Policy ForumInfectious Disease

The Stability of Malaria Elimination

Science  22 Feb 2013:
Vol. 339, Issue 6122, pp. 909-910
DOI: 10.1126/science.1229509

When the Global Malaria Eradication Programme (GMEP) was launched in 1955 (1, 2), all malaria-endemic countries outside of Africa were (or would soon be) eliminating malaria (3). The GMEP's design was based on a theory of malaria transmission dynamics and control that has become the standard for malaria elimination decisions today (46). When financial support for the GMEP collapsed in 1969, participating countries were caught at different stages of progress toward elimination (1). Examining their fate in the decades that followed provides a natural experiment that tests the theory. With a rise in funding (7) and renewed interest in eradication (8, 9), there is now a need to revisit the lessons learned from the GMEP. We identify changes in the epidemiology of malaria when elimination is reached that could explain its stability and discuss how this calls for a reassessment of strategies for eradication.

Standard Theory of Malaria Elimination

Malaria elimination involves stopping transmission in a defined region until no parasites remain (6, 10), through implementation of vector control, treatment of infected individuals, and other available interventions. Eradication is elimination at a global scale. Between 1945 and 2010, 79 countries eliminated malaria and 75 (95%) remained malaria-free (11), which shrank the geographical range of malaria (12). A recent review identified 75 resurgent malaria transmission events in 61 countries (13), which included marked increases in malaria incidence in 36 of the 49 countries (73%) that had participated in the GMEP but had failed to reach elimination (11, 14). Resurgent malaria after the GMEP confirms some predictions of the standard theory, but elsewhere, the apparent stability of elimination is counterintuitive and may require a reappraisal of the theory. It suggests that elimination could be a way to solve the Sisyphean problem of malaria (15).

Standard theory for malaria elimination is based on a threshold concept called R0 (4, 5), the expected number of secondary infections produced by each infected human in the absence of control or acquired immunity. R0 is a threshold condition because each case must cause at least one other case to have sustained transmission and endemic malaria. Mosquito aspects of R0 are described by vectorial capacity, a single number that summarizes mosquito population density, longevity, blood-feeding habits, human-feeding preferences, and parasite development rates (16). Human aspects of R0 are dominated by the long infectious period that is sustained by untreated asymptomatic infections lasting 6 to 8 months on average and, sometimes, longer than a year (17). Vector control reduces vectorial capacity; routine use of antimalarial drugs shortens infections; and, in the presence of such interventions, controlled reproductive numbers, denoted RC, describe the reduced number of malaria cases arising from each malaria case (9). Elimination requires reducing RC to <1, such that local transmission will eventually cease.

RC estimates by year (1998–2002) for 27 GMEP countries that eliminated malaria.

Elimination requires reducing RC to <1. Full data spanning 1980–2010, with countries spelled out are available in SM. Mean RC for available years spanning 1980–2010 is shown to the left.

After elimination, malaria importation poses a constant threat, because humans and mosquitoes carry malaria from endemic areas across international boundaries and within countries (18, 19). In such settings, RC describes probabilistically the total number of cases in an outbreak arising from an imported malaria case, so it is useful in planning the end of elimination (10) and the management of imported malaria.

According to standard malaria transmission theory, in order to eliminate malaria, RC must be reduced to below 1, and as long as importation continues, control measures must be maintained such that RC remains below 1 to prevent importation from reestablishing endemic malaria. Countries thus must retain the capability to stop imported malaria cases from transmitting, either by sustaining vector-control measures to keep RC low even after elimination, or through intense surveillance to identify and cure all imported and subsequent infections before they can lead to resumption of transmission (20, 21). If control measures are not sustained at sufficiently high levels, imported malaria could restart endemic transmission.

Malaria resurgence, which was common in those countries that tried, but failed, to eliminate malaria (13, 14), conforms to this theory. The causes of resurgence were poorly documented but, most frequently, resurgence was blamed on the failure to sustain high intervention coverage levels (13). If this pattern held everywhere, elimination would require long-term investments in vector control until eradication is achieved. The best hope for eradication would be a massive campaign with sufficient funding to bring about a coordinated end of malaria transmission such that importation is no longer possible and control efforts can end. If true, eradication would require significant worldwide collaboration.

What Explains the Stability of Elimination?

Elimination countries once had endemic malaria (i.e., R0 > 1), and they have remained malaria-free even though most of these countries have abandoned the elimination era vector-control programs. Without those interventions in place, why has transmission not resumed? Considering that all of these countries face ongoing malaria importation, the stability of elimination requires a closer look.

