Rinderpest Eradication: Appropriate Technology and Social Innovations

See allHide authors and affiliations

Science  14 Sep 2012:
Vol. 337, Issue 6100, pp. 1309-1312
DOI: 10.1126/science.1223805


Rinderpest is only the second infectious disease to have been globally eradicated. In the final stages of eradication, the virus was entrenched in pastoral areas of the Greater Horn of Africa, a region with weak governance, poor security, and little infrastructure that presented profound challenges to conventional control methods. Although the eradication process was a development activity rather than scientific research, its success owed much to several seminal research efforts in vaccine development and epidemiology and showed what scientific decision-making and management could accomplish with limited resources. The keys to success were the development of a thermostable vaccine and the application of participatory epidemiological techniques that allowed veterinary personnel to interact at a grassroots level with cattle herders to more effectively target control measures.

Rinderpest was declared eradicated by the World Organization for Animal Health in May 2011 (1), the second disease ever to be globally eradicated. The disease affected cattle and water buffalo with severe consequences for household livelihoods, and its long history in Europe and Asia was closely intertwined with war, drought, and famine. For millennia it was considered to be the animal disease with the greatest impact on human well-being (2, 3).

Rinderpest is an RNA virus and a member of the genus morbillivirus, which, as the name suggests, includes other notorious pathogens such as peste des petits ruminants (PPR), measles, and canine distemper. The classic presentation of rinderpest was a fulminating diarrhea leading to dehydration and death, often wiping out a family’s cattle in a matter of days. There is no treatment for rinderpest.

A series of control programs in Europe from the 17th century, often credited as the first attempts at organized disease control (4), sought to contain and eliminate rinderpest (5, 6). These efforts were associated with many technical and organizational innovations ranging from the founding of the first veterinary school in Lyon, France, in 1761 to the first use of the rectal thermometer in 1866 (7). Rinderpest’s relatively recent arrival in Africa via the port of Massawa, Eritrea, in the 1880s resulted in 90% mortality in cattle and cloven-hoofed ungulates (artiodactylids) as it swept southward to the Cape, causing widespread famine and ecological change (8) (Fig. 1). These events directly led to the founding of the Office International des Epizooties (OIE) in 1924 (9) and the African Union Interafrican Bureau for Animal Resources (AU-IBAR) in 1951.

Fig. 1

A timeline of major events in the history of rinderpest in Africa from its introduction in 1887, in military cattle brought to Eritrea to feed troops, to the declaration of rinderpest’s eradication in 2011. RP, rinderpest.

Early attempts to control rinderpest in Africa led to technical innovations such as the Koch’s serum/virus method of immunization in 1897 and Edwards’ goat-adapted vaccine in 1941 (7). The first tissue-culture–produced rinderpest vaccine was developed in 1960 by Plowright (10), and its availability inspired Joint Project 15 (JP15), a coordinated effort to eradicate rinderpest from Africa based on mass vaccination. By the 1970s, the range of rinderpest in Africa was reduced to two areas, the Niger Inland Delta in Mali in western Africa and southern Sudan in eastern Africa. With the decline of direct disease impact, the sense of urgency was lost, and the completion of eradication was left to national governments to handle individually through annual vaccination of calves. In the ensuing decade, rinderpest resurged across Africa, again causing up to 90% mortality in its wake. In the late 1980s, a second coordinated effort to eradicate rinderpest, the Pan-African Rinderpest Campaign (PARC), was launched. Initially, as with JP15, mass vaccination targeting entire national populations carried out by teams of government veterinary service personnel was the main approach employed. By the 1990s, rinderpest was largely, but not entirely, confined to remote endemic areas of eastern Africa such as southern Sudan, the Afar region of Ethiopia, the Karamojong region of Uganda, and the Somali rangelands, where insecurity, chronic conflict, and weak governance created challenging environments for vaccine delivery. The eradication effort was drifting toward indefinite disease containment of remote endemic areas (11), with donor fatigue looming on the horizon.

Here, we focus on the evolution of vaccination and surveillance approaches in Africa that enabled the eradication of rinderpest on a tight budget. The approach that ultimately led to success was pragmatic, with surveillance activities mainly identifying infected areas through detection and confirmation of a handful of index cases. Success of interventions was evaluated on the basis of their ability to cause the disappearance of disease. This approach could equally be applied to livestock disease control if a vaccine is available, transmission rates are not overwhelming, and where community stakeholders want eradication and will participate in achieving it and thereby facilitate access to remote areas. We explore the potential of applying these principles to other livestock diseases in less-developed countries.

