The number of malaria cases has dropped in some Southern Africa countries, but others still remain seriously affected. When people travel within and between countries, they can bring the parasites that cause the disease to different areas. This can fuel local transmission or even lead to outbreaks in a malaria-free area.

When new malaria patients are diagnosed, they are often asked to report their recent travel history, so that the origin of their infection can be tracked. In theory, this would help to spot regions where the disease is imported from, and design targeted interventions.

However, it is difficult to know exactly where the parasites come from based on self-disclosed travel history. At best, this history can provide information about that persons infection but nothing further in the past; at worst this history can be completely incorrect. Parasite DNA, on the other hand, has the potential to bring with it an indelible record of the past. To address the problem of determining where malaria infections came from, Tessema, Wesolowski et al. focused on Northern Namibia, a region where malaria persists despite being practically absent from the rest of the country. Patients movements were assessed using mobile phone call records as well as self-reported travel history In addition, samples a single drop of blood were taken so that the genetic information of the parasites could be examined.

Combining genetic data with travel history and phone records, Tessema, Wesolowski et al. found that, in Northern Namibia, most people had gotten infected by malaria locally. However, the genetic analyses also revealed that certain infections came from places across the Angolan and Zambian borders, information that was much more difficult to obtain using self-report or mobile phone data. A new, separate study by Chang et al. also supports these results, showing that, in Bangladesh, combining genetic data with travel history and mobile phone records helps to track how malaria spreads.

Overall, the work by Tessema, Wesolowski et al. indicate that, in Northern Namibia, it will be necessary to strengthen local interventions to eliminate malaria. However, different countries in the region may also need to coordinate to decrease malaria nearby and reduce the number of cases coming into the country. While genetic data can help to monitor how new malaria cases are imported, this knowledge will be most valuable if it is routinely collected across countries. New tools will also be required to translate genetic data into information that can easily be used for control and elimination programs.

Renewed efforts against malaria have resulted in substantial gains in malaria control, with active plans to eliminate malaria from 35 countries (Newby et al., 2016). Malaria elimination requires that national and regional strategies consider the impact of local and cross-border importation on local transmission (Cotter et al., 2013; Marshall et al., 2016a; Wangdi et al., 2015; WHO, 2017). This is particularly important for eliminating countries that share porous borders with areas of higher transmission, where importation can play a major role in sustaining or reestablishing local transmission (Sturrock et al., 2015). Identifying within-country and cross-border blocks of high parasite connectivity and coordinating elimination strategies accordingly will likely be required for national and regional success.

Coordinating the optimal interventions to deploy when and where depends on understanding the impact of imported malaria infections on local transmission. If transmission is self-sustained locally, local control measures such as vector control will be necessary. If importation strongly connects the local parasite population to an external one, then interventions aimed at reducing malaria in these sources of importation or otherwise reducing vulnerability to importation may additionally be needed or even take precedence (Cotter et al., 2013).

Currently, the extent of importation is estimated primarily by taking recent travel histories of malaria cases (Sturrock et al., 2015) and, less commonly, from more general estimates of human mobility. However, routine collection of travel data is not universal, even in areas nearing elimination, and requires a robust surveillance system. When these data are collected, they are often incomplete (e.g. only the most recent travel is reported), or are otherwise inaccurate (e.g. due to disincentives such as reduced access to free healthcare when a patient reports foreign nationality) (Marshall et al., 2016b; Pindolia et al., 2012). Even with an accurate travel history, it can be difficult to tell with confidence whether malaria parasites were acquired locally or during travel. Beyond these factors, obtaining travel history only from those presenting with symptomatic malaria does not address the contribution of asymptomatic carriers, which may be the population primarily responsible for importation. Thus, travel data alone are often unable to accurately capture the relative contribution of malaria importation to local transmission. Since travel data are often limited, approaches based on movement of the overall human population using anonymized mobile phone data have been developed to create a generalizable and scalable framework for estimating movement of malaria parasites. However, these methods rely heavily on modeling assumptions, assume that the movement patterns of mobile phone owners and the at risk population are similar, and have not been used to measure international travel (Pindolia et al., 2012; Ruktanonchai et al., 2016; Tatem, 2014; Tatem et al., 2014; Wesolowski et al., 2012; Zhao et al., 2016).

Since travel history and other data on human movement are limited in their ability to provide reliable information on malaria parasite connectivity, it may be valuable for control programs to additionally collect data on parasite genetics (Wesolowski et al., 2018). In principle, data on the genetics of malaria parasites have the potential to provide the most direct measure of parasite connectivity and to identify relevant sources and sinks of parasite movement (Auburn and Barry, 2017; Escalante et al., 2015). However, there have been few efforts to systematically collect and genotype malaria infected individuals at sufficient spatial and temporal scale or density to be useful in this regard. In addition, it has been difficult to detect relevant spatial signals in parasite genetic data using existing population genetic methods, particularly in areas such as sub-Saharan Africa that have high levels of population diversity and polyclonal infections (Anderson et al., 2000; Mobegi et al., 2012). Ideally, multiple complementary sources of human and parasite data would be compared and integrated to better understand the movement of malaria parasites and contribution to transmission at various spatial scales.

