Abstract
Since the year 2000, a concerted campaign against malaria has led to unprecedented levels of intervention coverage across sub-Saharan Africa. Understanding the effect of this control effort is vital to inform future control planning. However, the effect of malaria interventions across the varied epidemiological settings of Africa remains poorly understood owing to the absence of reliable surveillance data and the simplistic approaches underlying current disease estimates. Here we link a large database of malaria field surveys with detailed reconstructions of changing intervention coverage to directly evaluate trends from 2000 to 2015, and quantify the attributable effect of malaria disease control efforts. We found that Plasmodium falciparum infection prevalence in endemic Africa halved and the incidence of clinical disease fell by 40% between 2000 and 2015. We estimate that interventions have averted 663 (542–753 credible interval) million clinical cases since 2000. Insecticide-treated nets, the most widespread intervention, were by far the largest contributor (68% of cases averted). Although still below target levels, current malaria interventions have substantially reduced malaria disease incidence across the continent. Increasing access to these interventions, and maintaining their effectiveness in the face of insecticide and drug resistance, should form a cornerstone of post-2015 control strategies.
This is a preview of subscription content, access via your institution
Access options
Subscribe to this journal
Receive 51 print issues and online access
$199.00 per year
only $3.90 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
References
World Health Organization. World Malaria Report 2014 (World Health Organization, 2014)
Roll Back Malaria Partnership/World Health Organization. Global Malaria Action Plan 1 (2000–2015) (World Health Organization, 2008)
World Health Organization. Global Technical Strategy for Malaria 2016–2030 (World Health Organization, 2015)
Roll Back Malaria Partnership/World Health Organization. Action and Investment to Defeat Malaria 2016–2030 (World Health Organization on behalf of the Roll Back Malaria Partnership Secretariat, 2015)
Rowe, A. K. et al. Caution is required when using health facility-based data to evaluate the health impact of malaria control efforts in Africa. Malar. J. 8, 209 (2009)
Chizema-Kawesha, E. et al. Scaling up malaria control in Zambia: progress and impact 2005–2008. Am. J. Trop. Med. Hyg. 83, 480–488 (2010)
Lim, S. S. et al. Net benefits: a multicountry analysis of observational data examining associations between insecticide-treated mosquito nets and health outcomes. PLoS Med. 8, e1001091 (2011)
Lengeler, C. Insecticide-treated bed nets and curtains for preventing malaria. Cochrane Database Syst. Rev. 2, CD000363 (2004)
Giardina, F. et al. Effects of vector-control interventions on changes in risk of malaria parasitaemia in sub-Saharan Africa: a spatial and temporal analysis. Lancet Glob. Health 2, e601–e615 (2014)
Hay, S. I. et al. A world malaria map: Plasmodium falciparum endemicity in 2007. PLoS Med. 6, e1000048 (2009)
Gething, P. W. et al. A new world malaria map: Plasmodium falciparum endemicity in 2010. Malar. J. 10, 378 (2011)
Noor, A. M. et al. The changing risk of Plasmodium falciparum malaria infection in Africa: 2000–10: a spatial and temporal analysis of transmission intensity. Lancet 383, 1739–1747 (2014)
Patil, A. P. et al. Defining the relationship between Plasmodium falciparum parasite rate and clinical disease: statistical models for disease burden estimation. Malar. J. 8, 186 (2009)
Griffin, J. T., Ferguson, N. M. & Ghani, A. C. Estimates of the changing age-burden of Plasmodium falciparum malaria disease in sub-Saharan Africa. Nat. Commun. 5, 3136 (2014)
Smith, T. et al. Ensemble modeling of the likely public health impact of a pre-erythrocytic malaria vaccine. PLoS Med. 9, e1001157 (2012)
Hay, S. I. et al. Estimating the global clinical burden of Plasmodium falciparum malaria in 2007. PLoS Med. 7, e1000290 (2010)
Diggle, P. & Ribeiro, P. Model-based Geostatistics (Springer, 2007)
Weiss, D. J. et al. Re-examining environmental correlates of Plasmodium falciparum malaria endemicity: a data-intensive variable selection approach. Malar. J. 14, 68 (2015)
Smith, D. L., Guerra, C. A., Snow, R. W. & Hay, S. I. Standardizing estimates of the Plasmodium falciparum parasite rate. Malar. J. 6, 131 (2007)
