The use of a large-scale surveillance system in Planned Parenthood Federation of America clinics to monitor cardiovascular events in users of combination oral contraceptives

Abstract

Background: The publication of a series of papers based on three large, case-controlled studies raised questions about the safety of oral contraceptives (OCs) containing either desogestrel or gestodene as the progestational agent. In light of these studies, the Planned Parenthood Federation of America (PPFA) assessed the risk of thromboembolic events in patients who attended its clinics in the United States (U.S.). These PPFA patients were prescribed contraceptives containing one of the following progestins: desogestrel, levonorgestrel, norethindrone, or norgestimate. All of the contraceptives used contained either 30 or 35 microg of estrogen.

Methods: In this retrospective analysis, the association between the occurrence of a thromboembolic event, OC use, and progestin type was assessed in a group of PPFA patients. These patients were prescribed 2,265,087 woman-years (WY) of OCs from January 1, 1993 to December 31, 1995.

Results: Seventy events (deep vein thromboses [DVTs], pulmonary emboli [PEs], myocardial infarctions [MIs], and cerebral-vascular accidents [CVAs] were reported; a rate of 3 events per 100,000 WY of use. Five deaths occurred, at a rate of 0.22 per 100,000 WY of use. This low rate of incidents appears to be related to the younger age of PPFA patients and to the use of uniform guidelines for the selection and follow-up patients. Only the lower relative risk of PE reached statistical significance, when either norethindrone or norgestimate was compared to desogestrel-containing pills, but the absolute risk of these events was lower than in other published studies. Though the overall risk varied from a low of 1.895 events per 100,000 WY for norgestimate OC users to a high of 3.969 events per 100,000 WY for desogestrel OC users, these differences were not statistically significant.

Conclusions: This study supports the conclusion that OCs containing 30 or 35 microg of estrogen, combined with one of the four previously identified progestins, carry an extremely low risk of a thromboembolic event, particularly when prescribing is based on a defined medical protocol. Overall, these four groups of low-dose estrogen-containing OCs appear safer than any previously published study has indicated.