Background
Malaria elimination is rapidly becoming a tempting alternative to malaria control in many malaria-endemic African countries[
1]. This paradigm shift is largely a consequence of effective malaria control initiatives, which have decreased markedly the malaria burden across the African continent[
2]. Countries along the southernmost fringe of malaria transmission, such as Angola, Botswana, Swaziland and South Africa, already meet the WHO pre-elimination criteria and have therefore been earmarked for malaria elimination by 2020. Swaziland embarked on an elimination campaign in 2011[
3], while South Africa is preparing to implement its elimination strategy during the 2012/2013 malaria season.
Currently in South Africa, malaria is restricted to low-altitude border regions (below 1,000 m above sea level) of three provinces, namely Limpopo, Mpumalanga and KwaZulu-Natal, with limited transmission along the Molopo and Orange Rivers in the North West and Northern Cape Provinces, respectively. Approximately 10% (4.9 million) of South Africa’s total population resides in a malaria risk area[
4], with the predominant malaria parasite,
Plasmodium falciparum, primarily transmitted by the
Anopheles arabiensis mosquito vector. Malaria transmission is meso-endemic, occurring between September and May but peaking in March.
Following the last major epidemic in 2000, where more than 60,000 malaria cases were reported, a two-pronged intervention, focussing on both the vector and parasite, was implemented across all three malaria-endemic provinces. Indoor residual spraying (IRS) formed the principal vector control measure while timely diagnosis and effective treatment with artemisinin-based combination therapy (ACT) were used to control the parasite. These effective, well-structured, sustainable control strategies have resulted in marked reductions in the malaria burden, to the extent that the current malaria incidence in South Africa is less than one case per 1,000 population at risk.
This seemly low malaria incidence has prompted international and governmental organizations to call for the urgent adoption and implementation of elimination agenda by South Africa. Unfortunately these calls have generally been based solely on annual incidence data rather than a rigorous interrogation of the available scientific data. As the consequences of failure are likely to be costly in monetary terms and more importantly with respect to the loss of human life, the decision to move from control to elimination should not be taken lightly. It is therefore essential the appropriateness, timing as well as technical, operational and financial feasibility of implementation are thoroughly assess prior to embarking upon an elimination programme[
5].
The World Health Organisation (WHO) recommends major programmatic reorientation occurs when transitioning from malaria control to elimination. High coverage interventions must become more geographically targeted while laboratory and clinical services together with case reporting and surveillance are substantially up-scaled[
6]. This study aimed to determine the readiness, appropriateness as well as technical and operational feasibility of implementing an elimination programme in present day South Africa. The factors assessed included numbers of malaria cases and deaths, IRS coverage rates, levels of resistance to insecticides used for IRS, vector distribution, prevalence of anti-malarial resistance markers and the efficiency of the malaria information system (MIS). It is hoped data from this study will be used to inform the current South African malaria elimination agenda.
Discussion
Malaria has long been regarded as a major public health burden in South Africa, affecting vast regions of the country. During the late 1920s, malaria epidemics spread as far south as Port St Johns[
22], with more than 20,000 deaths reported in a single season[
23]. To address the malaria threat, two interventions, residual spraying with insecticides (initially DDT and later pyrethroids) and surveillance were introduced in 1946. These interventions, particularly IRS with DDT resulted in malaria being eliminated from large areas of the country while reducing it to low levels in the three malaria-endemic provinces. Seasonal focal outbreaks associated with favourable climatic conditions and migrant movement, still occurred[
24]. The impact of IRS on malaria distribution and cases numbers in South Africa was so dramatic because IRS operations eliminated an extremely efficient malaria vector
An. funestus and greatly reduced the density and distribution of the main malaria vector
An. arabiensis. Despite the existence of a very effective, integrated malaria control programme comprising effective vector control and good case management, malaria continues to linger in restricted areas within the malaria-endemic provinces.
Embarking on a malaria elimination agenda is costly, time consuming and extremely labour intensive. It is therefore imperative to ensure the timing of an elimination intervention is appropriate and that all necessary resources are in place as failure would have a deleterious effect on the gains made against malaria transmission. Following 10 years of intensive, integrated, malaria control interventions, malaria incidence in South Africa has declined markedly. The three endemic provinces are at different stages in the malaria elimination continuum, proposed by WHO[
6].
