## Main

On 14 May 1796, 73 years before the first issue of Nature, and inspired by Lady Montagu’s “variolation” concept, Edward Jenner inoculated eight-year-old James Phipps with cowpox pus to prove that the less virulent cowpox would protect against smallpox. This experiment was a game changer in medicine and health. For the first time, it was possible to medically prevent infection in a healthy person. Although vaccines have been widely introduced in high-income countries since the late 1950s, it took 180 years after Jenner before the Expanded Programme on Immunization (EPI) was launched in 1974, promoting access to six essential vaccines in all countries worldwide. Today, vaccines against 26 infectious diseases are internationally available according to the World Health Organization (WHO)1, although more have been licensed worldwide, changing the future of the human species. Others are in experimental public health use, such as Ebola vaccines, or pilot implementation such as the RTS,S malaria vaccine, and about 240 vaccine candidates are in development2 (Table 1). The US Centers for Disease Control and Prevention declared vaccination the number one success story for public health in the twentieth century3.

However, progress in vaccine coverage remains highly uneven—both between and within countries—which threatens hard-won progress and raises uncertainty about how to make further advances. Vaccine-preventable diseases such as measles are on the rise, and episodes of vaccine reluctance and refusal are occurring globally, questioning one of the most transformative interventions for survival and health.

This Review focuses on preventive immunization in humans and its impact (rather than on the vaccines themselves), including in low-, middle- and high-income countries. We discuss the current status of vaccine coverage, as well as unmet needs, four hurdles to overcome to ensure sustainable immunization programmes starting with the discovery of a new vaccine, the growing issue of vaccine confidence, and conclude with several opportunities and needed actions to ensure the full potential of immunization for human health and society. Developmental challenges for vaccine production for low- and middle-income countries, which were recently discussed in separate articles4,5, and therapeutic vaccines are not discussed.

Vaccines are biological products that induce protective immunity against infection and disease; they consist of sub-components, killed or inactivated organisms or live-attenuated viruses that train the immune system for a future response to a natural infection. They are probably the only medical intervention that is recommended for every single individual on the planet. Unlike therapeutics, vaccines are used in healthy people, and demand a very high standard of safety and require continuous monitoring for potential side effects. Besides considerations of safety, effectiveness, impact and cost, this raises complex governance, regulatory and public trust issues. All countries have a national immunization plan, often with goals inspired by the Global Vaccine Action Plan (GVAP) framework for 2011–20206.

## How immunization has crucially benefited society

It is hard to imagine a world without vaccines. A decade ago, the WHO, UNICEF and the World Bank estimated that routine childhood immunization programmes were preventing more than 2.5 million deaths every year7. With the increase in vaccine coverage, the growth of populations, and the introduction of new life-saving vaccines, immunization is ever more important for survival. In addition to preventing deaths, vaccines prevent disease and disability, including in adults and the elderly. In a high-income country such as the United States, for a single birth cohort, vaccines prevent nearly 20 million cases of disease, and more than 40,000 deaths8.

A vaccine has for the first time in history eradicated a human disease, smallpox. Efforts to eradicate polio are in the final stages, with only two countries, Afghanistan and Pakistan, still experiencing wild transmission of the polio virus. All countries with the exception of 13 have eliminated neonatal and maternal tetanus. Without vaccination, there would be far more infections that require antibiotic therapy, exacerbating the major problem of drug-resistant infections.

Between 1990 and 2017, immunization contributed to a 55% global decline in under-five mortality rates, with a drop from 87 to 39 deaths per 1,000 live births9. More than 14 million deaths are estimated to have been prevented by measles vaccination alone between 2011 and 20206.

Vaccination benefits not only those who are vaccinated, but also others in their family and community. This population-wide benefit, known as ‘herd immunity’, reduces the exposure of unvaccinated individuals to pathogens through a reduction or interruption of the chains of transmission. A recent study in Kenya showed that the introduction of a pneumococcal vaccine resulted in not only a major reduction in invasive pneumococcal disease, but also a nearly 100% decline in incidence among infants too young to be vaccinated, and a more than 74% reduction among unvaccinated children10. Community or herd immunity is an important consideration when estimating the full public health value of immunization. The threshold to achieve such community protection can be as high as 95% for measles, but as low as 80% for rubella, and 60% in high-income settings for the effect to begin for pneumococcal vaccination, which means that the programme strength required to derive additional impact varies substantially by vaccine11,12,13. These differences in the required critical vaccination coverage rates are due to the basic reproductive ratio of an infection (R0)14, which can vary greatly among various infectious diseases. The R0 of a specific infection indicates the average number of cases one case generates in a population—in the case of measles it is 12–18, which is among the highest15. It is an indicator of how contagious an infection is, and determines the minimum level of vaccination coverage needed to generate herd immunity.

