Potent neutralization of Rift Valley fever virus by human monoclonal antibodies through fusion inhibition.

Rift Valley fever virus (RVFV), an emerging arboviral and zoonotic bunyavirus, causes severe disease in livestock and humans. Here, we report the isolation of a panel of monoclonal antibodies (mAbs) from the B cells of immune individuals following natural infection in Kenya or immunization with MP-12 vaccine. The B cell responses of individuals who were vaccinated or naturally infected recognized similar epitopes on both Gc and Gn proteins. The Gn-specific mAbs and two mAbs that do not recognize either monomeric Gc or Gn alone but recognized the hetero-oligomer glycoprotein complex (Gc+Gn) when Gc and Gn were coexpressed exhibited potent neutralizing activities in vitro, while Gc-specific mAbs exhibited relatively lower neutralizing capacity. The two Gc+Gn-specific mAbs and the Gn domain A-specific mAbs inhibited RVFV fusion to cells, suggesting that mAbs can inhibit the exposure of the fusion loop in Gc, a class II fusion protein, and thus prevent fusion by an indirect mechanism without direct fusion loop contact. Competition-binding analysis with coexpressed Gc/Gn and mutagenesis library screening indicated that these mAbs recognize four major antigenic sites, with two sites of vulnerability for neutralization on Gn. In experimental models of infection in mice, representative mAbs recognizing three of the antigenic sites reduced morbidity and mortality when used at a low dose in both prophylactic and therapeutic settings. This study identifies multiple candidate mAbs that may be suitable for use in humans against RVFV infection and highlights fusion inhibition against bunyaviruses as a potential contributor to potent antibody-mediated neutralization.

Spatiotemporal overlapping of dengue, chikungunya, and malaria infections in children in Kenya.

Malaria, chikungunya virus (CHIKV), and dengue virus (DENV) are endemic causes of fever among children in Kenya. The risks of infection are multifactorial and may be influenced by built and social environments. The high resolution overlapping of these diseases and factors affecting their spatial heterogeneity has not been investigated in Kenya. From 2014-2018, we prospectively followed a cohort of children from four communities in both coastal and western Kenya. Overall, 9.8% were CHIKV seropositive, 5.5% were DENV seropositive, and 39.1% were malaria positive (3521 children tested). The spatial analysis identified hot-spots for all three diseases in each site and in multiple years. The results of the model showed that the risk of exposure was linked to demographics with common factors for the three diseases including the presence of litter, crowded households, and higher wealth in these communities. These insights are of high importance to improve surveillance and targeted control of mosquito-borne diseases in Kenya.

Pupal productivity of larval habitats of Aedes aegypti in Msambweni, Kwale County, Kenya.

Aedes aegypti is an important vector of several arboviruses including dengue and chikungunya viruses. Accurate identification of larval habitats of Ae. aegypti is considered an essential step in targeted control. This study determined Ae. aegypti productivity in selected larval habitats in Msambweni, Kwale County, Kenya. Three sequential larval habitat surveys were conducted. The first survey was habitat census (baseline) through which 83 representative larval habitats were identified and selected. The second and third surveys involved estimating daily productivity of the 83 selected larval habitats for 30 consecutive days during a wet and a dry season, respectively. Of 664 larval habitats examined at baseline, 144 larval habitats (21.7%) were found to be infested with Ae. aegypti larvae. At baseline, majority (71%) of the pupae were collected from two (2/6) larval habitat types, tires and pots. Multivariate analysis identified habitat type and the habitat being movable as the predictors for pupal abundance. During the 30-day daily pupal production surveys, only a few of the habitats harbored pupae persistently. Pupae were found in 28% and 12% of the larval habitats during the wet and dry seasons, respectively. In the wet season, drums, tires, and pots were identified as the key habitat types accounting for 85% of all pupae sampled. Three habitats (all drums) accounted for 80% of all the pupae collected in the dry season. Predictors for pupal productivity in the wet season were habitat type, place (whether the habitat is located at the back or front of the house), habitat purpose (use of the water in the habitat), and source of water. Although the multivariate model for habitat type did not converge, habitat type and habitat size were the only significant predictors during the dry season. Drums, pots, and tires were sources of more than 85% of Ae. aegypti pupae, reinforcing the "key container concept." Targeting these three types of habitats makes epidemiological sense, especially during the dry season.

