Global health policymakers should focus on the way entire healthcare systems work together - improving one element does not always create better outcomes if the wider system is not ready.
Publishing their findings today (21 June) in BMJ Global Health, an international research team led by the University of Birmingham, Nottingham Trent University and Stellenbosch University reveals how well-intentioned changes to one part of a healthcare system can lead to worse patient outcomes.
In one of the first studies to capture the full complexity of a health system delivering injury care, researchers mapped how patients move through care, from seeking and reaching help, to receiving treatment, and remaining in care for recovery.
They found almost 1,000 interconnected factors that influence patient survival after injury in low- and middle-income countries (LMICs), which account for around 85% of injury-related deaths worldwide. However, health systems are highly complex and interconnected – improving one part can unintentionally create pressure elsewhere.
Researchers discovered that, among several possible interventions, enhancing trust in the healthcare system had the greatest impact on clinical outcomes. When trust is high more people seek care and outcomes can improve, but if too many people come in and services cannot cope, quality drops and trust can fall again.
Our work clearly shows that improving care results in increased demand on the health system, but this must be matched by greater capacity and higher-quality services to prevent services becoming overwhelmed. Increased demand quickly can lead to delays, reduced quality of care, and declining trust, ultimately undermining the very improvements those changes were intended to achieve."
Justine Davies, Lead Author, Professor of Global Health Research, University of Birmingham
The study analyses interactions which span multiple pathways of care and their interactions including: patient and community factors, such as beliefs, ability to pay, and trust; health service factors, including quality of care, staffing, and funding; and wider societal context, such as poverty, infrastructure, and policy.
Senior author Antuela Tako, Professor of Operations Research at Nottingham Business School, said: "Patient trust, perceptions of care quality, and people's willingness to seek treatment were among the most influential factors shaping health outcomes. However, addressing these factors in isolation is not enough. For example, increasing trust can encourage more people to seek care earlier, improving recovery and reducing mortality and disability."
The research calls for a fundamental shift in how health systems are understood and strengthened through multi-sector approaches, linking health policy with investments in workforce capacity, transport, education, communities, and economic development.
Kathryn Chu, Professor of Global Surgery at Stellenbosch University, said: "Conventional approaches are insufficient to deliver sustained improvement. Things can only be truly improved through a whole-system approach that consider wider social and economic factors and places patient trust, system design and equity at the centre of reform."
The research involved researchers from the UK and South Africa – including University of Aberdeen, UK; Umeå University, Sweden; University of the Witwatersrand, South Africa; National Health Service (NHS), Grampian, UK; University of York, UK; Western Cape Government, South Africa; University of Cape Town, South Africa; and South African Medical Research Council.
The study was supported by a prestigious two-month residential team fellowship awarded by the Stellenbosch Institute of Advanced Studies (STIAS) to Professor Justine Davies, Professor Kathryn Chu, Dr Lucia D'Ambruoso (University of Aberdeen), Professor Laura Bojke (York University) and Professor Antuela Tako.
Source:
Journal reference:
Davies, J. I., et al. (2026). A systems approach to understanding injury care in LMICs using causal loop diagrams. BMJ Global Health. DOI: 10.1136/bmjgh-2025-023069. https://gh.bmj.com/content/11/6/e023069