High-quality health systems in the Sustainable Development Goals era: time for a revolution

The key findings of this section are shown in panel 12.

System-wide improvements in quality of care will require effort from providers, health system administrators, and communities, but they begin with a political commitment from heads of state and ministers. Global development partners can and should assist, but they should not drive this agenda. Contributions from across the health system, including the private sector, and from sectors outside of health will be crucial. Early gains in quality are likely to be visible within a few years, though meaningful improvement might take longer. People everywhere have a right to receive effective and respectful care—the time to get started is now.

We propose a new improvement approach that addresses the scope of the quality challenge and recognises the complex adaptive nature of health systems. This approach emphasises macro-level reforms—what we call universal actions—that can not only establish and cascade systemic change across all levels of the health system, but also include a role for targeted meso-level and micro-level strategies. Macro-level strategies are best able to directly tackle the social, political, economic, and organisational structures that shape a health system. Meso-level (subnational) interventions address quality of care through the coordination and management of a network of facilities and communities. Interventions at this level are also well positioned to improve communication and learning between facilities and across levels of the health system. Micro-level interventions aim to directly influence the performance of the staff or the operations of a facility. Appendix 2 includes examples of interventions at the three levels of the health system.

A transformative quality improvement agenda is based on the recognition that health systems are complex adaptive systems, defined as systems in which many component parts interact in unexpected ways and often produce unanticipated results. 221 , 222 Complex adaptive systems are resistant to change, and diffuse and isolated interventions, especially at the micro-level, are unlikely to result in large-scale improvements. 221 , 223 An example of this is the proliferation of point-of-care technologies for health, few of which have been taken to scale or shown to have had an effect on health in LMICs. At the same time, evidence 23 from health and other sectors shows that complex adaptive systems can thrive if actors within the system have a shared vision, clear rules, and space to allow evolution and learning. Research 224 in behavioural economics noted that successful systems create a choice architecture that supports intended goals and reduces harmful variation. Choice architecture comprises the elements of a system that influence choices and behaviour, including information flow, incentives, presentation of choices, and decision-making contexts. 224 Nudging, or steering people in a particular direction while preserving their choice, is a common behavioural economics strategy, but the broader notion is to align motivations, incentives, oversight, and management across levels to promote the best actions.

Research on strategies to directly improve health worker and facility performance found that most micro-level solutions have modest effect sizes. Studies tend to be small and brief, limiting conclusions about sustainability and effects at scale. This Commission recommends that selected meso-level and micro-level interventions be implemented alongside efforts to improve the foundations of health systems.

However, strategies for quality improvement in LMICs have generally focused on a narrow set of solutions, such as increasing health system inputs and changing people’s behaviours and routines at the point of care—ie, the lowest (micro) level of the health system. A 2018 review of primary care quality found that, globally, 72% of strategies targeted the micro level ( ; appendix 1). Although interventions aimed directly at facilities and staff can be motivational and promote local commitment to quality, 218 people tend to revert to entrenched ways of doing things, especially when surrounding systems do not support transformation. 23 The application of multiple micro-level interventions might lead to deleterious effects, with interventions clashing at the point of care because implementing them consumes a large amount of attention from managers, potentially detracting from other priorities. 219 , 220 This raises the challenge of how to situate micro-level efforts as part of broader reforms that will improve health systems.

Despite some impressive health gains in LMICs in the past several decades, this Commission’s analysis showed that health systems are beset by poor-quality care. The pervasiveness of poor quality suggests that the cause is not a few weak providers or clinics, but rather that whole health systems are underperforming. To successfully address the endemic nature of poor-quality care and to give providers the right support to deliver the competent and respectful care that people deserve, this Commission calls for an ambitious improvement agenda that moves beyond targeting the manifestations of poor quality and aims to transform health systems.

