Malaria surveillance reveals parasite relatedness, signatures of selection, and correlates of transmission across Senegal.

We here analyze data from the first year of an ongoing nationwide program of genetic surveillance of Plasmodium falciparum parasites in Senegal. The analysis is based on 1097 samples collected at health facilities during passive malaria case detection in 2019; it provides a baseline for analyzing parasite genetic metrics as they vary over time and geographic space. The study's goal was to identify genetic metrics that were informative about transmission intensity and other aspects of transmission dynamics, focusing on measures of genetic relatedness between parasites. We found the best genetic proxy for local malaria incidence to be the proportion of polygenomic infections (those with multiple genetically distinct parasites), although this relationship broke down at low incidence. The proportion of related parasites was less correlated with incidence while local genetic diversity was uninformative. The type of relatedness could discriminate local transmission patterns: two nearby areas had similarly high fractions of relatives, but one was dominated by clones and the other by outcrossed relatives. Throughout Senegal, 58% of related parasites belonged to a single network of relatives, within which parasites were enriched for shared haplotypes at known and suspected drug resistance loci and at one novel locus, reflective of ongoing selection pressure.

Malaria surveillance reveals parasite relatedness, signatures of selection, and correlates of transmission across Senegal.

Parasite genetic surveillance has the potential to play an important role in malaria control. We describe here an analysis of data from the first year of an ongoing, nationwide program of genetic surveillance of Plasmodium falciparum parasites in Senegal, intended to provide actionable information for malaria control efforts. Looking for a good proxy for local malaria incidence, we found that the best predictor was the proportion of polygenomic infections (those with multiple genetically distinct parasites), although that relationship broke down in very low incidence settings (r = 0.77 overall). The proportion of closely related parasites in a site was more weakly correlated ( r = -0.44) with incidence while the local genetic diversity was uninformative. Study of related parasites indicated their potential for discriminating local transmission patterns: two nearby study areas had similarly high fractions of relatives, but one area was dominated by clones and the other by outcrossed relatives. Throughout the country, 58% of related parasites proved to belong to a single network of relatives, within which parasites were enriched for shared haplotypes at known and suspected drug resistance loci as well as at one novel locus, reflective of ongoing selection pressure.

Two decades of molecular surveillance in Senegal reveal changes in known drug resistance mutations associated with historical drug use and seasonal malaria chemoprevention.

Drug resistance in Plasmodium falciparum is a major threat to malaria control efforts. We analyzed data from two decades (2000-2020) of continuous molecular surveillance of P. falciparum parasite strains in Senegal to determine how historical changes in drug administration policy may have affected parasite evolution. We profiled several known drug resistance markers and their surrounding haplotypes using a combination of single nucleotide polymorphism (SNP) molecular surveillance and whole-genome sequence (WGS) based population genomics. We observed rapid changes in drug resistance markers associated with the withdrawal of chloroquine and introduction of sulfadoxine-pyrimethamine in 2003. We also observed a rapid increase in Pfcrt K76T and decline in Pfdhps A437G starting in 2014, which we hypothesize may reflect changes in resistance or fitness caused by seasonal malaria chemoprevention (SMC). Parasite populations evolve rapidly in response to drug use, and SMC preventive efficacy should be closely monitored.

Survival of eight LLIN brands 6, 12, 24 and 36 months after a mass distribution campaign in rural and urban settings in Senegal.