To assess contemporary malaria importation and to quantify RC in elimination countries, national malaria reporting data were sought from 1980 onward for all countries that participated in the GMEP and eliminated malaria [see supplementary materials (SM)]. In many cases, the data were either sporadically or not collected or not publicly reported. Data were found for 30 countries that eliminated malaria, documenting 249,250 imported malaria cases over 596 country-years versus 4993 introduced (acquired by mosquito transmission from an imported case) or otherwise locally acquired. Formulas from branching theory were used to estimate RC (10). The overall yearly average was RC ≈ 0.04, and ∼85% (506 out of 596) of year-by-country RC estimates were less than 0.01 (see the chart). Because these countries once sustained endemic transmission, this analysis suggests that, for 85% of country-years, transmission is proportionally lower by a factor of more than 100 (i.e., R0/RC > 100). Elimination has become highly stable.

This familiar, but often overlooked, pattern suggests that elimination has substantial advantages over the alternative strategy of minimizing the burden indefinitely without attempting to eliminate malaria, which faces the challenge of financing and distributing interventions to suppress transmission and to manage an ever-present but inapparent threat (15). Understanding the causes of very low RC values in elimination countries is crucial to an assessment of the potential for malaria elimination to become a stable endpoint elsewhere. Important questions remain as to how the 99 countries with endemic malaria should weigh their technical and operational challenges (20, 21) and assess the feasibility of and formulate plans for malaria elimination (22).

There are six main nonexclusive hypotheses for the stability of elimination after top-down control was ceased. (i) R0 might have decreased because of indirect effects of economic development wholly unrelated to disease control, such as changes in mosquito ecology or human demography resulting in reduced human-vector contact; (ii) a decline in RC might have caused economic development, which then reduced R0 (23); (iii) RC is low because countries have maintained vector-control programs (this is likely true only in few countries); (iv) RC is low because country health systems have exerted highly effective control through routine surveillance and case management combined with outbreak control; (v) vector control permanently altered vector ecology and R0; and (vi) the stability of elimination is partly due to travel patterns and the destinations of imported malaria. All six hypotheses could explain some portion of the 100-fold difference in RC to varying degrees in different countries.

The appropriate plan for countries contemplating elimination depends on the causes of the stability. A critical determination is whether elimination was caused by external changes in R0 unrelated to control efforts or whether elimination itself caused the permanent declines in R0 and/or reductions in RC. If malaria elimination requires an external change in R0 (such as exogenous economic development), then control measures will need to be maintained until a transition occurs. If, instead, elimination itself causes stability through economic development and/or other means, then elimination is a highly desirable endpoint and should be aggressively pursued and supported. It is possible that one part of the mechanism by which development produces stability is through improved health systems and surveillance; if so, then improving health systems may be a necessary part of elimination.

Our analysis from post-elimination countries documenting large declines in R0 or RC—whatever the cause—suggests that eradication of malaria differs from smallpox and other vaccine-preventable, directly transmitted diseases (24). It has been argued that eradication is all-or-nothing, that interventions must remain in place until eradication has been achieved, and that there is no such thing as partial success (6). These patterns show that malaria elimination has been a partial success.

The possibility that the complete absence of ongoing malaria transmission can become a highly stable state is relevant for policy because it suggests that, before achieving global eradication, some countries could eliminate, scale back control measures, and rely on their health systems. Projected economic costs of elimination are dominated by management of imported malaria (22, 25), but if elimination is stable, then it could save costs before achieving eradication. Objections to elimination have been drawn, ironically, from the experiences of countries that did not eliminate. This evidence suggests that elimination on path to eradication may be less risky than currently supposed, and that elimination could have substantial advantages over indefinite control.

If malaria elimination helps cause its own stability, then eradication may benefit from regional coordination, but it does not require a globally coordinated campaign. Malaria elimination can proceed like a ratchet, country-by-country and region-by-region culminating in global eradication.

References and Notes

  1. Acknowledgments: Authors received funding from the Bloomberg Family Foundation (D.L.S., G.J.), National Institute of Allergy and Infectious Diseases, NIH (U19AI089674; D.L.S., A.J.T., G.J.), the Bill & Melinda Gates Foundation (49446: D.L.S., A.J.T.; 1032350: A.J.T.; 1013170: J.M.C., R.G.), the RAPIDD program of the Science and Technology Directorate, Department of Homeland Security, and the Fogarty International Center, NIH (D.L.S., A.J.T., S.I.H.), and a Senior Research Fellowship from the Wellcome Trust (095066; S.I.H.). Authors also acknowledge support from the Malaria Elimination Group.


Navigate This Article