Universal actions for improving quality

This Commission recommends four universal actions to improve health system quality: governing for quality, redesigning service delivery to optimise quality, transforming the health workforce, and igniting people’s demand for quality ( ). These actions are based on successes and failures from all countries, best practices from high-performing health systems, research and evaluation, and the experience and deliberation of the Commissioners. This Commission sees these universal actions as the start of a paradigm shift towards a more ambitious health system improvement agenda. Beyond the universal actions, countries can select additional targeted opportunities that fit their needs and context. All universal and targeted actions are predicated on having adequate health system inputs, such as staff, medication, and equipment. The optimal composition, design, and implementation of the improvement agenda will vary by country, because approaches that work in one setting might not work in another. Countries need to monitor the implementation of this agenda to permit adaptation and assess the effects on health and other valued outcomes.

Universal action 1: Govern for quality

Health-system-wide change demands that the improvement and maintenance of quality be woven into the fabric of a health system. Governing for quality means reframing the pursuit of quality health care from a peripheral activity to the mandate of a health system, and making sure that a commitment to quality is actually translated from paper to actions that improve the health of people.225,226 Governing for quality includes several elements: adopting a national quality policy and strategy, improving capacity for management at all levels of the health system, strengthening regulation and accountability, and collecting and learning from health system data.

Governing for quality requires high-level political commitment to a shared vision for improving quality of care and translating this commitment into action across the health system (panel 13). Well aligned policies and strategies should be based on this vision, locally accepted definitions of quality, and national goals for improved outcomes.212 In response to requests from countries for guidance on how to design and implement these healthcare policies and strategies, WHO produced the National Quality Policy and Strategy Handbook.212 The handbook outlines eight elements of the strategy and argues that quality must be elevated nationally and become a priority across sectors. These policies, and the strategy linked to them, should ideally outline the roles and responsibilities of the organisational bodies and actors that participate in sustaining and improving quality of care. A plan for coordinating these elements is also needed, so that quality improvement programmes are harmonised to maximise learning and results at the system level.227 For example, an analysis228 of surveys from 310 health system leaders in Mexico identified insufficient coordination of quality improvement agendas and an unclear system of roles and responsibilities as key barriers to the translation of federal policies into improved quality of care. The successful development of shared vision, policies, strategy, coordination, and implementation are needed to design a choice architecture for health systems that directs patients and providers towards decisions that produce quality care and good health outcomes.

Improving the quality of the health system requires action from multiple sectors and stakeholders. Governing for quality includes managing these relationships and convening stakeholders under the shared vision of making large-scale sustainable improvements in quality and health outcomes.229 Inclusive processes that bring a diversity of voices together to solve problems are complex and difficult to manage, but they help to make action on quality possible, they foster innovation, and they lead to more comprehensive solutions.230 Building partnerships means aligning all stakeholders, including international donors, with national needs and priorities, which is a challenging goal. For example, in 2016, only 16% of development assistance for health went to the strengthening of health systems, despite evidence showing that condition-specific funding can compromise overall quality of health care and crowd out existing health services.231–233

Adopting a national quality policy and strategy, and engaging stakeholders around it, requires not only strong leadership skills, but also good management at all levels of the health system to effectively use available resources to realise the vision of high-quality health care.234 Middle management at the district or regional level could play an important role at the intersection of policy and implementation, although management capacity interventions at all levels have been linked to better health sector performance.235,236 Although the literature consistently points towards the importance of good management across health system levels, insufficient attention has been paid to creating the capacity for health-care management in LMICs.235,236 Data from multiple LMIC settings showed that management is a key factor that differentiates between high-performing and low-performing facilities.237,238 Bradley and colleagues235 outlined eight key management competencies and recommended designing training programmes for management professionals to achieve them. These key management competencies are: strategic thinking and problem solving, human resource management, financial management, operations management, performance management and accountability, governance and leadership, political analysis and dialogue, and community and user assessment and engagement. Examples of effective training programmes239,240 exist in various settings, including Ethiopia,239 where hospital performance improved under the management of graduates of the Masters in hospital and health care administration programme.