As part of the Elimination 8 (E8) initiative, a regional effort to eliminate malaria from Southern Africa, Namibia has been successful in decreasing malarial morbidity and mortality (Elimination 8, 2015). However, this success has recently stalled, with the number of reported cases increasing in the last few years (Nghipumbwa et al., 2018; WHO, 2017). To achieve the national malaria elimination target of 2020, it will be critical to reassess the elimination strategy in northern Namibia, where nearly all cases in the country are reported. Of particular concern is determining the contribution and spatial scale of local transmission to malaria within northern Namibia, which should guide the geographic coverage and relative timing of local interventions, and the contribution of importation from neighboring Angola and Zambia, which should guide cross-border strategies. To address these concerns, we systematically collected parasite genetic data and human mobility data – travel history from confirmed malaria cases and national mobile phone call data records – from the region in 2015–16. From these data, we aimed to determine the importance of local transmission and imported malaria, and to compare estimates of parasite connectivity at various spatial scales obtained from the different data sources. Our results demonstrate strong evidence for local transmission in northern Namibia, provide insight into patterns of parasite connectivity within Namibia and across national borders, and demonstrate the feasibility of efficiently generating actionable information for malaria control by augmenting traditional surveillance data with a direct evaluation of the parasite population.

A total of 4643 RDT confirmed, symptomatic malaria cases were enrolled from 29 health facilities in northeastern Namibia; 23 from Kavango East and six from Zambezi regions. The Kavango East survey was conducted between March and June 2016; whereas the Zambezi study was conducted between February 2015 and June 2016 (Figure 1, Figure 1—figure supplement 1). A subset of these infections (n = 2585, Data found in Supplementary file 1) were successfully genotyped from used rapid diagnostic tests (RDTs, n = 2128, Kavango East) and dried blood spots (DBS, n = 457, Zambezi). These data were analyzed along with concomitantly collected travel survey data in these patients and mobile phone data collected from subscribers in the study area (Ruktanonchai et al., 2016).

Figure 1 with 1 supplement see all

Download asset Open asset

Within-host and population diversity are spatially variable and reflect transmission intensity

Within-host diversity and population level genetic diversity were assessed by health district (n = 4) and health facility catchment (n = 29). Across health districts, infections from Rundu and Andara had greater within-host diversity than those from Nyangana and Zambezi districts as demonstrated by higher multiplicity of infection (MOI) and lower within-host fixation index (Figure 2—figure supplement 1), consistent with the higher malaria incidence and higher proportion of imported malaria cases in these districts. Overall, the genetic diversity of the parasite population was high throughout the study area (median H_E = 0.79 [IQR: 0.60–0.85]), though lower in Zambezi than the other three health districts (Figure 2—figure supplement 1). When stratified by health facilities, the patterns of within host and population diversity showed variability within districts, providing supporting evidence of fine-scale heterogeneity of malaria transmission in the study area (Figure 2—figure supplement 2). For example, infections detected at Rundu district hospital had the highest within-host diversity, which may be due to the larger proportion of patients who traveled to or resided in Angola (reported by 13% of patients). Infections from Rundu also had the highest population diversity, which may be attributable to the large catchment area of this facility (48% came from beyond the study region). Cases were then classified as local or imported based on recent travel history and the location of residence. Infections from individuals with a history consistent with importation had higher within-host diversity than those without, despite having similar population-level diversity (Figure 2A–C). These data suggest a lower rate of superinfection and thus less local transmission in northeastern Namibia compared to the international source populations.

Figure 2 with 4 supplements see all

Download asset Open asset

Local transmission and genetic connectivity within northeastern Namibia

To evaluate the local connectivity of parasite populations, we computed the pairwise genetic connectivity between infections sampled from different health facilities. In Zambezi, 60% (9/15) of the pairwise connections were highly related. However, we observed overall lower connectivity and fewer highly related pairwise connections 39% (99/253) in the Kavango East region (Figure 3A, Figure 3—source data 1). The degree of parasite connectivity between health facility catchments was heterogeneous, with some very well-connected health facilities (i.e. highly related connections to most other facilities) and others relatively unconnected. For example, two health facilities in Zambezi and four in Kavango East were connected to most other health facilities in each region (connectivity score = 0.65–0.91, Figure 3B, Figure 3—source data 1). In Zambezi, the two most connected clinics also had the highest incidence of malaria (Mumbengegwi et al., 2018) and were in close proximity to the Angolan border and Kavango East region than the other clinics. In contrast, Rundu district hospital, the largest health facility in the study area, was only connected to a few other health facilities (connectivity score = 0.05), consistent with the high genetic diversity and large catchment of this hospital, extending well beyond the study area. Overall, 17% of the pairwise genetic connectivity measures between health facilities in Kavango East and Zambezi regions were highly related, providing evidence of mixing between these parasite populations. Health facilities that were the most genetically connected within a region were also the most connected between regions (Figure 3C), suggesting that specific localities may represent priority targets for interventions to efficiently reduce within and between region transmission.

Figure 3

Download asset Open asset

Figure 3—source data 1

Proportion of highly related infections and connectivity scores by health facility.

https://doi.org/10.7554/eLife.43510.011

Download elife-43510-fig3-data1-v2.xlsx

Mixing of parasite populations inferred from parasite genetic and human mobility data

To compare the various data sets at equivalent spatial scales, we aggregated the genetic data to mobile phone catchments (n = 14) and travel survey destination locations (n = 8). In the mobility data alone, both mobile phone catchments (n = 14) and travel survey destination locations (n = 8) were strongly connected to their neighbors, with few travelers between Kavango East and Zambezi (Figure 4). Clusters identified using modularity maximization of these networks also highlight a strong spatial signature, where contiguous locations formed clusters. Using the same procedure and spatial areas, clusters were identified for the aggregated parasite genetic data. Clusters identified from mobile phone data shared some similarity to those identified from the genetic data (grouping similarity measure: Rand Index = 0.76, Figure 4 and Figure 4—figure supplement 2, Figure 4—source data 1). However, genetic data identified a substantial amount of parasite connectivity between locations that was not detected by mobile phone data. There was little agreement between clusters identified from the travel survey data and genetic data (grouping similarity measure: Rand Index = 0.46), albeit limited by a smaller number of geographic units due to the coarser spatial scale of the travel data. The precision of results obtained from travel data was also limited by the relatively small number of individuals who reported travel within the study area. Overall, these results suggest that both travel survey and mobile phone data have limitations in capturing the structure of parasite connectivity within northeastern Namibia detected using genetic data. 