Potential long-term effects beyond direct protection against a specific pathogen or disease have been attributed to several vaccines, in particular the BCG vaccine against tuberculosis and the measles vaccine, in which observational studies suggested a survival advantage compared with children who had remained unvaccinated. These non-specific effects (also known as heterologous effects) would add to the disease-specific, proven benefits of vaccines, and have been attributed to epigenetic changes in innate immune cells as opposed to the adaptive immunity induced by the antigen-specific responses to the vaccine16,17. However, the importance of heterologous effects remains controversial, and plausible immunological findings still need to be validated in large-scale clinical trials.

The benefits of vaccines in general go beyond health, and include economic, educational, health security and other benefits18. Their full economic value is not sufficiently quantified in assessments of cost-benefit, or in investment terms, and is an increasing area of inquiry and empiric measurement19.

Vaccination is a sound investment. Thus, the return on investment from childhood immunization in low- and middle-income countries is high. For every US$1 invested in immunization against ten diseases,$16–$18 are saved in healthcare costs, and the net return is as high as$44 per dollar spent when the broad economic benefits are considered, although the return on the investment varies by individual vaccine20. This is compared with the cost per DTPcv3-vaccinated child of $27 (having received all three doses of diphtheria-tetanus-pertussis (DTP)-containing vaccine)21. In the United States, the net economic benefits of vaccination in one birth cohort amount to almost$69 million22.

Modelling and observational data suggest that in low- and middle-income countries, vaccination contributes to the alleviation of, and protection against, poverty. Financial risk protection provided by the benefits of vaccination are accrued by the poorest households by the reduction of catastrophic and impoverishing health expenditures23,24. There is also evidence that vaccination improves childhood physical development, educational outcomes, and equity in distribution of health gains25. Finally, without vaccines, absenteeism from school and work would be much higher, and periodic epidemics would disrupt society. The economic effects of periodic influenza epidemics, for example, are enormous26,27,28, and can be reduced by immunization29.

## Vaccination is a lifetime investment

In addition to being the backbone of maternal and child health, vaccines provide important health benefits for all stages in life (Table 2). Given adaptations of the immune system throughout life, not all vaccines work equally well at all stages of life or in all geographical regions30,31.

Starting in infancy, the presence of maternal antibodies in the newborn can impede the response to vaccines, as the neonatal immune system undergoes its own journey of ontogeny, which enables it to adapt from the ‘sterile’ in utero environment to the confrontation with colonizing and potentially pathogenic microorganisms32. Particular immunological pathways have been identified33.

Despite considerable progress in reducing the rates of under-five mortality, important gaps remain in addressing neonatal morbidity and mortality. Neonates are particularly vulnerable to infection with Gram-negative bacteria and group B streptococcus, for which no neonatal vaccines currently exist33,34. The gap in early protection can potentially be bridged by administering vaccines to women in pregnancy, relying on passively transferred antibodies to protect infants in the first few months of life, until vaccinations administered in infancy or later can provide protection. On the basis of this principle, tetanus, influenza and pertussis vaccinations are recommended for pregnant women to prevent neonatal infections such as neonatal tetanus35. This maternal immunization strategy may be expanded with promising vaccines against group B streptococcus and respiratory syncytial virus36.

For adolescents, life-saving vaccines against human papilloma virus (HPV; the cause of cervical, anal, penile and head and neck cancers) are being increasingly introduced and need to be administered before the likely acquisition of HPV via sexual contacts. Vaccines against meningococcal meningitis—a potentially lethal infection with a second peak in adolescence—have also been introduced into this age group in some countries. New platforms such as schools had to be engaged to administer these vaccines.

Outbreaks of mumps have very occasionally been seen in teenagers, despite a solid vaccination record. This highlights the need for surveillance of all age groups for disease outbreaks, and could be due to waning of protection induced by vaccines that are otherwise regarded as highly efficacious37,38,39.

Booster vaccines against diphtheria, tetanus and polio are required to guarantee long-lasting protection and are required throughout adulthood to maintain protective immunity levels—although recommendations may vary by country.

A life-course approach to vaccination has become ever more pressing with pneumonia, influenza and shingles differentially affecting older adults, and death rates from pneumonia and influenza 130 times higher for adults over 85 than for younger adults40. Vaccination of the elderly with existing vaccines could prevent up to 90,000 deaths per year in the United States alone41.

