Understanding sepsis recognition and management in a Ghanaian emergency department: a convergent mixed methods study

Prah, Angela (2024). Understanding sepsis recognition and management in a Ghanaian emergency department: a convergent mixed methods study. University of Birmingham. Ph.D.

Preview

Prah2024PhD_Redacted.pdf
Text - Redacted Version
Available under License All rights reserved.
Download (12MB) | Preview

Abstract

Background: Sepsis is a severe response to an overwhelming infection, resulting in inflammation, coagulation, multi-organ failure and potentially death when not recognised and treated promptly. Sepsis is recognised as a significant cause of hospital admission and preventable deaths globally, hence it is considered a medical emergency. Many high-income countries have prioritised sepsis; however, it has received less attention in adult populations in low-income healthcare contexts, including Ghana, except in children and pregnant women. This study, therefore, explored existing practices associated with the recognition and management of sepsis in a Ghanaian secondary level hospital emergency department (ED) in order to develop a context-sensitive evidence-based sepsis bundle and pathway for future implementation and testing.
Methods: A convergent multiphase mixed methods design was employed. This included a: (1) systematic literature review; (2) a retrospective case record review (n=75); and (3) process mapping of ED sepsis practices, including interviews with healthcare professionals (n=14). Quantitative data were analysed using SPSS version 28.0.0 and interviews and field notes after transcription were analysed using thematic analysis supported by NVIVO© version 14. Data were integrated and findings were (4) presented at a series of co-production workshops with stakeholders to develop a sepsis intervention and plan for future implementation.
Findings: Twenty-two papers met the inclusion criteria for the literature review. Most of the papers used the Surviving Sepsis Campaign (SSC) bundle (21/22): one adopted the integrated management of adolescent and adult illness (IMAI) tool. Prior to introducing the bundle, various engagement strategies were employed with local teams and bespoke training was developed for staff. Reduction in mortality was associated with timely interventions, however, one reported increased mortality as a consequence of oversimplification of the implemented bundle.
The retrospective case record review identified delays and inaccurate sepsis recognition at presentation, time to medical assessment, omission or delayed vital sign/deteriorating patient re-assessment, access to lactate estimation and speed of reporting of routine blood tests and blood cultures. Elements of the SSC bundle were embedded in practice but others were unavailable due to resource and financial constraints. Similar findings were uncovered in the process mapping interviews and workshops, including, not thinking of sepsis as a probable diagnosis until later whereas it was found that other conditions, such as malaria, contribute to targeted management delays and poorer outcomes. Integration using the capability, opportunity, motivation - behaviour (COM-B) model was used to illuminate findings which were discussed in the co- production workshops to improve the recognition of sepsis and implementation of appropriate interventions. In this case, a sepsis algorithm and educational package were designed.
Discussion: The literature review suggested the SSC bundle could be successfully implemented in LMICs if contextual needs were accommodated and engagement with local multidisciplinary teams occurred. With this background, the retrospective review of case notes and process mapping aided in identifying the current practices regarding sepsis recognition and care. With this, possible pathway components and processes were identified through the lens of COM-B and Kotter’s eight step change models and debated through integration and co-production workshops. These were contextualised, and a sepsis algorithm with a standard operating procedure was adapted to enhance sepsis identification and management (nurse led approach to identification), including regular monitoring of vital signs/deterioration, reorganisation of sample collection and reporting services and a policy approach for antimicrobial stewardship. These were developed to support implementation together with an education programme.
Conclusion: This study demonstrates that engagement with key stakeholders in the target site is complex and necessary to develop a culturally specific evidence-based sepsis pathway. Early phases identified potential barriers and facilitators to successful implementation, and these have been considered, and where appropriate, integrated into the proposed implementation model. Recommendations for the designed intervention and implementation plan are outlined for future testing in the target clinical setting.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):