Figure 4 with 2 supplements see all

Download asset Open asset

Figure 4—source data 1

The relationship between parasite connectivity estimated from the two sources of mobility data (i.e. mobile phone and travel history) and parasite genetic data.

https://doi.org/10.7554/eLife.43510.015

Download elife-43510-fig4-data1-v2.xlsx

Evidence of cross-border connectivity between Namibia, Angola and Zambia

To evaluate cross-border connectivity, geographic regions were aggregated to nine locations: four health districts in Namibia, three locations in Angola, and two provinces in Zambia. Mobile phone data were limited to Namibia, not allowing for evaluation of cross-border connectivity. Data from the travel survey identified some sources of cross-border importation into Namibia, with the most prominent connections being from Rundu to southern Angola (2.8% of cases reporting travel to this area) and Zambezi to Western Zambia (1.8% of cases, Figure 5A). However, these data were only able to identify symptomatic cases with a direct history of travel and would not identify any cases which failed to report relevant history or those cases which may have originated from importation via asymptomatic carriers or transmissible individuals otherwise not detected by the routine surveillance. Therefore, we augmented travel history with genetic data to estimate the underlying connectivity of the parasite populations from the same nine locations.

Figure 5

Download asset Open asset

Figure 5—source data 1

Importation estimates from genetic data.

https://doi.org/10.7554/eLife.43510.017

Download elife-43510-fig5-data1-v2.xlsx

To evaluate connectivity using parasite genetics, we analyzed genetic data collected from Namibia (this study) as well as additional data from Angola and Zambia. Infections with malaria parasites from Namibia and northern Angola were not closely related (mean proportion of highly related infections = 0.00004 [Range = 0–0.00015]). However, parasites between health districts of Namibia; between Namibia and southern Angola; and between Namibia and Western and Southern provinces of Zambia were more closely related (Figure 5B). Overall, infections from Namibia were on average 142 times more likely to be genetically related to those from southern Angola and 191 times to Zambia than to those from northern Angola, indicating substantial parasite mixing within the geographically connected Namibia-Angola-Zambia regional block. In contrast, this finding suggests limited parasite connectivity between northern and southern Angola, though limited, non-contemporaneous sampling within Angola makes it difficult to make more detailed conclusions about transmission in this country. Within Namibia, infections from Andara and Nyangana were 3 and 4 times more likely to be genetically related to Zambezi than infections sampled from nearby Rundu, respectively (Figure 5B).

When evaluating specific connections between the nine locations, we estimated the direction of parasite flow in addition to the degree of connectivity between areas by weighting the pairwise proportion of highly related infections by malaria incidence to account for the differential risk of infections in different areas. Results from this analysis provided substantially more information on regional connectivity than using travel history data alone. The four studied health districts within Namibia were connected to each other but had stronger connections to nearby cross-border locations than to each other (Figure 5C and D, Figure 5—source data 1). The most important source populations were Calai and Dirico (border towns in southern Angola), followed by Western and Southern provinces of Zambia. Although there was evidence of importation into both Kavango East and Zambezi regions from other countries, based on these data the Zambezi region was estimated to receive high rates of importation from a larger number of sources (i.e., it was a dominant sink population).

It is clear that achieving elimination of malaria will require strategic coordination of local and regional interventions guided by accurate intelligence on parasite movement; what has not been clear is how to best obtain this information. In this study, we demonstrated that augmentation of traditional malaria surveillance with parasite genetic data added substantially to the understanding of transmission epidemiology in a critical region of Southern Africa straddling the border between Namibia, Angola, and Zambia. First, parasite genetics provided positive evidence that the majority of malaria cases observed in northeastern Namibia were due to local transmission, evidenced by the strong fine-scale spatial structure in the genetic data. It would be difficult to explain such consistent spatial clustering of highly related parasites and the observed decay with distance if local transmission did not predominate. This is a key piece of programmatically relevant information that would have been difficult to confirm with negative evidence, that is merely based on a lack of history consistent with importation, especially given potential disincentives for individuals to report living outside of Namibia. Second, the addition of genetic data to travel histories provided more detailed and extensive evidence of parasite connectivity over hundreds of kilometers, both within Namibia and across borders from Angola and Zambia. Conclusions regarding the origins and relative magnitude of malaria importation from genetic data were distinct from those obtained from travel history alone, which were likely limited by sparsity and bias, and from mobility from the mobile phone data, which were unable to inform cross-border movement in this study and appeared to underestimate the importance of long-distance connections within Namibia. Although cross-border movement is possible to obtain from mobile phone calling data, for example if the handset ID was used instead as an anonymized ID, data available for this study were limited to national travel patterns.

