Background: Although common risk factors for venous thromboembolism (VTE) are well known, little data exist concerning the clinical impact of VTE in postoperative patients outside of controlled studies. This study evaluated prospective perioperative demographic and clinical variables associated with occurrence of postoperative symptomatic VTE.
Methods: Demographic and clinical data were collected on surgical patients undergoing nine common general, vascular, and orthopedic operations presenting to the Veterans Health Administration Hospitals between 1996 and 2001 as part of the National Surgical Quality Improvement Program (NSQIP). The association between covariates and the incidence of postoperative symptomatic VTE (includes deep venous thrombosis and pulmonary embolism) was assessed using bivariable and multivariable regression.
Results: Complete demographic and clinical information for analysis were available for 75,771 patients. The mean patient age was 65 years, and 96.6% were men. Major comorbidities included diabetes mellitus (DM), 25%; chronic obstructive pulmonary disease (COPD), 18.3%; and congestive heart failure (CHF), 3.9%. Symptomatic VTE was diagnosed in 805 patients (0.68%), varied significantly with procedure (0.14% for carotid endarterectomy vs 1.34% for total hip arthroplasty), and was associated with increased 30-day mortality (16.9% vs 4.4%, P < .0001). The incidence of VTE did not decline substantially between 1996 and 2001 (0.72% vs 0.68%). Preoperative factors associated with symptomatic VTE were older age, male gender, corticosteroid use, COPD, recent weight loss, disseminated cancer, low albumin, and low hematocrit (all P < .01) but not DM. Postoperative factors associated with VTE were myocardial infarction (MI), blood transfusion (>4 units), coma, pneumonia, and urinary tract infection (UTI), whereas those with hemodialysis-dependent renal failure were less likely to experience VTE (all P < .01). In multivariable analysis, adjusting for age and the variables significant by bivariable analysis, the strongest positive predictors of symptomatic VTE included UTI (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.3 to 2.5), acute renal insufficiency (OR, 1.9; 95% CI, 1.1 to 3.2), postoperative transfusion (OR, 2.3; 95% CI, 1.4 to 3.7), perioperative MI (OR, 2.4; 95% CI, 1.5 to 3.9), and pneumonia (OR, 2.7; 95% CI, 2.1 to 3.5). In contrast, hemodialysis (OR, 0.3; 95% CI, 0.07 to 0.71), DM (OR, 0.75; 95% CI, 0.61 to 0.93), and higher preoperative albumin levels (OR, 0.8; 95% CI, 0.74 to 0.96, per mg/dL change) were protective from symptomatic VTE.
Conclusions: Although the overall incidence of symptomatic VTE is low in surgical patients, it is associated with significantly increased 30-day mortality. In addition to previously recognized risk factors, patients who have postoperative complications of an infectious nature, bleeding, or MI are at particular risk.