True versus apparent malaria infection prevalence: the contribution of a Bayesian approach

Assessment of malaria risk in Southeast Asia: a systematic review

BACKGROUND

Several countries in Southeast Asia are nearing malaria elimination, yet eradication remains elusive. This is largely due to the challenge of focusing elimination efforts, an area where risk prediction can play an essential supporting role. Despite its importance, there is no standard numerical method to quantify the risk of malaria infection. Thus, there is a need for a consolidated view of existing definitions of risk and factors considered in assessing risk to analyse the merits of risk prediction models. This systematic review examines studies of the risk of malaria in Southeast Asia with regard to their suitability in addressing the challenges of malaria elimination in low transmission areas.

METHODS

A search of four electronic databases over 2010-2020 retrieved 1297 articles, of which 25 met the inclusion and exclusion criteria. In each study, examined factors included the definition of the risk and indicators of malaria transmission used, the environmental and climatic factors associated with the risk, the statistical models used, the spatial and temporal granularity, and how the relationship between environment, climate, and risk is quantified.

RESULTS

This review found variation in the definition of risk used, as well as the environmental and climatic factors in the reviewed articles. GLM was widely adopted as the analysis technique relating environmental and climatic factors to malaria risk. Most of the studies were carried out in either a cross-sectional design or case-control studies, and most utilized the odds ratio to report the relationship between exposure to risk and malaria prevalence.

CONCLUSIONS

Adopting a standardized definition of malaria risk would help in comparing and sharing results, as would a clear description of the definition and method of collection of the environmental and climatic variables used. Further issues that need to be more fully addressed include detection of asymptomatic cases and considerations of human mobility. Many of the findings of this study are applicable to other low-transmission settings and could serve as a guideline for further studies of malaria in other regions.

Spatial and temporal distribution of Taenia solium and its risk factors in Uganda

OBJECTIVES

The lack of subnational mapping of the zoonotic cestode Taenia solium in endemic countries presents a major challenge to achieving intensified T. solium control milestones, as outlined in the "World Health Organization neglected tropical disease roadmap by 2030". We conducted a mapping study in Uganda, considered to be endemic, to identify subnational high-risk areas.

METHODS

T. solium prevalence data, adjusted for diagnostic sensitivity and specificity in a Bayesian framework, were identified through a systematic review. Spatial autocorrelation and interpolation techniques were used to transform demographic and health survey cluster-level sanitation and poverty indicators, overlaid onto a pig density map for Uganda into modelled  porcine cysticercosis (PCC) risk maps.

RESULTS

A total of 16 articles (n = 11 PCC and n = 5 human cysticercosis (HCC) and/or human taeniasis) were included in the final analysis. The observed HCC prevalence ranged from 0.01% to 6.0% (confidence interval range: 0.004-11.4%), whereas the adjusted PCC ranged from 0.3 to 93.9% (uncertainty interval range: 0-99.8%). There was substantial variation in the modelled  PCC risk factors and prevalence across Uganda and over time.

CONCLUSION

The high PCC prevalence and moderate HCC exposure estimates indicate the need for urgent implementation of T. solium control efforts in Uganda.

Quantitative bias analysis of prevalence under misclassification: evaluation indicators, calculation method and case analysis

Prevalence estimates are fundamental to epidemiological studies. Although they are highly vulnerable to misclassification bias, the risk of bias assessment of prevalence estimates is often neglected. Quantitative bias analysis (QBA) can effectively estimate misclassification bias in epidemiological studies; however, relatively few applications are identified. One reason for its low usage is the lack of knowledge and tools for these methods among researchers. To expand existing evaluation methods, based on the QBA principles, three indicators are proposed. One is the relative bias that quantifies the bias direction through its signs and the bias magnitude through its quantity. The second is the critical point of positive test proportion in case of a misclassification bias that is equal to zero. The third is the bound of positive test proportion equal to adjusted prevalence at misclassification bias level α. These indicators express the magnitude, direction and uncertainty of the misclassification bias of prevalence estimates, respectively. Using these indicators, it was found that slight oscillations of the positive test proportion within a certain range can lead to substantial increases in the misclassification bias. Hence, researchers should account for misclassification error analytically when interpreting the significance of adjusted prevalence for epidemiological decision making. This highlights the importance of applying QBA to these analyses. In this article, we have used three real-world cases to illustrate the characteristics and calculation methods of presented indicators. To facilitate application, an Excel-based calculation tool is provided.

Seroprevalence of SARS-CoV-2 on health professionals via Bayesian estimation: a Brazilian case study before and after vaccines

The increasing number of COVID-19 infections brought by the current pandemic has encouraged the scientific community to analyze the seroprevalence in populations to support health policies. In this context, accurate estimations of SARS-CoV-2 antibodies based on antibody tests metrics (e.g., specificity and sensitivity) and the study of population characteristics are essential. Here, we propose a Bayesian analysis using IgA and IgG antibody levels through multiple scenarios regarding data availability from different information sources to estimate the seroprevalence of health professionals in a Northeastern Brazilian city: no data available, data only related to the test performance, data from other regions. The study population comprises 432 subjects with more than 620 collections analyzed via IgA/IgG ELISA tests. We conducted the study in pre- and post-vaccination campaigns started in Brazil. We discuss the importance of aggregating available data from various sources to create informative prior knowledge. Considering prior information from the USA and Europe, the pre-vaccine seroprevalence means are 8.04% and 10.09% for IgG and 7.40% and 9.11% for IgA. For the post-vaccination campaign and considering local informative prior, the median is 84.83% for IgG, which confirms a sharp increase in the seroprevalence after vaccination. Additionally, stratification considering differences in sex, age (younger than 30 years, between 30 and 49 years, and older than 49 years), and presence of comorbidities are provided for all scenarios.

