Search

Background: Congenital anomalies have risen to become the 5th leading cause of death in under 5 year olds globally and 11th leading cause of years of life lost for the global population. This equates to over half a million annual deaths, 97% of which occur in low and middle-income countries (LMICs). Many gastrointestinal congenital anomalies are fatal without timely access to neonatal surgical care, but few studies have been done on these conditions in LMICs. This PhD aimed to 1) define the mortality from gastrointestinal congenital anomalies in low-, middle- and high-income countries (LICs, MICs and HICs) globally, and 2) develop and implement an interventional care bundle to reduce mortality from gastroschisis in LMICs, a common gastrointestinal congenital anomaly known to have a high neonatal mortality in sub-Saharan Africa (76%) compared to HICs (2%). Methods: I established the Global PaedSurg Research Collaboration consisting of children's surgical care providers across the globe. I designed and led our multicentre, international, prospective cohort study comparing mortality from the seven commonest gastrointestinal congenital anomalies in low-, middle-, and high-income countries globally (Chapter 4). The primary outcome was all-cause, in-hospital mortality. Chi-squared analysis was used to compare mortality between income strata for all study patients, and for each condition separately. Multivariable penalised Lasso regression analysis was used to identify factors associated with mortality. I designed and undertook a 2-year pre-post Gastroschisis Interventional Study, with study team members at seven tertiary paediatric surgery centres in Ghana, Zambia, Malawi, and Tanzania (Chapter 5). A systematic review, qualitative study and consensus process were used to develop the evidence-based care bundle (Chapters 6, 7 and 8, respectively). This consisted of core components, that were implemented in all seven study centres, and adaptable components unique to each centre to optimise the care bundle for each setting. Patients with simple gastroschisis (intact bowel that can be reduced at the cotside; without necrosis, perforation, atresia, or defect closure necessitating surgery) received the intervention. The primary clinical outcome was all-cause, in-hospital mortality; a difference in the pre- and post-intervention mortality was calculated using Chi-squared analysis. An interim analysis was undertaken where univariable analysis was used to identify factors associated with mortality. These preliminary findings are presented in this thesis (Chapter 9). Results: The Global PaedSurg cohort study included 3849 patients with 3975 study conditions (560 with oesophageal atresia, 448 with congenital diaphragmatic hernia, 681 with intestinal atresia, 453 with gastroschisis, 325 with exomphalos, 991 with anorectal malformation, and 517 with Hirschsprung's disease) from 264 hospitals (89 in HICs, 166 in MICs, and nine in LICs) in 74 countries. Mortality among all patients was 37 (39·8%) of 93 in LICs, 583 (20·4%) of 2860 in MICs, and 50 (5·6%) of 896 in HICs (p<0·0001 between all country income groups). Gastroschisis had the greatest difference in mortality between country income strata (nine [90·0%] of ten in LICs, 97 [31·9%] of 304 in MICs, and two [1·4%] of 139 in HICs; p≤0·0001 between all country income groups). Factors significantly associated with higher mortality for all patients combined included country income status, sepsis at presentation, higher American Society of Anesthesiologists score at primary intervention, surgical safety checklist not used, and unavailability of ventilation or parenteral nutrition when needed. Administration of parenteral nutrition and use of a peripherally inserted central catheter or percutaneous central line were associated with lower mortality. For the Gastroschisis Interventional Study, an evidence-based care bundle was established through a systematic review, qualitative study and consensus process, consisting of district hospital and tertiary level interventions. Core components consisted of outreach education on initial resuscitation and early transfer from referring hospitals, with the development of referral networks, and early resuscitation, preformed silo application, bowel reduction, and sutureless closure at the cotside without general anaesthesia, and an accelerated breastfeeding protocol, at the tertiary hospitals. Adaptable components included the use of locally sourced intravenous nutrition and type of intravenous access. In total, over 2000 multi-disciplinary team members were trained at the seven study centres, and their referring hospitals, and all centres successfully established the care bundles into routine clinical practice. Overall gastroschisis mortality was 97% (108/111) at baseline (during the 2-years prior to the study). Mortality from simple gastroschisis fell from 80% (47/59) during the pre-intervention phase to 65% (87/133) during the post-intervention phase (p=0.047). Of all patients with gastroschisis (simple and complex), 3/111 patients survived at the 7 study centres in the 2 years prior to the study and 68/281 survived during the 2-year study period. Hence, the number of patients presenting