Implementing an evidence-based sepsis treatment protocol to improve patient outcomes in the Tikur Anbessa specialized hospital emergency department: a multi-component qualitative methods process evaluation

Abstract Background Sepsis remains an important cause of mortality globally, with low- and middle-income countries (LMICs) disproportionately affected. We developed a sepsis treatment protocol tailored to the Tikur Anbessa Specialized Hospital Emergency Department (TASH-ED), Ethiopia. We observed a decreased trend for survival at 72 h and no change in 90-day mortality post-implementation. A concurrent process evaluation was undertaken to explore factors affecting implementation. Methods We conducted a pragmatic multi-component qualitative descriptive study, including: interviews and a document review of staff and study team meeting notes, to identify barriers and facilitators to implementation. Data were analyzed using qualitative content analysis and triangulated using multiple methods. Results Nineteen TASH-ED staff and trainees were interviewed. Material and human resource shortages were identified as the primary barriers to implementation. Material resources shortages included inconsistent availability and delays in accessing necessary medications, supplies and equipment. Human resource shortages were noted to impact implementation through increased provider work load, and by limiting the implementation team’s ability to support activities, due to competing demands. Provider knowledge of protocol content was high among physicians, but varied among nursing staff. Some participants felt that this variability was a result of staff shortages, leading providers to relay on prior practices, rather than engage with and internalize the new protocol content. The vast majority of participants agreed with the protocol’s content, and valued its format including the protocol reminders posted in large format in clinical areas and on pocket cards. However, due to lack of back-up coverage for implementation team member absences, a number of participants reported not receiving the pocket card, and therefore did not have ready access to the antibiotic recommendations. Conclusion Despite tailoring of our sepsis protocol and implementation strategy to context, our process evaluation identified important barriers impacting implementation. In addition to barriers recognized a priori that tailoring did not fully address, a number of previously unidentified barriers emerged. These findings highlight the importance of including process measures in monitoring and evaluating implementation activities. The pronounced impact of resource shortages highlights the need for health system strengthening for outcomes to be improved for sepsis patients in LMICs. Study registration Open Science Framework https://osf.io/8ux6gRegistered June 28, 2017.