Climate change and global health: A call to more research and more action.

There is increasing understanding, globally, that climate change and increased pollution will have a profound and mostly harmful effect on human health. This review brings together international experts to describe both the direct (such as heat waves) and indirect (such as vector-borne disease incidence) health impacts of climate change. These impacts vary depending on vulnerability (i.e., existing diseases) and the international, economic, political, and environmental context. This unique review also expands on these issues to address a third category of potential longer-term impacts on global health: famine, population dislocation, and environmental justice and education. This scholarly resource explores these issues fully, linking them to global health in urban and rural settings in developed and developing countries. The review finishes with a practical discussion of action that health professionals around the world in our field can yet take.

High Frequency of Antibiotic Prescription in Children With Undifferentiated Febrile Illness in Kenya.

BACKGROUND: In low-resource, malaria-endemic settings, accurate diagnosis of febrile illness in children is challenging. The World Health Organization (WHO) currently recommends laboratory-confirmed diagnosis of malaria prior to starting treatment in stable children. Factors guiding management of children with undifferentiated febrile illness outside of malaria are not well understood. METHODS: This study examined clinical presentation and management of a cohort of febrile Kenyan children at 5 hospital/clinic sites from January 2014 to December 2017. Chi-squared and multivariate regression analyses were used to compare frequencies and correlate demographic, environmental, and clinical factors with patient diagnosis and prescription of antibiotics. RESULTS: Of 5735 total participants, 68% were prescribed antibiotic treatment (n = 3902), despite only 28% given a diagnosis of bacterial illness (n = 1589). Factors associated with prescription of antibiotic therapy included: negative malaria testing, reporting head, ears, eyes, nose and throat (HEENT) symptoms (ie, cough, runny nose), HEENT findings on exam (ie, nasal discharge, red throat), and having a flush toilet in the home (likely a surrogate for higher socioeconomic status). CONCLUSION: In a cohort of acutely ill Kenyan children, prescription of antimalarial therapy and malaria test results were well correlated, whereas antibiotic treatment was prescribed empirically to most of those who tested malaria negative. Clinical management of febrile children in these settings is difficult, given the lack of diagnostic testing. Providers may benefit from improved clinical education and implementation of enhanced guidelines in this era of malaria testing, as their management strategies must rely primarily on critical thinking and decision-making skills.

High Dengue Burden and Circulation of 4 Virus Serotypes among Children with Undifferentiated Fever, Kenya, 2014-2017.

Little is known about the extent and serotypes of dengue viruses circulating in Africa. We evaluated the presence of dengue viremia during 4 years of surveillance (2014-2017) among children with febrile illness in Kenya. Acutely ill febrile children were recruited from 4 clinical sites in western and coastal Kenya, and 1,022 participant samples were tested by using a highly sensitive real-time reverse transcription PCR. A complete case analysis with genomic sequencing and phylogenetic analyses was conducted to characterize the presence of dengue viremia among participants during 2014-2017. Dengue viremia was detected in 41.9% (361/862) of outpatient children who had undifferentiated febrile illness in Kenya. Of children with confirmed dengue viremia, 51.5% (150/291) had malaria parasitemia. All 4 dengue virus serotypes were detected, and phylogenetic analyses showed several viruses from novel lineages. Our results suggests high levels of dengue virus infection among children with undifferentiated febrile illness in Kenya.

Strengthening the Interaction of the Virology Community with the International Committee on Taxonomy of Viruses (ICTV) by Linking Virus Names and Their Abbreviations to Virus Species.

