There is growing interest in local elimination of soil-transmitted helminth (STH) infection in endemic settings. In such settings, highly sensitive diagnostics are needed to detect STH infection. We compared double-slide Kato-Katz, the most commonly used copromicroscopic detection method, to multi-parallel quantitative polymerase chain reaction (qPCR) in 2,799 stool samples from children aged 2–12 years in a setting in rural Bangladesh with predominantly low STH infection intensity. We estimated the sensitivity and specificity of each diagnostic using Bayesian latent class analysis. Compared to double-slide Kato-Katz, STH prevalence using qPCR was almost 3-fold higher for hookworm species and nearly 2-fold higher for Trichuris trichiura. Ascaris lumbricoides prevalence was lower using qPCR, and 26% of samples classified as A. lumbricoides positive by Kato-Katz were negative by qPCR. Amplicon sequencing of the 18S rDNA from 10 samples confirmed that A. lumbricoides was absent in samples classified as positive by Kato-Katz and negative by qPCR. The sensitivity of Kato-Katz was 49% for A. lumbricoides, 32% for hookworm, and 52% for T. trichiura; the sensitivity of qPCR was 79% for A. lumbricoides, 93% for hookworm, and 90% for T. trichiura. Specificity was ≥ 97% for both tests for all STH except for Kato-Katz for A. lumbricoides (specificity = 68%). There were moderate negative, monotonic correlations between qPCR cycle quantification values and eggs per gram quantified by Kato-Katz. While it is widely assumed that double-slide Kato-Katz has few false positives, our results indicate otherwise and highlight inherent limitations of the Kato-Katz technique. qPCR had higher sensitivity than Kato-Katz in this low intensity infection setting.
Soil-transmitted helminth infections (STH) (e.g., Ascaris, hookworm, Trichuris) contribute to a large burden of disease among children in low- and middle-income countries. There is increasing interest in implementing large-scale deworming programs to eliminate STH in certain settings. Efforts to monitor whether local elimination has occurred require sensitive diagnostic tests that will not miss positive cases. Kato-Katz, a microscopy-based diagnostic test, has commonly been used to identify STH eggs in stool, but in settings where infection intensity is low, this method frequently misses positive samples because it requires visual identification of small numbers of eggs, and hookworm eggs may degrade prior to visualization. Quantitative polymerase chain reaction (qPCR) is a molecular diagnostic method for detecting STH. It may detect more low intensity infections than Kato-Katz because it identifies STH DNA in stool; DNA can be detected in very small quantities and is less likely to degrade than STH ova. Thus, qPCR is likely to be more accurate than Kato-Katz. This study compared the performance of double-slide Kato-Katz and qPCR using 2,799 stool samples from children aged 2–12 years in a setting in rural Bangladesh with predominantly low STH infection intensity. qPCR was more sensitive than Kato-Katz for hookworm and Trichuris infections. 26% of samples were classified as Ascaris positive by Kato-Katz and negative by qPCR. DNA sequencing of 10 samples confirmed that Ascaris was absent in samples classified as positive by Kato-Katz and negative by qPCR. We conclude that Kato-Katz likely produced false positive results for Ascaris and that qPCR had a higher sensitivity than double-slide Kato-Katz in this low infection intensity setting.
Citation: Benjamin-Chung J, Pilotte N, Ercumen A, Grant JR, Maasch JRMA, Gonzalez AM, et al. (2020) Comparison of multi-parallel qPCR and double-slide Kato-Katz for detection of soil-transmitted helminth infection among children in rural Bangladesh. PLoS Negl Trop Dis 14(4): e0008087. https://doi.org/10.1371/journal.pntd.0008087
Editor: Matthew C. Freeman, Emory University, UNITED STATES
Received: May 1, 2019; Accepted: January 24, 2020; Published: April 24, 2020
Copyright: © 2020 Benjamin-Chung et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data and replication scripts are available through the Open Science Framework (https://osf.io/agk6w/).