Absence of multistakeholder commitment leads to minimal quality improvement in Nepal

In 2007, Nepal endorsed the Policy on Quality Assurance in Health-care Services, with the objective of ensuring “quality of services provided by governmental, non-governmental, and private sector according to set standards” and to establish an “autonomous body to ensure impartial decision regarding health services.” 11 years later, the success of this policy remains mixed.

Why did the policy have low impact? The policy was created without a shared vision and buy-in from stakeholders, including the Ministry of Health. Important partners, such as the Ministry of Education, did not provide critical inputs. The policy designers also did not create consensus on a definition of quality or agree on indicators against which to measure progress. A centrepiece of the policy—to establish an autonomous body for quality of care— never materialised. A quality assurance section was established in the Department of Health Services, but it has little leverage over other units in the Ministry.

The absence of a political commitment and involvement of all stakeholders has meant that the objectives of the Policy on Quality Assurance in Health-care Services have been largely unrealised, and health institutions continue to deliver subpar care quality.A90

Governing for quality through strong accountability in Argentina

In 2005, Argentina implemented a public supplementary insurance program, SUMAR, designed to increase access to quality health care for uninsured children and pregnant women and to address large disparities in infant and maternal mortality rates.A91,A92 The programme is credited with decreasing the probability of low birthweight among beneficiaries by 19%.A91

In the setting of Argentina’s national decentralised health system, SUMAR’s success was dependent on high-level political commitments, buy-in from provincial governments, and well designed reporting pathways to ensure accountability. A presidential decree established the programme and provincial governments confirmed it under a collaborative agreement with their respective providers. The agreement is renewed yearly with review of procedures for expenditures and goals to be achieved. Federal commitments and provincial implementations were aligned through clear standards, and multidisciplinary oversight bodies monitored performance. A provincial level programmatic office regularly reported to the federal level. Local accountability was increased through the centralised monitoring of transferred funds to the provinces. The provinces were then responsible for enrolling beneficiaries, organising the provision of services, and paying providers.

Source: Amit Aryal and Franziska Fuerst.

Source: Programme SUMAR, Argentina. *Panel references can be found in appendix 1.

To improve and guarantee quality care, good leadership and management competences must be buttressed by regulatory structures that create accountability. Strong regulatory mechanisms, ie, so-called regulation with teeth—and transparency through good monitoring, measurement, and reporting practices—support accountability both internally within the health sector and externally with civil society and citizens.225,241 The accountability mechanisms, in turn, should be operated by leadership and management that can pull together a complex array of regulatory domains (eg, workforce, facilities, products, and service delivery) that might be administered by multiple institutions. Lessons from the regulation of medicines suggest that multipronged collaborative approaches that include a suite of regulations, mechanisms for legal redress, and training of inspectors in the public and private sector are most likely to be effective in mixed health systems.242 These accountability mechanisms should also include monitoring of the flow of providers between private and public practice.243 Two first steps that are yet to be taken in many LMICs are gathering accurate descriptive data about private health care (see Section 4) and maintaining the capacity for ongoing monitoring. Local regulations that apply to private health care vary considerably and need to be explored in detail. Finally, regulatory bodies that can enforce compliance across public and private sector institutions are often severely under-resourced, do not have basic capacity, and will need to be strengthened.244

Governing for quality also means recognising the importance of, and making space for, civil society in regulating the quality of care. Professional organisations that regulate their members have an important role to play in health system quality by promoting high-quality performance of their members and by sanctioning them when they fail to meet minimum standards. Self-regulation is underused in LMICs, where professional organisations mainly advocate for their membership. Experience in high-income health systems has shown that the privilege and responsibility of self-regulation promotes professionalism, the sense of accountability among professionals to people, and reduces transaction costs for governments. For example, in Canada,245 physicians successfully self-govern all aspects of the profession, from setting nationally uniform entrance exams to monitoring and remediating substandard clinical practice among practising physicians. However, self-regulation is not without its challenges, as exemplified by the UK,246 which has moved towards joint government–professional oversight because of a series of widely publicised physician scandals. When professional groups have primary fiduciary responsibility, care should be taken to involve both practising clinicians and citizens in governance and to avoid unnecessary fragmentation of regulatory responsibilities.247 Professional organisations can also promote quality through continuing medical education and engaging directly with governments to address quality concerns. For example, the Philippine Medical Association has more than a century of experience in agitating for improvements in medical education, health facility infrastructure, and the regulation of pharmaceuticals.248

