Primary Health Care as a Foundation for Strengthening Health Systems in Low- and Middle-Income Countries

Primary health care (PHC) has been recognized as a core component of effective health systems since the early part of the twentieth century. However, despite notable progress, there remains a large gap between what individuals and communities need, and the quality and effectiveness of care delivered. The Primary Health Care Performance Initiative (PHCPI) was established by an international consortium to catalyze improvements in PHC delivery and outcomes in low- and middle-income countries through better measurement and sharing of effective models and practices. PHCPI has developed a framework to illustrate the relationship between key financing, workforce, and supply inputs, and core primary health care functions of first-contact accessibility, comprehensiveness, coordination, continuity, and person-centeredness. The framework provides guidance for more effective assessment of current strengths and gaps in PHC delivery through a core set of 25 key indicators (“Vital Signs”). Emerging best practices that foster high-performing PHC system development are being codified and shared around low- and high-income countries. These measurement and improvement approaches provide countries and implementers with tools to assess the current state of their PHC delivery system and to identify where cross-country learning can accelerate improvements in PHC quality and effectiveness.

Primary health care (PHC) has been recognized as a core component of effective health systems since the early part of the twentieth century. However, despite notable progress, there remains a large gap between what individuals and communities need, and the quality and effectiveness of care delivered. The Primary Health Care Performance Initiative (PHCPI) was established by an international consortium to catalyze improvements in PHC delivery and outcomes in low-and middle-income countries through better measurement and sharing of effective models and practices. PHCPI has developed a framework to illustrate the relationship between key financing, workforce, and supply inputs, and core primary health care functions of first-contact accessibility, comprehensiveness, coordination, continuity, and person-centeredness. The framework provides guidance for more effective assessment of current strengths and gaps in PHC delivery through a core set of 25 key indicators (BVital Signs^). Emerging best practices that foster high-performing PHC system development are being codified and shared around low-and high-income countries. These measurement and improvement approaches provide countries and implementers with tools to assess the current state of their PHC delivery system and to identify where crosscountry learning can accelerate improvements in PHC quality and effectiveness.

INTRODUCTION
In 1920, the British government commissioned a report to suggest ways to structure their expanding health system investments. The commission chairman, Lord Bertrand Dawson, borrowing from previous experience in education, proposed three hierarchical levels of care locations (primary, secondary, tertiary). He and the commission first identified Bprimary care^as the most basic level of a structured health system (akin to primary or elementary education), concerned with caring for simple, common problems in outpatient settings ( Fig. 1). 1 Since that time, a profound evolution has occurred toward understanding the central role of primary health care (PHC) in ensuring individual and population health, transforming PHC from responsibility for the lowestlevel basic tasks toward becoming the heart of an integrated, people-centered system of care. The development of community-oriented primary care in South Africa, India, and the US in the mid-twentieth century showed the early potential of strong PHC systems to produce promising population health results, but these vanguard programs were not met with uniform support for a common conceptual understanding or even shared definition of PHC. Competing notions varied from quite broad definitions of primary health care for all in the Alma-Ata Declaration of 1978, to the narrower definitions of selective primary care that followed in the 1980s, designed to address major causes of death in lower-resource settings at that time. 1,2 With converging forces including broader population demand for higher-quality services, rising health costs, and emergence of non-communicable diseases, health system leaders across the globe are increasingly recognizing the central importance of strengthening PHC to meet these evolving needs (Fig. 1). 3 This is particularly true in lowand middle-income countries (LMIC) in the midst of providing care for the Bdouble burden^of communicable and non-communicable disease, while transitioning from external donor assistance to sustainable domestic funding. However, a common understanding of how strong PHC systems are produced, and how to iteratively improve them, remains elusive. In this perspective, we introduce a framework designed to better understand if and how PHC is working in LMIC, and where lessons can be learned and spread globally.

TARGETING DISEASES, NOT PEOPLE
Seeking to capitalize on the influx of global health funding, in 2000, 189 countries ratified the Millennium Development Goals, initiating an unprecedented global effort to promote health and well-being across LMIC, largely through Bverticalp rograms focused on specific disease or care delivery areas. Since the conclusion of the Millennium Development Goals in 2015, it has become clear that this effort led to tremendous gains in health: over 15 years, deaths in children under 5 years of age declined by more than 50 %, maternal mortality dropped by 45 %, new HIV infections fell by 40 %, and over 6.2 million malaria-related deaths were prevented. 4 Despite this remarkable progress, unacceptable gaps in health outcomes and patient needs remain. Every day there are still 16,000 deaths among children under 5, and 830 women die from preventable causes related to pregnancy and childbirth. 4,5 The Millennium Development Goals' vertical approach toward healthcare has, in many countries, created fragmented, inefficient, often parallel health systems focused on treating specific diseases rather than promoting holistic health and well-being.

