Incidence of venous thromboembolism in hospitalized patients vs community residents

Objective: To estimate the incidence rates of deep venous thrombosis (DVT) and pulmonary embolism (PE) in hospitalized patients and to compare these with incidence rates in community residents. Patients and Methods: We performed a retrospective review of the complete medical records from a population-based inception cohort of patients who resided in Olmsted County, Minnesota, and had an incident DVT or PE from 1980 through 1990. Results: From 1980 through 1990, 911 Olmsted County residents experienced their first lifetime event of definite, probable, or possible venous thromboembolism. Of these residents, 253 had been hospitalized for some reason other than a diagnosis of DVT or PE (in-hospital cases), and 658 were not hospitalized at onset of venous thromboembolism (community residents). The average annual age- and sex-adjusted incidence of in-hospital venous thromboembolism was 960.5 (95% confidence interval, 795.1-1125.9) per 10,000 person-years and was more than 100 times greater than the incidence among community residents at 7.1 (95% confidence interval, 6.57.6) per 10,000 person-years. The incidence of venous thromboembolism rose markedly with increasing age for both groups, with PE accounting for most of the age-related increase among in-hospital cases. Incidence rates in the 2 groups changed little over time despite a reduction in the average length of hospital stay between 1980 and 1990. Conclusions: Venous thromboembolism is a major national health problem, especially among elderly hospitalized patients. This finding emphasizes the need for accurate identification of hospitalized patients at risk for venous thromboembolism and a better understanding of the mechanisms involved so that safe and effective prophylaxis can be implemented.