Evaluation of an Antibody Detecting Point of Care Test for Diagnosis of Taenia solium Cysticercosis in a Zambian Rural Community: A Prospective Diagnostic Accuracy Study

The lack of cheap, easy-to-use, rapid diagnostic tests has led to the development of several rapid diagnostic tests for cysticercosis. The new prototype two-strip, Taenia solium point of care test (TS POC) detects antibodies against taeniosis (TS POC T) and cysticercosis (TS POC CC). This study evaluated the diagnostic performance of the TS POC CC in the Sinda district in eastern Zambia. A sample of 1254 participants was recruited and tested with the TS POC. Out of the 1249 participants with a valid TS POC result, 177 (14%) tested positive while 1072 (86%) tested negative. All individuals with a positive TS POC and a subset of negative TS POC participants were selected for serum sampling, and were subjected to the recombinant glycoprotein T24H enzyme-linked immunoelectrotransfer blot (rT24H EITB) and the serum B60/158 (serum Ag) enzyme-linked immunosorbent assay (Ag ELISA). Performance characteristics were estimated using a Bayesian approach with probabilistic constraints. Based on 255 complete cases, the estimated sensitivity and specificity of the TS POC CC test were 35% (95% CI: 14–63%) and 87% (95% CI: 83–90%), respectively. The diagnostic performance needs to be improved, possibly by titrating antigen and other reagents’ concentration in the strip to produce a performance similar to existing cysticercosis tests such as the rT24H EITB.

Biogeographical characteristics of Schistosoma mansoni endemic areas in Ethiopia: a systematic review and meta analysis

BACKGROUND

In Ethiopia, schistosomiasis is caused by Schistosoma mansoni and S. haematobium with the former being widespread and more than 4 million people are estimated to be infected by S. mansoni annually with 35 million at risk of infection. Although many school- and community-based epidemiological surveys were conducted over the past decades, the national distribution of schistosomiasis endemic areas and associated socio-environmental determinants remain less well understood. In this paper, we review S. mansoni prevalence of infections and describe key biogeographical characteristics in the endemic areas in Ethiopia.

METHODS

We developed a database of S. mansoni infection surveys in Ethiopia through a systematic review by searching articles published between 1975 and 2019 on electronic online databases including PubMed, ScienceDirect, and Web of Science. A total of 62 studies involving 95 survey locations were included in the analysis. We estimated adjusted prevalence of infection from each survey by considering sensitivity and specificity of diagnostic tests using Bayesian approach. All survey locations were georeferenced and associated environmental and geographical characteristics (e.g. elevation, normalized difference vegetation index, soil properties, wealth index, and climatic data) were described using descriptive statistics and meta-analysis.

RESULTS

The results showed that the surveys exhibited a wide range of adjusted prevalence of infections from 0.5% to 99.5%, and 36.8% of the survey sites had adjusted prevalence of infection higher than 50%. S. mansoni endemic areas were distributed in six regional states with the majority of surveys being in Amhara and Oromia. Endemic sites were found at altitudes from 847.6 to 3141.8 m above sea level, annual mean temperatures between 17.9 and 29.8 ℃, annual cumulative precipitation between 1400 and 1898 mm, normalized difference vegetation index between 0.03 and 0.8, wealth index score between -68 857 and 179 756; and sand, silt, and clay fraction in soil between 19.1-47.2, 23.0-36.7, and 20.0-52.8 g/100 g, respectively.

CONCLUSIONS

The distribution of S. mansoni endemic areas and prevalence of infections exhibit remarked environmental and ecological heterogeneities. Future research is needed to understand how much these heterogeneities drive the parasite distribution and transmission in the region.

A review of dengue diagnostics and implications for surveillance and control

Dengue is the world’s most common arboviral infection, with almost 4 billion people estimated to be living at risk of dengue infection. A recently introduced vaccine is currently recommended only for seropositive individuals in a restricted age range determined by transmission intensity. With no effective dengue vaccine for the general population or any antiviral therapy, dengue control continues to rely heavily on vector control measures. Early and accurate diagnosis is important for guiding appropriate management and for disease surveillance to guide prompt dengue control interventions. However, major uncertainties exist in dengue diagnosis and this has important implications for all three. Dengue can be diagnosed clinically against predefined lists of signs and symptoms and by detection of dengue-specific antibodies, non-structural 1 antigen or viral RNA by reverse transcriptase–polymerase chain reaction. All of these methods have their limitations. This review aims to describe and quantify the advantages, uncertainties and variability of the various diagnostic methods used for dengue and discuss their implications and applications for dengue surveillance and control.