Funding: This study was supported by an award from the Task Force for Global Health (NTD-SUSTAIN4 089G). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Soil-transmitted helminths (STH) infect an estimated 1.5 billion individuals, almost one fifth of the global population . These infections contribute to a substantial burden of disability and disease, particularly for children . To control the burden of STH infection, mass drug administration programs deliver preventive chemotherapy to populations at risk, such as school children. In recent years, approximately 33% of children in endemic settings received deworming medication for STH in the past six months according to caregiver report . Such programs have expanded dramatically following the London Declaration on Neglected Tropical Diseases in 2012. Mass drug administration (MDA) programs have yielded large reductions in STH infection prevalence and intensity to date . While the World Health Organization has set goals for STH focused on morbidity control , there is an active body of research focused on whether local elimination of STH is possible in endemic settings through community-wide MDA for deworming [6–9]. As STH infection intensity and prevalence wane, increasingly sensitive diagnostic methods are required to determine whether STH transmission has been interrupted and when interventions can be scaled back or discontinued .
Diagnostic methods for STH include copromicroscopic methods, such as the Kato-Katz method [11,12], formol-ether concentration , FLOTAC , mini-FLOTAC [15–18], and FECPAKG2 [18,19]. Of these, Kato-Katz is the most commonly used diagnostic method for STH surveillance because it is inexpensive and relatively easy to perform in low-resource settings . A significant limitation of this method is that samples must be evaluated within an hour of preparation in order to detect fragile hookworm ova. In addition, Necator americanus and Ancylostoma duodenale cannot be distinguished using Kato-Katz because of their morphological similarities. Because egg excretion is highly variable over time , there can be substantial variation in the count of eggs per gram (epg) of stool collected on different days. Kato-Katz has improved sensitivity when performed on multiple samples from defecation events on different days, but such samples are often logistically prohibitive to collect. Double-slide Kato-Katz examines stool using two separate aliquots from a single defecation event prepared as separate slides in order to help reduce the variability and increase sensitivity. Furthermore, a meta-analysis estimated the sensitivity of double-slide Kato-Katz to be 64.6% for Ascaris lumbricoides, 84.8% for Trichuris trichiura, and 63.0% for hookworm . Another limitation of Kato-Katz is that its sensitivity is lower in low infection intensity settings (55.2% for A. lumbricoides, 79.8% for T. trichiura, and 52.6% for hookworm) . Alternative copromicroscopic STH diagnostic methods, such as formol-ether concentration, FLOTAC, mini-FLOTAC, and the McMaster method also have moderate to low sensitivity for different types of STH in low-intensity settings .
Multiplex [23,24] and multi-parallel [25,26] quantitative polymerase chain reaction (qPCR) assays have been developed to detect STH DNA in stool. Multiplex and multi-parallel assays allow for detection of multiple helminths from a single stool sample. Growing evidence suggests that these methods may be used to estimate infection intensity, although qPCR-based cutoffs for light, moderate, and heavy intensity infections have not yet been defined [24,27,28]. To our knowledge, six studies have compared diagnostic performance of copromicroscopic methods and qPCR; these studies reported that analyses classifying STH status using qPCR as a diagnostic tool showed higher prevalence compared to those using single and/or double-slide Kato-Katz because of the higher sensitivity of the qPCR [18,23,25,27–29].
The greater sensitivity, specificity, and precision of qPCR relative to Kato-Katz makes it an attractive diagnostic for monitoring the successes of STH control programs deploying mass deworming since infection intensity tends to decline after several years of intervention . As a population reaches an STH transmission breakpoint, worm burdens decrease, and the comparatively low sensitivity of Kato-Katz may preclude its use in monitoring STH prevalence. In addition, a growing body of work suggests that qPCR may also be used to estimate infection intensity; one study found that qPCR is approximately 3.6 times more precise than Kato-Katz in estimating A. lumbricoides egg intensity . Such additional precision could increase the statistical power available to detect whether the STH prevalence meets the threshold for elimination.
The objectives of this study were to estimate the prevalence of STH infection and infection intensity using double-slide Kato-Katz and multi-parallel qPCR and to estimate the sensitivity and specificity of each method using Bayesian latent class models and archived samples collected from children aged 2–12 years in rural Bangladesh.
The study protocol was approved by the Ethical Review Committee at icddr,b (PR-14105), the Committee for the Protection of Human Subjects at the University of California, Berkeley (2014-08-6658), and the institutional review board at Stanford University (27864). Prior to enrollment, all adult subjects provided written informed consent. Parents or guardians of children provided written informed consent on behalf of children.
