Objective: To assess the use of venous thromboembolism prophylaxis in surgical patients.
Design: Retrospective multicentre study.
Setting: Eight acute-care teaching hospitals with more than 400 beds, Spain.
Patients: Medical records of all consecutive patients undergoing operations in the general surgical and trauma and orthopaedic services during the month of April, 1997, were randomly selected.
Intervention: The sample size for each type of operation (general, trauma-orthopaedic) was calculated from the number of operations done at each hospital (with an absolute precision of 5%, and an alpha error of 5%) and the prevalence of the use of venous thromboembolism prophylaxis obtained from a random sample of 50 records (25 from patients in general surgery and 25 from patients in orthopaedic surgery) from each centre.
Main outcome measures: Appropriate and inappropriate pharmacological prophylaxis defined according to a combination of risk categories for venous thromboembolism, doses of antithrombotic agents given, time of starting prophylaxis, and its duration.
Results: A total of 1848 medical records (general surgery, n = 1025; trauma-orthopaedic surgery, n = 823) were included. Physical methods (elastic stockings, intermittent pneumatic compression) were used in only 0.3% of patients. Pharmacological prophylaxis consisted of low molecular weight heparin in 99% of cases. The percentage given heparin-based prophylaxis was 54%. Overall, appropriate prophylaxis was given in 1175 patients (64%). Use of thromboprophylaxis ranged from 27% to 70% among the participating hospitals. Prophylaxis was more likely to be appropriate in orthopaedic patients (577, 70%) than in general surgical patients (598, 58%) in both the high and moderate risk categories.
Conclusions: Given the large variability between the participating hospitals, more specific protocols and recommendations about prophylaxis of thromboembolism in surgical patients are needed.