Against All Odds: A Mixed-Methods Evaluation of a Behavioral Postpartum Family Planning Intervention in Tigray, Ethiopia

"[I]t appears that the (re)solve intervention is able to move postpartum women along the intention-action continuum to actualize changes in key contraceptive behavior."

Postpartum Ethiopian women have an unmet need for contraceptives 2.1 times higher than Ethiopian women overall (35% vs. 17%). Despite progress in addressing nonuse of contraception through traditional social and behaviour change (SBC), these programmes can be limited by assumptions about what prevents women from using contraception. In response, the (re)solve project (described at Related Summaries, below) used expertise from consumer insights, behavioural design, and public health to design and customise data-informed solutions and services to the needs, motivations, and lived experiences of women and girls in Burkina Faso, Ethiopia, and Bangladesh. The aim of this mixed-methods evaluation of the Ethiopia iteration of (re)solve was to evaluate whether the package of solutions changed postpartum women's intention to use contraception (primary outcome). It also explored associations with behavioural outcomes like modern contraception use and perceptions such as contraceptive self-efficacy (secondary outcomes).

With Bill & Melinda Gates Foundation funding, (re)solve collaborators Pathfinder International, Camber Collective, the International Center for Research on Women (ICRW), and ideas42 undertook a 4-stage process in Ethiopia (click here [PDF] to learn more) in order to address 3 dominant barriers:

• Women have heard from friends or relatives that they are protected from pregnancy by breastfeeding until their menses return.
• Women hear about or see other women experiencing severe side effects due to contraceptives and contraceptive-induced infertility, and they do not hear about any downsides from breastfeeding.
• Providers explain long-acting methods to postpartum women and name them by their duration of efficacy (e.g., three years), and women want to have children before that duration.

The final set of solutions designed to reduce or eliminate these barriers was comprised of the following 4 tools: (i) an antenatal care (ANC) Planning Prompt card that providers and clients use to plan for and record dates for future family planning (FP) counseling visits; (ii) a Postpartum Family Planning (PPFP) Counseling Sheet, which provides clear talking points for providers to use during PPFP counseling; (iii) a Risk Referral Card, which providers complete with clients during immunisation visits to assess their risk of becoming pregnant based on their own circumstances at that time; and (iv) a Home Visit Tracking Log notebook that provides a structured and comprehensive way to systematically track women at risk of pregnancy during postpartum home visits.

From April to December 2021, (re)solve implemented the solution set in 7 primary health care units (PHCUs) that were already working with the United States Agency for International Development (USAID)-funded TRANSFORM: Primary Health Care project, which Pathfinder International implements with a consortium of partners. A total of 183 male and female providers were oriented to the (re)solve project, objectives, and process and trained on proper implementation of each tool. Implementation began in a staggered fashion, with each PHCU starting implementation of the tools as soon as its providers completed the training. For evaluation purposes, the evaluation team assigned 14 PHCUs to receive the package of facility-based (re)solve solutions or to serve as comparison PHCUs.

After approximately 6 months of implementation, the researchers conducted a cross-sectional, facility-based survey of women who were 4-6 months (16-24 weeks) postpartum at intervention and comparison PHCUs (N=321). In addition, they conducted qualitative interviews with intervention providers (N=28).

The majority of postpartum women were exposed to the ANC Planning Prompt during any ANC visits (96%). Exposure to the Risk Referral Card was 99% at the 45-day immunisation visit but dropped to 27% at the 10-week immunisation visit (27%). Exposure to the PPFP Counseling Sheet was very high at both the 45-day immunisation visit (97%) and 10-week immunisation visit (100%). Because the Home Visit Tracking Log was only used by health extension workers (HEWs), there were no survey questions related to exposure to it.

The results indicate that the (re)solve intervention had a positive non-significant association with contraceptive intention (adjusted odds ratio (aOR) 4.3 [confidence interval (CI) 0.8-23.0]). The researchers explain this finding in part by pointing to the fact that intention to use contraception was very high among this study population (93% in comparison group vs. 98% in the intervention group), which means there was not a lot of room for improvement. The intervention had a statistically significant association with other outcomes, including modern contraceptive use (aOR 19.4 [CI 9.2-41.1]), long-acting reversible contraceptive (LARC) use (aOR 7.8 [CI 3.2-18.8]), contraceptive confidence and self-efficacy (aOR=6.1 [2.0-19.4]), and accurate pregnancy risk assessment (aOR=6.9 [1.7-28.4]).

Qualitative data indicate that all interviewees had a strong understanding of how to properly implement the different intervention tools, and there was near-universal appreciation of their ease of use. The availability of different tools in the local language, Tigrigna, was lauded as an important element that facilitated the implementation of the tools. Among the positive feedback: In the context of social and gender norms in this setting that limit women's ability to decide on and use contraception in an equitable manner, a few providers noted how the tools likely helped women start important conversations with their husbands. Among the limitations cited: The (re)solve project experienced major implementation challenges and disruptions during the regional conflict in Tigray (late 2020); for example, multiple HEWs mentioned not being able to complete their usual home visits during various points of the conflict. (The COVID-19 pandemic also impacted implementation.)

Recommendations and considerations for future design and scale include:

• Make tools sturdier, and increase font size.
• Integrate additional reminders, such as ANC follow-up schedules and maternal and child health danger signs to watch for.
• Incorporate additional sexual and reproductive health (SRH) information from clients, such as resumption of sexual activity and contraceptive method of choice.
• Integrate tools with national health guidelines to streamline the training and deployment of these tools and to address the difficulties providers face with working with multiple tools.
• Explore application of digital solutions - e.g., through mHealth interventions that can digitise and streamline these tools for providers.

In conclusion: "This evaluation of the (re)solve intervention for postpartum Ethiopian women adds to the evidence base of studies from low- and middle-income countries that suggest PPFP interventions can have a positive effect on contraceptive intention and use.....It also adds to the evidence that approaches that are designed with and for women and frontline workers can have a positive effect on contraceptive behavior and uptake."