The International Committee on Taxonomy of Viruses (ICTV) is tasked with classifying viruses into taxa (phyla to species) and devising taxon names. Virus names and virus name abbreviations are currently not within the ICTV's official remit and are not regulated by an official entity. Many scientists, medical/veterinary professionals, and regulatory agencies do not address evolutionary questions nor are they concerned with the hierarchical organization of the viral world, and therefore, have limited use for ICTV-devised taxa. Instead, these professionals look to the ICTV as an expert point source that provides the most current taxonomic affiliations of viruses of interests to facilitate document writing. These needs are currently unmet as an ICTV-supported, easily searchable database that includes all published virus names and abbreviations linked to their taxa is not available. In addition, in stark contrast to other biological taxonomic frameworks, virus taxonomy currently permits individual species to have several members. Consequently, confusion emerges among those who are not aware of the difference between taxa and viruses, and because certain well-known viruses cannot be located in ICTV publications or be linked to their species. In addition, the number of duplicate names and abbreviations has increased dramatically in the literature. To solve this conundrum, the ICTV could mandate listing all viruses of established species and all reported unclassified viruses in forthcoming online ICTV Reports and create a searchable webpage using this information. The International Union of Microbiology Societies could also consider changing the mandate of the ICTV to include the nomenclature of all viruses in addition to taxon considerations. With such a mandate expansion, official virus names and virus name abbreviations could be catalogued and virus nomenclature could be standardized. As a result, the ICTV would become an even more useful resource for all stakeholders in virology.

Factors associated with early childhood stunted growth in a 2012-2015 birth cohort monitored in the rural Msambweni area of coastal Kenya: a cross-sectional study.

BACKGROUND: Chronic malnutrition, often measured as stunted growth, is an understudied global health problem. Though poor nutritional intake has been linked to stunted growth, there is evidence suggesting environmental exposures may have a significant role in its occurrence. Here, we characterize the non-nutritional prenatal and postnatal factors that contribute to early childhood stunted growth in rural coastal Kenya. METHODS: Overall, 232 women and 244 children from a 2012-2015 maternal-child cohort in Msambweni, Kenya were included. Women were tested for parasitic infections during the prenatal period and at the time of delivery. Children were tested for parasitic infections and assessed for stunted growth using height-for-age Z-scores (HAZ) at 6-month intervals after birth. Socioeconomic status (SES) was evaluated using both a simplified water, asset, maternal education, and income (WAMI) index and a principal component analysis (PCA) asset score. Multivariate logistic regression analysis was used to determine the relative influence of prenatal and postnatal factors on the occurrence of stunted growth. RESULTS: Of the 244 children (ages 6-37 months), 60 (25%) were stunted at the study endpoint. 179 mothers (77%) had at least one parasitic infection during pregnancy and 94 children (38%) had at least one parasitic infection during the study period. There was no significant association between maternal parasitic infection and child stunted growth (p = 1.00). SES as determined using the WAMI index was not associated with HAZ in linear regression analysis (p = 0.307), however, the PCA asset score was (p = 0.048). Multivariate logistic regression analysis identified low birth weight (AOR: 3.24, 95% CI: [1.38, 7.57]) and child parasitic infectious disease burden (AOR: 1.41, 95% CI: [1.05, 1.95]) as independent predictors of stunted growth, though no significant association was identified with PCA asset score (AOR: 0.98, 95% CI: [0.88, 1.10]). CONCLUSIONS: Stunted growth remains highly prevalent in rural Kenya, with low birth weight and child parasitic infectious disease burden demonstrated to be significantly associated with this indicator of chronic malnutrition. These results emphasize the multifaceted nature of stunted growth and the need to address both the prenatal and postnatal environmental factors that contribute to this problem.

Cord Blood Antiparasite Interleukin 10 as a Risk Marker for Compromised Vaccine Immunogenicity in Early Childhood.

