Pharmacovigilance Unit and Clinical Trial UnitMedical Products Agency
Department of Medical Sciences, Clinical PharmacologyUppsala University
Department of Public Health and Clinical Medicine, Family MedicineUmeå University
Original Research Article
Cite this article as:
Hedenmalm, K. & Samuelsson, E. Drug-Safety (2005) 28: 907. doi:10.2165/00002018-200528100-00007
Background: Fatal venous thromboembolism (VTE) is a rare complication of combined oral contraceptive (COC) treatment. This study aims to determine incidences of fatal VTE in relation to the type of COC and the percentage of cases reported to the Swedish Adverse Drug Reactions Advisory Committee (SADRAC). A further aim is to compare the characteristics of reported and not reported cases.
Methods: This retrospective study is a separate analysis using data from a larger study that included women aged 15–44 years between 1990 and 1999 with VTE coded as the underlying or contributory cause of death in the Swedish Cause of Death Register. COC use within 2 months of the date of symptom onset or death was identified in 28 cases. Sales data were obtained from the National Corporation of Swedish Pharmacies. Reported cases were identified in the SADRAC database.
Results: After excluding two cases where the type of COC was unknown, the crude incidences of fatal VTE were 5.1 (95% CI 2.3, 9.6), 8.6 (95% CI 4.3, 15.4) and 9.1 (95% CI 3.3, 19.8) cases per million women per year for levonorgestrel-, desogestrel- and norethisterone-containing COCs, respectively. Age-adjusted incidences were approximately twice as high for desogestrel- and norethisterone-containing COCs compared with levonorgestrel-containing COCs, although differences were not statistically significant. Thirty-six percent of cases were reported. Reporting was positively associated with information in medical records relevant to the VTE diagnosis that the patient was a COC user and was significantly higher in northern Sweden.
Conclusion: Results from this study support a higher incidence of fatal VTE with desogestrel-containing COCs than with levonorgestrel-containing COCs.