Search

Background Reactive case detection (RACD) is a widely practiced malaria elimination intervention whereby close contacts of index cases receive malaria testing to inform treatment and other interventions. However, the optimal diagnostic and operational approaches for this resource-intensive strategy are not clear. Methods We conducted a 3-year prospective national evaluation of RACD in Eswatini, a malaria elimination setting. Loop-mediated isothermal amplification (LAMP) was compared to traditional rapid diagnostic testing (RDT) for the improved detection of infections and for hotspots (RACD events yielding ≥1 additional infection). The potential for index case–, RACD-, and individual-level factors to improve efficiencies was also evaluated. Results Among 377 RACD events, 10 890 participants residing within 500 m of index cases were tested. Compared to RDT, LAMP provided a 3-fold and 2.3-fold higher yield to detect infections (1.7% vs 0.6%) and hotspots (29.7% vs 12.7%), respectively. Hotspot detection improved with ≥80% target population coverage and response times within 7 days. Proximity to the index case was associated with a dose-dependent increased infection risk (up to 4-fold). Individual-, index case–, and other RACD-level factors were considered but the simple approach of restricting RACD to a 200-m radius maximized yield and efficiency. Conclusions We present the first large-scale national evaluation of optimal RACD approaches from a malaria elimination setting. To inform delivery of antimalarial drugs or other interventions, RACD, when conducted, should utilize more sensitive diagnostics and clear context-specific operational parameters. Future studies of RACD’s impact on transmission may still be needed., Optimal approaches to find malaria infections and hotspots in reactive case detection are not clear. Here, molecular testing compared to standard rapid diagnostic testing and operational factors (≥80% coverage, response ≤7 days, and 200-m screening radius) maximized efficiency and yield.

Background Inference of person-to-person transmission networks using surveillance data is increasingly used to estimate spatiotemporal patterns of pathogen transmission. Several data types can be used to inform transmission network inferences, yet the sensitivity of those inferences to different data types is not routinely evaluated. Methods The influence of different combinations of spatial, temporal, and travel-history data on transmission network inferences for Plasmodium falciparum malaria were evaluated. Results The information content of these data types may be limited for inferring person-to-person transmission networks and may lead to an overestimate of transmission. Only when outbreaks were temporally focal or travel histories were accurate was the algorithm able to accurately estimate the reproduction number under control, Rc. Applying this approach to data from Eswatini indicated that inferences of Rc and spatiotemporal patterns therein depend upon the choice of data types and assumptions about travel-history data. Conclusions These results suggest that transmission network inferences made with routine malaria surveillance data should be interpreted with caution.

IntroductionTo reduce malaria transmission in very low-endemic settings, screening and treatment near index cases (reactive case detection (RACD)), is widely practiced, but the rapid diagnostic tests (RDTs) used miss low-density infections. Presumptive treatment near index cases (reactive focal mass drug administration (rfMDA)) may be safe and more effective.MethodsWe conducted a cluster-randomised controlled trial in Eswatini, a very low-endemic setting. 77 clusters were randomised to rfMDA using dihydroartemisin-piperaquine (DP) or RACD involving RDTs and artemether lumefantrine (AL). Interventions were delivered by the local programme. An intention-to-treat analysis was used to compare cluster-level cumulative confirmed malaria incidence among clusters with cases. Secondary outcomes included safety and adherence.ResultsFrom Sept 2015–Aug 2017, 220 index cases from 47 clusters triggered 49 RACD events and 68 rfMDA events. RACD and rfMDA were delivered to 1696 and 1932 individuals, respectively. Index case and target population intervention coverages for both arms were 75.6%–81.4% and adherence to DP was 98.7%. For rfMDA versus RACD, cumulative incidences (per 1000 person-years) of all malaria were 2.11 (95% CI 1.73–2.59) and 1.97 (1.57–2.47), respectively; and of locally acquired malaria, they were 1.29 (95% CI 1.00–1.67) and 0.97 (0.71–1.34), respectively. Adjusting for imbalance in baseline incidence, incidence rate ratio (aIRR) for rfMDA versus RACD was 0.93 (95% CI 0.54–1.60) for all malaria and 0.77 (95% CI 0.38–1.56) for locally acquired malaria. No serious adverse events occurred.ConclusionIn a very low-endemic, real-world setting, this trial is the first to evaluate rfMDA using DP. rfMDA was safe and resulted in lower cumulative incidence compared to RACD, but we were unable to confirm its effectiveness, potentially due to insufficient power. To assess impact of interventions in very low-endemic settings, multi-site, adaptive trials and use of complementary interventions may be needed.What is already known?Reactive case detection (RACD), or malaria testing and treatment in the vicinity of passively detected malaria cases, is a standard of care intervention used in low and very low transmission settings aiming for malaria elimination.Despite the use of RACD, progress toward malaria elimination has stalled in many countries and new strategies are needed.Reactive focal mass drug administration (rfMDA) is a transmission reducing strategy that has been shown to be effective in a low transmission setting, but there are no trial data from a very low transmission setting.What are the new findings?In a pragmatic, cluster-randomised controlled trial of rfMDA using dihydroartemisinin-piperaquine compared to RACD, we found that rfMDA was safe.rfMDA resulted in lower cumulative incidence, but we were unable to confirm its effectiveness compared to RACD, potentially due to insufficient power (we expected 63 total clusters would have incident cases, but observed 47).What do the new findings imply?When implemented in a real-world, very low transmission setting, rMDA was safe but evidence regarding its effectiveness to reduce transmission was weak.The challenge to show a statistically significant impact of a targeted community-based intervention in a very low transmission setting highlights the need for such trials to be multi-site, adaptive, and consider use of complementary interventions.