Seasonal variations in Plasmodium falciparum parasite prevalence assessed by varying diagnostic tests in asymptomatic children in southern Ghana

Plasmodium falciparum infections presenting either as symptomatic or asymptomatic may contain sexual stage parasites (gametocytes) that are crucial to malaria transmission. In this study, the prevalence of microscopic and submicroscopic asexual and gametocyte parasite stages were assessed in asymptomatic children from two communities in southern Ghana. Eighty children aged twelve years and below, none of whom exhibited signs of clinical malaria living in Obom and Cape Coast were sampled twice, one during the rainy (July 2015) and subsequently during the dry (January 2016) season. Venous blood was used to prepare thick and thin blood smears, spot a rapid malaria diagnostic test (PfHRP2 RDT) as well as prepare filter paper blood spots. Blood cell pellets were preserved in Trizol for RNA extraction. Polymerase chain reaction (PCR) and semi-quantitative real time reverse transcriptase PCR (qRT-PCR) were used to determine submicroscopic parasite prevalence. In both sites 87% (95% CI: 78-96) of the asymptomatic individuals surveyed were parasites positive during the 6 month study period. The prevalence of asexual and gametocyte stage parasites in the rainy season were both significantly higher in Obom than in Cape Coast (P < 0.001). Submicroscopic gametocyte prevalence was highest in the rainy season in Obom but in the dry season in Cape Coast. Parasite prevalence determined by PCR was similar to that determined by qRT-PCR in Obom but significantly lower than that determined by qRT-PCR in Cape Coast. Communities with varying parasite prevalence exhibit seasonal variations in the prevalence of gametocyte carriers. Submicroscopic asymptomatic parasite and gametocyte carriage is very high in southern Ghana, even during the dry season in communities with low microscopic parasite prevalence and likely to be missed during national surveillance exercises.

Vector-Borne Disease is a Common Cause of Hospitalized Febrile Illness in India

Acute febrile illness (AFI) is a major cause of morbidity and mortality in India and other resource-limited settings, yet systematic etiologic characterization of AFI has been limited. We prospectively enrolled adults (N = 970) and children (age 6 months to 12 years, N = 755) admitted with fever from the community to Sassoon General Hospital in Pune, India, from July 2013 to December 2015. We systematically obtained a standardized clinical history, basic laboratory testing, and microbiologic diagnostics on enrolled participants. Results from additional testing ordered by treating clinicians were also recorded. A microbiological diagnosis was found in 549 (32%) participants; 211 (12%) met standardized case definitions for pneumonia and meningitis without an identified organism; 559 (32%) were assigned a clinical diagnosis in the absence of a confirmed diagnosis; and 406 (24%) had no diagnosis. Vector-borne diseases were the most common cause of AFI in adults including dengue (N = 188, 19%), malaria (N = 74, 8%), chikungunya (N = 15, 2%), and concurrent mosquito-borne infections (N = 23, 2%) occurring most frequently in the 3 months after the monsoon. In children, pneumonia was the most common cause of AFI (N = 214, 28%) and death. Bacteremia was found in 68 (4%) participants. Central nervous system infections occurred in 58 (6%) adults and 64 (8%) children. Etiology of AFI in India is diverse, highly seasonal, and difficult to differentiate on clinical grounds alone. Diagnostic strategies adapted for season and age may reduce diagnostic uncertainty and identify causative organisms in treatable, fatal causes of AFI.

Impact of misclassification error in the estimation of maternal major depression disorder prevalence in home visitation programs

The objective of this study was to measure the bias introduced by misclassification error when estimating the prevalence of MDD among mothers in two community-based studies. Baseline data were collected from mothers participating in two home visitation study sites in South Central United States between 2010 and 2014. The operational definition of MDD was a Center of Epidemiological Studies-Depression - Short Form (CESD-SF) score of 10 or higher. Misclassification error was adjusted for using CESD-SF sensitivity and specificity priors that were either antepartum or postpartum specific or non-specific. Bias was measured as the difference between the observed and misclassification error-adjusted prevalence estimates using a Binomial Bayesian Latent Class model. The proportion of mothers in the antepartum and postpartum periods confounded the level of bias in estimating MDD prevalence. When using antepartum and postpartum specific sensitivity and specificity of the CESD-SF, misclassification error led to nearly no bias in prevalence estimates. In contrast, ignoring differences in CESD-SF sensitivity and specificity between these periods showed considerable MDD prevalence bias. The use of period of assessment (antepartum versus postpartum) specific case-finding instrument diagnostic performance values is critical to the valid estimation of MDD prevalence among mothers. Studies using other case-finding instruments are needed to support this conclusion.

True malaria prevalence in children under five: Bayesian estimation using data of malaria household surveys from three sub-Saharan countries

BACKGROUND

Malaria is one of the major causes of childhood death in sub-Saharan countries. A reliable estimation of malaria prevalence is important to guide and monitor progress toward control and elimination. The aim of the study was to estimate the true prevalence of malaria in children under five in the Democratic Republic of the Congo, Uganda and Kenya, using a Bayesian modelling framework that combined in a novel way malaria data from national household surveys with external information about the sensitivity and specificity of the malaria diagnostic methods used in those surveys-i.e., rapid diagnostic tests and light microscopy.

METHODS

Data were used from the Demographic and Health Surveys (DHS) and Malaria Indicator Surveys (MIS) conducted in the Democratic Republic of the Congo (DHS 2013-2014), Uganda (MIS 2014-2015) and Kenya (MIS 2015), where information on infection status using rapid diagnostic tests and/or light microscopy was available for 13,573 children. True prevalence was estimated using a Bayesian model that accounted for the conditional dependence between the two diagnostic methods, and the uncertainty of their sensitivities and specificities obtained from expert opinion.

