In the 1970s, the WHO embarked on an ambitious project to promote primary health care worldwide. The Expanded Programme on Immunization (EPI) was one of the most successful parts of that effort, yet some national EPIs struggled to increase vaccination coverage while others were very successful. Drawing on documentary sources from the WHO Archives and Library, this paper traces the historical development of global EPI policy and compares the development of two programs: the high-performing EPI in Malawi and the low-performing one in Cameroon. Global advisers’ rigid adherence to then-current global policy and blindness to local conditions and historical legacies exacerbated problems faced by Cameroon's EPI, helping explain that program's weakness. In Malawi, in contrast, the similarity of global policy and local practices helped strengthen the EPI. Greater flexibility in pursuing program goals and attention to historical legacies could help future programs avoid similar counterproductive dynamics.
In 1978, the International Conference on Primary Health Care issued the Declaration of Alma Ata (UNICEF and WHO 1978), which affirmed the importance of health (broadly defined as “a state of complete physical, mental and social wellbeing”) as a human right and called for the development of primary health care (PHC). It defined PHC as appropriate community-based care to address the main health problems faced by communities through intersectoral cooperation to promote health and to prevent and treat diseases. This vision of PHC served as the basis for the WHO's Health for All by the Year 2000 campaign, which lasted until 1994, when the WHO concluded that it couldn't achieve its goals (Chan 2008:865).
The PHC campaign was a precursor to major global development efforts that have followed, like the Millennium Development Goals (MDGs) and the Sustainability Development Goals (SDGs). Many of the goals it espoused informed subsequent efforts, including efforts to promote child and maternal health (MDGs 4 and 5), to end hunger (MDG 1 and SDG 2), and to ensure access to clean water and sanitation (SDG 6). Recent years have seen new interest in the promises of PHC; WHO director-general Margaret Chan called for a return to Alma Ata in 2008 (Chan 2008). The renewed emphasis on PHC is the foundation for new global priorities, including efforts to encourage universal health coverage and renewed interest from many donors and health organizations, like the GAVI Alliance and the Global Fund to Fight AIDS, Tuberculosis and Malaria, in strengthening health systems as a key part of promoting development.
The literature on PHC is vast, including debates about selective/vertical versus comprehensive/horizontal approaches (Gish 1982; Newell 1988; Walsh and Warren 1979), community participation (Rosato et al. 2008), and integration (Bhutta et al. 2008; Ekman, Pathmanathan, and Liljestrand 2008). However, this extensive literature does not examine how the global PHC program was implemented and the effect this had on participating countries. Instead, most examinations tend to focus on national (or even sub-national) PHC programs (see e.g. Rohde et al. 2008). This neglect continues despite growing attention to the effect global health programs and organizations have on participating communities (Biesma et al. 2009; Doyle and Patel 2008) and the impact of aid on receiving communities (OECD 2008). There is also a large literature on how the World Bank and other international financial institutions influence health and health policy in the developing world, particularly through structural adjustment programs (Babb 2005; Noy 2015, 2017), yet little attention to the WHO. Given the renewed global interest in PHC and the key role global development campaigns continue to play in promoting the development of health services worldwide, we must ask: What can global development efforts learn from previous attempts to promote PHC during the WHO's Health for All campaign?
In this paper, I address this question through an examination of one part of the Health for All effort: the Expanded Programme on Immunization (EPI). While the Health for All program ultimately failed, the EPI met its goal, “universal childhood immunization,” defined as 80% global coverage (“Expanded Programme on Immunization and Vaccine Quality: Progress Report by the Director General,” 1992). However, this broader success masked large differences in country experiences. I examine the implementation of the EPI in two countries: Malawi, which was particularly successful at extending immunization coverage, and Cameroon, which was not. Focusing on vaccine delivery strategies, I examine the interplay between changing global recommendations and national program development, as national EPIs strove to meet ambitious program goals.
Global-level recommendations regarding vaccine delivery strategies changed considerably between 1974, when the EPI began, and 1992, when the program achieved its target. Global-level changes profoundly influenced national-level practices, with global advisers and periodic program evaluations playing a key role in transmitting new global program recommendations to national programs. In doing so, global advisers rarely showed a deep understanding of local context or history, even when global recommendations called for radical changes in national programs. However, national legacies proved tenacious in practice, with local-level personnel resisting radical change even while national policy quickly changed to conform to global recommendations. I argue that global advisers’ rigid adherence to global policy recommendations and blindness to local historical legacies exacerbated problems faced by Cameroon's EPI, helping explain that program's weak performance. In Malawi, in contrast, the fortuitous congruence of global policy and local practices created synergistic conditions that helped strengthen the EPI there.
In the following sections, I review the data and the methods I use in choosing my two cases and analyzing the development of their national EPIs. I then provide an overview of the history of vaccination services in Malawi and Cameroon prior to the EPI. Next, I examine program developments over the course of the EPI, focusing on three distinct periods of global policy recommendations. I conclude by reflecting on Malawi and Cameroon's experiences and the lessons they hold for global efforts to promote health and the development of strong health systems.