Adult immunization does not have a clear prioritization in low- and middle-income countries, and is a complex programme across high-income countries. It is different from paediatric immunization, which has a global programme and focused, substantial funding. As the demographics are shifting across the world to an older distribution, a focus on adult immunization will become increasingly relevant, as advocated by the World Coalition on Adult Vaccination42. Despite national recommendations43,44, vaccine coverage among adults in high-income countries is uneven45 (vaccine coverage for herpes zoster, which causes shingles, among adults aged 60 or over in the United States was 24% compared with 65% for influenza among those aged 65 or over), and very low or not even available in most low- and middle-income countries46. Yet, several studies have shown good cost-effectiveness of adult vaccinations against influenza, pneumococcal infection, shingles, HPV and tetanus-diphtheria-pertussis47.

Important gaps also exist in our understanding of the fundamental biology of adult immunization. Owing to ‘immunosenescence’—the gradual decline of the immune system associated with ageing—vaccination of older adults is in general not as effective as in younger people, but the reasons for poorer responsiveness are not well defined, and require a new effort in terms of strategies and products for immunization of adults. However, it is likely that several compartments of the immune system are affected48.

There are three areas in which alterations to increase vaccine efficacy in the elderly could be considered: (i) increased vaccine potency; (ii) the use of adjuvants to enhance immunity; and (iii) application of immune modulators or other interventions to alter host immunity generally.

As populations age across the world, it will be increasingly important to identify how to integrate immunization programmes in health and care services to reach all age groups.

In addition, vaccinations are needed for travel, particular professions or specific health conditions49,50,51, and international travel has had a role in the resurgence of measles in areas such as the United States52.

## From discovery to impact: four hurdles to overcome

There are still major infectious diseases that required an effective vaccine for control and ultimate elimination, such as HIV infection and tuberculosis. Therefore, the continuing development of new vaccines is a public health imperative. Unfortunately, most early vaccine candidates in the discovery phase never make it as a safe and effective product. Development and deployment of vaccines is a long and complex process. We briefly describe here four hurdles that need to be overcome from the discovery phase of a new vaccine to sustainable population impact (Table 3).

The first hurdle is a ‘valley of death’ from discovery to early clinical development, when a potential antigen, adjuvant or new vaccine formulation developed in the laboratory is further tested for clinical proof-of-concept and safety in humans, in addition to optimizing production elements. Real progress has been made in recent years owing to several public and private initiatives that are helping partly to overcome this first major challenge, such as the Coalition for Epidemic Preparedness Innovation (CEPI)53, which was created after the 2014–2015 Ebola epidemic in West Africa to accelerate the development of vaccines against epidemic pathogens2,4,54.

The second hurdle in vaccine development, also referred to as the ‘second valley of death’, relates to the shift from early clinical development to the large and very expensive efficacy trials most often needed4, unless a previous similar vaccine is already developed and a new product can be licensed using an established correlate of immunity or protection. This is also the most expensive phase of vaccine development, absorbing more than two-thirds of the total costs of development of a new vaccine, including the building of special manufacturing facilities and conducting phase 3 trials in several countries, ideally with independent research partners. Often, this major financial effort is beyond the means of smaller biotech companies, and in general only big pharmaceutical companies and large foundations or public institutions have the financial bandwidth to support such trials that can cost as much as hundreds of millions of dollars. For vaccine candidates without a prospect of a high-income market to ensure a return on investment, and when the potential market for the new vaccine is limited to low- and middle-income countries, there is an almost unsurmountable valley of death unless philanthropic and public funding intervene2.

The needs and unique challenges of vaccines against epidemic pathogens demand innovation in product development pathways. The Merck recombinant vesicular stomatitis virus–Zaire Ebola virus (rVSV–ZEBOV) vaccine was deployed on a large scale during the recent Ebola outbreak in eastern Democratic Republic of the Congo before the product was licensed—even for indications for which no efficacy data were available such as primary prevention in healthcare workers. A second experimental vaccine, Ad26.ZEBOV/MVA-BN-Filo, is now also deployed for the same outbreak and in Rwanda55. Well-informed country leadership and transparent governance of such use are crucial, as is genuine community involvement. The ‘animal efficacy rule’ that applies when efficacy trials in humans are not feasible or ethical56 should also be considered for vaccines against epidemic pathogens. The development of Ebola vaccines has shown how this type of ‘learning by doing’ model can offer early access in humanitarian situations55,57, although it should be stressed that nearly five years after the first Ebola vaccine clinical trials in West Africa, no Ebola vaccine is licensed despite well-documented immunogenicity, safety, and human and/or non-human primate efficacy data. When a crisis such as Ebola is no longer the headline news, the sense of urgency is lost, and regulators and normative committees go back to often extraordinarily long processes.