BACKGROUND: Long lasting insecticidal nets (LLIN) are one of the core components of global malaria prevention and control. The lifespan of LLIN varies widely depending on the population or environment, and randomized studies are required to compare LLIN inaccording to arbitrary thresholds households under different field conditions. This study investigated survival of different LLIN brands in Senegal. METHODS: Ten thousand six hundred eight LLINs were distributed in five regions, each stratified by rural and urban setting. As part of the longitudinal follow-up, 2222 nets were randomly sampled and monitored from 6 to 36 months. Using random effects for households, Bayesian models were used to estimate independent survival by net type (Interceptor®, Life Net®, MAGNet™, Netprotect®, Olyset® Net, PermaNet® 2.0 R, PermaNet® 2.0 C, Yorkool® LN) and by area (rural/urban). In addition to survival, median survival time and attrition of each LLIN brand was determined. Attrition was defined as nets that were missing because they were reported given away, destroyed and thrown away, or repurposed. RESULTS: Three net types had a proportion of survival above 80% after 24 months: Interceptor®87.8% (95% CI 80-93.4); conical PermaNet® 2.0 86.9% (95% CI 79.3-92.4) and Life Net® 85.6% (95% CI 75-93). At 36 months, conical PermaNet® 2.0 maintained a good survival rate, 79.5% (95% CI 65.9-88.8). The attrition due to redistributed nets showed that the two conical net types (PermaNet® 2.0 and Interceptor®) were more often retained by households and their median retention time was well above 3 years (median survival time = 3.5 years for PermaNet® 2.0 and median survival time = 4 years for Interceptor®). Despite this good retention, Interceptor® had weak physical integrity and its median survival due to wear and tear was below 3 years (median survival time = 2.4 years). The odds ratio of survival was 2.5 times higher in rural settings than in urban settings (OR 2.5; 95% CI 1.7-3.7). CONCLUSIONS: Differences in survival among LLIN may be driven by brand, shape or environmental setting. In this study in Senegal, conical PermaNet® 2.0 were retained in households while rectangular PermaNet® 2.0 had lower retention, suggesting that net shape may play a role in retention and should be further investigated. Distribution of preferred LLIN shape, accompanied by good communication on care and repair, could lead to increased effective lifespan, and allow for longer intervals between universal coverage campaigns.

Spatiotemporal Dynamic of the RTS,S/AS01 Malaria Vaccine Target Antigens in Senegal.

The RTS,S/AS01 malaria vaccine confers only moderate protection against malaria. Evidence suggests that the effectiveness of the RTS,S/AS01 vaccine depends upon the parasite population genetics, specifically regarding the circumsporozoite protein haplotypes in the population. We investigated Plasmodium falciparum circumsporozoite protein (PfCSP) gene sequences from two endemic sites in 2018 in Senegal. The PfCSP sequences were compared with those retrieved from the Pf3k genome database. In the central repeat region of PfCSP, the distribution of haplotypes differed significantly between the two study sites (Fisher's exact test, P < 0.001). No 3D7 vaccine strain haplotype was observed in this locus. In the C-terminal region, there was no significant difference in haplotypes distribution between Kedougou and Diourbel (Fischer's exact test, P = 0.122). The 3D7 haplotype frequency was 8.4% in early samples (2001-2011), but then it contracted in the subsequent years. The extensive plasticity of the P. falciparum genes coding the RTS,S/AS01 vaccine target antigens may influence the immune responses to circulating alleles. Monitoring the genetic diversity baseline and its dynamics over time and space would be instrumental in rationally improving the malaria RTS,S/AS01 vaccine and/or its implementation schedule.

Genetic surveillance for monitoring the impact of drug use on Plasmodium falciparum populations.

The use of antimalarial drugs is an effective strategy in the fight against malaria. However, selection of drug resistant parasites is a constant threat to the continued use of this approach. Antimalarial drugs are used not only to treat infections but also as part of population-level strategies to reduce malaria transmission toward elimination. While there is strong evidence that the ongoing use of antimalarial drugs increases the risk of the emergence and spread of drug-resistant parasites, it is less clear how population-level use of drug-based interventions like seasonal malaria chemoprevention (SMC) or mass drug administration (MDA) may contribute to drug resistance or loss of drug efficacy. Critical to sustained use of drug-based strategies for reducing the burden of malaria is the surveillance of population-level signals related to transmission reduction and resistance selection. Here we focus on Plasmodium falciparum and discuss the genetic signatures of a parasite population that are correlated with changes in transmission and related to drug pressure and resistance as a result of drug use. We review the evidence for MDA and SMC contributing to malaria burden reduction and drug resistance selection and examine the use and impact of these interventions in Senegal. Throughout we consider best strategies for ongoing surveillance of both population and resistance signals in the context of different parasite population parameters. Finally, we propose a roadmap for ongoing surveillance during population-level drug-based interventions to reduce the global malaria burden.

Genomic investigation of atypical malaria cases in Kanel, northern Senegal.