IntroductionTo reduce malaria transmission in very low-endemic settings, screening and treatment near index cases (reactive case detection (RACD)), is widely practised, but the rapid diagnostic tests (RDTs) used miss low-density infections. Reactive focal mass drug administration (rfMDA) may be safe and more effective.MethodsWe conducted a pragmatic cluster randomised controlled trial in Eswatini, a very low-endemic setting. 77 clusters were randomised to rfMDA using dihydroartemisin–piperaquine (DP) or RACD involving RDTs and artemether–lumefantrine. Interventions were delivered by the local programme. An intention-to-treat analysis was used to compare cluster-level cumulative confirmed malaria incidence among clusters with cases. Secondary outcomes included safety and adherence.ResultsFrom September 2015 to August 2017, 222 index cases from 47 clusters triggered 46 RACD events and 64 rfMDA events. RACD and rfMDA were delivered to 1455 and 1776 individuals, respectively. Index case coverage was 69.5% and 62.4% for RACD and rfMDA, respectively. Adherence to DP was 98.7%. No serious adverse events occurred. For rfMDA versus RACD, cumulative incidences (per 1000 person-years) of all malaria were 2.11 (95% CI 1.73 to 2.59) and 1.97 (95% CI 1.57 to 2.47), respectively; and of locally acquired malaria, they were 1.29 (95% CI 1.00 to 1.67) and 0.97 (95% CI 0.71 to 1.34), respectively. Adjusting for imbalance in baseline incidence, incidence rate ratio for rfMDA versus RACD was 0.93 (95% CI 0.54 to 1.62) for all malaria and 0.84 (95% CI 0.42 to 1.66) for locally acquired malaria. Similar results were obtained in a per-protocol analysis that excluded clusters with ConclusionIn a very low-endemic, real-world setting, rfMDA using DP was safe, but did not lower incidence compared with RACD, potentially due to insufficient coverage and/or power. To assess impact of interventions in very low-endemic settings, improved coverage, complementary interventions and adaptive ring trial designs may be needed.Trial registration numberNCT02315690.

