Implementation Challenges and Successes of a Multicenter, Randomized, Controlled Trial of Patients With Hypoxemic Respiratory Failure at Five Hospitals in East Africa

Abstract Rationale: Although most critical care and pulmonary research is conducted in high-income countries, the barriers and facilitators of randomized controlled trials (RCTs) in resource-constrained settings are poorly characterized. Understanding RCT implementation in these settings can facilitate future studies and ultimately improve equity and generalizability of research. Methods: We analyzed qualitative data from Building Respiratory support in East Africa Through High flow versus standard flow oxygen Evaluation (BREATHE), an RCT comparing clinical outcomes of high-flow oxygen therapy to standard care at resource-constrained, predominantly RCT-inexperienced hospitals in Rwanda, Malawi, and Kenya (ClinicalTrials.govNCT05754034). Employing the Consolidated Framework for Implementation Research (CFIR) and Proctor's Implementation Outcomes Framework, we conducted rapid CFIR-based inductive analysis on 34 weekly staff meetings and 8 clinician interviews to identify barriers, facilitators, and other features of trial implementation. Results: Trial acceptance was facilitated by preexisting trust in local leaders, attentiveness of research assistants, appreciation of increased patient monitoring, education sessions for clinical and study staff, and perception of intervention benefit. Clinicians appreciated receiving the results of diagnostic testing performed by the study, and hospital leadership at one site requested study protocols to adapt for clinical care. Concerns about harm for the intervention were rare. Consent processes were adapted to cultural preferences for extensive family participation, particularly at two sites. Consent was facilitated by language and/or tribe concordance between patient and study staff at three sites. Suspicion of the study intervention occasionally arose surrounding the formal consent process and the switch from hospital device to study device. A government-employed doctors’ strike halted enrollment at one (government) site when few patients were presenting for care, while another (non-government) site was overwhelmed by an influx of critically ill patients, making study enrollment difficult. A public health emergency halted enrollment at two sites when study devices were needed for patients with Marburg virus. Infrastructure challenges included delays in replacing and repairing equipment, as well as adaptations around variable internet connectivity. Both oxygen and electricity interruptions occurred at some sites. Regulatory, data-sharing, and patient protection requirements were substantial, varied by country and over time, and required significant resources to anticipate and manage. These findings are preliminary, pending analysis of additional meetings and interviews. Conclusions: CFIR-guided analysis of BREATHE revealed unique barriers and facilitators to implementation relating to acceptability, environment, consent, communication, critical incidents, infrastructure, and regulatory oversight. These insights could streamline future RCTs in East Africa and other resource-constrained contexts.