GOING BEYOND VERTICAL PROGRAMS, AND PRIORITIZING USER-CENTERED APPROACHES
To build upon the successes of the Millennium Development Goals, in 2015, 193 countries adopted the Sustainable Development Goals, to define the next iteration of worldwide development targets for the year 2030. 6 The health-related Sustainable Development Goal is to achieve universal health coverage, defined as Bfinancial risk protection, access to quality essential health-care services and access to safe, effective, quality and affordable essential medicines and vaccines for all.^4 Strong PHC is the foundation of efficient, equitable, and resilient health systems, and can address the majority of care needs for the majority of people, regardless of where they live. 7,8 Achieving quality universal health coverage in a sustainable way will require moving beyond vertical programming, toward integrated health systems, in large part by prioritizing primary health care. [7][8][9] HOW BIG IS THE GAP?
Unfortunately, in many LMIC, PHC capacity is lacking, and health outcomes are poor. Across the world, these gaps are being exposed and exacerbated by the increasing burden of non-communicable diseases and concomitant increases in care complexity. 10 Acute threats also put a spotlight on weak PHC systems, as was apparent during the 2014 Ebola outbreak in West Africa. The crisis highlighted not only the severe shortage of healthcare infrastructure, human resources, and essential supplies in affected countries, but also poor access to and low quality of care, and a consequent lack of trust and connection between systems and the populations they serve. 10 Even in the absence of acute threats, recent illustrative studies on the quality of routine PHC in LMIC show that it is far from adequate, with wide variation between and within regions. 11 A large study in northern India found that providers on average spent 3.6 min with a patient, completing only onethird of recommended history and physical exams. Diagnoses were provided in only 36 % of cases, and only 12 % of these were correct. Harmful or unnecessary treatments were more common than correct ones (42 % vs. 30 %). 12 Similarly poor results were found in Tanzania, Paraguay, and Indonesia. 11 In China, clinicians completed only 18 % of items on a recommended checklist, provided a correct diagnosis in just 26 % of patients, and dispensed unnecessary or harmful medications in nearly 70 % of interactions. 13 In Afghanistan, direct observations revealed gaps not just in technical quality, but also communication and time spent with patients. 14   activities are derived from a Conceptual Framework for PHC in LMIC (Fig. 2). This framework describes the components necessary for building strong PHC systems, informs better assessment, and identifies modifiable gaps in performance.
A CONCEPTUAL FRAMEWORK FOR GLOBAL PHC: The PHCPI framework was developed through an extensive literature review, global stakeholder consultations with leading experts, advocates, and policymakers, and a review of over 40 existing health systems frameworks. 15 This process revealed that most frameworks used in LMIC prioritize health systems inputs such as funds, human resources, supplies, facilities, and information systems, with less attention to the interactions between providers, families, communities, and patients and the quality of services they provide and receive. Building off of widely known frameworks, [16][17][18][19] the PHCPI framework offers a stronger emphasis on people-and community-centered care, supply and demand functions, and integrated service delivery through effective organization and management. The PHCPI framework focuses not just on the traditional system inputs needed to achieve good health outcomes, but also on identifying and mapping the key functions of service delivery that underlie outcome achievement.
The heart of the framework is the Service Delivery domain that captures the interaction of systems and supplies with providers and patients at the moment of care delivery. The first two sub-domains describe aspects of population health management such as local priority setting, community engagement, empanelment, and proactive outreach, as well as teambased organization, facility management capability, information system use, and performance measurement. Together, these sub-domains create the pre-conditions for whether patients have meaningful financial and geographic access to timely care (sub-domain 3) and whether competent and motivated providers are present to deliver safe, respectful care (sub-domain 4). The fifth sub-domain-high-quality PHC-includes the classic primary health care functions described by Barbara Starfield and others, including first-contact accessibility, comprehensiveness, coordination, and continuity. 20 In addition, the person-centeredness function captures important and often lacking interpersonal and relational dimensions of primary health care, including trust, respect, communication, and improved patient self-management and activation.
The successful combination of systems, inputs, and service delivery contribute to PHC Outputs. Importantly, this domain is focused not just on crude coverage of services, but on effective coverage of high-quality services focused on both prevention and treatment. The final domain, Outcomes, is influenced by all preceding domains. Drawing from the Universal Health Coverage Monitoring Framework 21 and the Global Reference List of 100 Core Health Indicators, 22 the domain is focused on morbidity and mortality, as well as key outcomes of people-centered care such as responsiveness to Figure 2 The Primary Health Care Performance Initiative Conceptual Framework. Key terms: Surveillance: the capacity of health systems to identify new and emerging health-related threats; population health: Bthe health outcomes of a group of individuals, including the distribution of such outcomes within the group^2 7 ; empanelment: the process by which all patients in a given facility and/or geographic area are assigned to a primary care provider or care team; community engagement: Binvolving communities in decision-making and in the planning, design, governance and delivery of health care services^2 8 ; resilience: Bthe capacity of health actors, institutions, and populations to prepare for and effectively respond to crises; maintain core functions when a crisis hits; and, informed by lessons learned during the crisis, reorganise if conditions require it.^2 9 people, equity, efficiency, and resilient health systems. The overall framework explicitly recognizes that PHC systems lie within wider political, cultural, demographic, and socioeconomic contexts that play a critical role in determining their success.