RESULTS

The estimated true malaria prevalence was 20% (95% uncertainty interval [UI] 17%-23%) in the Democratic Republic of the Congo, 22% (95% UI 9-32%) in Uganda and 1% (95% UI 0-3%) in Kenya. According to the model estimations, rapid diagnostic tests had a satisfactory sensitivity and specificity, and light microscopy had a variable sensitivity, but a satisfactory specificity. Adding reported history of fever in the previous 14 days as a third diagnostic method to the model did not affect model estimates, highlighting the poor performance of this indicator as a malaria diagnostic.

CONCLUSIONS

In the absence of a gold standard test, Bayesian models can assist in the optimal estimation of the malaria burden, using individual results from several tests and expert opinion about the performance of those tests.

A comparison of PCR assays for beak and feather disease virus and high resolution melt (HRM) curve analysis of replicase associated protein and capsid genes

Beak and feather disease virus (BFDV) threatens a wide range of endangered psittacine birds worldwide. In this study, we assessed a novel PCR assay and genetic screening method using high-resolution melt (HRM) curve analysis for BFDV targeting the capsid (Cap) gene (HRM-Cap) alongside conventional PCR detection as well as a PCR method that targets a much smaller fragment of the virus genome in the replicase initiator protein (Rep) gene (HRM-Rep). Limits of detection, sensitivity, specificity and discriminatory power for differentiating BFDV sequences were compared. HRM-Cap had a high positive predictive value and could readily differentiate between a reference genotype and 17 other diverse BFDV genomes with more discriminatory power (genotype confidence percentage) than HRM-Rep. Melt curve profiles generated by HRM-Cap correlated with unique DNA sequence profiles for each individual test genome. The limit of detection of HRM-Cap was lower (2×10^-5ng/reaction or 48 viral copies) than that for both HRM-Rep and conventional BFDV PCR which had similar sensitivity (2×10^-6ng or 13 viral copies/reaction). However, when used in a diagnostic setting with 348 clinical samples there was strong agreement between HRM-Cap and conventional PCR (kappa=0.87, P<0.01, 98% specificity) and HRM-Cap demonstrated higher specificity (99.9%) than HRM-Rep (80.3%).

Hierarchical spatial modelling of pneumonia prevalence when response outcome has misclassification error: Applications to household data from Malawi

Pneumonia remains a major cause of child mortality in less developed countries. However, the accuracy of its prevalence and burden remains a challenge because disease data is often based on self-reports, resulting in measurement error in a form of under- and over-reporting. We propose hierarchical disease mapping approaches that permit measurement error, through different prior distributions of sensitivity and specificity. Proposed models were used to evaluate spatial variation of risk of pneumonia in children in Malawi. Results show that the true prevalence was 0.50 (95 CI: 0.4-0.66), however, estimates were dependent on sensitivity and specificity parameters. The estimated sensitivity was 0.76 (95% CI: 0.68-0.95), whereas specificity was 0.84 (95% CI: 0.72-0.93). A lower specificity underestimated the true prevalence, while sensitivity and specificity of greater or equal to 0.75 provided reliable and stable prevalence estimates. The spatial variation in disease risk changed little; however, misclassification of areas as high risk was visible.

Bias in logistic regression due to imperfect diagnostic test results and practical correction approaches

Background

Logistic regression is a statistical model widely used in cross-sectional and cohort studies to identify and quantify the effects of potential disease risk factors. However, the impact of imperfect tests on adjusted odds ratios (and thus on the identification of risk factors) is under-appreciated. The purpose of this article is to draw attention to the problem associated with modelling imperfect diagnostic tests, and propose simple Bayesian models to adequately address this issue.

Methods

A systematic literature review was conducted to determine the proportion of malaria studies that appropriately accounted for false-negatives/false-positives in a logistic regression setting. Inference from the standard logistic regression was also compared with that from three proposed Bayesian models using simulations and malaria data from the western Brazilian Amazon.

Results

A systematic literature review suggests that malaria epidemiologists are largely unaware of the problem of using logistic regression to model imperfect diagnostic test results. Simulation results reveal that statistical inference can be substantially improved when using the proposed Bayesian models versus the standard logistic regression. Finally, analysis of original malaria data with one of the proposed Bayesian models reveals that microscopy sensitivity is strongly influenced by how long people have lived in the study region, and an important risk factor (i.e., participation in forest extractivism) is identified that would have been missed by standard logistic regression.

Conclusion

Given the numerous diagnostic methods employed by malaria researchers and the ubiquitous use of logistic regression to model the results of these diagnostic tests, this paper provides critical guidelines to improve data analysis practice in the presence of misclassification error. Easy-to-use code that can be readily adapted to WinBUGS is provided, enabling straightforward implementation of the proposed Bayesian models.

Electronic supplementary material

The online version of this article (doi:10.1186/s12936-015-0966-y) contains supplementary material, which is available to authorized users.

Diagnosis of Leptospirosis: Comparison between Microscopic Agglutination Test, IgM-ELISA and IgM Rapid Immunochromatography Test

Background

Leptospirosis is diagnosed on clinical grounds, and confirmed by microscopic agglutination test (MAT). IgM-ELISA (Serion-Virion) and immunochromatography test (Leptocheck-WB) are two immunodiagnostic assays for leptospirosis. Their sensitivity, specificity and applicability in Sri Lanka have not been systematically evaluated.