Background: Antenatal exposure to parasites can affect infants' subsequent responses to vaccination. The present study investigated how maternal prenatal infections and newborns' antiparasite cytokine profiles relate to immunoglobulin G (IgG) responses to standard vaccination during infancy. Methods: A total of 450 Kenyan women were tested for parasitic infections during pregnancy. Their newborns' responses to Plasmodium falciparum, schistosome, and filaria antigens were assessed in cord blood lymphocytes. Following standard neonatal vaccination, this infant cohort was followed biannually to age 30 months for measurement of circulating IgG levels against Haemophilus influenzae b (Hib), diphtheria toxoid (DT), hepatitis B virus (HBV), and tetanus toxoid. Results: Trajectories of postvaccination IgG levels were classified by functional principal component (PC) analysis to assess each child's response profile. Two main components, PC1, reflecting height of response over time, and PC2, reflecting crossover from high to low responses or from low to high responses, were identified. Cord blood cytokine responses to schistosome and filarial antigens showed a significant association between augmented antihelminth interleukin 10 and reduced antibody levels, particularly to DT and HBV, and a more rapid postvaccination decline in circulating IgG levels against Hib. Conclusion: Antenatal sensitization to schistosomiasis or filariasis and related production of antiparasite interleukin 10 at birth are associated with reduced antivaccine IgG levels in infancy, with possibly impaired protection.

Clinical, Serological, and Molecular Observations from a Case Series Study during the Asian Lineage Zika Virus Outbreak in Grenada during 2016.

This paper describes the spatial and temporal distribution of cases, demographic characteristics of patients, and clinical manifestations of Zika virus (ZIKV) during the 2016 outbreak in Grenada. The first reported case was recorded in St. Andrew Parish in April, and the last reported case was seen in November, with peak transmission occurring in the last week of June, based on test results. Data were collected from a total of 514 patients, of whom 207 (40%) tested positive for ZIKV. No evidence was found that testing positive for ZIKV infection was related to age, gender, or pregnancy status. Clinical presentation with rash (OR = 2.4, 95% CI = 1.5 to 3.7) or with lymphadenopathy (OR = 1.7, 95% CI = 1.0 to 2.9) were the only reported symptoms consistent with testing positive for ZIKV infection. During the Zika outbreak, the infection rate was 20 clinical cases per 10,000 in the population compared to 41 cases per 10,000 during the chikungunya outbreak in Grenada in 2014 and 17 cases per 10,000 during the dengue outbreak in 2001-2002. Even though the country has employed vector control programs, with no apparent decrease in infection rates, it appears that new abatement approaches are needed to minimize morbidity in future arbovirus outbreaks.

Unrecognized Dengue Virus Infections in Children, Western Kenya, 2014-2015.

We detected a cluster of dengue virus infections in children in Kenya during July 2014-June 2015. Most cases were serotype 1, but we detected all 4 serotypes, including co-infections with 2 serotypes. Our findings implicate dengue as a cause of febrile illness in this population and highlight a need for robust arbovirus surveillance.

Principles, practices and knowledge of clinicians when assessing febrile children: a qualitative study in Kenya.

