Timing and perioperative risk factors for in-hospital and post-discharge venous thromboembolism after colorectal cancer resection

Introduction: We postulated that the risk of venous thromboembolic disease (VTE) may persist after discharge and tested this hypothesis in patients undergoing colorectal resection for cancer. Methods: The American College of Surgeons National Surgery Quality Improvement Program database was queried for patients undergoing colorectal resections for cancer from 2005 to 2009. The outcome analyzed was a 30-day deep vein thrombosis (DVT) and/or pulmonary embolism (PE). Multivariable forward stepwise regression was used to identify independent predictors of VTE. Results: The database contained 21 943 colorectal cancer resections. The 30-day DVT rate was 1.4% (306 of 21 943), 29% (89 of 306) were diagnosed post-discharge. The 30-day PE rate was 0.8% (180 of 21 943), 33% (60 of 180) was diagnosed post-discharge, the combined DVT/PE rate was 2.0% (446 of 21 943). The median time to diagnosis of VTE was 9 days (interquartile range 4-16) after surgery. Post-discharge VTE rates in patients with length of stay (LOS) less than 1 week (0.6%) were similar to patients with LOS greater than 1 week (0.7%, Fisher exact P not significant). Independent risk factors for post-discharge VTE were preoperative steroid use for chronic condition (odds ratio [OR] 2.90, 95% confidence interval [CI] 1.51-5.57, P =.001) and preoperative systemic inflammatory response syndrome (OR 2.26, 95% CI 1.24-4.10, P =.008). Conclusions: Diagnosis of almost one third of postoperative VTE in this patient population occurred after discharge. The duration of the prothrombotic stimulus of surgery is not well defined, and patients with malignancy are at high risk of VTE; thromboprophylaxis after discharge should be considered for these patients.