Background: Implementation Research (IR) in comparison to other research domains, is demand-driven and research questions are based on the needs identified by the implementers in the health system. It is context-specific and is mindful of cultural and community-based influences. Furthermore, although IR is dynamic and adaptive, it takes place within real-life settings and there is no attempt to manipulate the setting within which the intervention is taking place. It engages with relevant stakeholders including the beneficiaries. Since IR is especially concerned with the users of the research and not purely the production of knowledge, it aims to promote the uptake of research findings into routine practice. The process of knowledge translation is promoted through the active involvement of the relevant actors in the identification, design and execution of research and should not be used only as a target for the dissemination of study findings.

Example of an IR project: To inform a planned mass drug administration (MDA) for lymphatic filariasis (LF) in two districts of Indonesia, a micro-narrative survey tool was developed to capture community members’ experiences with MDA and the social realms where drug delivery and compliance occur. The goal of the project was to enhance coverage and compliance in MDA for the elimination of LF in two ‘endgame’ districts. It was a three-phase study involving a baseline survey, implementation of the identified recommendations and an end-line survey. The systematic approach began with the multidisciplinary research team collaborating with the stakeholders and programme implementers to identify barriers related to the delivery of MDA. The relevant stakeholders were involved in the selection of the study sites, development of the survey tool, analysis of both the baseline and end-line surveys, discussion of research findings and resulting recommendations, dissemination of research findings and identification of feasible actions to improve delivery and access.

The barriers to effective coverage of MDA identified included: Men and 15–24 years old youths lacked appropriate information about the programme; misconceptions about drug safety were common; ineligibility criteria were not clear; and there were limited distribution points. The findings were discussed with the relevant stakeholders and feasible recommendations and interventions were executed using existing health system structures. The recommended interventions were implemented within the local sociodemographic context. For example, social media and texting were used for reaching young people, specific messaging was developed for ‘systematic non-compliers’, and flow charts were produced to guide drug distributors. The eligibility criteria was adapted to the local context. Specific messages addressing drug safety, drug-taking procedure, information on eligibility, benefits of compliance by all people and where to go for assistance, were carefully crafted on the packaging of the medicines. Both districts that were responsible for implementing the identified recommendations and the end-line survey showed an improvement in coverage of MDA.

Conclusion: The research conducted was demand-driven and the findings were used by the local health offices to improve delivery and access of MDA services. Furthermore, the research did not manipulate the routine health services. Active involvement enhanced stakeholders’ ownership and enabled them to mobilize local resources and relevant networks to promote drug uptake, improving compliance.

Lessons: The research team profile should reflect the skill sets required to address an implementation challenge and the team should actively engage relevant stakeholders to fully understand the context where the intervention occurs.

Background: Although services to prevent mother-to-child HIV transmission (PMTCT) have increased in sub-Saharan Africa throughout the past decade, with the improvement of HIV testing and anti-retroviral treatment (ART) improving, retention in PMTCT care remains a challenge. Kenya, one of the countries in the region facing this barrier, has committed to eliminating new paediatric HIV infections. In 2014, the country had a 5.6% national HIV prevalence, including an estimated 75 000 women living with HIV who become pregnant annually. Although the percentage of pregnant women tested for HIV is >90%, only 64% of HIV-exposed infants (HEI) received ART for PMTCT. To increase the proportion of infants protected from HIV exposure, the barriers preventing pregnant women and their infants from being identified, linked to and followed up/referred to care services need to be tackled.

The US National Institutes of Health (NIH), the President’s Emergency Plan for AIDS Relief (PEPFAR) and the Implementation Science (IS) Alliance funded the current study (MIR4H). A combination intervention was designed to reduce loss-to-follow-up for women entering PMTCT services in ten health facilities in Kenya using an individual randomized trial approach. Their aim was to evaluate the effectiveness of standard of care (SOC) with active patient follow-up among pregnant women living with HIV and their infants at six months postpartum. The SOC included antenatal care (ANC) and HIV services, while the interventions delivered by lay counsellors included four additional components: individualized health education; retention and adherence support; SMS appointment reminders; and follow-up and tracking of missed clinic visits. Routine data and questionnaires were used to collect the data for the study. The study results highlighted that pregnancy complications, infant deaths, and transfer out of specific facilities increased loss-to-follow-up among women and infants in PMTCT care.

Conclusion: This study encountered many of the realities encountered on the ground when conducting implementation research. The MIR4H study faced real-life challenges – such as delays in funding, health-care worker strikes, shortage of rapid HIV test kits, slow uptake of new HIV guidelines – that together led to evident delays and resulted in an adaptation to the project implementation.

Lessons: Implementation research must be adaptive to any challenges.