Risk factors for pulmonary embolism and the effects of fondaparinux after total hip and knee arthroplasty: a retrospective observational study with use of a national database in Japan.

BACKGROUND

Clinical evidence demonstrating the effectiveness of pharmacological and mechanical thromboprophylaxis for the prevention of pulmonary embolism is limited because the prevalence of postoperative pulmonary embolism following total hip and knee arthroplasty is very low. Our purposes were to characterize a patient population with in-hospital pulmonary embolism, to identify perioperative risk factors associated with pulmonary embolism, and to analyze the effect of combining fondaparinux with mechanical prophylaxis on the prevalence of pulmonary embolism following total hip and knee arthroplasty.

METHODS

We retrospectively identified 27,542 patients who underwent total hip or knee arthroplasty at 793 hospitals, using data from the Diagnosis Procedure Combination database, collected from July 1 to December 31 in 2007 and 2008. We extracted data on patient sex, age, primary diagnoses, and comorbidities that could potentially affect the prevalence of pulmonary embolism. The dates of pharmacological and mechanical thromboprophylaxis were identified for each patient. Logistic regression analysis was performed to analyze the concurrent effects of various factors on the prevalence of postoperative pulmonary embolism.

RESULTS

The mean age (and standard deviation) of the patients at the time of arthroplasty was 69.9 ± 10.3 years, and 23,783 patients (86.4%) were diagnosed as having osteoarthritis. The overall mean duration of anesthesia was 159 ± 84 minutes. The overall prevalence of postoperative pulmonary embolism was 0.55% (151 of 27,542). Significant risk factors for postoperative pulmonary embolism included age, number of comorbidities, diagnosis of rheumatoid arthritis, type of anesthesia, and duration of anesthesia. Multivariate analysis found that the prevalence of postoperative pulmonary embolism was significantly reduced when fondaparinux was used in combination with mechanical prophylaxis, compared with the use of mechanical prophylaxis alone (0.40% versus 0.66%; odds ratio, 0.60; 95% confidence interval, 0.42 to 0.84; p = 0.003).

CONCLUSIONS

These findings could help to identify patients at higher risk of postoperative pulmonary embolism after total hip or knee arthroplasty. Our results demonstrate the effectiveness of fondaparinux in combination with mechanical prophylaxis for the prevention of postoperative pulmonary embolism after total hip or knee arthroplasty.