After a successful phase 3 trial, there is a complex path to the licensing of any new vaccine, which requires reproducibility and safety tests of several batches of vaccines, while manufacturing facilities are finalized. Many countries still request clinical trial data conducted locally, delaying country licensing and implementation considerably, while further raising the costs of development. In Europe, there is advanced harmonization in the regulatory approval of vaccines through the European Medicines Agency (EMA), and in sub-Saharan Africa, the Africa Vaccine Regulatory Forum (AVAREF) is aiming to strengthen regulatory capacity for clinical trials and harmonization of regulatory practices58.

Following all of these activities, which can take as long as ten years or more, a new vaccine is now ready for deployment, but a third hurdle can occur between the licensing of a vaccine and broad-scale implementation, which is dependent on both a policy recommendation and the ability to implement. Many years can go by before important new vaccines reach communities in need, the cost of which is measured in human lives that could have been saved as well as money for their development.

There are many contributors to this third hurdle: first is cost, which is especially relevant for countries that are neither wealthy enough to procure vaccines at high cost nor poor enough to receive funding assistance from Gavi, the Vaccine Alliance. However, when a Gavi-eligible country transitions out of the programme owing to an increase in its gross national income per capita, it needs to increasingly mobilize domestic resources or other development assistance59. Even when the broader value proposition of a new vaccine is substantial, there remains the question of affordability. Second is the question of country capacity to take on new vaccines; the past decade has been a remarkable era for vaccine introduction, with 113 countries having introduced at least one new vaccine, which represents a real success story60. Country capacity to introduce and sustain ever growing programmes involves human and financial resources, and time to build political support and community demand. Both the pneumococcal conjugate and the rotavirus vaccines now have coverage in low-income Gavi countries that meets or exceeds the global average; however, this reflects the fact that not all countries in any income strata have yet introduced these vaccines in spite of their availability61. Even high-income countries can experience delays. Thus, in the United Kingdom, a meningococcal B vaccine was licensed in January 2013, recommended for introduction in March 2014, and finally announced for introduction in May 2015. It then took more than 12 months to resolve procurement discussions to enable implementation62.

For products that address priority diseases for low-income countries, the uncertainty of the market may risk products collapsing unless a full end-to-end product solution is articulated, with non-commercial support. Inclusion of the new vaccine in the WHO’s pre-qualification list is a requirement for procurement through funders such as UNICEF and Gavi. Some of these are vaccines against parasitic diseases, which are much more complex than bacterial or viral vaccines owing to the wide range of antigens with often a complex life cycle that exhibit different antigens relevant for vaccine protection. Thus, the RTS,S vaccine—the first ever malaria vaccine used in a routine immunization system63—took nearly 30 years since its creation by GlaxoSmithKline in 198764 before the EMA issued a positive scientific opinion in 2015, and the WHO recommended large-scale pilot programmes in 2016. These programmes took another three years to start in several African countries, and demonstrate the sometimes incredibly long development, licensing, and introduction times. The RTS,S malaria vaccine is also an example of a vaccine for which the clinical trial performance of partial protection led to a policy decision to advance in a step-wise manner rather than full programmatic deployment. This may become a more common pathway for future products, in part because these vaccines have performance and implementation characteristics that are more complex than those of current vaccines.

We are entering an era in which the path from vaccine licensing to routine implementation requires more than safety and efficacy data. Policy recommendations for new vaccines may only be realized after implementation research to determine how to ensure use and impact most effectively. Deliberations about cost effectiveness, the full value of vaccine assessments, and country priorities in the face of constrained resources remain drivers for delays associated with the third hurdle. National Immunization Technical Advisory Groups (NITAGS) will be increasingly important to guide evidence-based decision making.

Even after the lengthy and costly trajectory to introduce a new vaccine, ensuring sustainable impact faces a fourth set of hurdles that need to be overcome. These include supply and demand sustainability, and resilience and acceptance of immunization. Logistical issues such as the in-country ‘cold chain’ system of transporting and storing vaccines at recommended temperatures, procurement management, and the organization of vaccination clinics in remote areas, vaccine hesitancy, and equity of access can all present challenges. In addition, the misuse of vaccination campaigns as political tools has seriously damaged vaccine confidence in areas such as the Philippines, Nigeria, Afghanistan, Italy and Pakistan65. Some side effects or limitations of duration of protection may only become obvious after larger scale use, such as for live oral rotavirus vaccination in high-mortality settings66, pertussis vaccine67 and others68. A recent example is the results from a retrospective analysis of long-term efficacy trials that show that although there is a clear overall population benefit of the Dengvaxia vaccine against dengue, the vaccine also caused an excessive risk of severe dengue in seronegative vaccinees (that is, those not exposed to dengue virus69). In the Philippines, this new risk was reported after more than 800,000 school children were vaccinated, prompting a marked reaction by the public in 201870.