Nevertheless, many malaria-endemic districts now meet the incidence limit required to move into the malaria elimination phase. Unfortunately the data from this study show the decrease in malaria incidence has not been uniform across all districts in the malaria-endemic provinces. KwaZulu-Natal is the only province in which all the districts have achieved the minimum elimination incidence level. Therefore only in KwaZulu-Natal can all the efforts be focused solely on malaria elimination. The picture is very different and rather more complex in the remaining two malaria-endemic provinces.
Both Limpopo and Mpumalanga have districts at each of the four categories on the elimination continuum. The border areas of these two provinces are most affected by malaria, with Mozambique largely influencing the malaria incidence in Mpumalanga and the transmission intensity in Zimbabwe impacting on the malaria incidence in Limpopo. Given the wide range of malaria incidence in districts of Limpopo and Mpumalanga, intervention efforts focussing on control and pre-elimination would be most appropriate in these provinces.
In addition to highlighting the complex malaria status pattern, this study found several other areas where information, essential for the formulation of an effective elimination strategy, was lacking. The first area of concern is vector control. Although IRS operations have been rigorously adhered to, the routine assessment of vector distribution and efficacy evaluation of insecticides used for IRS operations have not been conducted in the three malaria-endemic provinces. As no current vector distribution data are available, it is not be possible to meet the elimination requirement of targeted vector control measures.
This issue is further compounded by the lack of current insecticide efficacy data. Data available are over a decade old and more than likely no longer valid given the changes in insecticide pressure over the past 10 years. The emergence of pyrethroid-resistant vectors, a major factor contributing to the 1999/2000 malaria epidemic, led to the re-introduction of DDT-based IRS operations in 2000. This policy change was crucial to controlling the malaria epidemic, particularly in KwaZulu-Natal[
25]. While no epidemics have been reported since 2000, DDT resistant
An. arabiensis populations were found in KwaZulu-Natal in 2003[
18]. Unfortunately no new studies on the emergence and or spread of insecticide resistance have been conducted in South Africa.
In the past five years, all malaria-endemic provinces have reported spray coverage rates greater than 80%, the level of coverage recommended by the WHO to achieve maximum impact. Since the IRS campaign is not achieving the desired outcome (in terms of disease burden reduction) in all provinces, it is possible there are flaws in IRS programme. Possible confounders include the use of incorrect insecticides, poor drug administration, importation of malaria or general programme failures. Further, data for IRS coverage prior to 2005 are not readily available making the assessment of IRS impact prior to 2005 virtually impossible. Although, at present the SIS appears to be functioning effectively, there is a lag period in case reporting which needs to be improved upon in order for case reporting to be made in real-time.
Despite the existence of the MIS in all three malaria-endemic provinces, data collected across all three provinces are not uniform. Even more concerning is the fact that essential data such as quantities of Coartem® dispensed, numbers of RDTs used and numbers of positive smears/RDTs obtained are not being captured by the MIS in any of three provinces. Probable sources of infection were routinely captured by the MIS in Mpumalanga, while it appeared more ad hoc in Limpopo and KwaZulu-Natal. Active case detection data were incomplete across all three provinces. These issues need to be addressed urgently as the country has already embarked on an elimination agenda.
Routine anti-malarial resistance marker surveillance across all three malaria-endemic provinces revealed a high prevalence of
P. falciparum parasites carrying mutant alleles associated with SP and chloroquine resistance. This finding was surprising given the limited current use of both these drugs in South Africa. Chloroquine was replaced by SP as first line treatment in 1991 following reports of growing chloroquine resistance. Sulphadoxine-pyrimethamine remained effective for approximately 10 years before it was phased out of use in KwaZulu-Natal in 2000 and in Mpumalanga and Limpopo in 2003 owing to high levels of SP treatment failures. The resistance data generated by this study supports the South African Department of Health’s decision to replace SP with the ACT, Coartem
®. On a cautionary note however, given that tolerance to Coartem
® has been detected along the Thailand-Myanmar border area[
26], it is essential more rigorous sentinel site-based drug resistance surveillance is conducted. Currently, only a small subsets of all positive RDTs are surveyed for the presence of markers associated with lumefantrine resistance. Greater efforts must be made to improve surveillance in this regard.