BACKGROUND: The diagnosis of malaria cases in regions where the malaria burden has decreased significantly and prevalence is very low is more challenging, in part because of reduced clinical presumption of malaria. The appearance of a cluster of malaria cases with atypical symptoms in Mbounguiel, a village in northern Senegal where malaria transmission is low, in September 2018 exemplifies this scenario. The collaboration between the National Malaria Control Programme (NMCP) at the Senegal Ministry of Health and the Laboratory of Parasitology and Mycology at Cheikh Anta Diop University worked together to evaluate this cluster of malaria cases using molecular and serological tools. METHODS: Malaria cases were diagnosed primarily by rapid diagnostic test (RDT), and confirmed by photo-induced electron transfer-polymerase chain reaction (PET-PCR). 24 single nucleotide polymorphisms (SNPs) barcoding was used for Plasmodium falciparum genotyping. Unbiased metagenomic sequencing and Luminex-based multi-pathogen antibody and antigen profiling were used to assess exposure to other pathogens. RESULTS: Nine patients, of 15 suspected cases, were evaluated, and all nine samples were found to be positive for P. falciparum only. The 24 SNPs molecular barcode showed the predominance of polygenomic infections, with identifiable strains being different from one another. All patients tested positive for the P. falciparum antigens. No other pathogenic infection was detected by either the serological panel or metagenomic sequencing. CONCLUSIONS: This work, undertaken locally within Senegal as a collaboration between the NMCP and a research laboratory at University of Cheikh Anta Diop (UCAD) revealed that a cluster of malaria cases were caused by different strains of P. falciparum. The public health response in real time demonstrates the value of local molecular and genomics capacity in affected countries for disease control and elimination.

Proactive community case management in Senegal 2014-2016: a case study in maximizing the impact of community case management of malaria.

The Senegal National Malaria Control Programme (NMCP) introduced home-based malaria management for all ages, with diagnosis by rapid diagnostic test (RDT) and treatment with artemisinin-based combination therapy (ACT) in 2008, expanding to over 2000 villages nationwide by 2014. With prise en charge à domicile (PECADOM), community health workers (CHWs) were available for community members to seek care, but did not actively visit households to find cases. A trial of a proactive model (PECADOM Plus), in which CHWs visited all households in their village weekly during transmission season to identify fever cases and offer case management, in addition to availability during the week for home-based management, found that CHWs detected and treated more cases in intervention villages, while the number of cases detected weekly decreased over the transmission season. The NMCP scaled PECADOM Plus to three districts in 2014 (132 villages), to a total of six districts in 2015 (246 villages), and to a total of 16 districts in 2016 (708 villages). A narrative case study with programmatic results is presented. During active sweeps over approximately 20 weeks, CHWs tested a mean of 77 patients per CHW in 2014, 89 patients per CHW in 2015, and 90 patients per CHW in 2016, and diagnosed a mean of 61, 61 and 43 patients with malaria per CHW in 2014, 2015 and 2016, respectively. The number of patients who sought care between sweeps increased, with a 104% increase in the number of RDTs performed and a 77% increase in the number of positive tests and patients treated with ACT during passive case detection. While the number of CHWs increased 7%, the number of patients receiving an RDT increased by 307% and the number of malaria cases detected and treated by CHWs increased 274%, from the year prior to PECADOM Plus introduction to its first year of implementation. Based on these results, approximately 700 additional CHWs in 24 new districts were added in 2017. This case study describes the process, results and lessons learned from Senegal's implementation of PECADOM Plus, as well as guidance for other programmes considering introduction of this innovative strategy.

Analysis of anti-Plasmodium IgG profiles among Fulani nomadic pastoralists in northern Senegal to assess malaria exposure.