Stock-out events and vaccine manufacturing capacity have been problematic for particular vaccines, even in high-income countries. Manufacturers emphasize the time needed to build and commission a factory71. Although manufacturers in middle-income country are now supplying most low-cost vaccines globally, they face low profit margins, ferocious tenders, and often unpredictable procurement schemes. More efficient and modular production technologies may enable decentralized production with lower capital costs.

Each of the four hurdles can be overcome, although the fourth one should be a continuing concern for every national immunization programme. Depending on the phase, they may require different sets of policy actors, and are sometimes a matter of policy, management and leadership, rather than money.

Throughout the development and use of vaccines, vaccine safety is an overriding concern, and requires a continuous and careful scientific and societal assessment. Safety monitoring during manufacturing typically occupies a major part of the process and costs of a vaccine, and is a key element of any vaccine programme. In specific high-income populations, such as in the elderly, personalized medicine approaches have been proposed to maximize both immunogenicity and safety in the presence of chronic conditions and changes related to older age, but large-scale applicability is still questionable at present72,73,74.

## Persistent unmet needs for vaccination

The extraordinary achievement of vaccines is reflected in countries having vaccinated more than 116 million infants in 2018 alone75—which represents the largest number ever—and a comparable number of infants were also estimated to have been vaccinated in 2017. The global and regional coverage of diphtheria–tetanus–pertussis (DPT3) vaccination between 1980 and 2018 in Fig. 1 shows overall high coverage with regional variations, but also some stagnation in coverage over the past 10 years76. Despite the high coverage, there still remained 19.4 million under-vaccinated or unvaccinated children, who were vulnerable to diseases that they could and should have been protected from. Substantial improvements in coverage have been achieved in some countries, whereas coverage is regressing in others, often because of social disruption, conflict, or political upheaval, which highlights the extremely dynamic nature of vaccine programme performance.

Around 60% of all children who did not receive basic immunization in 2018 live in ten countries: Angola, Brazil, the Democratic Republic of the Congo, Ethiopia, India, Indonesia, Nigeria, Pakistan, the Philippines and Vietnam77. To achieve rapid change in this situation requires the full commitment of governments, supported by international organizations. The Gavi Alliance provides funding for vaccination programmes in low- and low- to middle-income countries, and has had substantial impact. The technical support provided by Gavi partners will be essential to address persistent gaps in vaccine coverage. Consistently delivering vaccines with high coverage, reaching at least the minimum coverage required to achieve herd immunity in line with the basic reproductive ratio of an infection as mentioned above, remains a struggle in many other countries including in middle- and high-income settings, with poor children not being reached78,79. For example, in 2017 in the United States, 100,000 children under the age of two (1.3% of the population of that age) were not immunized against DPT and MMR (measles, mumps and rubella), which represents a fourfold increase since 200179,80.

Of particular concern are countries in which vaccination coverage has declined. There are 19 countries that had more than 80% coverage for first-dose measles at some point between 2011 and 2017, but with coverage in 2018 at least 10% lower than their peak coverage. The measles vaccine coverage of those 19 regressing countries now ranges from 38% to 88%, with 10 countries with well below 80% coverage61. Some of the regression on vaccine coverage may represent improvements in data rather than actual slippage in coverage. The data systems to monitor both the number of children born and the number of children vaccinated accurately are highly variable in quality81,82. In some settings, management and reward systems probably incentivize inaccurate reporting of coverage data to meet targets, rather than incentivizing accurate reporting.

Outbreaks of measles, diphtheria and yellow fever are the result of what happens when the world is complacent and immunization coverage declines. Diphtheria outbreaks surged in Russia in the early 1990s; outbreaks of meningitis occurred among Rohingya refugees from Myanmar in refugee camps in 2017; and the transmission of polio persists in parts of Afghanistan and Pakistan83. Measles outbreaks are occurring in all regions of the world. The recent 80-fold increase in reported measles cases in the WHO European Region over four years to more than 82,000 cases in 2018 with 72 deaths84,85 is a result of a mixture of vaccine refusals, cultural beliefs, and access issues that include interruptions in vaccine supply, such as in Ukraine86, and have led to a WHO declaration of a grade 2 health emergency87. In the Americas, thousands of cases have been reported in Venezuela owing to the political and economic crisis, with cases also appearing in Brazil, Colombia and Ecuador, and four countries in the WHO European Region (United Kingdom, Albania, Greece and the Czech Republic) have now lost their measles elimination status. The United States is also at risk of losing their measles elimination status.