BACKGROUND: Clinicians in low resource settings in malaria endemic regions face many challenges in diagnosing and treating febrile illnesses in children. Given the change in WHO guidelines in 2010 that recommend malaria testing prior to treatment, clinicians are now required to expand the differential when malaria testing is negative. Prior studies have indicated that resource availability, need for additional training in differentiating non-malarial illnesses, and lack of understanding within the community of when to seek care play a role in effective diagnosis and treatment. The objective of this study was to examine the various factors that influence clinician behavior in diagnosing and managing children presenting with fever to health centres in Kenya. METHODS: A total of 20 clinicians (2 paediatricians, 1 medical officer, 2 nurses, and 15 clinical officers) were interviewed, working at 5 different government-sponsored public clinic sites in two areas of Kenya where malaria is prevalent. Clinicians were interviewed one-on-one using a structured interview technique. Interviews were then analysed qualitatively for themes. RESULTS: The following five themes were identified: (1) Strong familiarity with diagnosis of malaria and testing for malaria; (2) Clinician concerns about community understanding of febrile illness, use of traditional medicine, delay in seeking care, and compliance; (3) Reliance on clinical guidelines, history, and physical examination to diagnose febrile illness and recognize danger signs; (4) Clinician discomfort with diagnosis of primary viral illness leading to increased use of empiric antibiotics; and (5) Lack of resources including diagnostic testing, necessary medications, and training modalities contributes to the difficulty clinicians face in assessing and treating febrile illness in children. These themes persisted across all sites, despite variation in levels of medical care. Within these themes, clinicians consistently expressed a need for reliable basic testing, especially haemograms and bacterial cultures. Clinicians discussed the use of counseling and education to improve community understanding of febrile illness in order to decrease preventable deaths in children. CONCLUSION: Results of this study suggest that since malarial testing has become more widespread, clinicians working in resource-poor environments still face difficulty when evaluating a child with fever, especially when malaria testing is negative. Improving access to additional diagnostics, continuing medical education, and ongoing evaluation and revision of clinical guidelines may lead to more consistent management of febrile illness by providers, and may potentially decrease prescription of unnecessary antibiotics. Additional interventions at the community level may also have an important role in managing febrile illness, however, more studies are needed to identify targets for intervention at both the clinic and community levels.

Clinical aspects of Zika virus.

PURPOSE OF REVIEW: Zika virus (ZIKV) is a mosquito-borne flavivirus that has caused a sudden and explosive outbreak in South America and the Caribbean in the last year, and has been declared a public health emergency by the WHO. As ZIKV afflicts previously naive populations, more severe clinical presentations and sequelae have been observed. A specific emphasis has been placed on the neurological effects in infants resulting from viral exposure in utero. RECENT FINDINGS: Acute onset of ZIKV disease is seen in approximately 20% of cases, whereas most individuals (80%) exposed are asymptomatic. Presentation of illness is typically mild, with disease spectrum ranging from arthralgia and rash to encephalitis, myelitis, and Guillain-Barré syndrome. Infants have been uniquely impacted by the current outbreak with significant congenital exposure resulting in permanent neurological defects and developmental complications. SUMMARY: The current ZIKV outbreak has illustrated the emergent capabilities of mosquito-borne viruses and the teratogenic nature of ZIKV. Causality and risk factors associated with severe manifestations, as well as chronic sequelae, have yet to be determined. Extensive research is required to understand the molecular mechanisms of infection, develop improved assays for differential diagnosis, and improve overall knowledge of the spectrum of ZIKV disease in order to develop modes of prevention and treatment.

Top 5 Things Health Professions Students Should Know About Ecology and Waste Management.

The environments in which we live affect individual and community risk for disease transmission and illness severity. Communities' and neighborhoods' waste stream management designs and health care organizations' spatial and structural architecture also influence individuals' and communities' pathogenic vulnerabilities and how well health sector industrial hygiene practices support them. This article describes a One Health approach to planetary environmental health and suggests strategies for implementing a One Health or Planetary Health approach in the context of climate change.

Sofosbuvir Off-label Treatment of Yellow Fever Patients During an Outbreak in Brazil, 2018: A Cohort Study.

We enrolled 21 patients with laboratory-confirmed yellow fever (YF), hospitalized at Eduardo de Menezes Hospital, Brazil, to be treated with sofosbuvir, a drug approved for hepatitis C. Given the absence of specific YF antiviral treatments, the off-label nonrandomized sofosbuvir treatment aimed to address high disease severity and the risk of fatal outcomes. Patients received a daily dose of 400 mg sofosbuvir from 4 to 10 days post-symptom onset. YF viral load (VL) comparisons were made between treated and nontreated patients who either survived or died. The genomic VL for the treated group steadily decreased after day 7 post-symptom onset, suggesting that sofosbuvir might reduce YF VL. This study underscores the urgent need for YF antiviral therapies, advocating for randomized clinical trials to further explore sofosbuvir's role in YF treatment.