For the next six years at least (until 2018), it is vital that the South African Government continues to implement its hugely successful malaria control programme while encouraging the creation and roll out of effective control programmes in neighbouring countries, a ground breaking example being the Lubombo Spatial Development Initiative[
27]. This regional collaboration between Swaziland, Mozambique and South Africa, aimed at strengthening vector control and case management in these countries, contributed significantly to the substantial reductions in malaria case numbers in southern Mozambique as well as the border areas of KwaZulu-Natal and Mpumalanga. Over the 10 year lifespan of this initiative, malaria prevalence in southern Mozambique declined from over 60% pre-intervention to below 15% by 2011[
28].
Malaria elimination in South Africa is only possible if elimination is also successfully occurring in neighbouring countries, particularly Mozambique and Zimbabwe. The current social, political and economic upheavals in Zimbabwe have contributed to the breakdown of the Zimbabwean malaria control programme. This has resulted in a large number of cases imported from Zimbabwe being reported in Limpopo, preventing the province from attaining a pre-elimination classification.
None the less, South Africa has embarked on programme re-orientation in areas fulfilling the WHO elimination criteria. Vector based interventions, like IRS operations are being scaled down, limited to areas of confirmed local transmission. In malaria endemic areas no longer under control it is hoped that intensive surveillance will elicit rapid outbreak responses in the event malaria cases (local or imported) area detected. The extreme risk associated with re-orientation is severe malaria epidemics in areas were no malaria has been detected for a number of years, particularly if the response alert thresholds are not sensitive enough and/or surveillance not rigorous enough. At present in South Africa there is a great paucity of essential information especially pertaining to the current status of tools being used to implement the elimination agenda[
29]. This information is vital for determining the success or failure of the elimination effort and must be collected as soon as possible.
Conclusions
Data generated by this study suggest that malaria elimination in South Africa is feasible. The IRS programme has successfully decreased malaria transmission, both spatially and temporally. The malaria case fatality rate has been reduced to extremely low levels through the implementation of effective case management protocols. However, there are many gaps in knowledge that must be filled before the country can successfully transition into the WHO defined malaria elimination phase for a sustained period of time. Surveillance of malaria morbidity and mortality, the delivery and use of diagnostic kits and the dispensing and use of ACT needs to be improved. The surveillance for insecticide resistance must become routine across all malaria endemic provinces, while drug resistance surveillance needs to be strengthened in light of decreasing case numbers. The issue of migration and malaria must to be addressed as a matter of priority, particularly as a very large proportion of the malaria cases reported in South Africa are imported from neighbouring countries and the rest of Africa.
It is of paramount importance that financing to achieve and sustain elimination is secured. Underfunding could result in the successful rebounding of malaria as seen in countries like Sri Lanka and India. Crude estimates suggest that the annual costs of achieving elimination in South Africa would be approximately USD6 per person at risk. However, this figure excludes the cost of sustaining elimination which needs to be urgently researched and factored into elimination cost estimates.
These gaps in knowledge and processes are not insurmountable and can be overcome relatively easily with the appropriate research and development of surveillance tools. Malaria elimination in South Africa by 2018 is conceivable provided there is adequate sustainable financing, the appropriate operational research is conducted and there is continued political commitment at all levels, both nationally and internationally.
Competing interests
The authors declare they have no competing interests.
Authors’ contributions
RM reviewed the literature, defined the study and drafted the manuscript. JR reviewed the literature, drafted and edited the manuscript and conducted the molecular analyses. RM, JR, IS and NM performed the statistical analysis. NM drafted and edited the manuscript and generated the spatial data and outputs. IS generated the IRS data. PK, DM, AM and ER were involved in data collection and drafting of the manuscript. All authors read and approved the final draft of the manuscript.