Development of the Thermostable Rinderpest Vaccine

The early rinderpest vaccines were superseded by Plowright’s tissue-culture–produced vaccine. This vaccine was successful because a single tissue culture infectious dose conferred lifelong protective immunity to all strains of the virus; the vaccine virus had no reported adverse reactions, and it was inexpensive and relatively easy to produce (12). However, its major shortcoming was that it required strict, expensive, and logistically difficult refrigeration (known as a “cold chain”). Hence, in 1981, the development of a thermostable rinderpest vaccine was identified as a research priority (13).

The thermostability of the Plowright vaccine was enhanced by improved production and lyophilization techniques (14) and adapted to the Vero cell line to further simplify production. Modified lyophilization resulted in a vaccine with a shelf life of more than 8 months at 37°C, which was sufficient to recommend the vaccine for use in the field for up to 30 days without a cold chain. Registration of the thermostable product, Thermovax, was based on an existing vaccine virus in wide use, and the new vaccine was in commercial-scale production and available for purchase by 1992, just 4 years from the start of the research (14, 15).

Technical Innovation as a Driver of Institutional Change

Now a vaccine was available that could be delivered on foot, by animal transport, or by bicycle and sent to remote areas. Vaccination with thermostable vaccine in marginalized communities became a cornerstone of the end-game eradication strategy (16, 17) and required a rethinking of service delivery that allowed new partnerships, such as community animal health programs (18), to achieve full vaccination coverage.

Institutional change was accomplished by public-private-community partnerships in which veterinarians delivered services through community members using approaches adapted to local institutions. The local intermediaries, called community-based animal health workers (CAHWs), were willing to travel on foot and able to work in remote areas (Fig. 2). They were trained to vaccinate livestock (19), and thus more remote communities were able to implement vaccination campaigns. Not surprisingly, some members of both official and private veterinary services saw this new paradigm as a threat to their livelihoods and opposed the implementation of such field programs. Incentives that were mutually favorable to the communities and professionals had to be carefully crafted to allow communities to vaccinate with the shared oversight of the veterinarians and community elders (19). In this way, rinderpest vaccination was integrated into broad community-based animal health programs that addressed a range of animal health priorities identified by local communities (20, 21). Local knowledge of the sizes of cattle herds, their location, seasonal movement patterns, and the optimum time for vaccination was essential for planning and extending effective rinderpest vaccination coverage to areas not covered by public programs. Because the CAHWs had participated in planning vaccination, and the thermostable vaccine allowed flexibility, the CAHWs could adjust their work to allow for changes in cattle movement patterns or unexpected events, in sharp contrast to inflexible vaccination campaigns that resulted in poor vaccination coverage. Community-based vaccination programs thus achieved herd immunity levels greater than 80% and were at least as effective as the best public veterinary service programs in Africa conducted in more accessible areas (19, 22, 23).

Fig. 2

Community animal health workers equipped with thermostable vaccine eradicated rinderpest from the chronically insecure regions of East Africa. A Karamojong community animal worker, Tom Olaka, is shown vaccinating his community’s cattle with thermostable rinderpest vaccine in 1994. Olaka also identified and reported the last outbreak of rinderpest from this insecure area and enabled the completion of the eradication of rinderpest from Uganda. [Photo by Christine Jost]

Participatory Approaches to Gathering Epidemiological Intelligence

Before the PARC, surveys of animal health knowledge systems of pastoral communities uncovered a rich source of information for planning animal health activities (24). Participatory rural appraisal evolved into a more comprehensive and effective tool kit of methods for understanding local knowledge that could be integrated into the design of the community programs. The advent of community-based rinderpest vaccination activities in PARC brought mainstream animal health personnel into direct contact with community health knowledge and methods for its study. The outcome was a real-time field intelligence approach to the collection of epidemiological information to guide decision-making on rinderpest control options.

Pastoral communities have a detailed understanding of the presentation, distribution, epidemiology, and history of key livestock diseases in their area. (25, 26). Livestock owners provided information that led to the identification of several of the final outbreaks of rinderpest where conventional surveillance activities had failed to disclose disease (27). In the later stages of disease eradication, participatory disease surveillance was an important tool in the recognition of freedom from rinderpest (28). The World Organization for Animal Health guidelines for recognition of disease freedom focused on randomized serosurveillance as a final objective check to validate eradication, but due to its high cost, it was usually only implemented after there was a consensus among program advisors that the disease had been eradicated. To the authors’ knowledge, serosurveillance never identified any occult circulation of virus, thus validating pastoralists’ participation in the eradication strategy.

This mainstream recognition of livestock owner knowledge and the development of participatory epidemiology and surveillance have now been extended to the control of other diseases, such as PPR (29), Rift Valley fever, highly pathogenic avian influenza (30, 31), and foot-and-mouth disease (FMD) (32).