METHODS AND DATA
My analysis combines an examination of dynamics at the global and national levels using documentary analysis methods. For case selection, I use the nested case method (Lieberman 2005). I focus on the period from 1974, when the EPI began, to 1992, when the program met its “universal childhood immunization” goal. For the national cases, I cover the period from 1974 to 1990 due to data limitations.
The global-level analysis focuses on global-level policy changes in the EPI program and draws on five sources: official EPI progress reports; debate and resolutions concerning the EPI from the annual meetings of the World Health Assembly, which sets official WHO policy; reports from the annual meetings of the EPI's Global Advisory Group; technical documents related to the EPI; and articles related to the EPI in the Weekly Epidemiological Record.
To choose cases, I calculate regression residuals based on the results of five separate fixed effects regressions of immunization coverage in low- and middle-income countries. Quantitative analyses focused on vaccination coverage with the four vaccines targeted by the EPI (DPT, polio, measles, and BCG) and an index variable including the average of coverage with all four vaccines. Control variables include infant mortality rate (IMR), GDP per capita, population, two measures of ties to international organizations, a measure of state durability, an autocracy index, and an indicator of armed conflict.1
Based on these regressions, I calculated regression residuals and then the average residuals by country. I then created lists of the 10 countries with the largest positive and negative residuals for each regression. Table 1 lists the countries with largest positive and negative residuals for each model, respectively. Countries with large positive residuals have higher vaccination coverage, on average, than predicted by my models. Countries with large negative residuals have lower vaccination coverage than predicted.
|Polio Vaccine Coverage .||Measles Vaccine Coverage .||DPT Vaccine Coverage .||BCG Vaccine Coverage .||Vaccine Coverage Index .|
|Countries with largest negative average residuals for all vaccine coverage series|
|Countries with largest positive average residuals for all vaccine coverage series|
|Polio Vaccine Coverage .||Measles Vaccine Coverage .||DPT Vaccine Coverage .||BCG Vaccine Coverage .||Vaccine Coverage Index .|
|Countries with largest negative average residuals for all vaccine coverage series|
|Countries with largest positive average residuals for all vaccine coverage series|
Note: Average residual =∑(coverage - predicted coverage)/N where N = number of years of data in the series
Countries in bold appear on all lists
CAF = Central African Republic
DRC = Democratic Republic of the Congo
PNG = Papua New Guinea
African countries are overrepresented on lists of low-performing countries, with countries in francophone West and Central Africa performing particularly poorly. All three countries that appear on all such lists (Chad, Cameroon, and Gabon) are in this region. High-performing countries are more geographically diverse. Although several African states are included in at least one list, only Malawi and Gambia appear on all five lists. I focus on Malawi and Cameroon as my high-performing and low-performing cases, respectively. Both were early adopters of the EPI. This allows me to trace program development in both countries over the entire period and ensures that political commitment to the program is relatively similar. I draw on the WHO Archives for information about the national EPIs in Malawi and Cameroon.2 Documents come from Malawi and Cameroon EPI country files, the general Africa files, and African training files. My analysis focuses on program plans, evaluations, and correspondence from both countries.
IMMUNIZATION SERVICES IN MALAWI AND CAMEROON
Malawi and Cameroon were among the first countries to participate in the EPI. An experimental expanded immunization program began in Cameroon's capital city, Yaoundé, in 1975, and the program joined the global EPI effort in 1977. Malawi began its national EPI in 1978. Although both are in sub-Saharan Africa, Malawi and Cameroon differ on multiple dimensions, including their size, levels of ethnic diversity, colonial history, and governance. However, my analysis does not seek to gain analytic leverage from a strict comparison of programs in each country. Instead, each program is compared to predicted coverage based on the regression model.
Figure 1 compares predicted and actual vaccination coverage with the vaccination coverage index variable in Malawi and Cameroon for the years 1980 through 1992. We see that immunization coverage is consistently at or near the predicted levels in Malawi, although overall coverage levels stagnate around 1983 to 1984. In Cameroon, coverage is consistently below predicted levels, with distinct trends. Coverage increases rapidly in the early period before stagnating in 1983 and experiencing a major drop from 1985 to 1988. Thereafter, coverage increases rapidly, although it remains well below predicted levels.
The analysis that follows will seek to understand some of the dynamics at work to explain these trends. Before this, I provide an overview of immunization services in Malawi and Cameroon before the EPI began. Immunization services were organized along distinctly different lines in the two countries, and these historical legacies influenced the national EPIs Malawi and Cameroon developed, and continued to influence national practices in subsequent years.
Pre-EPI Immunization Services
In Malawi, public and private (mainly Christian/missionary) health providers have long cooperated extensively. In 1978, when the EPI began, private missionary groups ran almost half of Malawi's hospitals, a third of its primary health centers, and a fifth of its sub-centers (see “Expanded Programme on Immunization: Malawi,” 1978, for information on immunization services in Malawi prior to 1978). Missionary health services have a long history in Malawi, as does cooperation between missionaries and government health services (Beck 1970; Good 2004; Hokkanen 2007).