Social participation in health care, especially for the most marginalised, has intrinsic value as a human right and instrumental value in improving health care and keeping systems accountable.249 People and communities are experts in their local experience and, with skilled support, can wield this knowledge to help create highly valued solutions to health-care problems.250,251 Social participation can also increase the uptake and sustainability of services.252,253 Although the composition of civil society varies by country, it is their diversity of perspectives, the opportunities for participation and action, and the availability of accurate and understandable information that will make this sector effective in holding governments accountable for high-quality health care.243,249,252,253 Civil society can be particularly powerful when adopting a human rights framework for advocacy.253 For example, in Uganda,254 the Center for Health, Human Rights, and Development regularly uses legal avenues to challenge policy makers on issues such as essential medicines, safe and respectful maternity care, and fair treatment of patients with disabilities.

Institutional accreditation uses external evaluators to assess facility performance against health-care standards. Although frequently cited as a quality accountability mechanism, a scoping review of reviews done by this Commission found that the direct effect of institutional accreditation on quality of care is uncertain (appendix 1). In a systematic review of improvement strategies, median effect sizes for institutional accreditation were modest: 7·1 percentage point improvements in quality outcomes were reported (appendix 1).255 However, accreditation can indirectly affect quality through improved management, professional development, and capacity of facilities to promote change.256

Improvement entails the continuous production of relevant data, which measures performance and outcomes, and the translation of those data into action—a learning system.226,257 This learning system facilitates the development of programmes and reforms based on the best available evidence (whether global, regional, or local data) and best practices. New initiatives should embed measurement, evaluation, and plans for how the results could be disseminated effectively to the people responsible for ongoing data use to inform adaptation of services. Learning systems should also identify best performers, as discussed in Section 2, and determine the basis for their success.

This set of intentional processes for actively learning and improving the health system is a goal that should be articulated and demonstrated first by the actions of senior leadership and subsequently echoed by middle management and the front-line staff. This system goal should become the primary guiding principle that creates the motivation for system improvement over time and for which health system actors hold themselves accountable.258 Planners should design better systems on the basis of lessons learned and then link back to system managers, supervisors, and front-line staff to support improvement. Developing well functioning learning systems is especially important because of the imperfect evidence base for quality improvement interventions and the large variation in effect sizes found between studies and contexts. Learning systems ensure that planners can make course corrections based on context-specific data. A meso-level strategy that illustrates this approach is the quality improvement collaborative, which we describe in the following subsection.

An analysis done by this Commission regarding five country experiences on governing for quality revealed practical lessons for operationalising the described principles (methods are described in appendix 1). District and facility-level health workers might be unaware of national quality policies and strategies or might not understand the implications of those on their daily work. The dissemination and translation of policies and strategies needs to be formally assigned, built into the job descriptions of public sector administrators, and included in performance reviews of these individuals. Additionally, the workforce might experience distracting and overwhelming policy crowding, with poorly coordinated and sometimes conflicting mandates. Countries are encouraged to review all policies affecting front-line workers; overlapping or conflicting policies can then be pruned, leaving a policy set that is coherent from the perspective of the service provider. For example, a nurse in primary care seeing a patient with diabetes and latent tuberculosis would benefit from having a single quality policy, not separate documents on diabetes and tuberculosis.

Informants from all levels of the health system discussed the challenges of good system-wide data use in the Commission analysis. Data generation and translation must start at the local level, but for system-wide improvements to occur, these data need to be coordinated centrally. We suggest the creation of planned spaces for information exchange, such as district-led meetings to learn from the evidence generated. Success stories of improvements made possible by accurate data collection and skilled data translation can be shared with front-line health workers to motivate continued quality care and improvement.