Cameron, E. et al. Defining the relationship between infection prevalence and clinical incidence of Plasmodium falciparum malaria. Nat. Commun. 6, 8170 (2015)
Wenger, E. A. & Eckhoff, P. A. A mathematical model of the impact of present and future malaria vaccines. Malar. J. 12, 126 (2013)
Battle, K. E. et al. Global database of Plasmodium falciparum and P. vivax incidence records from 1985–2013. Sci. Data 2, 150012 (2015)
WorldPop. Gridded population distributions. http://www.worldpop.org.uk (2015)
Cibulskis, R. E., Aregawi, M., Williams, R., Otten, M. & Dye, C. Worldwide incidence of malaria in 2009: estimates, time trends, and a critique of methods. PLoS Med. 8, e1001142 (2011)
Liu, L. et al. Global, regional, and national causes of child mortality in 2000–13, with projections to inform post-2015 priorities: an updated systematic analysis. Lancet 385, 430–440 (2015)
GBD 2013 Mortality and Causes of Death Collaborators. Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 385, 117–171 (2015)
Acknowledgements
The authors acknowledge assistance from M. Renshaw in providing information from the Roll Back Malaria Harmonization Working Group Programmatic Gap Analysis and other guidance in the interpretation of our results. We thank members of the Roll Back Malaria Monitoring and Evaluation Reference Group and the World Health Organization Surveillance Monitoring and Evaluation Technical expert Group for their feedback and suggestions. We thank C. Burgert of the DHS (Demographic and Health Surveys) Program for her assistance with DHS Survey access and interpretation. P.W.G. is a Career Development Fellow (no. K00669X) jointly funded by the UK Medical Research Council (MRC) and the UK Department for International Development (DFID) under the MRC/DFID Concordat agreement and receives support from the Bill and Melinda Gates Foundation (BMGF; nos OPP1068048, OPP1106023). These grants also support E.C., S.B., B.M., U.D., D.J.W., D.B. and A.H. The Swiss TPH component was supported through the project no. OPP1032350 funded by the BMGF. D.L.S. is funded by the BMGF (OPP1110495). S.I.H. is funded by a Senior Research Fellowship from the Wellcome Trust (no. 095066), which also supports K.E.B., and grants from the BMGF (nos. OPP1119467, OPP1106023 and OPP1093011). S.I.H. and D.L.S. also acknowledge funding support from the RAPIDD program of the Science & Technology Directorate, Department of Homeland Security, and the Fogarty International Center, National Institutes of Health. J.T.G. is funded by an MRC Fellowship (no. G1002284). E.A.W. and P.A.E. are funded by the Global Good Fund.
Author information
Authors and Affiliations
Contributions
Conceived of and designed the research: P.W.G. and S.B. Drafted the manuscript: P.W.G. and S.B. Drafted the Supplementary Information: S.B., D.J.W., E.C., D.B., U.D., B.M. Prepared data: S.B., D.J.W., B.M., U.D., K.B., C.L.M., A.H., A.B., J.Y., T.P.E. Conducted the analyses: S.B., D.J.W., E.C., D.B., C.A.F., M.L., R.E.C. Supported the analyses: P.A.E., E.A.W., O.B., M.A.P., T.A.S., J.T.G., C.A.F., M.L., F.L., D.L.S. Supported interpretation and policy contextualization: S.B., A.B., T.P.E., J.Y., C.A.F., M.L., J.M.C., C.L.J.M., D.L.S., S.I.H., R.E.C., P.W.G. All authors discussed the results and contributed to the revision of the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Extended data figures and tables
Extended Data Figure 1 Schematic overview of main input data, model components, and outputs.
Each component is detailed in the Supplementary Information.
Extended Data Figure 2 Fitted function representing effect of ITNs.
Curves illustrate the predicted effect of ITNs as a function of coverage (five example coverage levels are shown, specified as mean coverage over preceding 4-year period) and baseline transmission. The baseline PfPR is shown on the horizontal axis and the suppressed PfPR given the ITN coverage level shown on the vertical axis. The diagonal line (representing zero ITN effect) is shown in black, and parameter uncertainty around each ITN effect line is illustrated by the semi-transparent envelopes. Results shown are derived from a Bayesian geostatistical model fitted to n = 27,573 PfPR survey points; n = 24,868 ITN survey points; n = 96 national survey reports of ACT coverage; n = 688 country-year reports on ITN, ACT and IRS distribution by national programs; and n = 20 environmental and socioeconomic covariate grids.