Methods

Clinically diagnosed leptospirosis patients (n = 919) were recruited from three hospitals in the Western Province of Sri Lanka, during June 2012 to December 2013. MAT, IgM-ELISA and Leptocheck-WB were performed on all patient sera. MAT titer of ≥400 in single sample, four-fold rise or seroconversion ≥100 in paired samples were considered as positive for MAT. For diagnostic confirmation, MAT was performed during both acute and convalescent phases. Anti-leptospiral IgM ≥20 IU/ml and appearance of a band in the test window were considered as positive for IgM-ELISA and Leptocheck-WB test respectively. Patients with an alternative diagnosis (n = 31) were excluded. Data analysis was performed using two methods, i) considering MAT as reference standard and ii) using Bayesian latent class model analysis (BLCM) which considers each test as imperfect.

Results

MAT, IgM-ELISA and Leptocheck-WB positivity were 39.8%, 45.8% and 38.7% respectively during the acute phase. Acute-phase MAT had specificity and sensitivity of 95.7% and 55.3% respectively, when compared to overall MAT positivity. IgM-ELISA and Leptocheck-WB had similar diagnostic sensitivity when compared with acute-phase MAT as the gold standard, although IgM-ELISA showed higher specificity (84.5%) than Leptocheck-WB (73.3%). BLCM analysis showed that IgM-ELISA and Leptocheck-WB had similar sensitivities (86.0% and 87.4%), while acute-phase MAT had the lowest sensitivity (77.4%). However, acute-phase MAT had high specificity (97.6%), while IgM-ELISA and Leptocheck-WB showed similar but lower specificity (84.5% and 82.9%).

Conclusions

Both IgM-ELISA and Leptocheck-WB shows similar sensitivities and specificities. IgM-ELISA may be superior to MAT during the acute phase and suitable for early diagnosis of leptospirosis. Leptocheck-WB is suitable as a rapid immunodiagnostic screening test for resource limited settings.

Bayesian imperfect information analysis for clinical recurrent data

In medical research, clinical practice must often be undertaken with imperfect information from limited resources. This study applied Bayesian imperfect information-value analysis to realistic situations to produce likelihood functions and posterior distributions, to a clinical decision-making problem for recurrent events. In this study, three kinds of failure models are considered, and our methods illustrated with an analysis of imperfect information from a trial of immunotherapy in the treatment of chronic granulomatous disease. In addition, we present evidence toward a better understanding of the differing behaviors along with concomitant variables. Based on the results of simulations, the imperfect information value of the concomitant variables was evaluated and different realistic situations were compared to see which could yield more accurate results for medical decision-making.

Estimation of diagnostic test accuracy without full verification: a review of latent class methods

The performance of a diagnostic test is best evaluated against a reference test that is without error. For many diseases, this is not possible, and an imperfect reference test must be used. However, diagnostic accuracy estimates may be biased if inaccurately verified status is used as the truth. Statistical models have been developed to handle this situation by treating disease as a latent variable. In this paper, we conduct a systematized review of statistical methods using latent class models for estimating test accuracy and disease prevalence in the absence of complete verification.

Impact of intermittent screening and treatment for malaria among school children in Kenya: a cluster randomised trial

BACKGROUND

Improving the health of school-aged children can yield substantial benefits for cognitive development and educational achievement. However, there is limited experimental evidence of the benefits of alternative school-based malaria interventions or how the impacts of interventions vary according to intensity of malaria transmission. We investigated the effect of intermittent screening and treatment (IST) for malaria on the health and education of school children in an area of low to moderate malaria transmission.

METHODS AND FINDINGS

A cluster randomised trial was implemented with 5,233 children in 101 government primary schools on the south coast of Kenya in 2010-2012. The intervention was delivered to children randomly selected from classes 1 and 5 who were followed up for 24 months. Once a school term, children were screened by public health workers using malaria rapid diagnostic tests (RDTs), and children (with or without malaria symptoms) found to be RDT-positive were treated with a six dose regimen of artemether-lumefantrine (AL). Given the nature of the intervention, the trial was not blinded. The primary outcomes were anaemia and sustained attention. Secondary outcomes were malaria parasitaemia and educational achievement. Data were analysed on an intention-to-treat basis. During the intervention period, an average of 88.3% children in intervention schools were screened at each round, of whom 17.5% were RDT-positive. 80.3% of children in the control and 80.2% in the intervention group were followed-up at 24 months. No impact of the malaria IST intervention was observed for prevalence of anaemia at either 12 or 24 months (adjusted risk ratio [Adj.RR]: 1.03, 95% CI 0.93-1.13, p = 0.621 and Adj.RR: 1.00, 95% CI 0.90-1.11, p = 0.953) respectively, or on prevalence of P. falciparum infection or scores of classroom attention. No effect of IST was observed on educational achievement in the older class, but an apparent negative effect was seen on spelling scores in the younger class at 9 and 24 months and on arithmetic scores at 24 months.

CONCLUSION

In this setting in Kenya, IST as implemented in this study is not effective in improving the health or education of school children. Possible reasons for the absence of an impact are the marked geographical heterogeneity in transmission, the rapid rate of reinfection following AL treatment, the variable reliability of RDTs, and the relative contribution of malaria to the aetiology of anaemia in this setting.

TRIAL REGISTRATION

www.ClinicalTrials.gov NCT00878007.

Using a web-based application to define the accuracy of diagnostic tests when the gold standard is imperfect

BACKGROUND

Estimates of the sensitivity and specificity for new diagnostic tests based on evaluation against a known gold standard are imprecise when the accuracy of the gold standard is imperfect. Bayesian latent class models (LCMs) can be helpful under these circumstances, but the necessary analysis requires expertise in computational programming. Here, we describe open-access web-based applications that allow non-experts to apply Bayesian LCMs to their own data sets via a user-friendly interface.