BACKGROUND: Northern Senegal is a zone of very low malaria transmission, with an annual incidence of < 5/1000 inhabitants. This area, where the Senegal National Malaria Control Programme has initiated elimination activities, hosts Fulani, nomadic, pastoralists that spend the dry season in the south where malaria incidence is higher (150-450/1000 inhabitants) and return to the north with the first rains. Previous research demonstrated parasite prevalence of < 1% in this Fulani population upon return from the south, similar to that documented in the north in cross-sectional surveys. METHODS: A modified snowball sampling survey of nomadic pastoralists was conducted in five districts in northern Senegal during September and October 2014. Demographic information and dried blood spots were collected. Multiplex bead-based assays were used to assess antibody responses to merozoite surface protein (MSP-119) antigen of the four primary Plasmodium species, as well as circumsporozoite protein (CSP) and liver stage antigen (LSA-1) of Plasmodium falciparum. RESULTS: In the five study districts, 1472 individuals were enrolled, with a median age of 22 years (range 1 to 80 years). Thirty-two percent of subjects were under 14 years and 57% were male. The overall seroprevalence of P. falciparum MSP-119, CSP and LSA-1 antibodies were 45, 12 and 5%, respectively. Plasmodium falciparum MSP-119 antibody responses increased significantly with age in all study areas, and were significantly higher among males. The highest seroprevalence to P. falciparum antigens was observed in the Kanel district (63%) and the lowest observed in Podor (28%). Low seroprevalence was observed for non-falciparum species in all the study sites: 0.4, 0.7 and 1.8%, respectively, for Plasmodium ovale, Plasmodium vivax and Plasmodium malariae MSP-1. Antibody responses to P. vivax were observed in all study sites except Kanel. CONCLUSION: Prevalence of P. falciparum MSP-119 antibodies and increases by study participant age provided data for low levels of exposure among this transient nomadic population. In addition, antibody responses to P. falciparum short half-life markers (CSP and LSA-1) and non-falciparum species were low. Further investigations are needed to understand the exposure of the Fulani population to P. vivax.

Serological Data Shows Low Levels of Chikungunya Exposure in Senegalese Nomadic Pastoralists.

The chikungunya virus (CHIKV) is spread by Aedes aegypti and Ae. albopictus mosquitos worldwide; infection can lead to disease including joint pain, fever, and rash, with some convalescent persons experiencing chronic symptoms. Historically, CHIKV transmission has occurred in Africa and Asia, but recent outbreaks have taken place in Europe, Indonesia, and the Americas. From September to October 2014, a survey was undertaken with nomadic pastoralists residing in the northeast departments of Senegal. Blood dried on filter paper (dried blood spots; DBS) were collected from 1465 participants of all ages, and assayed for Immunoglobulin G (IgG) antibodies against CHIKV E1 antigen by a bead-based multiplex assay. The overall seroprevalence of all participants to CHIKV E1 was 2.7%, with no persons under 10 years of age found to be antibody positive. Above 10 years of age, clear increases of seroprevalence and IgG levels were observed with increasing age; 7.6% of participants older than 50 years were found to be positive for anti-CHIKV IgG. Reported net ownership, net usage, and gender were all non-significant explanatory variables of seropositivity. These data show a low-level historical exposure of this pastoralist population to CHIKV, with no evidence of recent CHIKV transmission in the past decade.

Malaria prevalence, prevention and treatment seeking practices among nomadic pastoralists in northern Senegal.

BACKGROUND: Malaria transmission in Senegal is highly stratified, from low in the dry north to moderately high in the moist south. In northern Senegal, along the Senegal River Valley and in the Ferlo semi-desert region, annual incidence is less than five cases per 1000 inhabitants. Many nomadic pastoralists have permanent dwellings in the Ferlo Desert and Senegal River Valley, but spend dry season in the south with their herds, returning north when the rains start, leading to a concern that this population could contribute to ongoing transmission in the north. METHODS: A modified snowball sampling survey was conducted at six sites in northern Senegal to determine the malaria prevention and treatment seeking practices and parasite prevalence among nomadic pastoralists in the Senegal River Valley and the Ferlo Desert. Nomadic pastoralists aged 6 months and older were surveyed during September and October 2014, and data regarding demographics, access to care and preventive measures were collected. Parasite infection was detected using rapid diagnostic tests (RDTs), microscopy (thin and thick smears) and polymerase chain reaction (PCR). Molecular barcodes were determined by high resolution melting (HRM). RESULTS: Of 1800 participants, 61% were male. Sixty-four percent had at least one bed net in the household, and 53% reported using a net the night before. Only 29% had received a net from a mass distribution campaign. Of the 8% (142) who reported having had fever in the last month, 55% sought care, 20% of whom received a diagnostic test, one-third of which (n = 5) were reported to be positive. Parasite prevalence was 0.44% by thick smear and 0.50% by PCR. None of the molecular barcodes identified among the nomadic pastoralists had been previously identified in Senegal. CONCLUSIONS: While access to and utilization of malaria control interventions among nomadic pastoralists was lower than the general population, parasite prevalence was lower than expected and sheds doubt on the perception that they are a source of ongoing transmission in the north. The National Malaria Control Program is making efforts to improve access to malaria prevention and case management for nomadic populations.