This study analyzed archived stool samples that were collected from participants in the WASH Benefits Bangladesh trial between May 2015 and May 2016 (Clinicaltrials.gov NCT01590095) [32,33]. The trial was implemented in Gazipur, Mymensingh, Tangail and Kishoreganj districts of Bangladesh. Biannual school-based mass drug administration with mebendazole had been implemented nationally in Bangladesh since 2008 and was ongoing in the study area during the trial. The trial delivered interventions to improve water, sanitation, handwashing, and nutrition and included single and combined intervention arms. Full details of interventions and user adherence have been published elsewhere [33,34]. This study randomly selected 2,800 stool specimens collected from children aged 22 months to 12 years of age (mean = 57 months) enrolled in the individual improved water, sanitation, handwashing, and combined water+sanitation+handwashing (WSH) arms. This sample size was chosen in order to estimate sensitivity with precision of +/- 5% and 80% statistical power; we assumed STH prevalence estimates from a prior study in rural Bangladesh  and double-slide Kato-Katz sensitivity and specificity estimates from a prior meta-analysis of STH diagnostic accuracy .
Stool collection and storage
Field staff provided study participants’ primary caregivers with sterile containers and requested that they collect stool from the child’s defecation event the next morning. Field staff returned to the house at least three times to attempt to collect stool containers. As part of the WASH Benefits trial, all study participants received a single dose of albendazole following stool collection. Field staff transported stool specimens on ice to the field laboratory, where 1 g of stool was archived in 1 ml of 100% ethanol. Specimens were stored at -20°C prior to relocation to a -80°C freezer. In the WASH Benefits parasites sub-study, the median time that elapsed between the reported time of defecation and when the stool was first placed on ice was 2.4 hours (range: 0.03 to 18 hours), and the median time that elapsed between the date of specimen collection and the date when specimens were stored in a -80° C freezer was 11 days (range: 1 to 330 days).
Technicians were trained in double-slide Kato-Katz using the Vestergaard Frandsen protocol at the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b) parasitology laboratory. On the day of stool collection, trained technicians performed double-slide Kato-Katz [11,12] on fresh stool within 30 minutes of preparing each sample. For quality assurance, 10% of slides were evaluated independently by two technicians, and 5% were evaluated by the same experienced parasitologist who conducted Kato-Katz training. The quality assurance results have previously been published elsewhere . In brief, agreement between laboratory technicians was high (Kappa statistic > 0.99 for each STH). The Kappa statistic for agreement between laboratory technicians and experienced parasitologists was 0.92 for A. lumbricoides, 0.20 for hookworm, and 0.86 for T. trichiura. The agreement is likely lower for hookworm because experienced parasitologists examined slides up to a few days after laboratory technicians, and hookworm ova may have begun to disintegrate by that time. For the subset of samples for which quality control was performed on the same day, the agreement for hookworm was high. Agreement decreased as the time between the original assessment and the senior parasitologist’s assessment increased (S1 Table). Samples in which at least one STH egg was visualized during Kato-Katz were classified as positive. Kato-Katz technicians did not have access to any clinical information about study participants.
DNA isolation and qPCR procedures
Approximately 1–2 years after stool collection, preserved stool specimens were shipped to Smith College in Northampton, MA, United States for qPCR analyses. Prior to shipment, ethanol was evaporated from all samples for compliance with shipping regulations, and samples were shipped on dry ice. Upon receipt at Smith College, all samples were stored at -20°C until DNA extraction. DNA was extracted using the FastDNA Spin Kit for Soil (MP Biomedicals, Solon, OH) utilizing a previously published modified version of the manufacturer’s methodology . Following DNA extraction and prior to qPCR analyses, extractions were stored at -20°C. We assigned each sample a unique number following extraction, and this number was electronically linked to the initial barcode on the stool sample tube at the time of extraction. Results were then linked back to the initial sample using both the Smith-assigned number and the barcode, and samples underwent qPCR processing in the same batches that they were extracted in. An internal amplification control (IAC) plasmid was employed during the extraction of each sample to ensure successful isolation, and adequate recovery of DNA . Utilizing previously described reaction methods , any sample which failed to produce a positive qPCR result for one or more of the IAC replicates underwent re-extraction. Similarly, following the completion of all extractions, the mean quantitation cycle (Cq) value for the IAC results from all samples was calculated, as was the standard deviation of the mean. Any sample with an individual IAC result determined to be 3 or more standard deviations greater than the mean underwent re-extraction, as such a recovery was deemed suboptimal. Following re-extraction, only the extracted DNA sample producing a positive IAC result within the defined Cq range (not the previously conducted invalid extract) was used for experimental testing.