Universal action 2: Redesign service delivery to optimise quality

Most LMIC health systems were originally designed to provide basic episodic care, especially for infectious diseases. Many systems have not adapted to the changing landscape and challenges of caring for people with chronic diseases, mental health conditions, and more complex injuries and illnesses.20 Hospitals and healthcare facilities with advanced diagnostic and treatment capabilities are overcrowded with stable patients who could be treated in primary care facilities, whereas many first-level health clinics are expected to handle cases that are beyond their scope, with slow or non-functioning referral for emergencies.259,260 Poorly organised health systems lose lives, waste scarce resources, and squander the good will of populations.

To address this, this Commission calls for a quality-focused service delivery redesign: a reorganisation of services within the health system to efficiently maximise health outcomes and user confidence, rather than only geographic access to clinics. Service delivery redesign capitalises on existing health system assets to provide services at the appropriate level and achieve the highest quality of care possible.

First, some services should be shifted to primary care. Reflecting the core principles of continuity, coordination, comprehensiveness, and first contact, competent primary care is ideal for treatment of chronic and stable conditions that require sustained engagement with the health system (eg, non-communicable diseases and stable HIV or tuberculosis infection), preventive care (eg, immunisation, antenatal or routine child care, and growth monitoring), and low acuity and algorithmic services (eg, care of minor child and adult illnesses and injuries).20,36 Palliative care can also be expertly delivered close to home by primary care and in partnership with families, community caregivers, and spiritual supporters.36

Examples of the partial implementation of quality-focused service delivery in LMICs reveal the benefits of shifting these services to primary levels. In HIV care, stable patients are managed in primary care clinics with impressive results, and new patients can initiate treatment in their own communities.261 As a result, centralised specialty centres are less crowded, allowing higher-skilled providers to focus on more complicated cases, such as HIV treatment failures.260 A multicountry meta-analysis262 of 39 090 patients with HIV showed that patients in primary care were half as likely to be lost to follow-up than patients treated at a centralised HIV clinic. In tuberculosis care, community-based models are also substantially less costly to implement.263 Uncomplicated non-communicable diseases are especially well suited for care at the primary level, where providers can more effectively monitor chronic disease over time and build relationships that form the foundation for effective communication and counselling regarding crucial lifestyle modifications.20 An important caveat is that current primary care models in many LMICs are outdated and ill-suited for these new tasks. New thinking is needed on primary care functions, capacities, and connections with specialised services, especially in urban settings.20,264 For example, experience from high-income settings suggests that non-visit care, in the form of virtual or phone visits, has the potential to extend the reach of primary care for low-acuity conditions.265

Acute or chronic conditions with higher risk of mortality or severe morbidity are best assessed at a hospital with emergency capacity. The correct health system level for some surgeries should be determined on the basis of availability of specific technical skills, laboratory, imaging, and intensive care infrastructure, acuity of the condition and projected procedure volume. Complex or rare conditions are ideally managed in tertiary, highly specialised, care centres.

Childbirth is one situation that benefits from care at hospitals with surgical and specialised newborn care services, because complications can arise without warning and require rapid, highly skilled care.266 However, in low-income countries, a substantial proportion of obstetric and newborn care is provided in primary care facilities without adequate expertise or surgical capacity.267 For women and newborn babies who develop complications in primary care clinics, poorly functioning referral and transport to a higher level facility mean a much greater risk of morbidity and mortality.267,268 Guided by this logic, many high-income and middle-income countries mandate that all women deliver in, or next to, hospitals with surgical and advanced newborn care services.269 The structural deficits in highly skilled health workers and surgery at primary care levels might explain why the Better Birth trial,39,270 a large randomised controlled study, found that implementing a safe childbirth checklist and coaching for nurses and midwives at primary care centres in India did not reduce maternal and newborn morbidity or mortality.

