• Dmpa
• Iud
• Counselling
• Discontinuation

Following publication of our trial on the effects of postnatal depot medroxyprogesterone acetate (DMPA) compared with the copper intrauterine device (IUD) on postnatal depression in this journal in July 2016,1 we have sought to evaluate contraceptive discontinuation in our study sample. Postnatal contraception is promoted as part of the WHO strategy to reduce the unmet need for family planning in low- and middle-income countries (LMICs) and to reduce preventable maternal and child mortality.2 However, little is known about discontinuation rates associated with postnatal contraception use in these settings.

From the trial, 75 of 242 participants were contactable two or more years after randomisation and 54 consented to a follow-up interview, which was conducted by a Masters student from the University of Fort Hare (NDY). Twenty-three women had received DMPA and 31 women an IUD. In the DMPA and IUD arms, respectively, 48% (11/23) and 42% (13/31) had discontinued their contraceptive methods by the time of the interview. All participants who discontinued did so within the first year, 10 within 3 months of allocation (DMPA=4, IUD=6), and 17 within 6 months (DMPA=7, IUD=10). Reasons for discontinuation are shown in table 1. Six of the participants allocated to DMPA (26%) and five allocated to the IUD (16%) became pregnant following discontinuation.

Contraceptive discontinuation is a major problem in our setting and in LMICs generally, and leads to high rates of unintended pregnancy.3 This is supported by these exploratory data. Despite the limitations of this small study, the high discontinuation and pregnancy rates among our cohort highlight the difficulties that postnatal women, particularly, have in sustaining contraception utilisation in country settings like South Africa. IUD expulsion occurred in one-third of users in this follow-up study, which might be explained by the study limitations; however, this does not appear to be extraordinarily high: a 2015 systematic review of postnatal IUD insertion found IUD expulsion rates ranging from 2.4% to 72% across a variety of study designs.4 This suggests that some high-quality research is needed to clarify the risk of IUD expulsion for postnatal IUD users. Of further concern to us is that reinsertion was not performed in any of the women in our study and it is not clear whether they were aware of, or offered, this option. We believe this reflects the wider clinical practice relating to IUD insertion in our setting, which needs to be addressed urgently. Indeed, it is not surprising that women with infants, and possibly other young children at home, lack the time and resources to attend clinic for family planning-related matters (such as for IUD re-insertion, repeat contraception injections or help with side effects) during the first year postpartum.

A multicentre randomised trial (the Evidence for Contraceptive options and HIV Outcomes or ECHO trial), aimed at increasing long-acting, reversible contraceptive options for women in LMICs is currently ongoing.5 However it is apparent that for health services to effectively expand contraception options, careful consideration will need to be given to implementation of those options provided to postnatal women to ensure continuation of the chosen postnatal method. Such implementation strategies could include routine follow-up visits for IUD users during the first year of use, mobile community contraception services providing a variety of options, ensuring that health providers capable of IUD insertion are always available, as well as effective communication strategies to raise awareness of contraception choice. Given the unique family planning needs of postnatal women and the limited available evidence on discontinuation rates in this group of contraception users, we believe that more research is needed to determine how best to support their contraception needs.