METHODS/PRINCIPAL FINDINGS

Applications for Bayesian LCMs were constructed on a web server using R and WinBUGS programs. The models provided (http://mice.tropmedres.ac) include two Bayesian LCMs: the two-tests in two-population model (Hui and Walter model) and the three-tests in one-population model (Walter and Irwig model). Both models are available with simplified and advanced interfaces. In the former, all settings for Bayesian statistics are fixed as defaults. Users input their data set into a table provided on the webpage. Disease prevalence and accuracy of diagnostic tests are then estimated using the Bayesian LCM, and provided on the web page within a few minutes. With the advanced interfaces, experienced researchers can modify all settings in the models as needed. These settings include correlation among diagnostic test results and prior distributions for all unknown parameters. The web pages provide worked examples with both models using the original data sets presented by Hui and Walter in 1980, and by Walter and Irwig in 1988. We also illustrate the utility of the advanced interface using the Walter and Irwig model on a data set from a recent melioidosis study. The results obtained from the web-based applications were comparable to those published previously.

CONCLUSIONS

The newly developed web-based applications are open-access and provide an important new resource for researchers worldwide to evaluate new diagnostic tests.

Assessing malaria transmission in a low endemicity area of north-western Peru

BACKGROUND

Where malaria endemicity is low, control programmes need increasingly sensitive tools for monitoring malaria transmission intensity (MTI) and to better define health priorities. A cross-sectional survey was conducted in a low endemicity area of the Peruvian north-western coast to assess the MTI using both molecular and serological tools.

METHODS

Epidemiological, parasitological and serological data were collected from 2,667 individuals in three settlements of Bellavista district, in May 2010. Parasite infection was detected using microscopy and polymerase chain reaction (PCR). Antibodies to Plasmodium vivax merozoite surface protein-119 (PvMSP1₁₉) and to Plasmodium falciparum glutamate-rich protein (PfGLURP) were detected by ELISA. Risk factors for exposure to malaria (seropositivity) were assessed by multivariate survey logistic regression models. Age-specific antibody prevalence of both P. falciparum and P. vivax were analysed using a previously published catalytic conversion model based on maximum likelihood for generating seroconversion rates (SCR).

RESULTS

The overall parasite prevalence by microscopy and PCR were extremely low: 0.3 and 0.9%, respectively for P. vivax, and 0 and 0.04%, respectively for P. falciparum, while seroprevalence was much higher, 13.6% for P. vivax and 9.8% for P. falciparum. Settlement, age and occupation as moto-taxi driver during previous year were significantly associated with P. falciparum exposure, while age and distance to the water drain were associated with P. vivax exposure. Likelihood ratio tests supported age seroprevalence curves with two SCR for both P. vivax and P. falciparum indicating significant changes in the MTI over time. The SCR for PfGLURP was 19-fold lower after 2002 as compared to before (λ1 = 0.022 versus λ2 = 0.431), and the SCR for PvMSP1₁₉ was four-fold higher after 2006 as compared to before (λ1 = 0.024 versus λ2 = 0.006).

CONCLUSION

Combining molecular and serological tools considerably enhanced the capacity of detecting current and past exposure to malaria infections and related risks factors in this very low endemicity area. This allowed for an improved characterization of the current human reservoir of infections, largely hidden and heterogeneous, as well as providing insights into recent changes in species specific MTIs. This approach will be of key importance for evaluating and monitoring future malaria elimination strategies.

The mechanistic, diagnostic and prognostic utility of biomarkers in severe malaria

Malaria remains an important global cause of severe illness and mortality. This literature review summarizes available data on how biomarkers might be applied to diagnose, prognosticate and provide mechanistic insights in patients with severe malaria. Of the large number of candidate biomarkers, only PfHRP2 has consistently demonstrated clinical utility and, when incorporated into rapid antigen detection tests, has shown diagnostic sensitivity above 95%, which is at least as good as light microscopy. As a quantitative test, PfHRP2 also shows some promise in differentiating severe malarial from non-malarial disease in areas where asymptomatic carriage of malaria parasites is common, and possibly as a tool to estimate sequestered parasite burden and subsequent mortality. Biomarkers such as pLDH and panmalarial antigen have lower sensitivity for non-falciparum malaria in rapid antigen detection tests. There is an urgent need to discover and validate better biomarkers for incorporation into rapid antigen detection tests in countries where Plasmodium vivax is a common cause of severe disease. A large number of host-derived acute-phase reactants, markers of endothelial dysfunction and immune mediators have been proposed as biomarkers. Although they have provided mechanistic insights into the immunopathology of severe malaria, their roles as clinical tools remain uncertain.

Estimating the true accuracy of diagnostic tests for dengue infection using bayesian latent class models

Background

Accuracy of rapid diagnostic tests for dengue infection has been repeatedly estimated by comparing those tests with reference assays. We hypothesized that those estimates might be inaccurate if the accuracy of the reference assays is not perfect. Here, we investigated this using statistical modeling.