We examined the practical implications of shifting delivery care to hospitals in a geographic modelling that linked facilities with pregnant women in six LMICs (Malawi, Haiti, Tanzania, Kenya, Namibia, and Nepal; methods are described in appendix 1). We found that delivery care redesign would result in substantial gains in technical quality for care of pregnant women without reducing interpersonal quality and with minimal reductions in 2 h access to care. For example, in Tanzania, hospitals score twice as high as primary care facilities on a basic measure of childbirth quality and, therefore, quality of care would improve by moving all deliveries to hospitals. Although this would increase the average distance from a delivery facility for rural dwellers, only 27% of pregnant women would live more than 2 h away from a delivery facility in Tanzania, compared with a current 17%. In the remaining countries, 1% to 7% of women lost 2 h access to care. This redesign can also produce efficiency gains because resources could be redirected from providing obstetric care in thousands of facilities to improving quality in fewer hospitals, promoting care integration across facilities, working with communities, and enabling transport to hospitals.

Strong interfacility communication and referral networks are crucial to the success of quality-focused redesign, along with investments and participation from non-health-care sectors. Tools to facilitate redesign that warrant consideration include improved transportation (eg, community taxi services and ambulances),271 communication (district-led learning, discussed in the following subsection), measures to reduce access barriers to high-quality facilities (eg, vouchers and maternity waiting homes),272,273 and public education to enhance population understanding of the right place for care.274 Local context, with a focus on facilitating access to high-quality care for the most marginalised subpopulations, should drive the mix of interventions and incentives.

Planning for quality-focused service delivery redesign in any country would require analyses of patient volumes, bed and surgical capacity, provider competence in existing hospital facilities, and potential upgrades to existing health-care centres to permit high-quality care, as well as attention to transport, costs, and building community demand.275

Universal action 3: Transform the health workforce

The data in Section 2 showed that providers often do less than half of recommended evidence-based care measures and that rates of diagnostic accuracy are low across health conditions and countries. A Commission analysis showed that this is also true of providers in their first 3 years of practice, suggesting a probable role of poor preservice education in provider performance (appendix 1).276 Low knowledge and competence of the health workforce is at risk of worsening over the coming years because of the rapid expansion of health workforce training institutions, resulting in dilution of already insufficient faculty and curricular resources.277,278 Despite this threat to health-care quality in LMICs, improving the education of health-care professionals has not been a central part of the improvement discourse.279 In the previously mentioned review of primary care quality improvement, only 16 of 379 articles addressed the preservice education of health professionals ( ). Fixing these gaps through in-service training is not an effective antidote,280 and reforms in professional education are required to adequately equip these professionals to provide high-quality care.

The Lancet commissions277,281 on health professionals for a new century and on the future of health in sub-Saharan Africa highlighted key steps to address the quality gap of the health-care workforce. First, the education of health professionals should focus on achieving competence through active learning, early clinical exposure, and problem-based learning. Competency should be defined by the gaps and needs of each individual country and include domains beyond the technical skills of providers. Ethical, respectful, and compassionate care, and the fundamentals of systems thinking and quality improvement should be additional core competencies. Dysfunctional systems will continue unless the workforce is prepared to improve them.

Second, the chronic understaffing of many health-care professional schools in LMICs must be addressed, along with support of high-quality teaching, for the quality of clinical education to improve.278 Possible solutions include increasing salaries, expanding professional development opportunities, using state policy levers to require practising clinicians to teach trainees, and providing small incentives, such as free housing or telecommunications.278 Finally, health education institutions should establish student recruitment and retention policies to increase the representativeness of the student population.252,282,283 Evidence has shown that care interactions between providers and patients who are racially, culturally, ethnically, or linguistically similar are associated with higher perceived quality of care, satisfaction, and improved medical communication.284,285 These changes within institutions of higher learning must be supported by good governance and quality-informed policy making. Intersectoral coordination between ministries of health and education would create a more direct link between the production of a health workforce and the needs of the health system.281