Experimental qPCR reactions were performed using previously published multi-parallel assays targeting non-coding repetitive sequences to detect Ascaris lumbricoides , Trichuris trichiura, Strongyloides stercoralis, Necator americanus, Ancylostoma duodenale  and Ancylostoma ceylanicum  (S2 Table). Laboratory staff were blinded to Kato-Katz results for each sample during the initial qPCR analyses. However, upon completion of initial testing, all samples determined to be Kato-Katz positive but qPCR negative for A. lumbricoides were tested using a second qPCR assay targeting an unrelated, ribosomal DNA sequence in order to validate the negative qPCR result and confirm the absence of A. lumbricoides DNA [26,40].
For each assay, all samples were tested in replicate reactions and samples were classified as positive when amplification occurred in both reactions with a Cq value of <40, consistent with prior studies . Samples which produced Cq values ≥40 in both replicate reactions or failed to amplify in both replicate reactions were classified as negative. Consistent with common practice , samples that were positive in one of two replicate reactions were re-tested in duplicate and classified as positive if the Cq value was <40 in at least one of two re-test reactions. In all cases of re-testing, the Cq value reported as the sample’s final result was the re-test Cq value.
In addition to the IAC control, all experimental qPCR reaction plates were accompanied by the testing of a titration of the appropriate assay’s control plasmid. Plasmid controls for each target sequence were prepared as previously described . Each plasmid contained a single copy of the corresponding assay’s target sequence and 20 pg, 200 fg and 2 fg of plasmid were tested in duplicate reactions. Following the completion of all testing, the mean Cq value, across all experimental plates, for each plasmid concentration was determined. Any plate which produced a Cq value for any concentration of plasmid that was 3 or more standard deviations greater than the mean calculated for all plates was retested in its entirety, and all results from that plate were considered to be invalid. Only results from valid plates were reported (S3 Table).
To verify that Kato-Katz positive/qPCR negative samples were truly qPCR negative for A. lumbricoides, we performed sequencing on a subset of samples. Because <1% of samples tested negative for hookworm or T. trichiura using qPCR but positive using Kato-Katz, we only performed sequencing for samples with discordant A. lumbricoides classification. A subset of 10 samples was selected to undergo amplicon sequencing-based analysis. To facilitate the selection of samples, all samples that were double-slide Kato-Katz positive for the presence of A. lumbricoides but qPCR negative for A. lumbricoides were identified. From this list, a convenience sample (n = 7) was then selected, such that two of these samples would contain epg counts characteristic of moderate-intensity infections as determined by World Health Organization guidelines, while five samples would contain egg counts that were characteristic of light intensity infections . All seven of the Kato-Katz positive samples were negative by qPCR using both A. lumbricoides assays. Additionally, three Kato-Katz negative samples were chosen for inclusion in this panel. Two of these samples were selected due to their status as A. lumbricoides-positive as determined by qPCR analysis, while a single sample was selected that was both Kato-Katz negative and qPCR negative to serve as an uninfected control.
Samples were prepared for sequencing using a modified version of the Earth Microbiome Project’s 18S Illumina Amplicon Protocol available at: http://www.earthmicrobiome.org/protocols-and-standards/18s/. This protocol utilizes primers targeting the variable region 9 (V9) of the eukaryotic small subunit (SSU) rDNA . This protocol has been used in the analysis of helminth biodiversity in the gut of rats  and validated as a method for assessing global parasite diversity . Briefly, the targets within each sample were amplified in triplicate 25 μl reactions using a uniquely barcoded reverse primer coupled with a conserved forward primer (S4 Table). A “mammalian blocking” primer, based on a previously described strategy, was also included in all reactions to minimize the prevalence of mammal-derived amplicons . All reactions contained 10 μl of Platinum Hot Start PCR Master Mix (2X) (ThermoFisher, Waltham, MA), 0.2 μM forward and reverse primers, and 1.6 μM “mammalian blocking” primer (Integrated DNA Technologies, Coralville, IA). Cycling consisted of an initial denaturation step of 94°C for 3 minutes, followed by 35 cycles of 94°C for 45 seconds, 65°C for 15 seconds, 57°C for 30 seconds, and 72°C for 90 seconds. Samples then underwent a final extension step at 72°C for 10 minutes. Following amplification, triplicate reactions were combined, and pooled products were run on a 1.5% agarose gel to confirm the presence of a band of the expected size (~260 bp). 240 ng of each pooled product was then combined in preparation for sequencing, and this library was purified using the ZR-96 Clean & Concentrator purification kit (Zymo Research, Irvine, CA). The purified library was diluted to a 9 pM concentration, and 30% PhiX was added to improve diversity. Sequencing then occurred on the Illumina MiSeq platform, using a MiSeq Reagent Kit v2 (300-cycles) (Illumina, San Diego, CA).