for care increased during the study as well as the survival. Due to the nature of the condition, it can be assumed that patients who do not present cannot survive. The interim analysis highlighted improvements in care delivery at the district and tertiary levels. The leading cause of death was sepsis. Lower mortality was significantly (p < 0.05) associated with: not having associated anomalies, arriving at the study centre sooner, having a lower ASA score at primary intervention, having a surgeon present at primary intervention, receiving twice daily silo bowel reductions, receiving a sutureless closure, and having a surgeon, neonatologist or paediatrician, and nurse present at defect closure. Conclusion: The Global PaedSurg study highlighted that Sustainable Development Goal 3.2, to end preventable deaths in neonates and children under-5 by 2030, will not be achievable without urgent action to improve neonatal surgical care in LMICs. Interventions to improve survival should focus on 1) improved antenatal diagnosis and delivery at a paediatric surgery centre, 2) early resuscitation and transfer for those born at the district level, and 3) early resuscitation and improved perioperative care, with availability of central venous access, parenteral nutrition and ventilation for neonates, at paediatric surgery centres. The Gastroschisis Interventional Study demonstrates how an evidence-based interventional care bundle can be developed and implemented across multiple LMIC hospitals & countries to improve survival from gastroschisis within the existing healthcare systems; a model that could be used and developed further for a broader range of congenital anomalies.

In the world today, two out of every three people lack access to essential or emergency surgical care. Compared with high-income countries (HICs), surgery in low- and middle-income countries (LMICs) is generally of inferior quality and surgical outcomes are significantly worse. This creates gross inequity worldwide. Quality is a multi-‐dimensional concept and while complex to define, commonly agreed dimensions include: effectiveness, efficiency, equitability, timeliness, safety and patient-centeredness. My research focuses on the dimension of patient safety. Although many interventions to improve perioperative patient safety do exist, very few are widely implemented in LMICs. One of the greatest challenges therefore facing the global surgery community is how to take proven interventions and apply them successfully at scale in LMICs. The purpose of my research is to tackle this challenge head on, using the World Health Organisation (WHO) Surgical Safety Checklist (SSC) as an example. SSC usage rates in HICs are estimated at 88-89%, but less than 30% in LMICs. In all settings, the SSC has been repeatedly shown to reduce surgical mortality and morbidity. For clinicians, academics and policy-‐makers, the knowledge gap to be addressed is no longer, 'does the SCC work?' but, 'how can [we] make the SSC work in our setting?' And the relevant research questions are as follows: firstly, can the SSC be implemented successfully and sustainably at scale in LMICs; and secondly, is the effort required worth it? This thesis (incorporating five publications) aims to answer these questions by taking an implementation science-‐based approach to address the first question, and two different economic approaches to answer the second question. My overall goal in undertaking my research is to inform the knowledge gap in SSC implementation in LMICs, thereby 'making surgery safer in LMICs in sub-Saharan Africa'. The over-arching methodological approach to my research is grounded within the discipline of implementation science. Implementation science research is a form of health policy and systems research that can be used to study and support the scale-up of evidenced quality improvement interventions and their subsequent integration into health systems at the national level. To achieve my research objectives I used three implementation science frameworks: the Expert Recommendation for Implementing Change, the Consolidated Framework for Implementation Research, and the Knowledge-to-Action Framework, and eight internationally recognised implementation outcomes. I employed a range of study designs and data analyses: systematic review and meta-analysis (Paper I); mixed-methods design, multivariate linear regression and thematic analysis (Papers II-IV); cost-effective analysis and cost-benefit analysis (Paper V). The systematic review and meta-analysis on SSC implementation in LMICs (Paper I) showed that most SSC implementations to date were single centre studies with poor reporting of implementation strategies. Estimates of clinical effectiveness showed SSC use was associated with a 23% reduction in mortality and a 44-56% reduction in complications. Nationwide SSC implementations in Madagascar, Benin and Cameroon that I undertook in collaboration with the non-governmental organisation Mercy Ships (Papers II-‐IV), showed SSC use increased significantly after a multi-faceted implementation strategy with high usage rates (74%, 86% and 54% respectively) sustained over time. Overall, most operating room staff reported positive personal