As early as 1960, preventive care was available at some health centers during “under-five clinics” (geared to children under the age of five). In 1973 the Miniplan, a seven-year development plan for maternal and child health, called for the extension of under-five clinics and for an increased emphasis on preventive health, in particular “to increase the coverage of protection amongst children, especially against diseases for which effective antigens are available” (“Expanded Programme on Immunization: Malawi,” 1978).
By 1977, a year before the Malawi EPI began, 50% of the country had access to under-five clinics providing vaccination with BCG, DPT, smallpox, and polio vaccines (“Expanded Programme on Immunization: Malawi,” 1978; Letter dated 10/11/1978). Under-five clinics were held at some of the fixed health centers and run by mobile teams based at these centers. Malawi also ran mass immunization campaigns using specialized mobile vaccination teams to vaccinate against smallpox and tuberculosis, but these efforts were short-term, targeted campaigns, clearly distinguished from the “routine immunization” available through the under-five clinics.
Cameroon's immunization services differed from Malawi's in two major ways: how the services were provided and the groups involved. Cameroon built on the grandes endémies (major endemic diseases) system of mobile health teams that were common in francophone West and Central Africa. This excerpt from an unpublished manuscript describes how these teams worked:
In Yaoundé, as was traditional throughout urban or rural francophone Africa, vaccination services were delivered by a mobile vaccination team … which visited the city for 3 months every 2 years and attempted to vaccinate both child and adult populations against measles (6 month – 6 year age group), yellow fever (after 1 year-old), smallpox (all ages), and tuberculosis (BCG from 6 months – 20 years). (“Program of Multiple Antigen Childhood Immunization in Yaoundé, Cameroon,” 1977)
Thus, in Cameroon in 1974, the government provided immunization using national vaccination teams that traveled the country in a biannual circuit. Teams were permanent but were not integrated into curative health services, instead operating out of a separate government subdepartment dedicated to public health and preventive medicine. Private, missionary groups seem to have played a much smaller role than in Malawi.
The mobile grandes endémies teams were an innovation of the renowned colonial doctor, Eugène Jamot. He first developed this approach to address a sleeping sickness epidemic in French Equatorial Africa in 1917. In 1922, he established similar services in French Cameroon, where he continued to work for a decade (Headrick 1994; Pepin 2011). This method, which relied on highly specialized mobile health teams periodically visiting the entire population to detect and treat (or prevent) diseases of major public health concern, was used to address multiple health issues over decades, including malaria, leprosy, yaws, and smallpox (Bado 1996; Pepin 2011). After independence, a regional health organization, the Organisation de Coordination pour la lutte contre les Endémies en Afrique Central (OCEAC), assured continuation of such efforts (Bado 1996:374; OCEAC 2010).
IMPLEMENTING THE EPI IN MALAWI AND CAMEROON
EPI policy concerning vaccine delivery strategies changed significantly as the EPI developed. My analysis reveals three distinct approaches during three distinct periods: the experimental approach from 1974 to 1981, the PHC approach from 1981 to 1985, and the coverage approach from 1985 to 1992.
The Experimental Approach
Early on, the EPI emphasized the need for innovation and experimentation to develop vaccination strategies for the developing world. This was a key recommendation from the report of the WHO's 1974 Seminar on Expansion of Immunization Services in Developing Countries. Even though EPI progress reports highlighted the connection between routine childhood immunization and PHC as early as 1976, the emphasis on the need for experimentation, particularly with “mixed strategies” (mixing mobile and fixed services), continued (“Expanded Programme on Immunization: Progress Report by the Director General,” 1976:2). Even in 1977, as the EPI began its fully operational phase, the EPI simply recommended that national programs define a strategy, with little advice on what that strategy should be (“Expanded Programme on Immunization: Progress Report by the Director General,” 1977:7–8)
Malawi and Cameroon both adapted existing health systems in establishing their national EPIs. In Malawi, this effort grew organically from the existing health system, while in Cameroon it consciously responded to the call for developing countries to experiment.
Malawi formally began its EPI program with its 1978 EPI Plan, which it submitted to the WHO for feedback. The plan called for expanding the under-five clinic system and adding measles vaccine to the list of antigens provided at the clinics. Under-five clinics were conducted on a weekly or monthly basis in fixed health services, or by mobile teams in areas “not easily covered by a static unit.” At clinics, immunizations services were “carried out together with weight and growth monitoring, health/nutrition education, and early treatment at clinics.” The plan also called for a two-year mass measles vaccination campaign using mobile teams (akin to the smallpox eradication campaign) to rapidly increase measles vaccination coverage, which was scant because that vaccine had not previously been available in under-five clinics (“Expanded Programme on Immunization: Malawi,” 1978).
At first, EPI personnel at the WHO misunderstood the nature and intent of the mass measles campaign, believing it would only serve areas of the country not yet served by under-five clinics and that the campaign would be repeated on a two-year cycle. Based on this mistake, they criticized the limited number of antigens included and the two-year cycle, which wasn't sufficiently rapid to break transmission (Letter dated 17 Aug. 1978). A national-level health officer quickly clarified the plan (Letter dated 11 Oct. 1978), which received no further objections from the EPI program office in Geneva, and Malawi proceeded with the mass measles campaign. They added a second mass vaccination campaign targeting polio in 1980 (“Report on the Evaluation of the Expanded Programme on Immunization in Malawi,” 1980) based on recommendations from a 1978 survey of polio prevalence (Ward and Lungu 1978).