Inference of person-to-person transmission networks using routinely collected surveillance data is being used increasingly to estimate spatiotemporal patterns of pathogen transmission. Several data types can be used to inform transmission network inferences, yet the sensitivity of those inferences to different data types is not routinely evaluated. We evaluated the influence of different combinations of spatial, temporal, and travel-history data on transmission network inferences for Plasmodium falciparum, the pathogen responsible for most human malaria. After developing a new inference framework and applying it to simulated data, we found that these data types have limited utility for inferring transmission networks and, in some combinations, tend to overestimate transmission. Only when outbreaks were highly focal in time or when travel histories were highly accurate was the inference algorithm able to accurately estimate the reproduction number under control, Rc, a key metric of transmission. Applying this approach to surveillance data from Eswatini indicated that inferences of Rc and spatiotemporal patterns therein are sensitive to the choice of data types and assumptions about the accuracy of travel-history data. Taken together, these results suggest that transmission network inferences made with routinely collected surveillance data should be interpreted with caution. As we have done here, future studies inferring transmission networks should apply their algorithm to data simulated under alternative assumptions to assess the robustness of their inferences. Author Summary Person-to-person networks of pathogen transmission can reveal epidemiological patterns that are not obvious based on routinely collected surveillance data alone. Several types of data can be used to inform estimates of these networks, including the timing of when cases present, the locations where those individuals reside, and whether those individuals recently traveled to an area with higher transmission. Little is known about whether transmission network inferences may depend on the types of data used to inform those inferences. To address this question, we developed an algorithm to estimate transmission networks of Plasmodium falciparum malaria. We validated this algorithm on simulated data and then applied it to surveillance data from Eswatini. Our results demonstrate that transmission network inferences are sensitive to which data types are used to inform them and to assumptions about how to interpret those data, such as whether data about recent travel are accurate. These results suggest that future work should carefully consider the epidemiological setting, which data types are used to inform transmission network inferences, and how those inferences are interpreted.

Background: Swaziland is one of the southern African countries that aim to eliminate malaria by 2020. In 2010, the country introduced an Immediate Disease Notification System (IDNS) for immediate reporting of notifiable diseases, including malaria. Health facilities are to report malaria cases within 24 h through a toll-free telephone number (977), triggering an alert for case investigation at the patient's household within 48 h. We assessed the completeness of reporting in the IDNS, the subsequent case investigation, and whether it was done within the stipulated timelines. Methods: A cross-sectional study using routine country-wide data. Results: Of 1991 malaria cases notified between July 2011 and June 2015, 76% were reported in the IDNS, of which 68% were investigated-a shortfall of 24% in reporting and 32% in case investigations. Of the 76% of cases reported through the IDNS, 62% were reported within 24 h and 20% were investigated within 48 h. These shortcomings were most pronounced in hospitals and private facilities. Investigated cases (n = 1346) were classified as follows: 60% imported, 35% local and 5% undetermined. Conclusion: The utilisation of the IDNS for case reporting to trigger investigation is crucial for active surveillance. There is a need to address the reporting and investigation gaps identified to ensure that malaria cases receive appropriate interventions.

Background: Screening and treatment near malaria index cases (reactive case detection (RACD)), is widely practiced, but the rapid diagnostic tests (RDTs) used miss low density infections. Presumptive treatment near index cases (reactive focal mass drug administration (rfMDA)) may be safe and more effective but has not been evaluated in very low-endemic settings. Methods: A cluster-randomised controlled trial was conducted in Eswatini, a very low -endemic setting. 77 clusters were randomised to rfMDA using dihydroartemisinin-piperaquine (DP) or RACD involving RDTs and artemether lumefantrine (AL), administered within 200 or 500 m of index cases, respectively. Interventions were delivered by the local programme, and evaluated using an intention-to-treat analysis. Findings: From Sept 2015–Aug 2017, 220 index cases from 47 clusters triggered 49 RACD events and 68 rfMDA events. RACD and rfMDA were delivered to 1 696 and 1 932 individuals, respectively. For rfMDA versus RACD, cumulative incidences (per 1 000 person-years) of all malaria were 2·11 (95% CI 1·73–2·59) and 1·97 (1·57–2·47), respectively; and of locally acquired malaria, they were 1·29 (95% CI 1·00–1·67) and 0·97 (0·71–1·34), respectively. Adjusting for imbalance in baseline incidence, incidence rate ratio (aIRR) for rfMDA versus RACD was 0·93 (95% CI 0·54–1·60) for all malaria and 0·77 (95% CI 0·38–1·56) for locally acquired malaria. No serious adverse events occurred. Interpretation: In a very low-endemic, real-world setting, rfMDA was safe. There was limited statistical power to estimate its effectiveness compared to RACD. Trial Registration: This trial is registered with the ClinicalTrials.gov (NCT02315690). Funding Statement: This study was supported by the Bill & Melinda Gates Foundation (A122394) and the Horchow Family Fund (5300375400). Declaration of Interests: The authors have no competing interests to report. Ethics Approval Statement: Ethics approval was obtained from Eswatini Ministry of Health (MOH) (MH/599C) and University of California San Francisco (14-15226). Written informed consent was obtained from individual participants. For children less than 18 years, written informed consent from a parent or guardian was required, as was written assent for children 12–17 years.

