he BOHEMIA trial (Broad One Health Endectocide-based Malaria Intervention in Africa; ClinicalTrials.gov: NCT04966702) was a cluster-randomized controlled trial evaluating mass drug administration (MDA) with ivermectin and albendazole for malaria transmission reduction. It enrolled approximately 28,932 adults and 2,871 children aged 5–15 across 84 clusters in 69 villages in Kwale County, Kenya, with three rounds of MDA conducted between 2021 and 2023. The trial achieved a 26% reduction in malaria incidence in the intervention arm.

BOHEMIA was the trial that made community engagement visible—although through our different professional lenses—in a way it had not been before.

In Kenya, this visibility was partly structural: Community engagement liaisons were not outsiders brought in to interface with the community, but recognized community members—health workers already known, trusted, and embedded in the communities they served. Their presence did not require introduction; it was already legible to the people they engaged. This made community engagement not just visible to the trial team, but visible as the trial to community members—seamlessly integrated rather than announced.

The result was a trial that could achieve a 26% reduction in malaria incidence partly because the communities it depended on were already engaged, already informed, and already invested.

The lesson, visible from both our perspectives, is the same: Community engagement that is documented, structured, and integrated into the trial’s quality and operational architecture produces results that neither monitoring infrastructure nor community goodwill can produce alone.

More Than Trials: Clinical Research as a Care Option

One dimension of community engagement in these settings that has no real equivalent in the experience of most global sponsors is the role that research institutions play in the broader healthcare ecosystem of the communities they work with, and the trust relationships that develop over time when community engagement is sustained rather than project-specific.

At IDI in Uganda, Simon described a public engagement infrastructure that includes peer mothers: individuals with lived experience in maternal health who support participants in ways that clinical staff often cannot.

“For some clinical studies, participants actually trust peer mothers more than they trust a doctor or research nurse. They confide in them because they know this is a person who has similar experience. If I explain to them the challenge I’m facing, they will easily understand what I’m going through.”

This is not a recruitment strategy. It is a care relationship. The peer mother is not trying to enroll anyone. She is providing continuity of human connection to someone navigating a health system that may otherwise feel impersonal, inaccessible, or intimidating. Home visitors trained in counseling perform similar functions, such as following up with participants between visits, providing guidance, and sometimes identifying safety concerns before they escalate. From a clinical operations standpoint, this is a retention mechanism that no site visit schedule can replicate. From a quality standpoint, it is an additional node in the safety signal detection network, one that reaches participants in the spaces between formal data collection windows.

At KEMRI in Kenya, Mumba described what two decades of sustained community engagement produces: a population that understands what research is, knows its rights as participants, and in some cases actively seeks out participation. “We do community engagement for the sake of building relationships, such that researchers can come into our research institute and find communities that are not research-naïve. For us, one of the successes is that we’ve been able to continue carrying on research successfully for almost four decades in the same community,” Mumba explained.

In the Kenya arm of BOHEMIA, this was observable in a specific and meaningful way: Given the general low literacy levels of the population, participant rights—including the right to decline—were identified as a critical-to-quality factor during study design. Significant effort was invested in ensuring that participants genuinely understood the research and what their consent indicated.

The results of that investment were visible in practice: Although community leaders and the community at large had agreed to the trial’s presence in their communities, some individual families declined to participate. Rather than representing a failure of engagement, the study team recognized this as evidence of success: Participants understood not only what they had agreed to, but more importantly, that they retained the right to say no.

In settings where community trust in the research institution has been built over time, participation in a clinical trial is not experienced as an experiment being done to a community. It is understood as a relationship of mutual benefit, in which the community gains access to monitoring, support, and potentially therapeutic benefit, and the research institution gains the access and cooperation that make science possible: Clinical Research As A Care Option (CRAACO).

Professor Adegbola raised a further question sitting at the ethical frontier of this concept: What happens to communities after the trial ends? In his view, informed by his 19 years working in Gambian villages and his current role reviewing trial protocols at the national level, to conduct a study without guaranteeing that the people who participated will have access to the intervention if it proves effective approaches unethical:

“I think we are moving to a stage where it would be unethical to do a study where you cannot guarantee that the people participating will have access to the outcome. That is what the whole idea of community engagement is to avoid: people being used. It is about respecting the individual, and the idea of justice.”

This is a challenge that community engagement alone cannot solve; it implicates pricing, access agreements, and the fundamental economics of drug development. But it surfaces through community engagement, because community engagement creates the conditions under which communities can articulate what they need and expect from research. And it is increasingly on the agenda of regulators and ethics committees as a criterion of ethical review, as Professor Adegbola noted.

What Makes Community Engagement a Critical Success Factor

Several specific mechanisms explain why community engagement functions as a critical success factor rather than a supplementary activity, and why those mechanisms are visible from both a quality and a clinical operations perspective:

It enables protocol feasibility, not just recruitment. When community engagement begins at protocol development (as it does at KEMRI and IDI), the protocol is written with genuine knowledge of the population it will involve. Investigators learn about local social structures, health-seeking behaviors, cultural norms, and potential barriers before those barriers become protocol amendments or recruitment crises. From an operations standpoint, it means the study is scoped and resourced for the community it will actually, not theoretically, enroll. From a quality standpoint, it is the earliest possible intervention point; community-related risks identified before protocol IRB submission won’t become monitoring findings.

