Assessing trauma care health systems in low- and middle-income countries, a protocol for a systematic literature review and narrative synthesis

Trauma represents a major global health problem with injuries accounting for more deaths than TB, malaria and HIV combined and with 90% of these deaths occurring in low- and middle-income countries (LMICs) [1]. Along with other non-communicable diseases, death from trauma is set to increase with some projecting road traffic collisions to be the third leading cause of death by 2030 [2]. Non-fatal injuries are common, with 1 billion people sustaining an injury in 2013 that warranted health care [3]. There is also considerable global variation in injury-related morbidity. Disability-adjusted life years (DALYs) in children are 9 times greater in sub-Saharan Africa compared to those in high-income Asia Pacific counterparts [3], likely due to both differences not only in preventative measures but also in the injury care available from health systems. Indeed, if the survival rates following injury in LMICs were to be improved to the rates seen in HICs, the estimated one third of annual global trauma deaths could be avoided [4].

Considering and developing the whole system of trauma care from point of injury to rehabilitation services has resulted in significant improvements in trauma care in high-income country (HIC) settings. Such improvements were particularly amongst the most severely injured [5‐7]. Although prevention is rightly a key focus on reducing the global burden of injuries, better trauma care through system improvement could lead to major reductions in global trauma associated mortality and has been strongly advocated [4, 8]. Furthermore, trauma care has been considered a tracer condition that can be useful in assessing wider emergency health system performance [9, 10]. The World Health Organization (WHO) advocates that promoting essential trauma care will concurrently promote wider health care system improvements beneficial for other urgent surgical and non-surgical emergencies [11]. The Lancet Global Health Commission on High Quality Health Systems has highlighted the disparity between global burden of injuries and the lack of available data on care quality provided by health systems. Better assessment of such care is one of the commission’s stated research priorities [10].

The WHO defines health systems as the “organisations, people and actions whose primary intent is to promote, restore or maintain health” [12]. Health systems have been described as complex adaptive systems that may respond in non-linear, unpredictable ways to interventions [13]. Health systems consist of an intricate web of relationships between the component parts, embedded within social institutions with human behaviours influencing performance and function [14]. Understanding existing health systems, through their assessment, is important to inform impactful health system improvement efforts [15]. Many different frameworks for describing and understanding health systems exist. Such frameworks have their origins in differing paradigms of understanding and sociopolitical backgrounds [16]. Whilst a universal framework for understanding such complex systems may therefore not exist, choosing a particular health system framework of understanding should be done to fit a purpose [16]. One particular framework that has had success in improving understanding and outcomes in maternal mortality is the “Three Delays” model. It was developed to help evaluate the delays to care driving adverse outcome in maternal mortality in LMICs [17]. The framework has been widely adopted in the field of maternal, neonatal and child health in an attempt to evaluate and drive improvements in care [18‐22]. It has also been proposed as a framework through which to evaluate emergency healthcare in LMICs including trauma [23]. The Three Delays framework considers the barriers that result in delays in seeking care (delay 1), reaching care (delay 2) and receiving appropriate care (delay 3) [17].

In order to inform the development of future trauma care health system assessment, we will undertake a review of the existing literature on assessing trauma care health systems. Applying the Three Delays framework of analysis to this literature review will allow future development of a health system assessment strategy based on this approach.

The objective is to conduct a narrative synthesis of a systematic search of the literature on the assessment of trauma care health systems through a Three Delays model framework of understanding, in order to inform the development of trauma care health system assessment.

We will include primary quantitative, qualitative or mixed method studies and secondary literature reviews. No restriction will be placed on language or date. We will also include reports and publications identified from the grey literature including from relevant national and international health organisations (listed in Appendix 1). The inclusion and exclusion criteria are illustrated in Table 1.

A comprehensive search strategy has been developed to electronically search the following databases from inception: MEDLINE (Ovid), Global Health (Ovid), Embase (Ovid), Web of Science (Clarivate Analytics), Cochrane (Wiley), Global Index Medicus (WHO) and Africa Wide Information (Ebsco). The MEDLINE example can be found in Appendix 2, and the full strategy is available on request. The search strategy uses variations in text words found in the title, abstract or keyword fields, and relevant focused subject headings to retrieve articles combining the concepts of (1) trauma, including disaster planning, mass casualty incidents or emergency care along with (2) various types of assessments, evaluations, benchmarking or tools used to create or improve (3) health system programmes. Three-delay models, rapid assessments and verbal or social autopsies will also be verified. Animal studies will be excluded. Articles will be initially collated in Endnote X8 for de-duplication of results. Screening of abstracts and titles will be done collaboratively using the Rayyan QCRI online open-source web application [24]. There is no agreed gold standard for searching the grey literature; however, a four-stage approach has been advocated for seeking relevant grey literature articles [25]. These complementary approaches are searching of grey literature databases, a customised Google search, targeted websites and consultation with experts [25]. We will search the following grey literature databases using the search terms “trauma” OR “injury” AND “assessment” OR “evaluation” AND “health system” modified according to the database requirements. The grey literature databases to be searched are OpenGrey, WorldCat Dissertations and Theses (OCLC) and New York Academy of Medicine Grey Literature Report and Core. We will use advanced Google searches both with and without limiting the domains to .org, .edu, .int and .gov using combinations of the terms “trauma”, “injury”, “assessment”, “evaluation” and “health system”. The top 50 sites will be screened for relevant articles for each search. We will search the specific websites listed in Appendix 1 using the same search terms. We will also include any additional articles recommended by experts in health system research or trauma care that might inform our review. The reference lists of identified articles will be reviewed for any additional articles of relevance to include.

Key term screening within the Rayyan application will be used to remove any clearly identified as animal or cellular studies. Following piloting of the study selection process, two reviewers will independently screen the articles identified, firstly by title and then abstract. Disagreements over eligibility will be discussed in order to achieve consensus. Where disagreement persists, a third reviewer will arbitrate. Full texts of the abstracts will be obtained and assessed for eligibility by two reviewers. Articles not in English will be translated using Google translate where possible. Articles meeting eligibility criteria will proceed to data extraction. Reasons for exclusion will be recorded. Each grey literature database search, Google search, focused website search and expert request will be conducted by one reviewer with a second reviewer confirming eligibility of identified articles. The level of reviewer agreement will be presented in the final report.

This review is primarily focused on the methodological approach identified in the article and aims to understand a wide breadth of diverse research approaches used to assess trauma care health systems. The quality of conduct of each specific study and the trustworthiness of results and findings for each article are therefore less important.

A standardised extraction form will be developed and piloted. Information to be extracted will include author; publication year; study type; study clinical focus; conceptual framework used if applicable; which of the three delays are assessed if applicable; the country location; the methodological approach; and author reported strengths and limitations, including time, pragmatism and cost if reported. Two authors will extract the information independently with a third arbitrating in the case of unresolved disagreement.

The search will reveal heterogeneous studies. Meta-analysis of study findings is not a study objective. A narrative synthesis of the identified papers will be conducted to identify key methodologies used to assess trauma care health systems. The analysis will consider how the differing methodological approaches could be adopted to evaluate barriers and delays to care within a Three Delays framework to best understand trauma care health systems. The methods identified will also be assessed for their suitability to be employed in a rapid assessment, specifically their relative resource requirements including time taken to undertake, pragmatism and cost. An iterative approach will be adopted to categorise identified articles, with the results presented as both within and across study analysis.