Background Reactive focal mass drug administration (rfMDA), or presumptive treatment without malaria testing of household members and neighbours of a passively identified malaria case, is currently being explored as a possible malaria elimination strategy in low transmission settings. One of the primary factors determining the effectiveness of rfMDA on reducing or interrupting transmission is achieving high coverage of the target population with drug administration. This study aims to explore the acceptability of rfMDA and identify facilitators and barriers to its potential implementation, as well as the community’s general knowledge, attitudes and beliefs with regard to malaria elimination. Methods A qualitative study was performed using focus group discussions (FGDs) among villagers that received rfMDA through the National Malaria Control Programme in the low transmission setting of Eswatini as part of a 2-year clinical trial. FGDs were audio-recorded, transcribed and translated into English. All transcripts were managed in Dedoose and underwent qualitative content analysis. Results The majority of participants perceived their community to be at high risk of malaria. Witnessing others in their community suffer from malaria, proximity to Mozambique, various ecological factors, and the presence of mosquitoes contributed to this perception. The greatest motivator of participation in rfMDA was witnessing someone else suffer from malaria, since most participants had not personally experienced malaria themselves. Participants valued the education on rfMDA and on malaria in general, particularly when communicated by nurses and other health workers from the Ministry of Health. Participants were overwhelmingly motivated to participate in rfMDA in order to obtain protection from malaria. Most participants did not understand the concept of sub-clinical infection and, therefore, did not perceive the anti-malarial medication given in rfMDA to be a treatment medication. Conclusions Perceived risk for malaria was a major driver of acceptability; therefore, future intervention campaigns could aim to better quantify risk to inform interventions and encourage uptake. There were misunderstandings about the asymptomatic reservoir of parasites in humans. Given that this phenomenon is the rationale for rfMDA, this misunderstanding could threaten the uptake of the intervention if it persists in the community. Using local authorities to deliver messaging, additional education on this concept with re-inforcement that risk of malaria is ongoing, even in the absence of frequent cases, may help to maximize and maintain acceptability.

Background To better understand transmission dynamics, we characterized Plasmodium falciparum genetic diversity in Eswatini, where transmission is low and sustained by importation. Methods Twenty-six P. falciparum microsatellites were genotyped in 66% of confirmed cases (2014–2016; N = 582). Population and within-host diversity were used to characterize differences between imported and locally acquired infections. Logistic regression was used to assess the added value of diversity metrics to classify imported and local infections beyond epidemiology data alone. Results Parasite population in Eswatini was highly diverse (expected heterozygosity [HE] = 0.75) and complex: 67% polyclonal infections, mean multiplicity of infection (MOI) 2.2, and mean within-host infection fixation index (FWS) 0.84. Imported cases had comparable diversity to local cases but exhibited higher MOI (2.4 vs 2.0; P = .004) and lower mean FWS (0.82 vs 0.85; P = .03). Addition of MOI and FWS to multivariate analyses did not increase discrimination between imported and local infections. Conclusions In contrast to the common perception that P. falciparum diversity declines with decreasing transmission intensity, Eswatini isolates exhibited high parasite diversity consistent with high rates of malaria importation and limited local transmission. Estimates of malaria transmission intensity from genetic data need to consider the effect of importation, especially as countries near elimination., In contrast to the commonly held perception that P. falciparum diversity declines with decreasing transmission intensity, infections from Eswatini exhibited high parasite diversity consistent with high rates of malaria importation and limited local transmission.

