Research on high quality health care needs to move beyond what to how
Health system quality has received much attention since The Lancet Global Health Commission on high quality health systems in 2018, which proposed new ways to define, measure, and improve the performance of health systems.
Factors influencing quality of health systems are well known, both as inputs and processes, including WHO’s six building blocks (service delivery, health workforce, health information systems, access to essential medicines, financing, and leadership and governance), but we still need to think how to measure them.
Leadership and governance is arguably the most important building block, but few large-scale, quantitative studies include health system governance components when assessing primary care level health system quality in low-income and middle-income countries (LMICs).
The study by Todd P Lewis and colleagues in The Lancet Global Health aims to close this evidence gap by assessing health service quality in six LMICs, highlighting the system determinants of primary health care by analysing data from Service Provision Assessments (SPAs).
The authors apply a positive deviance approach comparing service provision in the best performing (top 10%) and worst performing (bottom 50%) health facilities (both hospitals and clinics). The aim is to identify facility-level factors that explain the performance gap across four dimensions of performance: population factors, governance factors, workforce factors, and tools. Although all four dimensions were significantly associated with best performance among the clinics, only tools were slightly better in best performing hospitals. The private-for-profit hospitals and clinics were significantly better performing than government owned facilities. The authors conclude that high quality governance, management, and community engagement are associated with best performance; yet the multivariable adjustment showed governance as a minimally contributing factor of best performance compared with workforce and tools.
Lewis and colleagues’ study further validates what is known about factors that influence health system quality.
It will be interesting to see whether these results will challenge implementation science practitioners on how to improve health systems with different methods or whether it will advance measurement approaches for monitoring improvement. There are different lenses through which to understand and measure health system quality drivers, some of which unpack the more tangible factors (service delivery lens) and some of which tackle the more complex factors, such as the attributes of the actors and the context (social and system lenses), while considering different health system levels.
Lewis and colleagues assessed the health system quality drivers and their association with system performance. The authors are to be commended for using existing datasets to avoid duplication and for applying a positive deviant approach, as we should be wary of investing in more large-scale studies that tell us what we already know.
There are several limitations to Lewis and colleagues’ study. First, more consideration is needed about using the SPA to measure health system quality, including if these measures can be validated and correctly used. For example, in Lewis and colleagues’ study external supervision was assigned as a measure of governance and supportive supervision was assigned to workforce, and both external supervision and supportive supervision measures were significantly different between best-performing and worst-performing facilities. Second, the authors did not analyse middle-performing health facilities, so there is no understanding of the transition of health system functions between performance levels. Third, this was a large-scale, cross-sectional survey, in which health system function measures, such as client feedback mechanisms, management meetings, quality assurance, and promotion opportunities, are assessed through semi-structured interviews rather than longitudinal observations and qualitative interviews. Fourth, the SPAs are limited to a checklist approach and do not capture the more nuanced realities of practice and governance.
Fifth, the SPAs do not use hierarchical modelling to quantify the change in gradients in clinical care when the different measure scores are adjusted. Sixth, the authors did not draw on the wealth of knowledge in health policy and system research, including seminal papers on governance, leadership, and management.
Delivering health care is challenging and complex, especially in low-resource settings. Investing in the everyday practice of governance will be crucial as human resources and tools can be translated into good processes and outputs through effective governance.
In the Sustainable Development Goal era, health system researchers need to go beyond identifying what factors influence quality of care and start to identify how to change quality of care. Using the suggested new approach of comparing best performers with worst performers and context-specific investigation could be an entry point. What is needed for high quality health systems is a contextual understanding of what works and why in order to determine how to better implement change within each setting.
Source : Lancet Publication