In 1975, a team from OCEAC with extensive backing from the Centers for Disease Control and Prevention and USAID began an expanded immunization program in Cameroon's capital city, Yaoundé. Although it wasn't affiliated with the EPI, the Yaoundé program was self-consciously set up as a model and testing ground for the EPI. The new program adapted the grandes endémies system to the needs of routine childhood immunization by combining “the advantages of both mobile teams and fixed centers.” A description of the program notes:
In the Yaoundé program, a mobile team circulates to 10 designated dispensaries and PMI's on a fixed calendar schedule … (e.g. each last Wednesday of the month the team will vaccinate at Centre A). These centers are situated throughout the city so that no area is more than about 1 km from a center. Only the trained team members handle vaccine, carry out technical aspects, do triage of vaccinees, etc. The chief of the team directs these activities. (“Program of Multiple Antigen Childhood Immunization in Yaoundé, Cameroon,” 1977)
While the new program made vaccination regularly available at fixed sites, it still used specialized mobile teams to carry out vaccinations. Instead of traveling the country in biannual drives, the team traveled the city in a continuous circuit to run monthly vaccination sessions at the fixed health centers. It self-consciously sought to experiment with “mixed strategies” to establish routine childhood immunization services in a developing country, in answer to recommendations from the WHO.
By 1977, Cameroon was ready to expand the program nationally. To this end the Ministry of Public Health, with technical support from OCEAC, prepared a national EPI plan for 1977 to 1982. The plan called for the expansion of immunization services to cover the entire country over 10 years, so that the whole country could “profit from Yaoundé's experience,” and built on the Yaoundé program, including plans to establish mobile vaccination teams in four new areas, along with a more general expansion of immunization services across the country (“Propositions et Projet de Budget, Programme de Vaccination Elargi, République Unie de Cameroun,” 1977).3 The WHO's response was largely positive. A 1978 evaluation of the program, which included a close study of the Yaoundé vaccination team, praised the team's motivation, expertise, organization, and punctuality. It declared, “This independent, highly specialized mobile vaccination team … should be regarded as the nucleus for EPI expansion,” and recommended that the team expand its activities to more health centers and play a key role in training personnel for the nascent national immunization program (“Review of the Expanded Programme on Immunization, Yaoundé, United Republic of Cameroon,” 1978).
In sum, during this period Malawi and Cameroon both built on existing health services or practices to implement routine childhood immunization services. The programs differed substantially. Immunization services were much more extensive in Malawi, and the same health personnel carried out immunization and other well-baby examinations. In contrast, childhood immunization services only existed in Cameroon's capital, Yaoundé, and they were carried out by specialized immunization personnel. However, there were some points of similarity. In both cases, childhood immunization services were only periodically available (at weekly or monthly clinics), and in both cases services were available at a combination of fixed health centers and other sites. Plans in both countries received support from the WHO.
The PHC Approach
The Global Advisory Group (GAG), which the WHO formed in 1978, quickly established more precise recommendations for the EPI. From the start, it emphasized the EPI's key position in the WHO's broader PHC program, advocating a PHC approach to immunization. This emphasis is clear in the action plan it endorsed at its 1981 meeting, the first point of which urged, “Promote EPI within the context of primary health care,” and identified two key aspects of this PHC approach: promoting community participation, and integrating immunization services into other health services, particularly maternal and child health services (“Report of the Expanded Programme on Immunization Global Advisory Group Meeting: 19-22 October 1981, Washington, D.C.,” 1981).
The priority placed on integration led to a focus on fixed health centers as the key sites for immunization services. Advice on national program expansion from the EPI's first and second Medium Term Programs emphasized that regular immunization services should first be established in fixed health centers, followed by outreach services operating out of those centers (“Global Medium Term Programme: Expanded Programme on Immunization,” 1980; “Global Medium Term Programme: Expanded Programme on Immunization,” 1983). This emphasis on integration and the primacy of fixed health centers caused little change in Malawi but had a dramatic impact on Cameroon's EPI.
Long before Alma Ata, Malawi's immunization services were already developing in a manner that corresponded well with the PHC approach the EPI adopted in 1980: integrated with maternal and child health services and based in fixed health centers. This is not surprising. The WHO's PHC program had its origins in developments in the Christian Medical Commission (Litsios 2004), and Malawi is a country with a very strong tradition of Christian (missionary) medicine and long-standing cooperation between Christian and government medical services.
