This study uses the PharMetrics and MarketScan US health care insurance claims database to estimate relative risks for non-fatal venous thromboembolism (including cerebral venous sinus thrombosis), ischemic stroke, and acute myocardial infarction (heart attack), in current users of ORTHO EVRA (norelgestromin and ethinyl estradiol contraceptive patch) compared to current users of oral levonorgestrel-containing oral contraceptives with 30 micrograms ethinyl estradiol, with special attention to duration of use.

This observational, retrospective study uses the PharMetrics and MarketScan US health care claim insurance database to assess the occurrence of venous thromboembolism (including cerebral venous sinus thrombosis), ischemic stroke, and acute myocardial infarction in current users of ORTHO EVRA (norelgestromin and ethinyl estradiol transdermal contraceptive patch) compared to current users of levonorgestrel-containing oral contraceptives with 30 micrograms ethinyl estradiol, with special attention to duration of use. PharMetrics is a United States based, ongoing longitudinal database with information on around 55 million covered lives going back as far as 1995. It is made up of data contributed by managed care plans throughout the United States and it contains information on paid claims for pharmaceuticals, medical diagnoses and procedures as well as demographic information on all subjects. There will be 3 sets of cases reflecting women who have a first-time recorded claim for an ICD-9 diagnosis of the following during the study period: (1) deep vein thrombosis (blood clot), pulmonary embolism, or cerebral venous sinus thrombosis, venous thrombotic event (VTE) with hospitalization, a visit to the emergency room or positive indication of VTE from diagnostic test results, at any time during the study period and who had subsequent multiple claims for anticoagulant treatment. The requirement for multiple prescriptions for anticoagulation therapy was used to provide evidence that the original diagnosis of VTE was confirmed. (2) ischemic stroke and who were hospitalized, or (3) acute myocardial infarction (heart attack) or acute coronary revascularization and who were hospitalized. Three separate sets of controls will be identified for each outcome. Planned analysis will estimate the relative risk of idiopathic (unknown cause) ischemic stroke or myocardial infarction, and another analysis will estimate the relative risk of idiopathic VTE. Conditional logistic regression will be used. Analyses will be stratified by calendar year. The analyses will be repeated including non-idiopathic cases of VTE, ischemic stroke, and acute myocardial infarction. Each transdermal patch containing 6 milligrams norelgestromin and 0.75 milligram ethinyl estradiol is worn for 1 week for 3 consecutive weeks; the fourth week is patch-free. Triphasic levonorgestrel-containing oral contraceptives with 30 micrograms ethinyl estradiol is taken for 21 consecutive days followed by no pill or an inert pill for 7 days. Duration of use can vary. NOTE: The study was conducted in two parts. This study, CR014383, represents the first portion of the study. Study CR012022, NCT00377988, represents the second portion of the extended study.

Inclusion Criteria:

• Users of ORTHO EVRA or first time users of levonorgestrel-containing oral contraceptive with 30 micrograms ethinyl estradiol between April 1, 2002 and March 31, 2006, who are identified in the MarketScan and PHARMetrics database using the National Drug Code (NDC) assigned by the FDA and modified by PharMetrics
• Start of study contraceptive use after April 1, 2002
• Six (6) months of enrollment in their health plan prior to the event date of their matched case
• At least four (4) months of history in their claims record before the first recorded study drug dispensing.

Exclusion Criteria:

• When women received both levonorgestrel-containing oral contraceptive with 30 micrograms ethinyl estradiol and levonorgestrel-containing oral contraceptive with 20 micrograms ethinyl estradiol, this analysis only includes the 30 micrograms ethinyl estradiol preparations.

PharmaMetrics and MarketScan US health care claims database identified women 15-44 who were users of ORTHO EVRA or first time users of levonorgestrel-containing oral contraceptive with 30 micrograms ethinyl estradiol between April 1, 2002 and March 31, 2006