Methods/Principal Findings

Data from a cohort study of 549 patients suspected of dengue infection presenting at Colombo North Teaching Hospital, Ragama, Sri Lanka, that described the application of our reference assay (a combination of Dengue IgM antibody capture ELISA and IgG antibody capture ELISA) and of three rapid diagnostic tests (Panbio NS1 antigen, IgM antibody and IgG antibody rapid immunochromatographic cassette tests) were re-evaluated using Bayesian latent class models (LCMs). The estimated sensitivity and specificity of the reference assay were 62.0% and 99.6%, respectively. Prevalence of dengue infection (24.3%), and sensitivities and specificities of the Panbio NS1 (45.9% and 97.9%), IgM (54.5% and 95.5%) and IgG (62.1% and 84.5%) estimated by Bayesian LCMs were significantly different from those estimated by assuming that the reference assay was perfect. Sensitivity, specificity, PPV and NPV for a combination of NS1, IgM and IgG cassette tests on admission samples were 87.0%, 82.8%, 62.0% and 95.2%, respectively.

Conclusions

Our reference assay is an imperfect gold standard. In our setting, the combination of NS1, IgM and IgG rapid diagnostic tests could be used on admission to rule out dengue infection with a high level of accuracy (NPV 95.2%). Further evaluation of rapid diagnostic tests for dengue infection should include the use of appropriate statistical models.

Estimating the prevalence of infections in vector populations using pools of samples

Several statistical methods have been proposed for estimating the infection prevalence based on pooled samples, but these methods generally presume the application of perfect diagnostic tests, which in practice do not exist. To optimize prevalence estimation based on pooled samples, currently available and new statistical models were described and compared. Three groups were tested: (a) Frequentist models, (b) Monte Carlo Markov-Chain (MCMC) Bayesian models, and (c) Exact Bayesian Computation (EBC) models. Simulated data allowed the comparison of the models, including testing the performance under complex situations such as imperfect tests with a sensitivity varying according to the pool weight. In addition, all models were applied to data derived from the literature, to demonstrate the influence of the model on real-prevalence estimates. All models were implemented in the freely available R and OpenBUGS software and are presented in Appendix S1. Bayesian models can flexibly take into account the imperfect sensitivity and specificity of the diagnostic test (as well as the influence of pool-related or external variables) and are therefore the method of choice for calculating population prevalence based on pooled samples. However, when using such complex models, very precise information on test characteristics is needed, which may in general not be available.

Rapid antigen detection tests for malaria diagnosis in severely ill Papua New Guinean children: a comparative study using Bayesian latent class models

Background

Although rapid diagnostic tests (RDTs) have practical advantages over light microscopy (LM) and good sensitivity in severe falciparum malaria in Africa, their utility where severe non-falciparum malaria occurs is unknown. LM, RDTs and polymerase chain reaction (PCR)-based methods have limitations, and thus conventional comparative malaria diagnostic studies employ imperfect gold standards. We assessed whether, using Bayesian latent class models (LCMs) which do not require a reference method, RDTs could safely direct initial anti-infective therapy in severe ill children from an area of hyperendemic transmission of both Plasmodium falciparum and P. vivax.

Methods and Findings

We studied 797 Papua New Guinean children hospitalized with well-characterized severe illness for whom LM, RDT and nested PCR (nPCR) results were available. For any severe malaria, the estimated prevalence was 47.5% with RDTs exhibiting similar sensitivity and negative predictive value (NPV) to nPCR (≥96.0%). LM was the least sensitive test (87.4%) and had the lowest NPV (89.7%), but had the highest specificity (99.1%) and positive predictive value (98.9%). For severe falciparum malaria (prevalence 42.9%), the findings were similar. For non-falciparum severe malaria (prevalence 6.9%), no test had the WHO-recommended sensitivity and specificity of >95% and >90%, respectively. RDTs were the least sensitive (69.6%) and had the lowest NPV (96.7%).

Conclusions

RDTs appear a valuable point-of-care test that is at least equivalent to LM in diagnosing severe falciparum malaria in this epidemiologic situation. None of the tests had the required sensitivity/specificity for severe non-falciparum malaria but the number of false-negative RDTs in this group was small.

Use of rapid diagnostic tests in malaria school surveys in Kenya: does their under-performance matter for planning malaria control?

Malaria rapid diagnostic tests (RDTs) are known to yield false-positive results, and their use in epidemiologic surveys will overestimate infection prevalence and potentially hinder efficient targeting of interventions. To examine the consequences of using RDTs in school surveys, we compared three RDT brands used during a nationwide school survey in Kenya with expert microscopy and investigated the cost implications of using alternative diagnostic approaches in identifying localities with differing levels of infection. Overall, RDT sensitivity was 96.1% and specificity was 70.8%. In terms of classifying districts and schools according to prevalence categories, RDTs were most reliable for the < 1% and > 40% categories and least reliable in the 1–4.9% category. In low-prevalence settings, microscopy was the most expensive approach, and RDT results corrected by either microscopy or polymerase chain reaction were the cheapest. Use of polymerase chain reaction–corrected RDT results is recommended in school malaria surveys, especially in settings with low-to-moderate malaria transmission.

Fool's gold: Why imperfect reference tests are undermining the evaluation of novel diagnostics: a reevaluation of 5 diagnostic tests for leptospirosis

BACKGROUND

We observed that some patients with clinical leptospirosis supported by positive results of rapid tests were negative for leptospirosis on the basis of our diagnostic gold standard, which involves isolation of Leptospira species from blood culture and/or a positive result of a microscopic agglutination test (MAT). We hypothesized that our reference standard was imperfect and used statistical modeling to investigate this hypothesis.