This fortuitous congruence meant that global advisers were generally pleased with Malawi's program. Only one aspect of the program caused concern: the ongoing mass campaigns against measles. While the 1980 program review expressed no disapproval of the ongoing polio and measles campaigns (“Report on the Evaluation of the Expanded Programme on Immunization in Malawi,” 1980), a 1982 evaluation was more critical of new “mini-campaigns” in areas surrounding reported measles cases. The government had started these mini-campaigns in the hopes of containing outbreaks, but global advisers criticized this approach on epidemiological grounds and because it could “lead to the public expecting and waiting for vaccinators to visit their village rather than attending a clinic.” Instead they recommended improving routine measles coverage through “a major propaganda campaign, through the media” in the hopes of “achieving … high coverage through under-5 clinics.” They also recommended that the country develop “appropriate strategies” for reaching areas without under-five clinics, such as “a three monthly visit by a well-equipped mobile team” (“Joint Evaluation of the Expanded Programme on Immunization, Malawi: September 20-October 8 1982,” 1982).
Due in large part to the program's failure to attract needed external financial support, Cameroon's attempt to expand its EPI in the late 1970s stalled, and the EPI was in crisis. In the face of this failure, the government submitted a new EPI proposal to the WHO in early 1980. Following WHO recommendations from 1978 to base its program on mobile vaccination teams and to use the Yaoundé team as “the nucleus for training to other health personnel for expansion of vaccination activities in Cameroon” (“Review of the Expanded Programme on Immunization, Yaoundé, United Republic of Cameroon,” 1978), the new plan focused on perfecting the services mobile teams offered in three “demonstration zones,” and then using those zones as centers for training as the program expanded throughout the country (“Projet du Programme Elargi de Vaccination du Cameroun,” 1980). With UNICEF providing an important new source of external funding, the new expansion fared much better than earlier efforts. By 1982, 62% of departments had EPI services, and this rose to over 80% in 1984 and to 100% by 1986 (“Prospects for Universal Immunization by 1990 in Cameroon,” 1986). Training personnel was a key priority, and at least 15 training courses were held between 1980 and 1986.
But in 1981, a new WHO-led program evaluation abruptly called the mobile-vaccination-team strategy on which the plan relied into question. It began by declaring, “The recommendation from the last evaluation concerning the reorientation of the Yaoundé vaccination team into a training team has not been realized. On the contrary, the team continues to carry out vaccinations according to the former mobile strategy system.” It advised a new strategy centered on four key points:
Integrating vaccination services into the primary health services dispensed at fixed health centers,
Reorienting the mobile team toward a supervision and training role,
Establishing a calendar to suppress the mobile team's vaccination activities and replace them with vaccination at primary health centers, and
Retraining mobile team personnel in management, logistics, evaluation and supervisory methods “so that they can give the fixed center adequate technical support.” (“Rapport d'une Evaluation du Programme Elargi de Vaccination (P.E.V.): République Unie du Cameroun,” 1981)
In other words, the report recommended reforming Cameroon's program to conform to the PHC approach.
The 1981 evaluation is remarkable because it simultaneously calls for an unprecedented reorganization of Cameroon's EPI and suggests that it is reiterating recommendations from the last program evaluation. A close reading of the 1978 evaluation to which it refers clearly shows that evaluators in 1978 were not demanding that the Yaoundé vaccination team take on training duties in the place of their vaccination activities, as the critique suggests, but rather in addition to them. The team was meant to help in the training of new teams that were meant to begin programs using a “mobile strategy” like theirs.
While the government instituted the suggested reforms, local and regional health personnel expressed profound reservations about the new system. Personnel in OCEAC advanced four critiques of the PHC approach.
Theoretical critique: specialized teams were “more effective and efficient” than the health personnel in fixed centers.
Empirical critique: in OCEAC countries, the PHC approach was less effective at improving vaccination coverage than the grandes endémies tradition.
Humanitarian critique: the focus on fixed centers left the rural majority with no access to services.
Personnel problem: health personnel in fixed centers, overburdened with curative care, “do not welcome being additionally burdened with vaccination activities,” and many staff didn't have adequate health training to undertake such duties. (“Compte-Rendu du Mission,” 1982)
A series of local studies from a mix of urban and rural areas written between 1983 and 1985 echo many of these objections (“Evaluation de la Couverture Vaccinale dans la Ville de Bafoussam (Cameroun), Decembre 1983,” 1983; “Evaluation de la Couverture Vaccinale dans le département du Noun en juillet 1984,” 1984; “Rapport de l'Evaluation de la Couverture Vaccinale dans la Ville d'Akonolinga en Novembre 1985,” 1985; “Résultats de l'Enquête Effectuée du 7 au 14 Novembre 1983 pour l'Evaluation de la Couverture Vaccinale dans le Nord-Est Benoué,” 1983; Untitled Report I8 370 2CAE R84, 1984).4 Although all the studies praised the PHC approach as the “ideal strategy,” they also all noted numerous ways this ideal failed in Cameroon. Studies from rural areas echoed the humanitarian critique, noting that mobile vaccination teams were necessary to reach rural populations due to lack of health centers or inadequate cold chains. Urban studies, in turn, referred to the personnel problem, noting health-center personnel's reluctance to undertake vaccination activities or incompetence in doing so. A particularly interesting series of studies from the department of Noun, including a coverage evaluation and cost-effectiveness study, addressed the empirical critique by comparing the efficiency and cost of fixed versus mobile vaccination services. They found that mobile teams were both more efficient and more cost-effective than fixed centers. Not only were children served by teams more likely to complete the full vaccination series than children served by fixed centers, but the cost per fully vaccinated child was also much lower: CFA 5,957 compared to CFA 12,354 (Untitled Report I8 370 2CAE R84, 1984).