Extended Data Figure 3 Changing incidence rate by country, 2000–2015.
Estimated country-level rates of all-age clinical incidence are shown for 2000 and 2015. For Sudan and South Sudan, we used the post-2011 borders throughout the time period to allow comparability. Results shown are derived from a Bayesian geostatistical model fitted to n = 27,573 PfPR survey points; n = 24,868 ITN survey points; n = 96 national survey reports of ACT coverage; n = 688 country-year reports on ITN, ACT and IRS distribution by national programs; n = 20 environmental and socioeconomic covariate grids; and n = 30 active-case detection studies reporting P. falciparum clinical incidence.
Extended Data Figure 4 Decline in infection prevalence attributable to main malaria control interventions.
a–d, Each map shows absolute decline in PfPR2–10 between 2000 and 2015 within areas of stable transmission attributable to the combined effect of ITNs, ACTs, and IRS (a); and the individual effect of ITNs (b); ACTs (c); and IRS (d). Note that the colour scaling differs between the panels. Results shown in all panels are derived from a Bayesian geostatistical model fitted to n = 27,573 PfPR survey points; n = 24,868 ITN survey points; n = 96 national survey reports of ACT coverage; n = 688 country-year reports on ITN, ACT and IRS distribution by national programs; and n = 20 environmental and socioeconomic covariate grids. Maps in this figure are available from the Malaria Atlas Project (http://www.map.ox.ac.uk/) under the Creative Commons Attribution 3.0 Unported License.
Supplementary information
Supplementary Information
This file contains Supplementary Text and Data, Supplementary Figures 1-5, Supplementary Tables 1-4 and Supplementary References. (PDF 13325 kb)
PowerPoint slides
Rights and permissions
About this article
Cite this article
Bhatt, S., Weiss, D., Cameron, E. et al. The effect of malaria control on Plasmodium falciparum in Africa between 2000 and 2015. Nature 526, 207–211 (2015). https://doi.org/10.1038/nature15535
Received:
Accepted:
Published:
Issue date:
DOI: https://doi.org/10.1038/nature15535
This article is cited by
-
Epidemiological behaviour and interventions of malaria in Niger, 2010–2019: a time-series analysis of national surveillance data
Malaria Journal (2024)
-
The effect of COVID-19 on malaria cases in Zambia: a mixed effect multilevel analysis
Malaria Journal (2024)
-
Mosquito odour-baited mass trapping reduced malaria transmission intensity: a result from a controlled before-and-after intervention study
BMC Medicine (2024)
-
LLIN Evaluation in Uganda Project (LLINEUP)–effects of a vector control trial on Plasmodium infection prevalence and genotypic markers of insecticide resistance in Anopheles vectors from 48 districts of Uganda
Scientific Reports (2024)
-
3D video tracking analysis reveals that mosquitoes pass more likely through holes in permethrin-treated than in untreated nets
Scientific Reports (2024)
Carlos Polanco
To the editor:
Malaria: common signs and alternatives for early identification
The article by Bhatt and colleagues (1) [The effect of malaria control on Plasmodium falciparum in Africa between 2000 and 2015, Nature], discusses the changes in infection prevalence of malaria disease in Africa during the past 15 years. . It highlights the need to implement strategies to decrease its dissemination, and minimize drug resistance.
Ebola, MERS-CoV, Malaria, and influenza have fever as a common sign. Fever is a measurable clinical sign through mobile devices or thermal imaging cameras that can be placed in strategic transit areas , in which clinical evaluation including lab tests can be carried out to determine the etiology. The voluntary use of electronic gadgets (2, 3) that enable the cameras to identify persons with fever, would be a major step as it would mark the beginning of a generation of robots allocated to perform triage.
Epidemiological alert methods seek to identify patient zero, as it substantially decreases the spread of the disease, and provides understanding of the microorganism that produces it. Only robots can efficiently carry out this work in real-time.