Background Swaziland aims to eliminate malaria by 2020. However, imported cases from neighbouring endemic countries continue to sustain local parasite reservoirs and initiate transmission. As certain weather and climatic conditions may trigger or intensify malaria outbreaks, identification of areas prone to these conditions may aid decision-makers in deploying targeted malaria interventions more effectively. Methods Malaria case-surveillance data for Swaziland were provided by Swaziland’s National Malaria Control Programme. Climate data were derived from local weather stations and remote sensing images. Climate parameters and malaria cases between 2001 and 2015 were then analysed using seasonal autoregressive integrated moving average models and distributed lag non-linear models (DLNM). Results The incidence of malaria in Swaziland increased between 2005 and 2010, especially in the Lubombo and Hhohho regions. A time-series analysis indicated that warmer temperatures and higher precipitation in the Lubombo and Hhohho administrative regions are conducive to malaria transmission. DLNM showed that the risk of malaria increased in Lubombo when the maximum temperature was above 30 °C or monthly precipitation was above 5 in. In Hhohho, the minimum temperature remaining above 15 °C or precipitation being greater than 10 in. might be associated with malaria transmission. Conclusions This study provides a preliminary assessment of the impact of short-term climate variations on malaria transmission in Swaziland. The geographic separation of imported and locally acquired malaria, as well as population behaviour, highlight the varying modes of transmission, part of which may be relevant to climate conditions. Thus, the impact of changing climate conditions should be noted as Swaziland moves toward malaria elimination. Electronic supplementary material The online version of this article (doi:10.1186/s12936-017-1874-0) contains supplementary material, which is available to authorized users.

Additional file 1. Annual malaria incidence (black) and climate variables (red) in Swaziland, 1985â 2015.

Background. Low-quality housing may confer risk of malaria infection, but evidence in low transmission settings is limited. Methods. To examine the relationship between individual level housing quality and locally acquired infection in children and adults, a population-based cross-sectional analysis was performed using existing surveillance data from the low transmission setting of Swaziland. From 2012 to 2015, cases were identified through standard diagnostics in health facilities and by loop-mediated isothermal amplification in active surveillance, with uninfected subjects being household members and neighbors. Housing was visually assessed in a home visit and then classified as low, high, or medium quality, based on housing components being traditional, modern, or both, respectively. Results. Overall, 11426 individuals were included in the study: 10960 uninfected and 466 infected (301 symptomatic and 165 asymptomatic). Six percent resided in low-quality houses, 26% in medium-quality houses, and 68% in high-quality houses. In adjusted models, low- and medium-quality construction was associated with increased risk of malaria compared with high-quality construction (adjusted odds ratio [AOR], 2.11 and 95% confidence interval [CI], 1.26–3.53 for low vs high; AOR, 1.56 and 95% CI, 1.15–2.11 for medium vs high). The relationship was independent of vector control, which also conferred a protective effect (AOR, 0.67; 95% CI, .50–.90) for sleeping under an insecticide-treated bed net or a sprayed structure compared with neither. Conclusions. Our study adds to the limited literature on housing quality and malaria risk from low transmission settings. Housing improvements may offer an attractive and sustainable additional strategy to support countries in malaria elimination.

The substantial impact of cross-border collaborative control efforts on the burden of malaria in southern Africa has previously been demonstrated through the successes of the Lubombo Spatial Development Initiative. Increases in malaria cases recorded in the three partner countries (Mozambique, South Africa, Swaziland) since termination of that programme in 2011 have provided impetus for the resuscitation of cooperation in the form of the MOSASWA malaria initiative. MOSASWA, launched in 2015, seeks to renew regional efforts to accelerate progress towards malaria elimination goals already established in the region. National malaria programmes, together with developmental partners, academic institutions and the private sector seek to harmonize policy, strengthen capacity, share expertise, expand access to elimination interventions particularly amongst migrant and border population groups, mobilize resources and advocate for long-term funding to ultimately achieve and sustain malaria elimination in the MOSASWA region.