The cost-effectiveness study also offered new theoretical justifications for mobile strategies: improved management and training and greater community mobilization. The study notes:
The mobile team doesn't only play a vaccination role but equally supervises our dispensaries and SSP centers, trains personnel in these centers, evacuates the ill, etc. Also, the arrival of a mobile team in an isolated village is a non-negligible factor in mobilizing the population. It would be foolhardy to suppress our mobile team in the long run only on financial criteria.
The focus on community mobilization is especially significant. Community participation was one of two key features of a PHC approach to immunization, according to the five-point plan developed by the GAG in 1981. However, advice from the GAG and the WHO in the early 1980s generally ignored this aspect of the PHC approach, instead focusing on integration.
In sum, during this period Malawi and Cameroon had very different experiences with global advisers. While Malawi met with praise and only modest critiques, Cameroon's program came under intense criticism, leading to a complete reorganization of the program during this period. The criticism Cameroon faced highlights global advisers’ blindness to local legacies. The authors of the WHO's 1981 program evaluation were apparently unaware of (or did not care about) the long history of mobile vaccination services in Cameroon. More surprisingly, they were not fully aware of the nature of feedback Cameroon had already received from previous WHO program evaluations.
The Coverage Approach
By mid-decade, problems in increasing vaccination coverage sufficiently to meet the universal childhood immunization goal led to innovations. While reaffirming the continued relevance of the PHC approach, the 1983 GAG report called for “experimentation with innovative strategies to increase vaccine delivery” (“Report of the Expanded Programme on Immunization Global Advisory Group Meeting: 31 October - 4 November, 1983, Manila,” 1982). In its 1985 meeting, the GAG issued five new recommendations concerning vaccine delivery strategies:
“Adopt a mix of complementary strategies for programme acceleration”;
Sustain increases by strengthening PHC;
“Provide immunization at every contact point”;
“Reduce drop-out rates between first and last immunization”; and
“Improve immunization services to the disadvantaged in urban areas.” (“Report of the Expanded Programme on Immunization Global Advisory Group Meeting: 4-8 November 1985, Copenhagen,” 1986)
These recommendations built on the PHC approach in important ways, including efforts to reduce the dropout rate and ensure immunization at every contact point. However, they also put new emphasis on “accelerated strategies,” particularly intensive immunization drives during national (or sub-national) immunization days (NIDs), marking a significant break with the PHC approach's emphasis on immunization as part of broader integrated services based in fixed health centers. In keeping with past program practices, Malawi and Cameroon responded differently to these recommendations. Malawi favored PHC-strengthening recommendations, like immunizing at every contact point, while Cameroon continued to favor specialized, mobile approaches, eagerly adopting NIDs as a campaign tactic.
Although Malawi's EPI was performing much better than Cameroon's, the mid-1980s posed some major challenges. The period started well, with high praise from Malawi's PHC program in a 1984 evaluation of maternal and child health (MCH) services: “The coverage of MCH services and the standard of work is good. Often there is a degree of fidelity to ideals and standards which is truly an inspiration.” Nevertheless, the review found that, while immunization coverage was good (55% of one-year-olds were fully immunized5), it had “plateaued,” leading to the conclusion that “new approaches will … be needed before the target of 80% fully vaccinated is met” (“Report of the Joint Programme Review: Maternal and Child Health, Expanded Programme on Immunization and Other Elements of Primary Health Care,” 1984). Concretely, it recommended greater community participation to encourage participation in programs, daily MCH/EPI clinics at health centers, and immunizing children who visited health centers for other reasons to reduce the dropout rate. MCH/EPI clinics had previously been held weekly or monthly, and there was no tradition of immunizing at other times. Nevertheless, all recommendations were incorporated in the 1985 to 1989 EPI Plan (“Malawi Epidemiological Quarterly: Expanded Programme on Immunization,” 1985).
Despite these plans, an evaluation in 1985 found that immunization coverage had dropped significantly, with only 35% of one-year-olds fully immunized. A survey of reasons for the vaccination failure found “obstacles” as the main reason (58% of total), with lack of vaccines or vaccinators accounting for 30% of cases (“Evaluation of National Immunization Coverage, Malawi - August 1985,” 1985). The 20% drop in coverage found in the 1985 evaluation was largely due to a major fuel crisis, which forced Malawi to cancel outreach services and remove vaccines from numerous fixed health centers because they lacked the kerosene needed to run their refrigerators (“Country Summary: Malawi,” 1985). Evaluators were apparently unaware of this: the evaluation made no mention of the fuel crisis and recommended, among other tactics, exploring “accelerated strategies” as a possible response to program decline, despite the high fuel needs of such programs (“Evaluation of National Immunization Coverage, Malawi - August 1985,” 1985).