Sincerely,
Carlos Polanco, PhD., (*,a)
Jorge Alberto Castañón González M.D. (b)
(a) Universidad Nacional Autónoma de México, México City, México.
(b) Hospital Juárez de México, México City, México
Carlos Polanco is an associate professor in the Department of Mathematics at Universidad Nacional Autónoma de México, México City, México. (polanco@unam.mx)
Jorge Alberto Castañón González is Professor of Medicine at the Departments of Critical Care Medicine and Biomedical Research, Hospital Juárez de México. México City, México.
References
1. Nature DOI: 10.1038/nature15535.
2. Polanco, C. (letter) Nature 520:620, 609-611 (2015).
3. Polanco, C. (letter) Sci Rep 5, 11401 DOI:10.1038/srep11401
Delmiro Fernandez-Reyes
To: Nature's Editor
From: Delmiro Fernandez-Reyes.
Paediatrics Infectious Diseases and Computational Statistics.
Department of Computer Science, University College London, London, UK
Department of Paediatrics, University College Hospital Ibadan, Nigeria.
delmiro.fernandez-reyes@ucl.ac.uk
It is good to observe the Nature's interest in a key and fundamental problem in one of the three Global Health Challenges. As we move to a 'big-data" driven scenario in all sciences and arts (see UK's Alan Turing Institute) it will be more assuring to see more data-driven, as opposed to model-driven work in these areas. There is a clear theoretical and empirical need to shift to benefit from the "digital revolution" in the last decade to drive science, medicine, interventions, policy and more importantly the individual using a data-centric paradigm. Despite I praise the presented work and its importance, I would rather focused on less model with more data & a analysis of why and how do we get to have the data that may well be needed to tackle the burden of disease.
In a more scientifically democratic scenario, and given the importance of the matter, the authors should have made the challenge available to other data-scientists so their model and other models can be put forward and discussed. I think this is very important, as it is very bias to just assume their conclusions by just accepting one model. NOTE: There is not correct or incorrect model. There is a plethora of models that can be constructed with a range of assumptions AND it is KEY to meta-compare their "suitability" by an open analysis of their limitations, their theory (bounds) and the empirical experimentation that takes into account the mathematical bounds.
I am worried that this article will them be used to shape policy and funding. Many colleges in the ground in large holoendemic areas of west Africa will certainly see the limitations and therefore implications to this.
Given the importance given by the journal Nature to these results we are ought to assess if the approach is really suited to the problem and the conclusions arising from it. So our team is now assessing the methodological supplementary material in detail.
My concerns in relation to the presented model is that violates a key principle in computational statistics: Ockham's razor: i.e "simplest of competing theories be preferred to the more complex" AND "explanations of unknown phenomena be sought first in terms of known quantities".
I also have great concerns in both the data aggregation and variable selection parts of the work.
Overcomplicated and overfitted models with distribution assumptions on complex ("likely non-linear") processes have been the norm among statistics practicioners in the health application domain and malaria is not exception.
Another fundamental limitation in the manus cript is the lack of behavioural factors and availability and accessibility is not synonymous of ussage i.e. ITNs..(the authors actually provide a good example of this fundamental limitation: (See Supplementary Material page 12, section 4.1.3.). It is a good example of an assumption that showed the relation is pretty important.
We are of the opinion that real data should be allowed to drive the conclusions. There are several large urban centres in the Pf African region with high-quality large cohort data of these parameters. Yes, despite there are difficulties on sampling some of these variables, this should not be the rationale to produce models that despite show trends (and all models do with any sort of data) are too complex to be amenable to be assessed for their validity (i.e. it is a difference between attempting to infer the concept embedded in a dataset to throw lots of maths to it – and believe me we can!). We are of the opinion that the model is too biased on the parametric assumptions presented among other bounds related assumptions. We suggest that a pattern-based non-parametric high-dimensional convex approach should be compared with.
As pointed-out by previous comment by Carlos Polanco, digital health and mobile devices are now playing a key role in real-time sampling of the disease processes in many pathologies across the Globe and should be the focus on our current strategies to tackle disease patterns discovery and interventions.
I would like to have the opportunity to point to a few points that I think may well be misleading and no-so-theoretical points which I would have like to have been addressed before the authors influence policy makers AND funding bodies... etc...