It creates an early-warning system that monitoring alone cannot replicate. In the BOHEMIA trial, community-facing personnel were equipped with structured tools to collect participant-reported information, including adverse events, and established channels to transmit that information to the study team, creating an additional pathway for safety-related signals to reach medical monitoring alongside formal safety reporting processes. As our understanding of the community engagement liaison responsibilities deepened and it became clear that these individuals would be interacting directly with participants, families, and caregivers (and potentially hearing about adverse or serious adverse events before any formal report reached the investigator), the study team recognized and rectified this structural gap. Community engagement liaisons were trained on the protocol and purpose-built data collection tools were developed so that, in the event a liaison became aware of an adverse event or serious adverse event (AE/SAE), there was a defined, structured method for capturing and transmitting that information through appropriate channels to the study team.

From a GCP quality perspective, this is a recognizable risk control: a supplementary signal detection mechanism that captures what formal data collection windows miss. From an operations perspective, it is the difference between learning about a problem at the next scheduled monitoring visit and learning about it in time to act. Communities talk to people they trust before they talk to investigators. A well-functioning community engagement infrastructure captures that conversation.

It supports retention, not just enrollment. In these communities, participation in a clinical trial is not typically an individual decision. It is a family decision, sometimes a community decision. When families and community members understand the study, they support their loved ones to remain in it. When they do not, participants often find themselves under significant social pressure to withdraw regardless of their own wishes. For operations teams managing dropout rates and protocol deviations attributable to early withdrawal, this is one of the most consequential dynamics in trial conduct, and yet one of the least visible in standard monitoring reports. From a quality standpoint, retention directly affects data completeness and reliability; community engagement is a data quality intervention as much as an operational one.

It builds the infrastructure that makes future research possible. The KEMRI model demonstrates what sustained community engagement produces over time: a community that understands informed consent, knows its rights, and actively participates in shaping how research is conducted—and an institution that carries that relationship forward across studies and decades. Professor Adegbola described a parallel dynamic from his Gambia experience: communities that would ask between studies, “When are we doing something again?” This is the opposite of the skepticism, resistance, and misinformation that characterize community-research relationships where engagement has been transactional or absent. It also produces inspectability: Community engagement governed by SOPs, trained personnel, and defined roles becomes an auditable element of the trial record, which is directly relevant to GCP inspection readiness and, in the authors’ view, a meaningful differentiator in the quality of sponsor oversight.

Lessons for the Global Sponsor Community

These experiences are specific to Kenya, Uganda, Nigeria, and The Gambia. We do not broadly extrapolate them to Africa or claim that every element of these models translates directly to every trial setting.

We do propose that the underlying design principles—community engagement built in from the start, resourced adequately, governed by clear roles and documented processes, and evaluated for its contribution to study outcomes—are transferable across sponsor organizations and trial settings. They are, in fact, precisely what ICH E6(R3) points toward; specifically, the guidance’s emphasis on stakeholder engagement, participant-centered trial design, and reducing unnecessary burden on sites and participants.

All the above collectively offer the following best practices framework:

• Budget for community engagement as a named line item in both the protocol and the study budget, not as an unfunded expectation that is delegated to the site. Adequate, dedicated resourcing is the most reliable predictor of whether community engagement is sustained or abandoned under operational pressure. It also signals to sites and communities that they are valued, not merely tolerated.
• Engage community engagement professionals during protocol development, before feasibility assessment and well before initiating recruitment. Protocol design is the earliest and most consequential point of intervention for identifying community-specific quality risks and for ensuring that the operational plan reflects the real conditions in which the study will run.
• Require context-specific community engagement strategies that reflect local social structures, cultural norms, and stakeholder landscape, not generic templates. What works in one community may not translate to another; the same logic that argues against one-size-fits-all monitoring plans applies here.
• Establish clear roles and boundaries between community engagement, informed consent, clinical monitoring, and safety reporting to prevent confusion, duplication, or gaps. These functions are distinct and complementary, and clarity in their responsibilities and boundaries is a quality management requirement and an operational necessity.
• Integrate community engagement into quality and risk management frameworks as a formal, recognized source of information about participant experience and emerging safety signals prospectively built into the study at the protocol stage rather than discovering them retrospectively during oversight reviews.
• Define meaningful success metrics. Study completion rates, retention rates, and community trust indicators are the primary outputs of effective community engagement, alongside the critical-to-quality factors established for community engagement at the protocol stage and monitored throughout.
• Support capacity building at investigator sites including investment in CAB infrastructure, community engagement SOPs, and trained personnel. Sites cannot deliver what they have not been resourced and trained to deliver.
• Engage the institutions that have been doing this for decades as genuine partners, not merely as access points to patient populations. KEMRI, IDI, and NHREC hold expertise and institutional memory from which the global sponsor community would greatly benefit. This engagement begins at study design, not at contract negotiation.

Conclusion

One phrase in quality management applies here with direct relevance: Quality cannot be inspected in; it must be designed in. The same is true of community engagement. It cannot be activated when recruitment falters, retrofitted when communities raise concerns, or satisfied by a checkbox at protocol submission. It is built into the architecture of the research settings described above from the moment a study concept is conceived because these practitioners have learned through experience, through failure, and through decades of relationship building that there is no other way to do it well.

From a GCP quality standpoint, this is a “culture of quality” argument: These organizations in these settings have embedded community engagement into their quality management systems, not bolted it on as an appendage. From a clinical operations standpoint, it is a “feasibility and execution” argument: Trials in these settings succeed not (only) because they are staffed by exceptional individuals but because the systems those individuals work within are designed for the specific environments in which they operate.

Global sponsors navigating new regulatory expectations, evolving trial models, and increasing pressure to demonstrate ethical and equitable research conduct do not need to build these systems from scratch. The knowledge exists. The practitioners who hold it are not asking to be discovered. They are asking to be taken seriously.

Community engagement, done well, is community engagement by design. The question is whether the global sponsor community—including GCP professionals and clinical operations leaders—is ready to design it in.