Effectiveness of Seasonal Malaria Chemoprevention in Children under Ten Years of Age in Senegal: A Stepped-Wedge Cluster-Randomised Trial.

BACKGROUND: Seasonal Malaria Chemoprevention (SMC) with sulfadoxine-pyrimethamine (SP) plus amodiaquine (AQ), given each month during the transmission season, is recommended for children living in areas of the Sahel where malaria transmission is highly seasonal. The recommendation for SMC is currently limited to children under five years of age, but, in many areas of seasonal transmission, the burden in older children may justify extending this age limit. This study was done to determine the effectiveness of SMC in Senegalese children up to ten years of age. METHODS AND FINDINGS: SMC was introduced into three districts over three years in central Senegal using a stepped-wedge cluster-randomised design. A census of the population was undertaken and a surveillance system was established to record all deaths and to record all cases of malaria seen at health facilities. A pharmacovigilance system was put in place to detect adverse drug reactions. Fifty-four health posts were randomised. Nine started implementation of SMC in 2008, 18 in 2009, and a further 18 in 2010, with 9 remaining as controls. In the first year of implementation, SMC was delivered to children aged 3-59 months; the age range was then extended for the latter two years of the study to include children up to 10 years of age. Cluster sample surveys at the end of each transmission season were done to measure coverage of SMC and the prevalence of parasitaemia and anaemia, to monitor molecular markers of drug resistance, and to measure insecticide-treated net (ITN) use. Entomological monitoring and assessment of costs of delivery in each health post and of community attitudes to SMC were also undertaken. About 780,000 treatments were administered over three years. Coverage exceeded 80% each month. Mortality, the primary endpoint, was similar in SMC and control areas (4.6 and 4.5 per 1000 respectively in children under 5 years and 1.3 and 1.2 per 1000 in children 5-9 years of age; the overall mortality rate ratio [SMC: no SMC] was 0.90, 95% CI 0.68-1.2, p = 0.496). A reduction of 60% (95% CI 54%-64%, p < 0.001) in the incidence of malaria cases confirmed by a rapid diagnostic test (RDT) and a reduction of 69% (95% CI 65%-72%, p < 0.001) in the number of treatments for malaria (confirmed and unconfirmed) was observed in children. In areas where SMC was implemented, incidence of confirmed malaria in adults and in children too old to receive SMC was reduced by 26% (95% CI 18%-33%, p < 0.001) and the total number of treatments for malaria (confirmed and unconfirmed) in these older age groups was reduced by 29% (95% CI 21%-35%, p < 0.001). One hundred and twenty-three children were admitted to hospital with a diagnosis of severe malaria, with 64 in control areas and 59 in SMC areas, showing a reduction in the incidence rate of severe disease of 45% (95% CI 5%-68%, p = 0.031). Estimates of the reduction in the prevalence of parasitaemia at the end of the transmission season in SMC areas were 68% (95% CI 35%-85%) p = 0.002 in 2008, 84% (95% CI 58%-94%, p < 0.001) in 2009, and 30% (95% CI -130%-79%, p = 0.56) in 2010. SMC was well tolerated with no serious adverse reactions attributable to SMC drugs. Vomiting was the most commonly reported mild adverse event but was reported in less than 1% of treatments. The average cost of delivery was US$0.50 per child per month, but varied widely depending on the size of the health post. Limitations included the low rate of mortality, which limited our ability to detect an effect on this endpoint. CONCLUSIONS: SMC substantially reduced the incidence of outpatient cases of malaria and of severe malaria in children, but no difference in all-cause mortality was observed. Introduction of SMC was associated with an overall reduction in malaria incidence in untreated age groups. In many areas of Africa with seasonal malaria, there is a substantial burden in older children that could be prevented by SMC. SMC in older children is well tolerated and effective and can contribute to reducing malaria transmission. TRIAL REGISTRATION: ClinicalTrials.gov NCT00712374.

Modeling malaria genomics reveals transmission decline and rebound in Senegal.

