PURPOSE: To examine the effectiveness of mechanical methods for preventing venous thromboembolism (VTE) in non-orthopedic surgical patients.
METHODS: We searched electronic databases and scanned reference lists of retrieved articles to identify randomized, controlled trials that examined mechanical prophylaxis in non-orthopedic surgical patients. Reviewers evaluated studies for inclusion, rated methodological quality and abstracted relevant data. We used random effects models to measure heterogeneity (I2) and pool odds ratios (OR) across studies for outcomes including surveillance-detected deep venous thrombosis detected by surveillance (SD-DVT), symptomatic DVT and pulmonary embolism (PE).
RESULTS: Of 485 potentially eligible studies, 50 met criteria for inclusion, including 8 studies that compared elastic stockings (ES) with no prophylaxis, 18 studies that compared intermittent pneumatic compression (IPC) with no prophylaxis, one that compared ES plus IPC with no prophylaxis, 9 that compared mechanical methods, 16 that compared mechanical with pharmacological prophylaxis and 7 that examined mechanical methods on a background of pharmacological prophylaxis. Compared with no prophylaxis, ES reduced SD-DVT by 67% (OR 0.33, 95% CI 0.24 to 0.46, I2=0%) and IPC reduced SD-DVT by 68% (OR 0.32, 95% CI 0.24 to 0.42, I2=53%). Compared with single modality mechanical prophylaxis, ES plus IPC reduced SD-DVT by 36% (OR 0.64, 95% CI 0.42 to 0.98, I2=55%). Differences between mechanical and pharmacological prophylaxis were neither confirmed nor excluded for SD-DVT (OR 0.98, 95% CI 0.79 to 1.21, I2=55%). Compared with pharmacological prophylaxis alone, combined mechanical and pharmacological prophylaxis reduced SD-DVT by 45% (OR 0.55, 95% CI 0.38 to 0.79, I2=37%). For all comparisons, differences between groups in the risk of symptomatic DVT or PE were neither confirmed nor excluded.
CONCLUSIONS: In non-orthopedic surgical patients, mechanical methods for thromboprophylaxis appear to reduce SD-DVT when used alone or in combination with pharmacological prophylaxis, but it is unclear whether they reduce symptomatic DVT or PE. It is not clear whether mechanical methods are as effective as pharmacological prophylaxis.
CLINICAL IMPLICATIONS: Future randomized, controlled trials should compare mechanical and pharmacological methods for VTE prevention in non-orthopedic surgical patients, and should be adequately powered to detect differences in symptomatic VTE events.
DISCLOSURE: The following authors have nothing to disclose: Hidenobu Shigemitsu, Marcus Ottochian, Peter Mestaz, Michael Gould
No Product/Research Disclosure Information