Malaria programs will require targeted interventions at sub-national scales to effectively achieve and sustain elimination (WHO, 2017). The success of such programs, for example targeted vector control, focal screening and treatment, and mass drug administration will largely be dependent on tailoring interventions to drivers of ongoing transmission (WHO, 2014). Using a novel analytic framework, we found that despite a signal of predominantly local transmission (within tens of kilometers), parasite populations in Namibia remain highly connected at longer scales within and between the two administrative regions (over hundreds of kilometers). In this context, restricting interventions to a relatively small area, such as a region, may result in improved malaria control but is unlikely to achieve elimination unless any malaria transmission from imported parasites is completely prevented. Indeed, limiting elimination efforts to national boundaries may be doomed to fail for the same reasons, for example it may be necessary for Namibia to coordinate efforts with Angola and Zambia to eliminate transmission within its own boundaries (Khadka et al., 2018). Consistent with this hypothesis, we found that parasite populations within Namibia were in many cases more closely connected to those across the border than to other parasites from the same country. At a more nuanced level, variations in the degree of genetic connectivity between areas we observed could be used to optimize the order in which interventions may be most efficiently implemented. For example, all else being equal, targeting interventions to locations with greater genetic connectivity than those with lower connectivity may be more effective in fragmenting the parasite population and reducing the influx of malaria from pockets of transmission. This finding is consistent with the previous observation that malaria cases were clustered in northeastern Namibia (Smith et al., 2017; Tatem et al., 2014). In addition, enhancing malaria surveillance and access to care, for example through additional clinics or border posts, may be effective in reducing the extent of cross-border importation if deployed in areas with high measured importation rates.

We estimated malaria parasite connectivity from mobile phone data, travel history, and parasite genetics, allowing us to compare estimates based on human population movement to more direct estimates of the connectivity of parasite populations. In principle, anonymized mobile phone data can provide a continuous and inexpensive source of human mobility data. Within Namibia, connectivity estimates derived from vast amounts of mobile phone data roughly mirrored those derived from genetics, though they were predominated by small scale movements and did not capture the extent of longer distance connections revealed by the genetic data. This difference could be due to differential patterns of movement of people with access to mobile phone and those who potentially transmit malaria, limitations in the accuracy of malaria incidence estimates used as inputs in the model, or the temporal difference between the two data sets (Ruktanonchai et al., 2016). An important current limitation of call data records analyzed was the inability to provide any information on movement beyond national boundaries.

In contrast to mobile phone data, travel data collected from symptomatic cases were sparser, estimated less travel, and were collected at a coarser spatial scale, limiting agreement with the genetic data within Namibia. However, travel data were able to provide low-resolution information and demonstrated evidence of cross-border importation by infected individuals, albeit possibly biased by patient omission on international travel or residence. Estimates derived from genetics are likely to be more comprehensive, as even with perfect accuracy travel history only captures movements of the interviewed patient while genetics can record evidence of movement through multiple generations of transmission. Importantly, information on travel allowed us to greatly extend the utility of genetic data, providing a means of ‘sampling’ parasites from beyond the study site in those with a definitive history of international travel. Travel history remains a critical part of routine surveillance, and when collected reliably and ideally at finer spatial scale than done here will likely provide important information on its own (Smith et al., 2017; Tejedor-Garavito et al., 2017) and in conjunction with genetic data.

The genetic data used in this study were generated via traditional methods – a panel of 26 microsatellites – but were well-suited for the intended application and captured strong spatial signal over local and regional scales. Particular strengths of these data were the ability to capture information from polyclonal infections (77% of the study population) given the multiallelic nature of the loci, and to obtain robust results from easily collected field samples (dried blood spots and used rapid diagnostic tests). Targeted deep sequencing of short, multiallelic haplotypes may provide similarly rich data from polyclonal infections, allowing greater flexibility in the number and location of loci, facilitating easier comparisons across data sets, and taking advantage of continual advances and cost savings in sequencing technology (Aydemir et al., 2018; Lerch et al., 2017). Generating P. falciparum whole genome sequence data from Southern Africa would facilitate rational selection of the most informative sequence targets for local and regional parasite movement.

Advances in analytical methods would likewise improve the quality of information obtained from parasite genetics (Wesolowski et al., 2018). Our methods for computing genetic relatedness, like the genetic data themselves, were relatively simple but provided useful information on relative connectivity between geographic areas. In this study site, classical methods for measuring or visualizing genetic differentiation (e.g., Gst, STRUCTURE, or phylogenetic trees) had limited utility due to the marginal differences in allele frequencies between geographically proximal locations in this relatively compact study site and the inability to utilize all information from polyclonal infections. Currently, there are few established analytical approaches to quantify fine-scale genetic connectivity between locations with predominantly polyclonal infections. Estimation of pairwise genetic relatedness between all infections in this study allowed the incorporation of data from all parasites detected in infections and extraction of useful signals of recent transmission created by recombination and cotransmission of multiple parasites. However, more sophisticated bioinformatics, statistical, and modeling tools that are designed to take advantage of genetic data from polyclonal infections and map these onto quantitative, calibrated estimates of migration rates would transform the utility of genetic data for understanding operationally relevant transmission patterns. The availability of such tools would provide a strong rationale for coordinated collection of regional data on parasite genetics, allowing for more systematic evaluation of malaria transmission and generalized utility.

The incorporation of parasite genetics added an important dimension to the understanding of local and cross-border malaria transmission epidemiology and connectivity in this area of the Elimination 8 region of Africa. Our results, showing strong connectivity between malaria parasite populations over hundreds of kilometers within Namibia and across national borders, calls for strengthening the simultaneous coordination of efforts between the Elimination eight countries. Furthermore, our data demonstrate the feasibility and added value of systematically integrating genetic data into national and regional surveillance efforts, particularly when the goal is elimination and movement of malaria parasites may threaten this goal or influence interventions. A combination of human mobility and parasite genetic data is proposed to mitigate limitations of each individual data source in isolation and to provide the most robust intelligence to guide local and regional strategy.