Epidemiological Targeting

Throughout the 1980s and into the 1990s, rinderpest programs used annual, pulsed, mass vaccination programs, where the goal was to cover entire national populations and surveillance was based on passive disease-reporting systems. Rather than cessation of virus circulation, mass vaccination programs were judged on the numbers of vaccinations performed relative to overall national populations, which were often inaccurately defined. Populations that were difficult to access were repeatedly overlooked for decades, resulting in highly variable herd immunity and reservoirs of virus persistence.

Mathematical modeling of measles has shown that maximal vaccination impact is achieved by focusing resources on those populations with the highest transmission rates (33) and with sizes larger than the community size required for sustained transmission (34). Building on this work, stochastic state transition models explored the transmission dynamics of the two lineages of rinderpest existing in Africa. This allowed estimation of the vaccine coverage levels needed to interrupt transmission and the minimum population size required to sustain transmission, and predictions of the impact of focused vaccination at different levels of coverage (35). The analysis made use of serological data from selected endemic communities where vaccination had not been practiced. The estimates of the mean basic reproductive number (R0), the number of new cases resulting from one infected animal introduced into a fully susceptible population, were 4.6 for lineage 1 (Sudan) and 1.5 for lineage 2 (Somali), which implied that the herd immunity thresholds for the interruption of transmission were 77 and 33%, respectively. The modeling indicated that the community size required for sustained transmission was in the order of 200,000 head for both lineages. This information reinforced the view that seeking greater than 80% immunity was adequate for remote Sudanese livestock in excess of 200,000 head. Managers were skeptical that a 33% vaccination rate could interrupt transmission of rinderpest in the Somali ecosystem, but rinderpest was conclusively eradicated from the Somali ecosystem with cattle herd immunity levels that probably never exceeded 50%, despite best efforts to achieve 80%.

Because available serologic tests could not differentiate antibodies induced by vaccination from those resulting from natural infection, serological approaches to the estimation of the effective reproductive number, or Re (the reproductive number in populations where mitigation measures have changed susceptibility or effective contact rates), were of little use to monitor the effectiveness of vaccination strategies. Likewise, it was not possible to use analysis of individual outbreaks for the estimation of Re, because rinderpest outbreaks developed rapidly in remote and insecure areas in mobile cattle populations. Rather, local intelligence of rinderpest infections gathered through participatory surveillance was the most successful approach to assessing the impact of vaccination methods and monitoring strategic progress toward eradication.

Livestock contact patterns are largely the result of human interactions. The use of participatory epidemiology and community animal health services heightened the awareness of how cultural differences and social relationships among pastoralists contributed to the patterns of rinderpest transmission (35) and influenced the transmission links between populations. However, a concerted effort using both field and modeling results was often required to convince decision-makers to adopt risk-based approaches (35). Effective coordination mechanisms—such as the Global Rinderpest Eradication Program at the United Nations Food and Agriculture Organization (FAO) and the Somali Ecosystem Rinderpest Eradication and Coordination Unit at the AU-IBAR—facilitated the uptake of targeted approaches. Once this focused approach was adopted in countries such as Ethiopia, Uganda, and Sudan, no infections were detected after 2 years.

Ethiopia provided a textbook example of the enhanced impact of targeted vaccination. From April 1989 to July 1990, at the peak of mass vaccination activity, 21.4 million vaccinations were performed. This represented up to 95% coverage in accessible areas but only 69% coverage nationally (36). Remote areas were not vaccinated year after year and acted as endemic reservoirs of infection where virus continually circulated. In 1993, based on disease intelligence, the country was divided epidemiologically into free, epidemic, or endemic zones, and vaccination was limited to endemic zones and some surrounding areas at high risk. The annual national vaccination target was reduced to less than 3 million, but these were among the most difficult cattle to reach in the country. Community animal health workers played a major role in reaching these remote cattle and eradicating rinderpest from this and other challenging areas of East Africa (Fig. 2). The result of the shift in strategic focus was that after decades of struggle to control rinderpest, the disease was not detected in Ethiopia after October 1995.

Comparisons Across Disease Eradication Programs

The rinderpest eradication effort succeeded even though the classic animal disease control options of movement control and culling of infected animals were not available for economic and social reasons. The core strategy that was finally successful identified infected communities and then interrupted transmission in those communities with thermostable vaccine delivered by local participants.

Pathogens with a low R0 are generally easier to control because each case results in only a limited number of secondary cases. A low R0 is why severe acute respiratory syndrome was much easier to contain than measles or FMD outbreaks. Rinderpest had an R0 of 4.6 for lineage 1, which meant that even virulent strains only required 77% herd immunity for elimination to occur and helped to explain why targeted vaccination alone was so effective. Epidemiological indicators other than R0 that should be considered in assessing the difficulty associated with controlling a disease and selecting control methods are the length of the generation interval of the infection between cases in chains of transmission and the life expectancy of hosts. The average life expectancy of a cow in Africa is on the order of 5 years; by adopting a series of annual vaccination cycles, this life span allowed the rinderpest program to build herd immunity. In contrast, although highly pathogenic avian influenza has a low R0, high poultry population turnover and the short duration of vaccine-induced immunity makes it impossible to build adequate flock immunity in free-ranging village poultry.

