysterectomy. The risk of VTE in women after hysterectomy was significantly higher compared to women who did not have a hysterectomy (unadjusted odds ratio (OR) 25.1 (95% confidence interval (CI) 20.0-31.5) and adjusted odds ratio (aOR) 11.2 (95% CI 8.7-14.5), p<0.001). Comparing VTE by mode of delivery, the unadjusted and adjusted incidences of VTE were: 6.9 (95% CI 6.5-7.3) and 7.4 per 10,000 deliveries (95% CI 6.9-7.8) among women after vaginal delivery without peripartum hysterectomy; 12.5 (95% CI 11.8-13.1) and 11.3 per 10,000 deliveries (95% CI 10.7-12.0) after cesarean delivery without hysterectomy; and 217.2 (95% CI 169.1-265.2) and 96.9 per 10,000 deliveries (95% CI 76.9-126.5) after hysterectomy, regardless of mode of delivery. Of the 110 diagnoses of VTE with peripartum hysterectomy, 89 (81%) occurred during the delivery admission. Of the remaining 21, 50% occurred within the first 10 days after discharge from the delivery encounter, and 75% had occurred by 25 days after discharge.\n\n\nCONCLUSION\nThese findings demonstrate that peripartum hysterectomy is associated with a significantly increased risk of VTE in the postpartum period even when controlling for other known risk factors for postpartum thromboembolic events. The incidence of VTE following peripartum hysterectomy observed in our study population (2.2%) meets some guideline-based risk thresholds for routine thromboprophylaxis, potentially for at least two weeks postpartum. Further investigation into the role of routine VTE prophylaxis during and beyond the delivery encounter is needed.