Factors Associated with Chikungunya Infection among Pregnant Women in Grenada, West Indies.

Neonates are vulnerable to vector-borne diseases given the potential for mother-to-child congenital transmission. To determine factors associated with chikungunya virus (CHIKV) infection among pregnant women in Grenada, West Indies, a retrospective cohort study enrolled women who were pregnant during the 2014 CHIKV epidemic. In all, 520/688 women (75.5%) were CHIKV IgG positive. Low incomes, use of pit latrines, lack of home window screens, and subjective reporting of frequent mosquito bites were associated with increased risk of CHIKV infection in bivariate analyses. In the multivariate modified Poisson regression model, low income (adjusted relative risk [aRR]: 1.05 [95% CI: 1.01-1.10]) and frequent mosquito bites (aRR: 1.05 [95% CI: 1.01-1.10]) were linked to increased infection risk. In Grenada, markers of low socioeconomic status are associated with CHIKV infection among pregnant women. Given that Grenada will continue to face vector-borne outbreaks, interventions dedicated to improving living conditions of the most disadvantaged will help reduce the incidence of arboviral infections.

Does Intra-Uterine Exposure to the Zika Virus Increase Risks of Cognitive Delay at Preschool Ages? Findings from a Zika-Exposed Cohort from Grenada, West Indies.

Maternal infection with Zika virus (ZIKV) is associated with a distinct pattern of birth defects, known as congenital Zika syndrome (CZS). In ZIKV-exposed children without CZS, it is often unclear whether they were protected from in utero infection and neurotropism. Early neurodevelopmental assessment is essential for detecting neurodevelopmental delays (NDDs) and prioritizing at-risk children for early intervention. We compared neurodevelopmental outcomes between ZIKV-exposed and unexposed children at 1, 3 and 4 years to assess exposure-associated NDD risk. A total of 384 mother-child dyads were enrolled during a period of active ZIKV transmission (2016-2017) in Grenada, West Indies. Exposure status was based on laboratory assessment of prenatal and postnatal maternal serum. Neurodevelopment was assessed using the Oxford Neurodevelopment Assessment, the NEPSY® Second Edition and Cardiff Vision Tests, at 12 (n = 66), 36 (n = 58) and 48 (n = 59) months, respectively. There were no differences in NDD rates or vision scores between ZIKV-exposed and unexposed children. Rates of microcephaly at birth (0.88% vs. 0.83%, p = 0.81), and childhood stunting and wasting did not differ between groups. Our results show that Grenadian ZIKV-exposed children, the majority of whom were without microcephaly, had similar neurodevelopmental outcomes to unexposed controls up to at least an age of 4 years.

Exploring potential risk pathways with high risk groups for urban Rift Valley fever virus introduction, transmission, and persistence in two urban centers of Kenya.

Rift Valley fever virus (RVFV) is a zoonotic arbovirus that has profound impact on domestic ruminants and can also be transmitted to humans via infected animal secretions. Urban areas in endemic regions across Africa have susceptible animal and human hosts, dense vector distributions, and source livestock (often from high risk locations to meet the demand for animal protein). Yet, there has never been a documented urban outbreak of RVF. To understand the likely risk of RVFV introduction to urban communities from their perspective and guide future initiatives, we conducted focus group discussions with slaughterhouse workers, slaughterhouse animal product traders, and livestock owners in Kisumu City and Ukunda Town in Kenya. For added perspective and data triangulation, in-depth interviews were conducted one-on-one with meat inspector veterinarians from selected slaughterhouses. A theoretical framework relevant to introduction, transmission, and potential persistence of RVF in urban areas is presented here. Urban livestock were primarily mentioned as business opportunities, but also had personal sentiment. In addition to slaughtering risks, perceived risk factors included consumption of fresh milk. High risk groups' knowledge and experience with RVFV and other zoonotic diseases impacted their consideration of personal risk, with consensus towards lower risk in the urban setting compared to rural areas as determination of health risk was said to primarily rely on hygiene practices rather than the slaughtering process. Groups relied heavily on veterinarians to confirm animal health and meat safety, yet veterinarians reported difficulty in accessing RVFV diagnostics. We also identified vulnerable public health regulations including corruption in meat certification outside of the slaughterhouse system, and blood collected during slaughter being used for food and medicine, which could provide a means for direct RVFV community transmission. These factors, when compounded by diverse urban vector breeding habitats and dense human and animal populations, could create suitable conditions for RVFV to arrive an urban center via a viremic imported animal, transmit to locally owned animals and humans, and potentially adapt to secondary vectors and persist in the urban setting. This explorative qualitative study proposes risk pathways and provides initial insight towards determining how urban areas could adapt control measures and plan future initiatives to better understand urban RVF potential.