All data generated or analyzed during this study are included in the manuscript and supplementary files.

1. 1. Alegana VA
   2. Atkinson PM
   3. Lourenço C
   4. Ruktanonchai NW
   5. Bosco C
   6. Erbach-Schoenberg EZ
   7. Didier B
   8. Pindolia D
   9. Le Menach A
   10. Katokele S
   11. Uusiku P
   12. Tatem AJ

   (2016) Advances in mapping malaria for elimination: fine resolution modelling of Plasmodium falciparum incidence

   Scientific Reports 6:29628.

   https://doi.org/10.1038/srep29628

   • PubMed
   • Google Scholar
2. 1. Anderson TJ
   2. Haubold B
   3. Williams JT
   4. Estrada-Franco JG
   5. Richardson L
   6. Mollinedo R
   7. Bockarie M
   8. Mokili J
   9. Mharakurwa S
   10. French N
   11. Whitworth J
   12. Velez ID
   13. Brockman AH
   14. Nosten F
   15. Ferreira MU
   16. Day KP

   (2000) Microsatellite markers reveal a spectrum of population structures in the malaria parasite Plasmodium falciparum

   Molecular Biology and Evolution 17:1467–1482.

   https://doi.org/10.1093/oxfordjournals.molbev.a026247

   • PubMed
   • Google Scholar
3. 1. Auburn S
   2. Campino S
   3. Miotto O
   4. Djimde AA
   5. Zongo I
   6. Manske M
   7. Maslen G
   8. Mangano V
   9. Alcock D
   10. MacInnis B
   11. Rockett KA
   12. Clark TG
   13. Doumbo OK
   14. Ouédraogo JB
   15. Kwiatkowski DP

   (2012) Characterization of within-host Plasmodium falciparum diversity using next-generation sequence data

   PLOS ONE 7:e32891.

   https://doi.org/10.1371/journal.pone.0032891

   • PubMed
   • Google Scholar
4. 1. Auburn S
   2. Barry AE

   (2017) Dissecting malaria biology and epidemiology using population genetics and genomics

   International Journal for Parasitology 47:77–85.

   https://doi.org/10.1016/j.ijpara.2016.08.006

   • PubMed
   • Google Scholar
5. 1. Aydemir O
   2. Janko M
   3. Hathaway NJ
   4. Verity R
   5. Mwandagalirwa MK
   6. Tshefu AK
   7. Tessema SK
   8. Marsh PW
   9. Tran A
   10. Reimonn T
   11. Ghani AC
   12. Ghansah A
   13. Juliano JJ
   14. Greenhouse BR
   15. Emch M
   16. Meshnick SR
   17. Bailey JA

   (2018) Drug-Resistance and population structure of Plasmodium falciparum across the democratic republic of Congo using High-Throughput molecular inversion probes

   The Journal of Infectious Diseases 218:946–955.

   https://doi.org/10.1093/infdis/jiy223

   • PubMed
   • Google Scholar
6. 1. Cotter C
   2. Sturrock HJW
   3. Hsiang MS
   4. Liu J
   5. Phillips AA
   6. Hwang J
   7. Gueye CS
   8. Fullman N
   9. Gosling RD
   10. Feachem RGA

   (2013) The changing epidemiology of malaria elimination: new strategies for new challenges

   The Lancet 382:900–911.

   https://doi.org/10.1016/S0140-6736(13)60310-4

   • Google Scholar
7. Book

   1. Elimination 8

   (2015)

   Elimination 8 Strategic Plan: 2015–2020

   Windhoek: Elimination 8.

   • Google Scholar
8. 1. Escalante AA
   2. Ferreira MU
   3. Vinetz JM
   4. Volkman SK
   5. Cui L
   6. Gamboa D
   7. Krogstad DJ
   8. Barry AE
   9. Carlton JM
   10. van Eijk AM
   11. Pradhan K
   12. Mueller I
   13. Greenhouse B
   14. Pacheco MA
   15. Vallejo AF
   16. Herrera S
   17. Felger I

   (2015) Malaria molecular epidemiology: lessons from the international centers of excellence for malaria research network

   The American Journal of Tropical Medicine and Hygiene 93:79–86.

   https://doi.org/10.4269/ajtmh.15-0005

   • PubMed
   • Google Scholar
9. 1. Hofmann N
   2. Mwingira F
   3. Shekalaghe S
   4. Robinson LJ
   5. Mueller I
   6. Felger I

   (2015) Ultra-sensitive detection of Plasmodium falciparum by amplification of multi-copy subtelomeric targets

   PLOS Medicine 12:e1001788.

   https://doi.org/10.1371/journal.pmed.1001788

   • PubMed
   • Google Scholar
10. Book

    1. Jacquard A
    2. Charlesworth B
    3. Charlesworth D

    (1974)

    The Genetic Structure of Populations, Biomathematics

    Springer Berlin Heidelberg.