Page 14 section 5.0: (VALIDITY)
1- Bayesian inference in itself is not the guarantor of a rigorous probabilistic framework. Also the term rigours what does it mean here: It will be adequate to state the bounds proven or assumed...
2- Bayesian inference is one way to do things but it is not more or less valid to non-Bayesian methods (so we need to make sure that the non-statistical reader does not gets mislead on this)
3- Again the Bayesian approach seems to be presented as the "magic solution" many pattern recognition methods have the same qualities. Similarly, the use of a Bayesian model or a pattern-based one DOES NOT gives VALIDITY to ANY model.
4- Overall, the literature is full of example of these misconceptions: using or NOT a Bayes approach is not synonymous of VALIDITY of the model NOR a guaranty on the GENERALISATION of the model. Lack of bounds in many of the parametric presented will certainly impair the generalisation as seen in many other application domains.
Page 17 Section 5.3:
1- All comments above show well in the problems that the author encounter in this section.
2- Also, for rigour (as the matter needs serious assessment) I would like to know what the authors mean with "extremely favourable computational properties"? What are those? Time or Space or Convexity? as far as I can see this problem as it is formulated is not convex = DOES NOT have a single solution and therefore prone to "local minima"
3- What "sample individual realisations in a computationally efficient manner" means?
4- The following statement: "allows for aggregations to different scales ..." does not follows from the mathematical analysis presented.
Page 17 Section 5.3.3:
1- Despite the disease ecology literature has indeed formalised the desired answer to the
problem, this is a ill-posed question that has not been solved. Several
assumptions made by the authors will not give VALIDITY to estimate the delta PR over time.
It is just a proposed way of justify the inference AND please remember that CORRELATION
IS NOT synonymous of CAUSALITY.
2- The following pages it should not be taken or interpreted as a bound for the validity nor the generalisation.
3- furthermore: and this is symptomatic of all my points above: The authors after a quite numerically justified model they move to "visual inspection" to show the "robustness" which we don't know if it is referring to training or testing (they speak of out-of-sample validation) BUT the whole paragraph DOES NOT have the mathematical rigour that the authors want to convey of the model. As far as I know this problem is not convex.
4- What do the authors mean with computational tractability? We know the computational complexity of the problem <del>no matter what</del>. Statements like this do not proof to the reader the
relation between the framework and the tractability, -let alone the complexity time-space-
5- "compare" How? and "nearly to be identical" = systematic overfitting anyone?!
6- I am not very keen to miss and match between theory and empiric results when some conceptual (known) problems clearly show.
Page 22
1- what are we talking here: "convergence is nearly assured as the sample size tends to infinity"
There is not such a thing as nearly assured. It is bound or it is not. As this problem is not convex, as far as my restricted knowledge goes, there is not bounds in sample size (NOTE: I am talking of SAMPLE not VARIABLES that define a SAMPLE). As this is a multivariate problem there is still difficulties to define a multivariate central limit theorem that sustains this.
2- I am in complete disagreement of the last paragraph on this page.
Look forward for comments and a working "think-tank" with other colleagues on how to assess the proposed model before it makes itself to be the "gold standard"....I think it is obvious that we have the responsibility to produce an open assessment.
Sincerely
Delmiro Fernandez-Reyes
delmiro.fernandez-reyes@ucl.ac.uk
Peter Gething
The paper describes how, for eleven lower-burden countries in Africa with more robust national reporting systems (accounting for around 3% of cases), we generated clinical incidence estimates using an earlier approach based on national case reports adjusted for care-seeking behaviour, low diagnostic testing rates, and under-reporting. These were as follows: Botswana, Eritrea, Gambia, Madagascar, Mauritania, Namibia, Rwanda, Senegal, South Africa, Swaziland, Zimbabwe.
back lock
I want to talk about something so important the old coconut oil now coconut oil is miraculous probably the probably one of the best oils on earth when it comes down to good health-preserving our body the nice thing about coconut oil is that it has antibacterial antimicrobial benefits in there it where it just does wonder for so many different things within our health but the big question that comes up is you may see that your coconut oil is getting hard than soft and liquids and hard again and liquid and you’re saying to yourself is that bad what mean is it ?
Does Coconut Oil Go Bad