Third, health-care providers also need a work environment in which they can succeed beyond graduation. Many health-care providers face challenging conditions, including inadequate and delayed salaries, heavy workloads, ambiguous responsibilities, no opportunities for growth, and poor treatment by colleagues and patients.276,286,287 Not only do these conditions result in burnout, mental distress, and poor retention for providers, but they also result in poorer quality care.287–289 Motivated providers are less likely to make poor decisions or medical errors and are more likely to be empathic towards patients.290 Good working conditions, regular pay, clinical support, and opportunities to learn and grow are essential to maintain a workforce that is motivated and committed to providing high-quality care.286,291,292

WHO recommended a set293 of decent employment policies to support providers, including ensuring occupational health and safety, fair terms for workers, merit-based career development, and a positive practice environment. In addition to broader policies, a review294 published in 2017 recommended a set of steps for facilities to foster joy and engagement in their own workforce. These include an initial process of inquiry to understand workforce priorities, followed by identifying and removing the primary annoyances, initiating simple fixes, and using improvement science methods to spur larger-scale change to create a fundamentally more satisfying and happier work environment. Although early reports suggest that sense of purpose can be strengthened through these approaches, much of this work has started in the past few years and the effectiveness of these interventions on improving quality of care in LMICs remains to be determined.

Universal action 4: Ignite population demand for high-quality care

High-quality health systems respond to people’s expectations, but if those expectations have been dampened by a history of disempowerment and poor-quality care, that response will not translate into better health care.295 Section 2 shows that when expectations are low, quality ratings of objectively poor care are high. This discrepancy lets health systems disregard issues of quality. Beyond putting pressure on systems to improve, generating demand for quality through information sharing would increase health system accountability (see universal action 1) and has an ethical foundation: for patients to be autonomous decision makers, they must have access to usable information about the quality of their care.296 This is imperative because of the information and power asymmetry that exists between patients and providers. Finally, this Commission’s recommendation is based on evidence that people who already demand higher quality in LMICs and actively make decisions can extract higher quality care from their health systems.118,119,297,298 National quality improvement strategists are encouraged to explore demand-side approaches that raise people’s expectations of quality.

Very few improvement programmes are explicitly designed to raise demand for quality care. We used those few programmes to draw lessons on this understudied improvement opportunity. Participatory women’s groups are a well documented299 example, and improved outcomes for women and children in communities with these groups are believed to be partly due to participants demanding better care, such as safe hygienic practices during childbirth. Community monitoring programmes can generate demand for quality, although few high-quality studies exploring this outcome exist (see Section 3).300 A programme301,302 in rural Uganda, for example, combined information sharing about quality care at local facilities with community participation and found reductions in neonatal deaths and improvements in measures of facility process quality 4 years after implementation. A study303 in Uttar Pradesh, India, showed that quality during prenatal visits was improved by sharing information about health and social service entitlements with pregnant women. A preliminary body of qualitative research304 also suggested that demand generation for quality might be especially well suited to improving user experience. Panel 14 includes examples of the use of advocacy to generate demand for high-quality care from the White Ribbon Alliance.

These interventions are based on sharing information with people and treating them as active agents in the health system. They are unlikely to work without system-level support that encourages patient-centredness, power-sharing, communication, and inclusion.300 Importantly, this supporting of people to be active agents should be done with careful attention to marginalised populations. The intersection of multiple sources of vulnerability is likely to make some groups less able and prepared to act on quality information than others. To prevent the exacerbation of existing disparities, particular attention must be paid to rural, less educated, and impoverished populations (see Section 3).

Interventions that might raise expectations and demand for quality often include social interaction through groups, committees, or meetings; this component is supported by social network science and evidence showing that people learn about quality from each other.305,306 This insight from social network science also suggests that demand generation interventions might take advantage of the increasing presence of interactive social media platforms in LMICs. gives an example of a people-facing dashboard that can be used to share information with populations. More country examples of improvement through the four universal actions can be found in appendix 2.