    • Google Scholar
11. 1. Jost L

    (2008) G(ST) and its relatives do not measure differentiation

    Molecular Ecology 17:4015–4026.

    https://doi.org/10.1111/j.1365-294X.2008.03887.x

    • PubMed
    • Google Scholar
12. 1. Khadka A
    2. Perales NA
    3. Wei DJ
    4. Gage AD
    5. Haber N
    6. Verguet S
    7. Patenaude B
    8. Fink G

    (2018) Malaria control across borders: quasi-experimental evidence from the Trans-Kunene malaria initiative (TKMI)

    Malaria Journal 17:224.

    https://doi.org/10.1186/s12936-018-2368-4

    • PubMed
    • Google Scholar
13. 1. Lerch A
    2. Koepfli C
    3. Hofmann NE
    4. Messerli C
    5. Wilcox S
    6. Kattenberg JH
    7. Betuela I
    8. O'Connor L
    9. Mueller I
    10. Felger I

    (2017) Development of amplicon deep sequencing markers and data analysis pipeline for genotyping multi-clonal malaria infections

    BMC Genomics 18:864.

    https://doi.org/10.1186/s12864-017-4260-y

    • PubMed
    • Google Scholar
14. 1. Marshall JM
    2. Bennett A
    3. Kiware SS
    4. Sturrock HJW

    (2016a) The hitchhiking parasite: why human movement matters to malaria transmission and what we can do about it

    Trends in Parasitology 32:752–755.

    https://doi.org/10.1016/j.pt.2016.07.004

    • PubMed
    • Google Scholar
15. 1. Marshall JM
    2. Touré M
    3. Ouédraogo AL
    4. Ndhlovu M
    5. Kiware SS
    6. Rezai A
    7. Nkhama E
    8. Griffin JT
    9. Hollingsworth TD
    10. Doumbia S
    11. Govella NJ
    12. Ferguson NM
    13. Ghani AC

    (2016b) Key traveller groups of relevance to spatial malaria transmission: a survey of movement patterns in four sub-Saharan african countries

    Malaria Journal 15:200.

    https://doi.org/10.1186/s12936-016-1252-3

    • PubMed
    • Google Scholar
16. 1. Mobegi VA
    2. Loua KM
    3. Ahouidi AD
    4. Satoguina J
    5. Nwakanma DC
    6. Amambua-Ngwa A
    7. Conway DJ

    (2012) Population genetic structure of Plasmodium falciparum across a region of diverse endemicity in west africa

    Malaria Journal 11:223.

    https://doi.org/10.1186/1475-2875-11-223

    • PubMed
    • Google Scholar
17. Software

    1. Molecular Module

    (2011)

    Preparation of rapid diagnostic tests (RDTs) for DNA extraction v1.1

    WWARN Procedure.
18. 1. Mumbengegwi DR
    2. Sturrock H
    3. Hsiang M
    4. Roberts K
    5. Kleinschmidt I
    6. Nghipumbwa M
    7. Uusiku P
    8. Smith J
    9. Bennet A
    10. Kizito W
    11. Takarinda K
    12. Ade S
    13. Gosling R

    (2018) Is there a correlation between malaria incidence and IRS coverage in western zambezi region, Namibia?

    Public Health Action 8:S44–S49.

    https://doi.org/10.5588/pha.17.0077

    • PubMed
    • Google Scholar
19. Software

    1. Murphy M

    (2018)

    Greenhouse-Lab/MicroSPAT: initial release

    Zenodo.
20. 1. Nei M
    2. Chesser RK

    (1983) Estimation of fixation indices and gene diversities

    Annals of Human Genetics 47:253–259.

    https://doi.org/10.1111/j.1469-1809.1983.tb00993.x

    • PubMed
    • Google Scholar
21. 1. Newby G
    2. Bennett A
    3. Larson E
    4. Cotter C
    5. Shretta R
    6. Phillips AA
    7. Feachem RGA

    (2016) The path to eradication: a progress report on the malaria-eliminating countries

    The Lancet 387:1775–1784.

    https://doi.org/10.1016/S0140-6736(16)00230-0

    • Google Scholar
22. 1. Newman ME

    (2006) Modularity and community structure in networks

    PNAS 103:8577–8582.

    https://doi.org/10.1073/pnas.0601602103

    • PubMed
    • Google Scholar
23. 1. Nghipumbwa MH
    2. Ade S
    3. Kizito W
    4. Takarinda KC
    5. Uusiku P
    6. Mumbegegwi DR

    (2018) Moving towards malaria elimination: trends and attributes of cases in Kavango Region, Namibia, 2010-2014

    Public Health Action 8:S18–S23.

    https://doi.org/10.5588/pha.17.0076

    • PubMed
    • Google Scholar
24. 1. Paradis E
    2. Claude J
    3. Strimmer K

    (2004) APE: analyses of phylogenetics and evolution in R language

    Bioinformatics 20:289–290.

    https://doi.org/10.1093/bioinformatics/btg412

    • PubMed
    • Google Scholar
25. 1. Pindolia DK
    2. Garcia AJ
    3. Wesolowski A
    4. Smith DL
    5. Buckee CO
    6. Noor AM
    7. Snow RW
    8. Tatem AJ

    (2012) Human movement data for malaria control and elimination strategic planning

    Malaria Journal 11:205.

    https://doi.org/10.1186/1475-2875-11-205

    • PubMed
    • Google Scholar
26. 1. Plowe CV
    2. Djimde A
    3. Bouare M
    4. Doumbo O
    5. Wellems TE

    (1995) Pyrimethamine and proguanil resistance-conferring mutations in Plasmodium falciparum dihydrofolate reductase: polymerase chain reaction methods for surveillance in africa

    The American Journal of Tropical Medicine and Hygiene 52:565–568.

