Background: Accurate population-based data are needed on the incidence of venous thromboembolism (VTE) in patients with different cancers in order to inform guidelines on which hospitalised and ambulatory cancer patients should receive VTE prophylaxis. Methods: We conducted a cohort study using data from the Clinical Practice Research Datalink, linked to Hospital Episode Statistics, Cancer Registry data and Office for National Statistics cause of death data. We determined the incidence rates (cases per 1000 person-years) of VTE separately for 24 cancer sites. To determine relative risk, incidence rates were compared to frequency-matched controls (by age) with no record of cancer. Findings: We identified 83,203 cancer patients and 577,207 controls. New cases of VTE were diagnosed in 3352 cancer patients, and 6353 controls. The absolute rate of VTE in all cancers was 13.9 per 1000 person-years (95% confidence interval [CI] 13.4-14.4), corresponding to an age, sex and calendar year adjusted hazard-ratio of 4.7 (CI 4.5-4.9) between cancer patients and the general population. Rates varied greatly by cancer site (range; 98 (CI 80-119) in pancreatic cancer to 3.1 (CI 1.5-6.5) in thyroid cancer), age (range; 16.9 for patients over 80 years to 4.9 for those under 30 years) and time from diagnosis (range; 75 in the first three months to 8.4, >1 year after diagnosis). Interpretation: VTE is strongly linked to cancer, but the annual rate varies greatly by cancer site, proximity to diagnosis and age. Prophylaxis guidelines should take account of cancer site and such intervention should also be targeted towards the three months following diagnosis. © 2012 Elsevier Ltd. All rights reserved.
European Journal of Cancer
1404 - 1413