We analyzed the data from 2005 to 2009 for patients undergoing colorectal resections for cancer based on the primary procedure CPT-4 code and operative ICD-9 diagnosis code. The primary outcome was 30-day deep vein thrombosis (DVT) and/or pulmonary embolism (PE). Intraoperative transfusion of RBC's was categorized as: none, 1-2 units, 3-5 units and 6 units or more. DVT/PE occurrences were analyzed by multivariable forward stepwise regression (p for entry < .05, for exit > .10) to identify independent predictors of DVT.
The database contained 21943 colorectal cancer resections. The DVT rate was 1.4%(306/21943) and the PE rate was 0.8%(180/21943). Patients were diagnosed with both only 40 times and the combined DVT or PE rate (VTE) was 2.0%(446/21943). After adjusting for age, gender, race, ASA (American Society of Anesthesiologists) class, emergency procedure, operative duration and complexity of the procedure (based on Relative Value Units, RVU's), along with six clinical risk factors, intraoperative blood transfusion was a significant risk factor for the development of VTE and the risk increased with increasing number of units transfused. Preoperative hematocrit did not enter the multivariable model as an independent predictor of VTE, nor did open versus laparoscopic resection or wound class.
In this study of 21943 patients undergoing colorectal resection for cancer, blood transfusion is associated with increased risk of VTE. Malignancy and surgery are known prothrombotic stimuli, the subset of patients receiving intraoperative RBC transfusion are even more at risk for VTE, emphasizing the need for sensible use of transfusions and rigorous thromboprophylaxis regimens.