    https://doi.org/10.4269/ajtmh.1995.52.565

    • PubMed
    • Google Scholar
27. 1. Pringle JC
    2. Tessema S
    3. Wesolowski A
    4. Chen A
    5. Murphy M
    6. Carpi G
    7. Shields TM
    8. Hamapumbu H
    9. Searle KM
    10. Kobayashi T
    11. Katowa B
    12. Musonda M
    13. Stevenson JC
    14. Thuma PE
    15. Greenhouse B
    16. Moss WJ
    17. Norris DE

    (2018) Genetic Evidence of Focal Plasmodium falciparum Transmission in a Pre-elimination Setting in Southern Province, Zambia

    The Journal of Infectious Diseases 526:.

    https://doi.org/10.1093/infdis/jiy640

    • Google Scholar
28. 1. Rand WM

    (1971) Objective criteria for the evaluation of clustering methods

    Journal of the American Statistical Association 66:846–850.

    https://doi.org/10.1080/01621459.1971.10482356

    • Google Scholar
29. Preprint

    1. Roh ME
    2. Tessema SK
    3. Murphy M
    4. Nhlabathi N
    5. Mkhonta N
    6. Vilakati S
    7. Ntshalintshali N
    8. Saini M
    9. Maphalala G
    10. Chen A
    11. Wilheim J
    12. Prach L
    13. Gosling R
    14. Kunene S
    15. Hsiang M
    16. Greenhouse B

    (2019) High genetic diversity of Plasmodium falciparum in the low transmission setting of the kingdom of eswatini

    bioRxiv.

    https://doi.org/10.1101/522896

    • Google Scholar
30. 1. Ruktanonchai NW
    2. DeLeenheer P
    3. Tatem AJ
    4. Alegana VA
    5. Caughlin TT
    6. Zu Erbach-Schoenberg E
    7. Lourenço C
    8. Ruktanonchai CW
    9. Smith DL

    (2016) Identifying malaria transmission foci for elimination using human mobility data

    PLOS Computational Biology 12:e1004846.

    https://doi.org/10.1371/journal.pcbi.1004846

    • PubMed
    • Google Scholar
31. 1. Smith JL
    2. Auala J
    3. Haindongo E
    4. Uusiku P
    5. Gosling R
    6. Kleinschmidt I
    7. Mumbengegwi D
    8. Sturrock HJ

    (2017) Malaria risk in young male travellers but local transmission persists: a case-control study in low transmission Namibia

    Malaria Journal 16:70.

    https://doi.org/10.1186/s12936-017-1719-x

    • PubMed
    • Google Scholar
32. 1. Sturrock HJW
    2. Roberts KW
    3. Wegbreit J
    4. Ohrt C
    5. Gosling RD

    (2015) Tackling imported malaria: an elimination endgame

    The American Journal of Tropical Medicine and Hygiene 93:139–144.

    https://doi.org/10.4269/ajtmh.14-0256

    • PubMed
    • Google Scholar
33. 1. Tatem AJ

    (2014) Mapping population and pathogen movements

    International Health 6:5–11.

    https://doi.org/10.1093/inthealth/ihu006

    • PubMed
    • Google Scholar
34. 1. Tatem AJ
    2. Huang Z
    3. Narib C
    4. Kumar U
    5. Kandula D
    6. Pindolia DK
    7. Smith DL
    8. Cohen JM
    9. Graupe B
    10. Uusiku P
    11. Lourenço C

    (2014) Integrating rapid risk mapping and mobile phone call record data for strategic malaria elimination planning

    Malaria Journal 13:52.

    https://doi.org/10.1186/1475-2875-13-52

    • PubMed
    • Google Scholar
35. 1. Tejedor-Garavito N
    2. Dlamini N
    3. Pindolia D
    4. Soble A
    5. Ruktanonchai NW
    6. Alegana V
    7. Le Menach A
    8. Ntshalintshali N
    9. Dlamini B
    10. Smith DL
    11. Tatem AJ
    12. Kunene S

    (2017) Travel patterns and demographic characteristics of malaria cases in Swaziland, 2010-2014

    Malaria Journal 16:359.

    https://doi.org/10.1186/s12936-017-2004-8

    • PubMed
    • Google Scholar
36. 1. Verity R
    2. Nichols RA

    (2016) Estimating the number of subpopulations (K) in structured populations

    Genetics 203:1827–1839.

    https://doi.org/10.1534/genetics.115.180992

    • PubMed
    • Google Scholar
37. 1. Wangdi K
    2. Gatton ML
    3. Kelly GC
    4. Clements AC

    (2015) Cross-border malaria: a major obstacle for malaria elimination

    Advances in Parasitology 89:79–107.

    https://doi.org/10.1016/bs.apar.2015.04.002

    • PubMed
    • Google Scholar
38. 1. Wesolowski A
    2. Eagle N
    3. Tatem AJ
    4. Smith DL
    5. Noor AM
    6. Snow RW
    7. Buckee CO

    (2012) Quantifying the impact of human mobility on malaria

    Science 338:267–270.

    https://doi.org/10.1126/science.1223467

    • PubMed
    • Google Scholar
39. 1. Wesolowski A
    2. Taylor AR
    3. Chang HH
    4. Verity R
    5. Tessema S
    6. Bailey JA
    7. Alex Perkins T
    8. Neafsey DE
    9. Greenhouse B
    10. Buckee CO

    (2018) Mapping malaria by combining parasite genomic and epidemiologic data

    BMC Medicine 16:.

    https://doi.org/10.1186/s12916-018-1181-9

    • PubMed
    • Google Scholar
40. Book

    1. WHO

    (2014)

    From Malaria Control to Malaria Elimination: A Manual for Elimination Scenario Planning

    Geneva: Global Malaria Programme, World Health Organization.