Background Countries remain reluctant to adopt the 2012 World Health Organization recommendation for single low-dose (0.25 mg/kg) primaquine (SLD PQ) for Plasmodium falciparum transmission-blocking due to concerns over drug-related haemolysis risk, especially among glucose-6-phosphate dehydrogenase-deficient (G6PDd) people, without evidence demonstrating that it can be safely deployed in their settings. Pharmacovigilance methods provide a systematic way of collecting safety data and supporting the rollout of SLD PQ. Methods The Primaquine Roll Out Monitoring Pharmacovigilance Tool (PROMPT), comprising: (1) a standardized form to support the surveillance of possible adverse events following SLD PQ treatment; (2) a patient information card to enhance awareness of known adverse drug reactions of SLD PQ use; and (3) a database compiling recorded information, was developed and piloted. Data on patient characteristics, malaria diagnosis and treatment are collected. Blood samples are taken to measure haemoglobin (Hb) and test for G6PD deficiency. Active follow-up includes a repeat Hb measurement and adverse event monitoring on or near day 7. A 13-month prospective pilot study in two hospital facilities in Swaziland alongside the introduction of SLD PQ generated preliminary evidence on the feasibility and acceptability of PROMPT. Results PROMPT was well received by nurses as a simple, pragmatic approach to active surveillance of SLD PQ safety data. Of the 102 patients enrolled and administered SLD PQ, none were G6PDd. 93 (91.2 %) returned on or near day 7 for follow-up. Four (4.6 %) patients had falls in Hb ≥25 % from baseline, none of whom presented with signs or symptoms of anaemia. No patient’s Hb fell below 7 g/dL and none required a blood transfusion. Of the 11 (11 %) patients who reported an adverse event over the study period, three were considered serious and included two deaths and one hospitalization; none were causally related to SLD PQ. Four non-serious adverse events were considered definitely, probably, or possibly related to SLD PQ. Conclusion Improved pharmacovigilance to monitor and promote the safety of the WHO recommendation is needed. The successful application of PROMPT demonstrates its potential as an important tool to rapidly generate locally acquired safety data and support pharmacovigilance in resource-limited settings. Electronic supplementary material The online version of this article (doi:10.1186/s12936-016-1410-7) contains supplementary material, which is available to authorized users.

In many parts of the world, malaria elimination— defined by the World Health Organization (WHO) as the absence of locally acquired malaria cases in the country—is being considered as a target because of recent successes in reducing disease burden (1, 2). Rigorous evaluation of malaria elimination programs is essential for financial and political support to be maintained. Yet such evaluation remains challenging, and appropriate metrics to ascertain “success” are needed. Although the long-term focus on elimination is commendable, evaluation of programs cannot rely on a dichotomous approach where “success” would correspond to no locally acquired malaria cases, and anything else would be seen as failure. Evaluation needs to take into account the local and regional epidemiological circumstances, because countries that manage to control local transmission to relatively low levels but receive large numbers of imported cases are likely to see locally acquired cases (3). Effective programs may therefore be wrongly perceived as unsuccessful, which would jeopardize their long term viability. In addition to the long-term objective of elimination, we must develop intermediate milestones that better capture and acknowledge these scenarios. Such a milestone should be the interruption of endemic transmission, meaning that the country would eventually see malaria go away if importation ceased. The country could then aim to reduce chains of transmission generated from imported cases down to zero. Although this sounds sensible, there are no operational measures and tools for evaluating local transmission as it approaches elimination, because incidence estimates require either surveillance data with perfect case detection or surveys with impractically large sample sizes (4). In addition, incidence does not measure local transmission in the context of importation. Here, we present simple, theoretically sound, and robust metrics to ascertain progress toward elimination. Aside from their importance for program evaluation, these metrics also allow more-refined assessment of the epidemiological situation, which could lead to more accurate evaluation of different elimination strategies (for example, differentiating between persistent low-level endemic transmission or repeated importation and extinction events that may require different public health approaches).

Additional file 1. Urine colour scale designed to assess the degree of haemoglobinuria. Dark-coloured urine with a colourimetric of 5 or above was considered evidence of haemoglobinuria [18]. With permission from the NEJM.