The Importance of Including Non-Household Environments in Dengue Vector Control Activities.

Most vector control activities in urban areas are focused on household environments; however, information relating to infection risks in spaces other than households is poor, and the relative risk that these spaces represent has not yet been fully understood. We used data-driven simulations to investigate the importance of household and non-household environments for dengue entomological risk in two Kenyan cities where dengue circulation has been reported. Fieldwork was performed using four strategies that targeted different stages of mosquitoes: ovitraps, larval collections, Prokopack aspiration, and BG-sentinel traps. Data were analyzed separately between household and non-household environments to assess mosquito presence, the number of vectors collected, and the risk factors for vector presence. With these data, we simulated vector and human populations to estimate the parameter m and mosquito-to-human density in both household and non-household environments. Among the analyzed variables, the main difference was found in mosquito abundance, which was consistently higher in non-household environments in Kisumu but was similar in Ukunda. Risk factor analysis suggests that small, clean water-related containers serve as mosquito breeding places in households as opposed to the trash- and rainfall-related containers found in non-household structures. We found that the density of vectors (m) was higher in non-household than household environments in Kisumu and was also similar or slightly lower between both environments in Ukunda. These results suggest that because vectors are abundant, there is a potential risk of transmission in non-household environments; hence, vector control activities should take these spaces into account.

Detection of acute dengue virus infection, with and without concurrent malaria infection, in a cohort of febrile children in Kenya, 2014-2019, by clinicians or machine learning algorithms.

Poor access to diagnostic testing in resource limited settings restricts surveillance for emerging infections, such as dengue virus (DENV), to clinician suspicion, based on history and exam observations alone. We investigated the ability of machine learning to detect DENV based solely on data available at the clinic visit. We extracted symptom and physical exam data from 6,208 pediatric febrile illness visits to Kenyan public health clinics from 2014-2019 and created a dataset with 113 clinical features. Malaria testing was available at the clinic site. DENV testing was performed afterwards. We randomly sampled 70% of the dataset to develop DENV and malaria prediction models using boosted logistic regression, decision trees and random forests, support vector machines, naïve Bayes, and neural networks with 10-fold cross validation, tuned to maximize accuracy. 30% of the dataset was reserved to validate the models. 485 subjects (7.8%) had DENV, and 3,145 subjects (50.7%) had malaria. 220 (3.5%) subjects had co-infection with both DENV and malaria. In the validation dataset, clinician accuracy for diagnosis of malaria was high (82% accuracy, 85% sensitivity, 80% specificity). Accuracy of the models for predicting malaria diagnosis ranged from 53-69% (35-94% sensitivity, 11-80% specificity). In contrast, clinicians detected only 21 of 145 cases of DENV (80% accuracy, 14% sensitivity, 85% specificity). Of the six models, only logistic regression identified any DENV case (8 cases, 91% accuracy, 5.5% sensitivity, 98% specificity). Without diagnostic testing, interpretation of clinical findings by humans or machines cannot detect DENV at 8% prevalence. Access to point-of-care diagnostic tests must be prioritized to address global inequities in emerging infections surveillance.