To study the effects of malaria-control interventions on parasite population genomics, we examined a set of 1,007 samples of the malaria parasite Plasmodium falciparum collected in Thiès, Senegal between 2006 and 2013. The parasite samples were genotyped using a molecular barcode of 24 SNPs. About 35% of the samples grouped into subsets with identical barcodes, varying in size by year and sometimes persisting across years. The barcodes also formed networks of related groups. Analysis of 164 completely sequenced parasites revealed extensive sharing of genomic regions. In at least two cases we found first-generation recombinant offspring of parents whose genomes are similar or identical to genomes also present in the sample. An epidemiological model that tracks parasite genotypes can reproduce the observed pattern of barcode subsets. Quantification of likelihoods in the model strongly suggests a reduction of transmission from 2006-2010 with a significant rebound in 2012-2013. The reduced transmission and rebound were confirmed directly by incidence data from Thiès. These findings imply that intensive intervention to control malaria results in rapid and dramatic changes in parasite population genomics. The results also suggest that genomics combined with epidemiological modeling may afford prompt, continuous, and cost-effective tracking of progress toward malaria elimination.

Safety monitoring of artemisinin combination therapy through a national pharmacovigilance system in an endemic malaria setting.

BACKGROUND: The National Malaria Control Programme in Senegal, introduced since 2006, artemisinin-based combination therapy (ACT administration) for the treatment of uncomplicated malaria cases. In this framework, an anti-malarial pharmacovigilance plan was developed and implemented in all public health services. This study investigated the occurrence of Adverse Drug Events (ADEs) after ACT. METHODS: The study was conducted between January 2007 and December 2009. It was based on spontaneous reports of ADEs in public health facilities. Data on patient demographic characteristics, dispensing facility, adverse signs and symptoms and causality were collected from a total of 123 patients. RESULTS: The age range of these patients was six months to 93 years with a mean of 25.9 years. Of the reported symptoms, 46.7% were related to the abdomen and the digestive system. Symptoms related to the nervous system, skin and subcutaneous tissue, circulatory and respiratory systems and general symptoms and signs were 7%, 9.7%, 3.5% and 31.3%, respectively. Causality results linked 14.3% of symptoms to Falcimon® (Artesunate-Amodiaquine) with certainty. Effects were classified as mild and severe in 69.1% and 7.3% of cases respectively while 23.6% were serious. All patients with serious ADEs were hospitalized. One death was reported in a patient who had taken 24 pills at once. CONCLUSION: These results confirm the need to develop and implement pharmacovigilance systems in malaria endemic countries in order to monitor the safety of anti-malarial treatments.

Major reduction in anti-malarial drug consumption in Senegal after nation-wide introduction of malaria rapid diagnostic tests.

BACKGROUND: While WHO recently recommended universal parasitological confirmation of suspected malaria prior to treatment, debate has continued as to whether wide-scale use of rapid diagnostic tests (RDTs) can achieve this goal. Adherence of health service personnel to RDT results has been poor in some settings, with little impact on anti-malarial drug consumption. The Senegal national malaria control programme introduced universal parasite-based diagnosis using malaria RDTs from late 2007 in all public health facilities. This paper assesses the impact of this programme on anti-malarial drug consumption and disease reporting. METHODS AND FINDINGS: Nationally-collated programme data from 2007 to 2009 including malaria diagnostic outcomes, prescription of artemisinin-based combination therapy (ACT) and consumption of RDTs in public health facilities, were reviewed and compared. Against a marked seasonal variation in all-cause out-patient visits, non-malarial fever and confirmed malaria, parasite-based diagnosis increased nationally from 3.9% of reported malaria-like febrile illness to 86.0% over a 3 year period. The prescription of ACT dropped throughout this period from 72.9% of malaria-like febrile illness to 31.5%, reaching close equivalence to confirmed malaria (29.9% of 584,873 suspect fever cases). An estimated 516,576 courses of inappropriate ACT prescription were averted. CONCLUSIONS: The data indicate high adherence of anti-malarial prescribing practice to RDT results after an initial run-in period. The large reduction in ACT consumption enabled by the move from symptom-based to parasite-based diagnosis demonstrates that effective roll-out and use of malaria RDTs is achievable on a national scale through well planned and structured implementation. While more detailed information on management of parasite-negative cases is required at point of care level to assess overall cost-benefits to the health sector, considerable cost-savings were achieved in ACT procurement. Programmes need to be allowed flexibility in management of these funds to address increases in other programmatic costs that may accrue from improved diagnosis of febrile disease.