    • Google Scholar
41. Book

    1. WHO

    (2017)

    A Framework for Malaria Elimination

    Geneva: Global Malaria Programme, World Health Organization.

    • Google Scholar
42. 1. Zhao Z
    2. Shaw S-L
    3. Xu Y
    4. Lu F
    5. Chen J
    6. Yin L

    (2016) Understanding the bias of call detail records in human mobility research

    International Journal of Geographical Information Science 30:1738–1762.

    https://doi.org/10.1080/13658816.2015.1137298

    • Google Scholar

Author details

1. Sofonias Tessema

   EPPIcenter program, Division of HIV, Infectious Diseases and Global Medicine, Department of Medicine, University of California, San Francisco, San Francisco, United States

   Contribution

   Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Amy Wesolowski

   For correspondence

   SofoniasK.Tessema@ucsf.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1057-5310

2. Amy Wesolowski

   Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, United States

   Contribution

   Data curation, Formal analysis, Validation, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Sofonias Tessema

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6320-3575

3. Anna Chen

   EPPIcenter program, Division of HIV, Infectious Diseases and Global Medicine, Department of Medicine, University of California, San Francisco, San Francisco, United States

   Contribution

   Investigation

   Competing interests

   No competing interests declared

4. Maxwell Murphy

   EPPIcenter program, Division of HIV, Infectious Diseases and Global Medicine, Department of Medicine, University of California, San Francisco, San Francisco, United States

   Contribution

   Software, Investigation

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0332-4388

5. Jordan Wilheim

   EPPIcenter program, Division of HIV, Infectious Diseases and Global Medicine, Department of Medicine, University of California, San Francisco, San Francisco, United States

   Contribution

   Investigation

   Competing interests

   No competing interests declared

6. Anna-Rosa Mupiri

   Multidisciplinary Research Center, University of Namibia, Windhoek, Namibia

   Contribution

   Investigation, Project administration

   Competing interests

   No competing interests declared

7. Nick W Ruktanonchai

   WorldPop Project, Geography and Environment, University of Southampton, Southampton, United Kingdom

   Contribution

   Investigation, Writing—review and editing

   Competing interests

   No competing interests declared

8. Victor A Alegana

   1. Multidisciplinary Research Center, University of Namibia, Windhoek, Namibia
   2. WorldPop Project, Geography and Environment, University of Southampton, Southampton, United Kingdom

   Contribution

   Investigation

   Competing interests

   No competing interests declared

9. Andrew J Tatem

   WorldPop Project, Geography and Environment, University of Southampton, Southampton, United Kingdom

   Contribution

   Investigation, Writing—review and editing

   Competing interests

   No competing interests declared

10. Munyaradzi Tambo

    Multidisciplinary Research Center, University of Namibia, Windhoek, Namibia

    Contribution

    Investigation

    Competing interests

    No competing interests declared

11. Bradley Didier

    Clinton Health Access Initiative, Boston, United States

    Contribution

    Investigation, Writing—review and editing

    Competing interests

    No competing interests declared

12. Justin M Cohen

    Clinton Health Access Initiative, Boston, United States

    Contribution

    Writing—review and editing

    Competing interests

    No competing interests declared

13. Adam Bennett

    Malaria Elimination Initiative, Institute of Global Health Sciences, University of California, San Francisco, San Francisco, United States

    Contribution

    Writing—review and editing

    Competing interests

    No competing interests declared

14. Hugh JW Sturrock

    Malaria Elimination Initiative, Institute of Global Health Sciences, University of California, San Francisco, San Francisco, United States

    Contribution

    Writing—review and editing

    Competing interests

    No competing interests declared

15. Roland Gosling

    1. Multidisciplinary Research Center, University of Namibia, Windhoek, Namibia
    2. Malaria Elimination Initiative, Institute of Global Health Sciences, University of California, San Francisco, San Francisco, United States

    Contribution

    Writing—review and editing

    Competing interests

    No competing interests declared

16. Michelle S Hsiang

    1. Malaria Elimination Initiative, Institute of Global Health Sciences, University of California, San Francisco, San Francisco, United States
    2. Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, United States
    3. Department of Pediatrics, UCSF Benioff Children's Hospital, San Francisco, United States

    Contribution

    Writing—review and editing

    Competing interests

    No competing interests declared

17. David L Smith

    Institute for Health Metrics and Evaluation, University of Washington, Seattle, United States

    Contribution

    Writing—review and editing

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0003-4367-3849

18. Davis R Mumbengegwi

    Multidisciplinary Research Center, University of Namibia, Windhoek, Namibia

    Contribution

    Conceptualization, Field supervision, Investigation, Project administration, Writing—review and editing

    Competing interests

    No competing interests declared

19. Jennifer L Smith

    Malaria Elimination Initiative, Institute of Global Health Sciences, University of California, San Francisco, San Francisco, United States

    Contribution

    Investigation, Project administration, Writing—review and editing

    Competing interests

    No competing interests declared

20. Bryan Greenhouse

    1. EPPIcenter program, Division of HIV, Infectious Diseases and Global Medicine, Department of Medicine, University of California, San Francisco, San Francisco, United States
    2. Chan Zuckerberg Biohub, San Francisco, United States

    Contribution

    Conceptualization and field supervision

    Competing interests

    No competing interests declared

• 3,813

  views

• 460

  downloads

• 90

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.