By 1986, coverage levels had rebounded to 1984 levels or better (“E.P.I. Malawi Country Report,” 1986), but evaluators were less satisfied with the program and staff performance than in 1984. The 1988 evaluation included a detailed analysis of vaccine misadministration (“National Evaluation of the Expanded Programme on Immunization,” 1988). Coverage had improved since 1985, with 58% of one-year-olds fully immunized, but vaccine misadministration greatly reduced coverage, particularly with measles vaccine. Sixty-eight percent of children would have been fully immunized had health workers correctly followed the vaccination schedule. Given this finding, it recommended that staff training and supervision be improved. The evaluation also found that many health units (16 out of 27 visited as part of the review) failed to provide immunization services daily and called for investigation of the reasons for these failures. It also highlighted the need for a better system for identifying “defaulters” (children who failed to complete the necessary immunization series) as a corollary to other social mobilization activities.
With its immunization program struggling, Cameroon was a leader in implementing new “accelerated strategies.” In 1986, services extended to 70% of the country, with 70% of vaccinations carried out at fixed centers and 30% by mobile teams, but only an estimated 35% of one-year-olds completed the recommended immunizations (“Plan d'Action, 1986-1990,” 1986). An evaluation of the prospects for achieving universal childhood immunization by 1990 was pessimistic, concluding that Cameroon needed to “radically improve on all aspects of the entire system” to meet the goal (Lantum 1986). Pressure for change was high due to the looming deadline and continent-wide efforts to improve immunization coverage as part of the African Immunization Year.
New international recommendations suggested several new strategies for improving coverage, and program research pointed to two main ways to do it: improving community participation in programs and improving immunization services at health centers. Two studies of reasons for the low participation in Youndé's EPI in 1979 and 1982 highlighted parents’ lack of knowledge about vaccines and vaccination services and indicated that outreach efforts needed to target different ethnic groups, as vaccination coverage was higher among some ethnic groups than others (Brown et al. 1982; “Expanded Programme on Immunization: Community Participation and Immunization Coverage,” 1984)
Cameroon responded to the need to mobilize parents by experimenting with new “accelerated strategies.” In 1984, Cameroon organized “special immunization days” in four areas. The evaluation showed a major impact on coverage in those areas (Lantum 1986), so the EPI organized a series of NIDs in 1986 for the African Immunization Year. It included substantial publicity, with the first lady starting off the campaign, and led to major improvements in coverage, although not substantial enough to reach the target of universal childhood immunization (“Rapid Assessment: Cameroon's National Vaccination Campaign,” 1988). Despite some logistical problems, the 1986 NIDs were broadly successful at increasing coverage, but less successful at strengthening routine services. As one report noted, the accelerated strategy became “an end in itself, instead of [a] means to improve [the] regular EPI” (“Country Notes: Cameroon,” 1988). After 1986, NIDs apparently became a regular feature of Cameroon's EPI (“République du Cameroun: Projet à ‘noter’ pour un financement supplémentaire,” 1986).
Improving services at health centers proved more difficult. A 1983 study showed that substantial numbers of partially or unimmunized children could be reached by immunizing children appearing at health centers for other reasons, like well-baby clinics or for illness. It recommended that health centers systematically review children's immunization status, that immunization services be included at all well-baby clinics, and that children with minor illnesses be immunized (“Expanded Programme on Immunization: Sick Children: Targets for Immunization,” 1983). However, acting on these recommendations would have required substantial cooperation from health center staff, and the “personnel problem” continued to be a major challenge for Cameroon's EPI. The 1986 to 1991 program plan noted that fixed-center staff “lacked motivation,” and included numerous plans for improving supervision, training, and motivation of staff, but did not include plans to immunize at every contact point or recommendations regarding sick children (“Plan d'Action, 1986-1990,” 1986). A 1986 study of the country's dropout rate found that illness was the main reason for failure to complete immunization series, despite recommendations to vaccinate children with minor illnesses (Keuzeta et al. 1986), and a 1988 rapid assessment noted that immunization sessions were still only held monthly and that supervision remained weak (“Rapid Assessment: Cameroon's National Vaccination Campaign,” 1988).
In sum, during this period Malawi's and Cameroon's EPIs both struggled to meet the ambitious goal of 80% vaccination coverage. However, national programs drew selectively on global recommendations, choosing the strategies that best fit with national legacies. Hence, Malawi focused on strategies meant to strength routine vaccination services, while Cameroon employed new mobile “accelerated strategies,” like NIDs. Despite a tepid response from global evaluators, who worried that NIDs were becoming “an end in themselves,” the return to mobile strategies proved popular in Cameroon. Efforts to strengthen routine services encountered more difficulty in Malawi, where health centers struggled to implement daily MCH/EPI clinics and to immunize at every contact point. The fuel crisis that shut down many vaccination services made implementing these new strategies particularly difficult.
