Objective: To analyse the frequency of autopsy-proven venous thromboembolism within a defined region and evaluate if certain risk groups not earlier recognized could be found. The incidence of objectively diagnosed venous thromboembolism was also calculated.
Setting: The city of Malmö, Sweden, with 230,000 inhabitants.
Main outcome measure: Analysis of 2,356 autopsies for the year 1987 of deceased inhabitants from the city of Malmö (2,981 deceased, autopsy frequency 79.0%) regarding venous thromboembolism.
Results: 25% of autopsies revealed venous thromboembolism. At the acute-care hospital 31%, at the chronic hospital 37%, but in forensic autopsies of non-hospital deaths only 5% (p<0.001) revealed venous thromboembolism. Major pulmonary embolism was seen in 13% and was more frequent in in-hospital deaths (p<0.001). Two in-hospital and two non-hospital deaths due to pulmonary embolism and fractures were found: two patients with knee fractures, one hip fracture and one ankle fracture patient. The incidence of objectively diagnosed venous thromboembolism (autopsy, phlebography, perfusion scintigraphy) was calculated and an incidence of 42.5/10,000 inhabitants/year was found, (strongly age-dependent).
Conclusion: Venous thromboembolism is a common finding in autopsies of hospitalized patients. Patients with fractures other than hip fractures are less well studied as regards venous thromboembolic complications. Further studies on these fracture patients are warranted.