An important aspect of the rinderpest story is that social change altered the efficacy of control options over time. The mass vaccination approach that worked fairly well across Africa during the time of JP15 did not work in areas affected by chronic conflict or failed infrastructure. Similarly, contagious bovine pleuropneumonia (CBPP) was eliminated from large parts of Africa in the colonial era using mass vaccination with a moderately effective vaccine and draconian movement control. With modern pluralistic governance, strict movement control is no longer feasible and CBPP has reentered many areas.

As a result of the rinderpest success, the international animal health community has become interested in adopting coordinated approaches to progressive control of disease in small ruminants. Small ruminants have a life expectancy on the order of 1 year, are traded heavily, and constitute an important disease transmission challenge. PPR, caused by a virus related to rinderpest, shares many attributes that make it a good prospect for eradication. As a disease of sheep and goats, PPR has a direct impact on the food security and well-being of the most impoverished communities. Against this backdrop, the Executive Council of the African Union and the participants of the FAO Global Rinderpest Eradication Symposium have urged that appropriate programs for the progressive control of PPR be undertaken (37). Strategies have been put forward to coordinate the current investment in PPR control as a progressive control program with the ultimate objective of eradication (38, 39). The development of more effective vaccination models that address the challenges of immunizing mobile populations of rapidly reproducing small ruminants is a major research objective. This time around, the research must integrate institutional and technical innovations from the start.

Similar considerations apply to FMD virus, for which the goal of elimination through a strategy of progressive control is increasingly discussed. The loss of production and export markets resulting from the presence of FMD make this disease a priority for commercially oriented livestock producers and national economies. A conservative estimate of 4.6 was given for R0 during the early stages of the FMD outbreak in the United Kingdom in 2001 (40), and in certain circumstances, such as intensive dairy farming in the Kingdom of Saudi Arabia, R0 might even exceed 70 (41), implying herd immunity thresholds of 98% and 99%, respectively. In practice, such herd immunity is not achievable. On the other hand, although FMD is a concern to almost all livestock farmers, many traditional producers find that they can cope with periodic FMD-induced production loss. Pastoralists and small-scale farmers place a much higher priority on the risk of diseases that cause catastrophic losses and failure of household livelihoods. Thus, there are serious issues of incentives for participation in FMD elimination among key livestock-keeping stakeholder groups who prioritize the well-being of their families over national economies.

In some ways, the rinderpest approach—searching for infected populations, followed by the application of intensive vaccination—was similar to that used for smallpox, but in the latter case the search extended to identifying and isolating infected individuals. The necessary distinction in value between human lives and that of their livestock shapes the control options available to the medical and veterinary professions and is often the elephant in the room in discussions on “one health” approaches based on integration of human, animal, and ecosystem health activities.

Lessons for the Future

The primary lesson was that the vision of a rinderpest-free world was both tangible and broadly valued by stakeholders. This easily understood goal stimulated the imagination and induced stakeholders to risk change in fundamental areas of animal health practice and human behavior.

The second lesson was that, despite the advent of the thermostable vaccine, this new technology alone could not solve institutional issues. The reluctance of researchers to champion the process of social change accompanying the deployment of new technologies might in part explain why many products of research fail to have impact. The technical research to develop a thermostable rinderpest vaccine required 2 years to complete, but the social innovation to capture the benefit required more than a decade. Ultimately, a clear institutional understanding that recognized the legitimate needs of each stakeholder group and the power relationships between groups was needed before the thermostable vaccine could achieve its full potential. An institutional analysis at the outset might have accelerated change.

The third lesson concerned the optimal gathering and use of information in decision-making. Resources for animal health surveillance in the developing world are limited. Conventional surveillance data for rinderpest were insufficient to reliably indicate the presence of disease. The use of pastoralists’ knowledge in surveillance was found to be critical for targeting eradication efforts. Ideally, a serological test to distinguish between vaccinal and natural immunity would have made epidemiological studies more robust but was not at all a prerequisite for success.

Internationally, the call is out to apply the lessons of rinderpest eradication to the notorious morbillivirus PPR as the distribution and impact of this infection has extended to most of Africa and across Asia to China.

References and Notes

  1. Acknowledgments: The eradication of rinderpest was a global undertaking that received support from a wide range of national governments and international agencies. Principal support for the later stages of eradication was provided by the European Union. The U.S. Agency for International Development supported research leading to the thermostable rinderpest vaccine. Preparation of this manuscript was supported by the International Livestock Research Institute and Tufts Cummings School of Veterinary Medicine.
View Abstract

Navigate This Article