These outbreaks reflect failures to achieve and maintain high vaccination coverage, community by community. Low vaccination coverage and high heterogeneity in coverage are most deeply seen among African countries where routine rates of immunization in many countries are well below the GVAP targets88.

Since 2010, routine immunization levels have either stagnated or decreased in 54 out of 85 middle-income countries, who do not qualify for support from the Gavi Alliance78. Vaccine expenditures per child are often lower in middle-income countries than in low-income Gavi countries. The issue may not be solely due to a lack of funding capability, but may also arise owing to a lack of prioritization of immunization, countries not participating in pooled procurement mechanisms such as via UNICEF, low volumes of vaccines, insufficient efforts to reach vulnerable populations, vaccine choices, and duplicative local regulatory requirements that delay the introduction of new vaccines.

Another unmet need concerns the introduction of new vaccines. Rapid progress has been made to scale up the introduction of vaccines through Gavi investments in low-income countries, but not all vaccines have progressed at the same rapid pace. The adolescent HPV vaccine has been particularly slow to be introduced outside of high-income settings because of programmatic challenges, public-access issues, supply constraints and pricing issues.

Addressing these unmet needs will require persistent implementation of strategies that have been shown to be effective—such as detailed microplanning of local efforts to assure all children are identified and immunized—and special campaigns and approaches such as drone delivery of vaccines in areas that are harder to reach89. Systematic evaluation and implementation research should be part of these efforts to develop a firm evidence base for overcoming such programmatic challenges. The WHO has elaborated guidance on implementing high impact immunization programmes (Global Routine Immunization Strategies and Practices, GRISP) to address these unmet needs. Middle-income countries that do not benefit from funding from the Gavi Alliance need procurement mechanisms that can secure more predictable tiered pricing. No set of strategies, however, will succeed without substantially enhanced domestic investment and local political commitment, which continue to limit progress in many parts of the world. As demand for services from communities increases, responsiveness to that demand from governments, the funder of such services in most countries, is more likely90.

In addition to the unmet needs related to existing vaccines, nearly half of all deaths from infectious diseases are caused by infections for which no vaccine is available (for example, more than 0.5 million deaths globally in children under 5 years from enteric infections for which there is no vaccine91). These should be the priorities for vaccine research and development, as well as improvements needed for particular vaccines such as those against rotavirus, pertussis, polio and yellow fever. Innovations in delivery devices are also important (for example, micropatches, temperature-stable vaccines, improved cold-chain equipment).

## The equity imperative

Equity has been a primary goal of immunization programmes. To reach those who are in greatest need means addressing issues of vaccine availability, affordability, accessibility, acceptability and financing. An effective immunization system that delivers vaccines with high equity across social and ethnic strata, maternal and community education, and geographies, is a purpose-built programme to deliver impact, and has been shown to be the crucial programmatic target.

Country-level vaccine coverage values mask subnational inequity, risking disease outbreaks and backsliding on achievements of vaccination. Immunization improvements should focus at the subnational level, as well as on other determinants of inequity, not all of which would be addressed by focused supplementary vaccine campaigns.

There is a special case for vaccine development for pathogens that cause epidemics. These diseases have little to no market incentive to drive product development, hence the need for innovative arrangements such as the CEPI53, US Biomedical Advanced Research and Development Authority (BARDA)92 and the European Innovative Medicines Initiative (IMI; https://www.imi.europa.eu)93.

Humanitarian crises are another increasing impediment to immunization. The number, size and duration of conflicts, the migration of refugees, and natural disasters have all caused major disruptions to immunization programmes and resulted in serious disease outbreaks. The persisting hurdles to the eradication of polio reveal how political, social and conflict situations can disrupt access to populations and risk violence targeting vaccinators such as in Pakistan and Afghanistan94. Nearly 100 polio vaccinators and their security guards have been targeted and killed while attempting to reach children for vaccination95.

## The growing challenge of vaccine confidence

Despite the success and wide acceptance of the importance of immunization, there are growing groups of people who delay or refuse vaccines. In 2013, the WHO Strategic Advisory Group of Experts (SAGE) established a working group to investigate the scope and scale of vaccine hesitancy96, the US National Vaccine Advisory Committee (NVAC) put together a Vaccine Confidence Working Group to investigate the situation in the United States (National Vaccine Advisory Committee, 2015), and the European Centre for Disease Prevention and Control (ECDC) published a review of the state of vaccine hesitancy in Europe97. In January 2019, the WHO named vaccine hesitancy as one of the top ten global health threats.

