Background It is known that the incidence of venous thromboembolism (VTE) is increased in people with cancer, but less is known about how the risk varies by cancer type or over time since initial cancer diagnosis. This study aims to describe the incidence of VTE in women after cancer and by type of malignancy.

Methods 1.3 million UK women were recruited into the Million Women Study in 1996–2001. These women are being followed by record linkage to routinely collected inpatient and day-case NHS hospital admissions data, cancer registrations and death records. The relative risks (RRs) and 95% confidence intervals (CIs) of an inpatient/day-case hospital admission or death record for VTE (ICD10 codes I26, I80-I82) were estimated in relation to a variety of malignancies. Analyses using Cox regression were adjusted for age, region of residence, socio-economic group, body mass index, menopausal hormone therapy, smoking and alcohol consumption.

Results Among 1,198,848 women with no previous VTE or cancer and an average of 11.7 years of follow-up per woman, 108,607 (9%) had a hospital admission for incident cancer and 21,501 (1.8%) had a hospital admission or death from VTE, including 5251 women whose VTE followed cancer. Overall, women were almost nine times more likely to have a hospital admission for VTE after cancer than without (or before) any cancer diagnosis (adjusted RR=8.7, 95% CI: 8.4–9.0). There was a 40–50 fold increased risk of VTE during the first three months, which decreased to 10-fold by one year, but still remained significantly elevated (around 2-fold) many years later. Cancers of the brain and pancreas were associated with the highest risks, at about 70 times the risk without cancer.

Discussion The risk of VTE is significantly increased following a cancer diagnosis, and there was considerable variation over time and by cancer type. We are now investigating the risks associated with different stages of disease and with different cancer treatments.