This paper examines the interaction between global advisers and local actors in implementing national EPIs in Malawi and Cameroon, with a focus on recommendations regarding vaccine delivery strategies. I identify three key phases of program development: the experimental phase from 1974 to 1981, the PHC phase from 1981 to 1985, and the coverage-oriented phase from 1985 to 1992. As shown in the above discussion, changes in global-level recommendations profoundly influenced national-level practices, with global advisers transmitting new policy recommendations through recommendations found in periodic program evaluations.
In Malawi the fortuitous congruence of global policy and local practices created synergistic conditions that helped strengthen the EPI there. In contrast, global policy increasingly diverged from practices in Cameroon, leading global advisers to call for radical change in Cameroon's EPI. While Cameroon quickly complied with these recommendations, national legacies proved tenacious in practice. Local-level personnel resisted radical change, calling into question the appropriateness of global advice when it diverged from historical practices.
These different dynamics help explain the relative strength of Malawi's EPI and the relative weakness of Cameroon's. A closer examination of how changing global–local tensions map onto program performance highlights this. In Figure 1 we saw that Malawi's EPI consistently had coverage near or above predicted levels. Similarly, there was consistently a good fit between global advisers’ recommendations and existing strategies. Even when advisers’ recommendations proved difficult to implement, as seen in the mid-1980s, the advice at least supported existing strategies or involved a sensible intensification of them. The worst moment for Malawi's EPI resulted from an acute external crisis: archival sources indicate that a fuel crisis in 1985 resulted in a precipitous drop in coverage. The record suggests that global evaluators were unaware of the cause of the drop and offered no advice on how to confront it.
In contrast, Cameroon's EPI never performed as well as expected. However, the degree to which it underperformed differed substantially across the period considered here. Early in the EPI, coverage was increasing rapidly, approaching predicted levels. However, around 1983 the trend changed, with coverage first plateauing and then dropping substantially until 1988. Finally, at the end of the period, coverage began increasing again, although it remained below predicted levels. This pattern maps onto changes in global-level recommendations, with a lag of approximately two years. During the experimental period (1974–1981), coverage increased and approached expectations. During the PHC period (1981–1985), coverage stagnated and then decreased, diverging sharply from predicted levels. Finally, during the coverage period (1985–1992), coverage again began to increase and approach expectations. These periods differed substantially in their openness to mobile strategies, such as those that had long been used in Cameroon. When mobile strategies were encouraged or tolerated (the experimental and coverage periods), Cameroon's program performed better. When they were anathema (the PHC period), the program performed worse.
Could other factors explain the different experiences of these two countries? One possibility is that different political or economic developments in the countries provide a better explanation. We saw, for example, that an economic crisis, in the form of a fuel crisis, had a profound effect on Malawi's program. Economic crises were common in Africa in the 1980s, and both Malawi and Cameroon faced economic crises during these periods. In theory, the predicted values account for this, as the regressions control for GDP per capita. Also, economic crisis seems unlikely to explain the dynamics in Cameroon, as the timing does not fit the observed trends well. Cameroon faced a severe economic crisis beginning in the mid-1980s, admitting to the crisis and introducing budget restrictions in 1987 (Delancey and Delancey 2000:104–05), just as vaccination coverage began to rebound. Perhaps general economic crisis was not a driver of poor performance but rather funding problems within the program itself. Cameroon's EPI had trouble attracting outside support early in the program, while Malawi's program benefitted from substantial outside support from the start. However, program problems deepened just as funding difficulties became less acute. UNICEF began supporting Cameroon's EPI in 1982, just as the program began to falter.
Another possibility is that the system in Malawi simply was better than Cameroon's. And this is undoubtedly the case. Malawi's EPI was stronger than Cameroon's from the very beginning. Services were much more widely available, government commitment to the program was higher, and supervision and training were better. And the integrated PHC approach Malawi favored had clear advantages over the mobile approaches Cameroon generally preferred. In particular, it was much easier to establish frequent (daily) services and to immunize at every contact point based on that approach. However, the regressions used to guide case selection also take this into account by including country fixed effects. Hence, Malawi's predicted and actual vaccination coverage are both higher than Cameroon's (Figure 1). Cameroon's program faced its severest problems when trying to establish a program like Malawi's. It was during the PHC period, when Cameroon attempted to replace mobile teams with vaccinations at fixed centers, that performance was worst. As locally produced documents suggest, the PHC approach may have been better in theory, but in practice, in Cameroon, it was not. Local pushback rooted in the long legacy of mobile services in Cameroon may be why.
The key lesson from the above analysis is: historical legacies matter. Even for new programs, as childhood immunization was in 1974, developing countries draw on their experience and adapt existing institutions to new needs. These differences mean that one-size-fits-all approaches are inappropriate. While they may be well adapted to some cases, they could be counter-productive in others.
Research on the World Bank has shown that international organizations can be flexible in their advice to nation-states, adapting advice to local contexts and concerns (Noy 2017). Similarly, for the EPI, program performance was strongest when global recommendations offered the most flexibility, such as during the first, experimental period and the third, coverage-oriented period. This flexibility enabled global advisers to work with national programs to find the approaches that worked best for them, without demanding an unrealistic degree of expertise in the history of health services in each country.