Since 2015, the Vaccine Confidence Index (VCI) has surveyed more than 300,000 respondents globally to detect early signals of waning public confidence in vaccine importance, safety and effectiveness, to prompt early intervention where needed (see Fig. 2 for world map of confidence in vaccine safety in 2018). The European Commission adopted the VCI as part of an effort in 2018 to strengthen cooperation against vaccine-preventable diseases98, and the Wellcome Trust used the VCI as part of their 144-country study into public confidence in vaccines (Wellcome Global Monitor 2018)99. Safety was identified as a key issue in both the 2018 European study and the Wellcome report, with public confidence in vaccine safety being consistently lower than the confidence in vaccine effectiveness and importance99.

Although a lack of familiarity by both physicians and parents with many childhood diseases because of years of successful vaccination programmes may have a role in a lack of interest in vaccines, the reasons for a decline in vaccine confidence are far more complex. Newer challenges to vaccine confidence include social media campaigns that have disrupted MMR vaccination efforts in southern India, collapsed HPV vaccination efforts in Japan, provoked false scares of vaccine poisoning in Pakistan, and undermined vaccination programmes in Indonesia.

Vaccine confidence issues are highly varied by setting and vaccine. In a three-year review (2015–2017) of the WHO/UNICEF Joint Reporting Form (JRF) completed annually by national immunization programmes, over 90% of the 194 countries reported that they experienced vaccine hesitancy. The top three reasons for hesitancy were: (1) ‘risk–benefit (scientific evidence)’—that is, safety concerns; (2) lack of knowledge on the benefits of immunization; and (3) religion, culture and socio-economic issues100.

Challenges around building confidence in vaccine safety are well beyond communication, although more accessible public communication around the complex issues of safety and risk benefit analysis are important. What needs to be addressed is not only better communication around the known, albeit sometimes misinterpreted, risks and benefits of vaccination, but also investing in more research in the areas in which the public is asking questions and the science is incomplete. Findings that the AS03-adjuvanted influenza vaccine Pandemrix was linked to increased cases of narcolepsy in Europe prompted further research, but a systematic review concluded that more research is needed101.

Although uncertainty is the norm in science, the political and social worlds of the public have become less tolerant of ambiguity and risk102. New modes of listening to the public, with rapidly evolving technologies to monitor social media, can collect emerging safety questions as well as detect signals of possible issues that need investigation. Working towards better aligned public questions and accessible, evidence-based answers should be a goal. The WHO Vaccine Safety Net initiative is an important resource and can be further built on to address new questions as they emerge, as well as to make new research accessible103.

Social and political contexts and the reliability of health services are important levers of trust, and a low trust setting will have less tolerance for risk than one with high trust. A 2015 study showed that high trust in immunization services clearly correlated with lower rates of vaccine hesitancy104. The public’s experience with health services and health workers is highly influential in vaccine decision, but both are needed. The Wellcome Global Monitor report showed that in Japan, for example, despite low trust in vaccines and low trust in government, confidence in health providers remained high.

Introducing new vaccines into populations requires adequate time to train and prepare front-line health workers and vaccinators to be ready to manage public questions, and continuing dialogue between scientists and the public will be important to build confidence from the start, as well as to anticipate and manage adverse events.

As mentioned above, reported risks of a recently introduced dengue vaccine105 in the Philippines amplified into public outrage mediated through Facebook pages, and were made more complex because the events occurred during political elections. The result was a marked drop in public confidence in vaccines more generally from 99.5% in 2015 to 76.2% in 2018, and confidence in vaccine safety plummeted from 99.5% to 65.2%65 (Fig. 3). The overall drop in public trust affected willingness to accept even the measles vaccine, prompting measles outbreaks with more than 25,000 measles cases and 355 deaths by March 2019106 and requiring considerable efforts to rebuild public confidence and increase vaccine uptake.

Conflict situations also affect confidence in vaccines and vaccinators owing to an environment of distrust and uncertainty, such as in Pakistan and Afghanistan, and in the Democratic Republic of the Congo, where local violence and conflict in the Ebola-affected areas has been an obstacle to vaccination efforts.

## The future of immunization

The contribution of immunization to human health, security and prosperity has been matched by few other activities in health and development, and has been crucial for progress in child survival. As immunization coverage among adults is generally low, it is another area in which greater advances can be made.