and organisational benefits of SSC use, and the commonest barriers to use were lack of time and obstructive leadership. The economic evaluation (Paper V) showed that my model of nationwide SSC implementation was 'very cost-effective' by WHO standards and provides a return on investment of between $9 - 62 dollars for every $1 dollar spent. The research presented in this thesis shows that, to date, most SSC implementations in LMICs have been limited with few widespread initiatives, and generally poor reporting of the implementation process. The approach I have presented in the thesis for nationwide SSC scale up is feasible, has a positive and sustainable impact on personal and organisational practice, and is both cost-effective and of net-positive benefit to society. I conclude the thesis with reflections on my dual role as an implementor-evaluator and potential avenues for further integration of implementation methods in global health to achieve high quality perioperative care in sub-Saharan Africa.

Trauma and injuries represent a vast and relatively neglected burden of disease, predominantly affecting low- and middle-income countries (LMIC) where outcomes are worse than high-income counterparts. A whole health system assessment approach offers a potential benefit in better researching, understanding and addressing this substantial, neglected and unequal burden of disease. Efficiently assessing injury health systems offers potential benefits for all stakeholders. In my thesis, I develop and use a Three Delays framed mixed-method assessment to evaluate the trauma health system serving Karonga HDSS, Northern Malawi. I also appraise the suitability of the methods for future use in rapid health system assessments elsewhere. In the first chapter, I set the scene for why a rapid assessment of trauma health systems is important and how the Three Delays framework (considering delays to seeking, reaching and receiving care) can help. This PhD is supported by the British Army and Defence Medical Services. Therefore, in chapter two, I provide the context and background for why global health research in general, and LMIC trauma health system research in particular, are of relevance and importance for militaries. The third chapter is a systematic literature review and evidence synthesis mapping the Three Delays framework to injury health system assessments within LMICs. I found few studies considered the whole health system as understood by the Three Delays framework, with assessments of facility capacity dominating the literature. Most studies used methods that could potentially be applied within a rapid assessment. However, innovating and combining methods to generate insight into Delays 1 and 2, and evaluating the quality of care processes and outcomes, require greater attention. In chapter four, I present a Delphi study to establish expert consensus on the most important barriers, within a Three Delays framework, to accessing quality injury care in LMICs, that should be considered when evaluating a health system. I identified 20 distinct barriers potentially delaying care following injury in LMICs. Categorised within a Three Delays framework, expert participants reached consensus of agreement that eleven barriers were easy to assess, likely to delay care for a significant proportion of injured persons, likely to cause avoidable death or disability for affected injured persons, and could potentially be readily changed to improve care for injured patients. I, therefore, drew on the systematic literature review and Delphi study findings to inform the selection and development of methods used in my injury healthcare system assessment. In chapter five, I describe the methods used to assess the injury health system serving the Karonga Health and Demographic and Surveillance Site (HDSS), Malawi. The study components are 1) household survey; 2) verbal autopsy database analysis; 3) community focus group discussions); 4) community photovoice study; 5) facility process mapping study of the care pathway and barriers following injury; 6) a survey of facility healthcare workers; 7) facility assessment survey; 8) clinical vignettes for care process quality assessment of healthcare workers; and 9) geographic information system analysis. I also describe the mixed-method integrated analytical approach used. In chapter six I report the non-integrated results from each of my study findings within a vertical study component analysis. In chapter seven, I present the integrated mixed-method analysis of the health system assessment. The novel approach to integrating nine different studies, including qualitative, quantitative, community, patient and healthcare worker derived data sources, offered a rich insight into this health system's functioning, using a Three Delays theoretical lens. This integrated analysis highlights the existence of important barriers to seeking, reaching and receiving care within the Karonga HDSS health system. The perceived financial cost associated with seeking care, despite a primarily government-funded health system with no user fees, was found to have strong convergent evidence as important barriers to seeking care. So too was a lack of community health and healthcare