HOW DO WE ASSESS PHC CARE DELIVERY SYSTEMS?
After developing the conceptual framework, PHCPI partners identified a set of 25 key performance indicators-BVital Signs^-from globally available data that could be used at the national level to assess the performance of PHC systems, identify gaps in need of improvement, and allow countries to benchmark themselves across their region, income peer group, or other relevant comparisons (Table 1). For example, the Vital Signs measure continuity of care in three vertical areas: antenatal care, childhood vaccination, and TB treatment. These indicators can be compared within and across countries to understand common challenges and positive outliers, as shown in Figure 3. This type of analysis shows both that continuity of care varies widely across countries, and that these three measures of continuity of care can vary widely within countries. The latter insight highlights the importance of strong PHC systems in providing a foundation for vertical programs to build upon in order to provide continuity for all users of the healthcare system, not just some.
The PHCPI Vital Signs indicate that much work is needed to strengthen PHC measurement across LMIC. Data availability is a clear problem: of 135 LMIC, only one-Uganda-has data available for all 25 Vital Signs. Furthermore, many subdomains of the PHCPI Conceptual Framework, particularly within the Service Delivery domain, are not represented by any Vital Signs, reflecting a lack of accepted, reliable, and globally comparable indicators for measuring the core functions of a primary health care system. Consistent with other work on PHC delivery in LMIC, 10,12,23 the Vital Signs also demonstrate that performance across most domains is weak in many LMIC, suggesting ample targets for improvement in order to meet the Sustainable Development Goals.

NEXT STEPS
At its core, the PHCPI framework is an explication of the relationship between common global health areas of focus-such as accessibility and availability of services-and Starfield's classic primary health care functions. 20 An emerging body of work suggests that these functions resonate in both low-and high-income country settings. 24 Focusing health system development on the production of these functions in LMIC settings (as opposed to simply meeting input targets) is critical for accelerating improvement in primary health care and moving from vertical programming toward integrated health systems. This lens of primary care functions provides the opportunity for shared learning between health systems in various stages of development. For example, high-performing primary care systems in high-income countries feature core components of effective management and team-based models of delivery such as proactive population management and risk-stratified care management. 25 These features can be targeted in countries with transitioning burdens of disease that require a strong focus on chronic disease management and patient/ community engagement. Similarly, high-income countries can learn from successful examples of LMIC PHC models that feature deep community participation and engagement using community health worker models to reduce burdens of disease in an efficient manner. 26 For example, Brazil's Family Health teams and Costa Rica's EBAIS model highlight the utility of using multidisciplinary teams, including community health workers, to improve health outcomes equitably. In Estonia and Cuba, a focus on increasing access to care through a better trained and widely distributed PHC workforce ensures that primary health care is the first point of contact, while Turkey and Thailand highlight the benefits of reformed financing mechanisms to ensure consistent universal financial coverage for PHC. PHCPI is profiling innovative and successful strategies such as these to build a repository of BPromising Practices^for improving PHC performance. 3 These bright spots of PHC provide useful lessons to countries as they embark on the path toward achieving universal health coverage. PHCPI is also working with a set of interested countries to catalyze PHC development efforts by making data on existing performance more available and useful at local levels, and linking it to a set of proven improvement pathways.
As all countries increasingly grapple with the challenges of providing high-quality, people-centered services at an affordable cost, primary health care is becoming a shared common focus. We present a framework for defined, detailed, and measurable primary health care performance that can be used Figure 3 Data from PHCPI Vital Signs Indicators showing wide variety within and between countries in delivering continuity of care for children (immunization), women (antenatal care), and those with infectious diseases (tuberculosis). The height of the bars indicates a country's performance on each indicator relative to all other low-and middle-income countries. The diphtheria-tetanus-pertussis (DTP) dropout rate is the percentage of children who do not receive three doses of DTP after receiving an initial dose. The antenatal care (ANC) dropout rate reflects the difference in the percentage of women who do not receive four ANC visits after receiving an initial visit. The tuberculosis (TB) treatment success rate is the percentage of all TB cases with successful completion of treatment. Data available at http://www.phcperformanceinitiative. org/tools/compare-tool#?ind=&loc=.
by stakeholders at the global, national, and local levels to drive iterative improvement. Though the contexts within which they are produced vary widely, we propose that core functions of primary health care are universal across the world. The extent to which a health system produces services that are accessible at the point of first contact, continuous, coordinated, comprehensive, and people-centered is directly linked to the ability to produce intended outcomes. Successful progress requires sustained leadership focus and investment in measuring these key PHC domains, and making this data transparent and relevant particularly at the front lines of service delivery. Countries have an opportunity and a responsibility to go beyond antiquated notions of Blevels^of care, simplified packages of care, or overly broad but under-defined concepts of primary health care, to move toward comprehensive, coordinated care that puts people and their needs at the center.