Addressing the following issues will be crucial to ensure that the effect of vaccination is optimized.

(1) Leadership and funding. Achieving immunization for all those in need should be a top priority for every country. This will require stronger political leadership and a continuing increase in investments in immunization, both domestically and internationally6. The power of immunization to achieve wider health and societal benefits should be further documented. The prioritization of vaccines is particularly crucial for middle-income countries that no longer benefit from support from the Gavi Alliance and for countries that are transitioning out of Gavi support.

A successful replenishment of Gavi resources in 2020 for the proposed Gavi 5.0 strategy107 is vital for the next decade of progress in child survival, and will be a test of the commitment of the international community to immunization and global health.

(2) Universal vaccine coverage and equity. Overcoming the stagnation in reaching all people in need with even the basic vaccines is an overriding priority in all countries, especially in those with the lowest coverage and the greatest number of unvaccinated children. As we look towards the next decade, ensuring that vulnerable people in all countries are not left behind should be a top concern, particularly in middle-income countries, as there will be more poor people living there than in poorer countries78.

Ensuring a sustainable and affordable supply of quality vaccines, with differential pricing according to the wealth of a country, is fundamental to achieving sustainability and equity of immunization. Only a few multinational companies are producing vaccines, and a growing number of middle-income manufacturers are major suppliers. There is a risk that continuous lowering of prices may lead to new monopolies, and possibly to higher prices. Healthy vaccine markets with sustainable supply are an important objective for vaccine programmes. Harmonization and strengthening of regulatory capabilities of low- and middle-income countries are essential. Initiatives such as the AVAREF58 deserve support. The fact that some countries require local clinical trials despite WHO pre-qualification can be a source of major delays in the introduction of vaccines.

(3) People-centred programmes. Immunization programmes can become more effective with a systems-driven and ‘precision public health’ approach, taking into account local variation in immunization levels, specific needs, cultural specifics, and circumstances of vulnerable populations. Quality data at administrative levels closer to communities should be collected to inform ‘micro-planning’ and adaptive programme delivery. Innovative efforts such as thoughtful integration of immunization into health services, education systems and elderly care are needed.

As most vaccines have incomplete efficacy, tailored approaches to optimize their impact will be needed, particularly for vaccines against malaria, influenza, dengue and probably HIV when it becomes available.

(4) Vaccine confidence. Vaccine confidence needs to be addressed up front and be an integral part of immunization programmes. Many approaches to increasing vaccine uptake do not take into account the social, historical and political realities of the public for whom information alone is not the antidote to vaccine reluctance. Instead of older demand-creation models, a new model and language of engaging with the public is needed, starting with better listening and prompt responding to concerns as well as building on local capacities. Inclusion of non-traditional partners, new modes of digital communication, social scientists, and religious and traditional leaders have been invaluable in addressing hesitancy around polio vaccination, and the engagement of teenage girls in co-designing social media outreach to address HPV vaccination concerns had positive effects on vaccine uptake in Denmark. With safety anxieties being reported as one of the top reasons for vaccine hesitancy, aligning vaccine safety research with dominant safety concerns will also be important for confidence building.

(5) Investment in research and innovation. Many issues mentioned in the other recommendations require further research in a wide range of disciplines. Product innovation as a result of the formidable progress in immunology and infection pathogenesis has been a strong driver of immunization programmes. There is reluctance of industry to develop vaccines when market incentives are limited, and licensing is uncertain. Although companies such as Merck and Johnson & Johnson invested considerably in the development of candidate Ebola vaccines, partly supported by public funds in North America and Europe, but without a prospect of a return on investment, it would be unrealistic to expect that industry will follow this example for each new emerging pathogen. There is a major role for the public sector and philanthropy to support mechanisms such as the CEPI to develop vaccines for low-income countries2. As discussed under the ‘second hurdle’ on the challenge to fund and conduct late clinical development through to the market introduction for vaccines for which there is no market incentive, there is an urgent need to address this gap, possibly via a specific global initiative or at least a concerted action of several funders. There is also a need for innovation in trial design (for faster trials with smaller sample sizes, and including collection of valuable biosamples to inform correlates of protection) and in trial analysis, as well as in vaccine delivery. Escalating antimicrobial resistance is a powerful incentive to develop vaccines against bacterial infections, malaria, tuberculosis and HIV infection108,109,110. Innovation in the delivery of vaccination programmes is as important as product innovation.

The world cannot afford to turn the clock back on immunization, and ever more innovative vaccines will offer additional opportunities to reduce mortality and improve the quality of life for every person on the planet. This will require the best of science, entrepreneurship, programme implementation on the ground, and politics.