literacy. In the absence of a formal prehospital emergency care system, quickly finding affordable, suitable transport to cover the long distances, particularly to referral facilities, involving unreliable roads susceptible to weather, were found to have strong convergent evidence as important barriers to reaching care. The expected limited availability of physical resources within facilities, and challenges with the availability of motivated, trained staff, including specialists, were strongly evidenced as important barriers to receiving quality facility-based care. These insights provided a rich understanding of and evidence for various health system barriers to consider when planning improvements to the health system response to the injured patient in this and similar contexts. In the final chapter, I appraise the generalisable methodological learning from the thesis. I critically reflect upon the utility of studied methodological approaches in a rapid health system assessment and potential ways to improve their application. My thesis sheds light on the relatively under-researched and less-resourced injury care health system in Northern Malawi. Focussing on the broader injury care health system inclusive of health-seeking and care access, helps fill a gap within Malawian trauma care research, which, as in many LMICs, is facility-centric. Further, my thesis makes an important contribution to the existing literature in applying mixed-methods approaches to such health system research. Integrating nine different methods through a joint matrix display to inform a deep system understanding of a single population is novel in both depth and focus and should encourage debate and use of such approaches for similar future studies. This is particularly the case in LMIC trauma care research, where mixed-methods studies are rare. This study also advances the health systems research application of the Three Delays framework into the fields of emergency care and trauma. Future post-doctoral use of this approach in trauma health system research is planned with funding applications in progress. Undertaking this PhD thesis has facilitated my academic growth and experience with multiple research approaches and methods. It has developed my cultural competence and boundary spanning skills and experience valuable to future academic pursuit in the arenas of both Health Systems Research and Military Global Health Engagement.

Global surgery is a slowly emerging field in global health. In 2006, the World Bank published a chapter on Surgery in the second edition of Disease Control Priorities for Developing Countries (DCP2)1 which used a survey of 18 surgeons to estimate the disability-adjusted life years (DALYs) attributable to surgically treatable conditions. The rudimentary methodology used in this study attracted widespread attention. Since then, efforts have been made to better quantify the burden of surgical disease that is currently untreated, and to gain political traction and raise the profile of surgery within global health. The publication of the Lancet Commission on Global Surgery2 (LCoGS) and the Essential Surgery volume of the third edition of Disease Control Priorities for Developing Countries (DCP3)3, together with the passing of the World Health Assembly Resolution on Essential and Emergency Surgical Care4 led to 2015 being described as a ‘banner year’ for global surgery5. Further progress was achieved in 2016 with the incorporation of the first of the LCoGS indicators into the World Bank Development Indicators6, and then, this year, with Zambia producing the first national surgical plan, with other countries like Ethiopia creating strong national programs to improve the quantity and quality of surgical care. The optimism within the global surgery community has not yet translated into any health gains for the poor and marginalised in LMIC settings, partly because only a few countries have begun to develop policies for surgical care, but crucially also because the funding commitments from the international community and national health budgets are not forthcoming. This inertia is unlikely to be overcome without better evidence to inform policy makers and without political prioritization for surgery within the donor community and by Ministers of Finance. At present, the evidence for the enormous burden of death and disability which could be prevented by access to essential surgical care in low and middle income countries (LMICs), is largely based on modelled data, with almost no primary national population level data on surgical conditions. The same applies for the economic data – it is largely based on modelling of the macroeconomic impact of untreated disease7 and modelling of the financial investments required to scale up care8. The motivation behind these different but related publications was the conviction that in order to realize the vision of universal access to surgical and anaesthesia care globally, there is a need to move from modelled data to primary and country specific data. This thesis brings together five papers all centred around the question: ‘is there any financial incentive for investing in the provision of surgical care in low income countries?’