The Safety of Intrauterine Contraception Initiation Among Women with Asymptomatic Cervical Infection or at High Risk for Sexually Transmitted Infections

Background: Women perceived as being at high risk of sexually transmitted infections (STIs) face barriers to initiating intrauterine devices (IUDs).

Methods:   We searched PubMed and Cochrane for studies that directly evaluated pelvic inflammatory disease (PID) risk among women with undiagnosed cervical infection or who were high risk of sexually transmitted infections (STIs), comparing those who did and did not initiate IUDs. We also searched for studies that provided relevant evidence indirectly related to this question.

Results:  Ten studies met inclusion criteria. Two studies provided direct evidence comparing PID rates in women undergoing IUD placement with women initiating other contraceptive methods: one examined women with asymptomatic gonococcal or chlamydial (GC/CT) infection, while one examined women at high risk for STIs. Neither study found a difference in PID rates between IUD users and non-users. Eight studies provided indirect evidence. One study found no difference in PID rates between levonorgestrel and copper IUD initiators. Five studies compared algorithms with laboratory screening for GC/CT. None of these algorithms adequately differentiated women at high risk of current asymptomatic cervical infection who should not undergo IUD placement from women at low risk who may safely undergo placement. Two studies found no difference in PID rates with IUD placement at the time of STI screening compared with delayed or no placement after screening, or no screening within one year.

Conclusions: Limited evidence suggests PID rates are not increased after IUD placement among women with asymptomatic GC/CT or at high risk of STIs. Algorithms to identify asymptomatic GC/CT did not have added benefit over laboratory screening prior to IUD initiation. Women screened for asymptomatic GC/CT according to guidelines on the same day as IUD placement had similarly low rates of subsequent PID as delayed placement after screening or no screening.