Long-term safety of extended levonorgestrel-containing oral contraceptives in the United States

Objectives: To compare the safety profile of Seasonique, a 91-day levonorgestrel-containing combined oral contraceptive (COC_LNG), to 28-day COC_LNG regarding the risk of venous thromboembolism (VTE) and arterial thromboembolism (ATE).

Study design: A new user cohort study was conducted in a US health care database from 2006 to 2017. Each 91-day COC_LNG treatment episode in females was matched to up to four 28-day COC_LNG treatment episodes by propensity score. We identified VTE cases in either (1) an inpatient setting with ICD-9 and ICD-10 diagnosis codes of PE and/or DVT in the primary position, or (2) an outpatient setting with ICD-9 or ICD-10 diagnosis codes of DVT in conjunction with an anticoagulant medication dispensing or alteplase (thrombolytic) during the 30-day period following the date of DVT diagnosis. VTE was validated using medical records. We assessed the study endpoints in the two cohorts using incidence rates and Cox proportional hazards models adjusted for potential confounders.

Results: Of the 25,593 treatment episodes in 91-day COC_LNG and 76,586 treatment episodes in 28-day COC_LNG, 35 and 68 patients had VTEs, respectively, corresponding to a hazard ratio (HR) of 1.40 (95% confidence interval [CI], 0.90-2.19). The VTE algorithm had a positive predictive value of 76.4% (95% CI, 66.2%-84.8%). ATEs were recorded in 13 and 28 episodes, respectively, with a corresponding HR of 1.21 (95% CI, 0.58-2.53).

Conclusions: These results do not indicate a significant difference between 91-day COC_LNG and 28-day COC_LNG in terms of VTE or ATE risk.

Implications: Compared to use of 28-day COC_LNG, use of 91-day extended COC_LNG was not associated with a significant difference in risk of venous and arterial thromboembolism.