Contraceptive Use and Pregnancy Incidence among VOICE Participants in Uganda

Abstract

Background: Recent HIV prevention trials have required use of effective contraceptive methods to fulfill eligibility for enrolment. We compared pregnancy rates in participants enrolled in the VOICE (MTN-003) trial who initiated depot medroxyprogesterone acetate (DMPA) or combined oral contraceptives (COCs) at enrollment (new users) relative to those already on DMPA or COCs. Methods: Data were analyzed from HIV-1 seronegative participants enrolled in Uganda. Prior to enrolment information on contraceptive type and initiation date was obtained; contraception was provided to new users. Participants received urine pregnancy tests at monthly follow-up visits. Cox proportional hazards models stratified by baseline contraceptive method were used to compare pregnancy incidence among new users (initiated 60 days prior to enrolment). Participants were censored after the first observed pregnancy on study. Results: Of 322 women enrolled 296 were COC or DMPA users; 82 (28%) were new users (50 [61%] initiated DMPA and 32[39%] initiated COCs) and 214 (72%) were established users. Overall pregnancy incidence was 13.35 per 100 person-years [p-yrs] (49/367 p-yrs) with an incidence in DMPA users of 5.39 per 100 p-yrs (13/241 p-yrs) compared to 28.62 per 100 p-yrs (36/126 p-yrs) in COC users. In new DMPA users pregnancy incidence was 10.20 per 100 p-yrs versus 3.48 per 100 p-yrs in established DMPA users (adjusted hazard ratio [aHR]=2.84; 95% confidence interval [CI] 0.92 - 8.74). Similarly in new COC users pregnancy incidence was 42.67 per 100 p-yrs versus 23.67 per 100 p-yrs in established COC users (aHR 2.08; 95% CI 1.02 - 4.23). Conclusions: New contraceptive users in VOICE had an increased pregnancy risk. Pregnancy incidence was high among both new and established COC users. New DMPA and all COC users participating in HIV prevention trials may benefit from intensive contraceptive counseling and provision of less user-dependent methods.