Trends in admission rates and in-hospital stay for venous thromboembolism

Background: Acute venous thromboembolism leads to significant morbidity and mortality. Advances in pharmacotherapy facilitate outpatient care in low-risk acute venous thromboembolism. The proportion of hospitalized acute venous thromboembolism cases and the average length of stay are not known. We sought to identify predictors of hospitalization, changes in hospitalization rates and length of stay of acute venous thromboembolism over a decade in Alberta, Canada.

Methods: Using linked administrative health databases, we identified adult patients diagnosed primarily with acute venous thromboembolism between April 2002 and March 2012. We measured trends using Poisson regression, adjusted length of stay using analysis of covariance. We identified predictors of hospitalization using multivariate logistic regression.

Results: 8198 out of 31,656 acute venous thromboembolism cases were hospitalized. The overall venous thromboembolism admission rates ranged between 23.7% and 27.8% with no evident temporal trend (P=0.10). The average admission rate was 51.9% for pulmonary embolism and 16.1% for deep vein thrombosis. The mean length of stay for deep vein thrombosis and pulmonary embolism remained unchanged with an adjusted mean for venous thromboembolism of 6.9±1.0days. Higher Charlson index, older age, male gender, pulmonary embolism at presentation and multiple comorbidities were associated with hospitalization. Hospitalization was associated with 30-day mortality (odds ratio:2.8, 95% CI: 2.2-3.5) whereas the length of stay was not (odds ratio:1.0, 95% CI: 0.99-1.0).

Conclusion: Hospitalization rates and mean length of stay for acute venous thromboembolism did not change significantly between 2002 and 2012. Advances in pharmacotherapy have not yet reduced hospitalization rates or length of stay for venous thromboembolism.