Eliminating malaria from a defined region involves draining the endemic parasite reservoir and minimizing local malaria transmission around imported malaria infections. In the last phases of malaria elimination, as universal interventions reap diminishing marginal returns, national resources must become increasingly devoted to identifying where residual transmission is occurring. The needs for accurate measures of progress and practical advice about how to allocate scarce resources require new analytical methods to quantify fine-grained heterogeneity in malaria risk. Using routine national surveillance data from Swaziland (a sub-Saharan country on the verge of elimination), we estimated individual reproductive numbers. Fine-grained maps of reproductive numbers and local malaria importation rates were combined to show ‘malariogenic potential’, a first for malaria elimination. As countries approach elimination, these individual-based measures of transmission risk provide meaningful metrics for planning programmatic responses and prioritizing areas where interventions will contribute most to malaria elimination. DOI: http://dx.doi.org/10.7554/eLife.09520.001, eLife digest Swaziland has set a national goal of eliminating malaria transmission in the very short term, which would make it the first country in sub-Saharan Africa to do so. More than half of the cases of malaria that are observed in Swaziland are caused by infections picked up by travelers while they were in other countries where the disease is much more prevalent. The other cases – people who became infected in Swaziland – are the cases that the government of Swaziland is trying to prevent. If Swaziland is going to eliminate malaria, it will need to identify any places where the malaria parasites are still spreading throughout the population so it can target those communities with effective prevention measures. It will also need to manage the risk that infections imported from abroad may re-start transmission in places where it has been stopped. To work out how likely it is that a malaria infection will be transmitted by mosquitoes in a particular place, researchers can look at past malaria data and calculate how many new infections are caused by each case. Reiner et al. have now produced a computer model that estimates how this number varies across Swaziland, highlighting places where the government is going to need to focus efforts to eliminate malaria. The model shows that in some rural areas near Mozambique, each individual infected with malaria is causing more than one other person to become infected. This confirms that the disease has not yet been eliminated from these areas. However, in other regions of the country, malaria rarely spreads between individuals. The detailed regional information from the model may help public health authorities in Swaziland better target their anti-malaria resources. In large cities where most cases are imported, Reiner et al. suggest focusing resources on providing preventive treatment to travelers who plan on visiting places where malaria is spreading. However, in rural areas where malaria continues to spread, preventively treating the whole population or providing them with tools to protect them from mosquitoes might be more appropriate. Similar considerations of regional differences in the spread of malaria could also help other countries to more effectively combat the disease. DOI: http://dx.doi.org/10.7554/eLife.09520.002

Background Mapping malaria risk is an integral component of efficient resource allocation. Routine health facility data are convenient to collect, but without information on the locations at which transmission occurred, their utility for predicting variation in risk at a sub-catchment level is presently unclear. Methods Using routinely collected health facility level case data in Swaziland between 2011–2013, and fine scale environmental and ecological variables, this study explores the use of a hierarchical Bayesian modelling framework for downscaling risk maps from health facility catchment level to a fine scale (1 km x 1 km). Fine scale predictions were validated using known household locations of cases and a random sample of points to act as pseudo-controls. Results Results show that fine-scale predictions were able to discriminate between cases and pseudo-controls with an AUC value of 0.84. When scaled up to catchment level, predicted numbers of cases per health facility showed broad correspondence with observed numbers of cases with little bias, with 84 of the 101 health facilities with zero cases correctly predicted as having zero cases. Conclusions This method holds promise for helping countries in pre-elimination and elimination stages use health facility level data to produce accurate risk maps at finer scales. Further validation in other transmission settings and an evaluation of the operational value of the approach is necessary. Electronic supplementary material The online version of this article (doi:10.1186/1475-2875-13-421) contains supplementary material, which is available to authorized users.

Background Swaziland has made great progress towards its goal of malaria elimination by 2015. However, malaria importation from neighbouring high-endemic Mozambique through Swaziland’s eastern border remains a major factor that could prevent elimination from being achieved. In order to reach elimination, Swaziland must rapidly identify and treat imported malaria cases before onward transmission occurs. Methods A nationwide formative assessment was conducted over eight weeks to determine if the imported cases of malaria identified by the Swaziland National Malaria Control Programme could be linked to broader social networks and to explore methods to access these networks. Results Using a structured format, interviews were carried out with malaria surveillance agents (6), health providers (10), previously identified imported malaria cases (19) and people belonging to the networks identified through these interviews (25). Most imported malaria cases were Mozambicans (63%, 12/19) making a living in Swaziland and sustaining their families in Mozambique. The majority of imported cases (73%, 14/19) were labourers and self-employed contractors who travelled frequently to Mozambique to visit their families and conduct business. Social networks of imported cases with similar travel patterns were identified through these interviews. Nearly all imported cases (89%, 17/19) were willing to share contact information to enable network members to be interviewed. Interviews of network members and key informants revealed common congregation points, such as the urban market places in Manzini and Malkerns, as well as certain bus stations, where people with similar travel patterns and malaria risk behaviours could be located and tested for malaria. Conclusion This study demonstrated that imported cases of malaria belonged to networks of people with similar travel patterns. This study may provide novel methods for screening high-risk groups of travellers using both snowball sampling and time-location sampling of networks to identify and treat additional malaria cases. Implementation of a proactive screening programme of importation networks may help Swaziland halt transmission and achieve malaria elimination by 2015.