Prevention of venous thromboembolism after injury: an evidence-based report--part I: analysis of risk factors and evaluation of the role of vena caval filters.
BACKGROUND: Trauma surgeons use a variety of methods to prevent venous thromboembolism (VT). The rationale for their use frequently is based on conclusions from research on nontrauma populations. Existing recommendations are based on expert opinion and consensus statements rather than systematic analysis of the existing literature and synthesis of available data. The objective is to produce an evidence-based report on the methods of prevention of VT after injury. METHODS: A panel of 17 national authorities from the academic, private, and managed care sectors helped design and review the project. We searched three electronic databases (MEDLINE, EM-BASE, and Cochrane Controlled Trial Register) to identify articles relevant to four key questions: methods of prophylaxis, methods of screening, risk factors for VT, and the role of vena caval filters. The initial 4,093 titles yielded 73 articles for meta-analysis. A random-effects model was used for all pooled results. Study quality was evaluated by previously published quality scores. In this article (part I), we report on the question ranked by the experts as the most important, i.e., Which is the best method to prevent VT?, and also on the incidence of deep venous thrombosis and pulmonary embolism in trauma patients. RESULTS: The incidence of deep venous thrombosis and pulmonary embolism reported in different studies varies widely. The pooled rates are 11.8% for deep venous thrombosis and 1.5% for pulmonary embolism. Only a few randomized controlled trials have evaluated the methods of VT prophylaxis among trauma patients, and combining their data is difficult because of different designs and preventive methods used. The quality of most studies is low. Meta-analysis shows no evidence that low-dose heparin, mechanical prophylaxis, or low-molecular-weight heparin are more effective than no prophylaxis or each other. However, the 95% confidence intervals of many of the comparisons are wide; therefore, a clinically important difference may exist. CONCLUSION: The trauma literature on VT prophylaxis provides inconsistent data. There is no evidence that any existing method of VT prophylaxis is clearly superior to the other methods or even to no prophylaxis. Our results cast serious doubt on the existing policies on VT prophylaxis, and we call for a large, high-quality, multicenter trial that can provide definitive answers.