METHODS

Data for 1652 patients with suspected leptospirosis recruited during three observational studies and one randomized control trial that described the application of culture, MAT, immunofluorescence assay (IFA), lateral flow (LF) and/or PCR targeting the 16S rRNA gene were reevaluated using Bayesian latent class models and random-effects meta-analysis.

RESULTS

The estimated sensitivities of culture alone, MAT alone, and culture plus MAT (for which the result was considered positive if one or both tests had a positive result) were 10.5% (95% credible interval [CrI], 2.7%-27.5%), 49.8% (95% CrI, 37.6%-60.8%), and 55.5% (95% CrI, 42.9%-67.7%), respectively. These low sensitivities were present across all 4 studies. The estimated specificity of MAT alone (and of culture plus MAT) was 98.8% (95% CrI, 92.8%-100.0%). The estimated sensitivities and specificities of PCR (52.7% [95% CrI, 45.2%-60.6%] and 97.2% [95% CrI, 92.0%-99.8%], respectively), lateral flow test (85.6% [95% CrI, 77.5%-93.2%] and 96.2% [95% CrI, 87.7%-99.8%], respectively), and immunofluorescence assay (45.5% [95% CrI, 33.3%-60.9%] and 96.8% [95% CrI, 92.8%-99.8%], respectively) were considerably different from estimates in which culture plus MAT was considered a perfect gold standard test.

CONCLUSIONS

Our findings show that culture plus MAT is an imperfect gold standard against which to compare alterative tests for the diagnosis of leptospirosis. Rapid point-of-care tests for this infection would bring an important improvement in patient care, but their future evaluation will require careful consideration of the reference test(s) used and the inclusion of appropriate statistical models.

Bayesian Latent Class Models in malaria diagnosis

Aims

The main focus of this study is to illustrate the importance of the statistical analysis in the evaluation of the accuracy of malaria diagnostic tests, without admitting a reference test, exploring a dataset (3317) collected in São Tomé and Príncipe.

Methods

Bayesian Latent Class Models (without and with constraints) are used to estimate the malaria infection prevalence, together with sensitivities, specificities, and predictive values of three diagnostic tests (RDT, Microscopy and PCR), in four subpopulations simultaneously based on a stratified analysis by age groups (, 5 years old) and fever status (febrile, afebrile).

Results

In the afebrile individuals with at least five years old, the posterior mean of the malaria infection prevalence is 3.2% with a highest posterior density interval of [2.3–4.1]. The other three subpopulations (febrile 5 years, afebrile or febrile children less than 5 years) present a higher prevalence around 10.3% [8.8–11.7]. In afebrile children under-five years old, the sensitivity of microscopy is 50.5% [37.7–63.2]. In children under-five, the estimated sensitivities/specificities of RDT are 95.4% [90.3–99.5]/93.8% [91.6–96.0] – afebrile – and 94.1% [87.5–99.4]/97.5% [95.5–99.3] – febrile. In individuals with at least five years old are 96.0% [91.5–99.7]/98.7% [98.1–99.2] – afebrile – and 97.9% [95.3–99.8]/97.7% [96.6–98.6] – febrile. The PCR yields the most reliable results in four subpopulations.

Conclusions

The utility of this RDT in the field seems to be relevant. However, in all subpopulations, data provide enough evidence to suggest caution with the positive predictive values of the RDT. Microscopy has poor sensitivity compared to the other tests, particularly, in the afebrile children less than 5 years. This type of findings reveals the danger of statistical analysis based on microscopy as a reference test. Bayesian Latent Class Models provide a powerful tool to evaluate malaria diagnostic tests, taking into account different groups of interest.

Enhanced understanding of infectious diseases by fusing multiple datasets: a case study on malaria in the Western Brazilian Amazon region

BACKGROUND

A common challenge to the study of several infectious diseases consists in combining limited cross-sectional survey data, collected with a more sensitive detection method, with a more extensive (but biased) syndromic sentinel surveillance data, collected with a less sensitive method. Our article describes a novel modeling framework that overcomes this challenge, resulting in enhanced understanding of malaria in the Western Brazilian Amazon.

METHODOLOGY/PRINCIPAL FINDINGS

A cohort of 486 individuals was monitored using four cross-sectional surveys, where all participants were sampled regardless of symptoms (aggressive-active case detection), resulting in 1,383 microscopy and 1,400 polymerase chain reaction tests. Data on the same individuals were also obtained from the local surveillance facility (i.e., passive and active case detection), totaling 1,694 microscopy tests. Our model accommodates these multiple pathogen and case detection methods. This model is shown to outperform logistic regression in terms of interpretability of its parameters, ability to recover the true parameter values, and predictive performance. We reveal that the main infection determinant was the extent of forest, particularly during the rainy season and in close proximity to water bodies, and participation on forest activities. We find that time residing in Acrelandia (as a proxy for past malaria exposure) decreases infection risk but surprisingly increases the likelihood of reporting symptoms once infected, possibly because non-naïve settlers are only susceptible to more virulent Plasmodium strains. We suggest that the search for asymptomatic carriers should focus on those at greater risk of being infected but lower risk of reporting symptoms once infected.

CONCLUSIONS/SIGNIFICANCE

The modeling framework presented here combines cross-sectional survey data and syndromic sentinel surveillance data to shed light on several aspects of malaria that are critical for public health policy. This framework can be adapted to enhance inference on infectious diseases whenever asymptomatic carriers are important and multiple datasets are available.