Sequencing results were analyzed using the Quantitative Insights into Microbial Ecology 2 (QIIME2) pipeline . Briefly, all interlaced read pairs which did not contain a valid barcode were filtered out and all read pairs whose best match was to bacterial 16S ribosomal sequence were also removed. This resulted in a list of reads exclusively matching eukaryotic 18S sequence. Reads with 99% or greater identity were then classified into operational taxonomic units (OTUs) and consensus sequences were built. Taxonomy was then assigned to all OTUs that had ≥ 95% identity to sequences within the Silva release 132 QIIME-compatible database.
All statistical analyses were pooled across intervention arms to ensure sufficient statistical power. Unless otherwise specified, analyses were conducted in R version 3.4.3. Data and replication scripts are available through the Open Science Framework (https://osf.io/agk6w/). The STARD Standard and Bayes Latent Class Models Checklists for this study are included in S1 Checklist and S2 Checklist.
Assessment of quality of Kato-Katz slide readings
For the 2,800 samples included in this analysis, we estimated concordance between individual Kato-Katz technicians and expert technicians by comparing the classification of positive/negative status for each STH on each slide as well as the mean difference in eggs counted per slide.
Prevalence and correlation between eggs per gram and Cq values
We estimated the prevalence of STH infection using each diagnostic method with robust sandwich standard errors to account for clustering at the village level . We calculated the agreement between prevalence estimates from double-slide Kato-Katz and qPCR using a kappa statistic. We tested whether prevalence differed between the two diagnostic methods using the Wilcoxon matched-pairs signed rank test that conditioned on randomized block in the original trial [32,50]. To classify infection intensity, we used the World Health Organization cutoffs (light intensity infections were defined as < 5,000 epg for A. lumbricoides, < 1,000 epg for T. trichiura, and < 2,000 epg for hookworm; moderate intensity infections were defined as 5,000 ≤ epg < 50,000 for A. lumbricoides, 1,000 ≤ epg < 10,000 for T. trichiura, and 2,000 ≤ epg < 4,000 for hookworm; heavy intensity infections were defined as ≥ 50,000 epg for A. lumbricoides, ≥ 10,000 epg for T. trichiura, and ≥ 4,000 epg for hookworm) . To assess the correlation between epg estimated by double-slide Kato-Katz and Cq value estimated by qPCR, we estimated Kendall’s tau and calculated two-sided p-values using a bootstrap with 1,000 replicates that resampled blocks to account for the study’s cluster-randomized design using Stata version 14.2. We compared mean Cq values within levels of infection intensity using a paired t-test paired within randomized blocks.
Estimation of sensitivity and specificity
A challenge in studies assessing accuracy of STH diagnostics is the lack of a gold standard diagnostic method . We thus used two different approaches to estimate sensitivity and specificity of each method, consistent with prior studies. We created a reference using both detection methods and defined positivity as qPCR Cq values < 40 or at least one egg detected by double-slide Kato-Katz. We also used Bayesian latent class analysis, which defines the true prevalence of STH infection, sensitivity, and specificity as latent variables that are estimated simultaneously from the data and assumes no gold standard [22,28,52–54]. We hypothesized that the number of STH worms in the intestines influences STH DNA concentration and epg in stool, each of which determines observed STH infection measured by qPCR and Kato-Katz. Because observed STH status measured by qPCR and Kato-Katz each depends upon true worm burden, we defined a model that allowed for conditional dependence between each test by including terms for covariance between diagnostic methods since both methods are dependent upon the same underlying biological process  (S1 Text).
Consistent with prior studies, our model did not include separate latent variables for worm burden, DNA concentration and epg, which were unobserved, because the model would have been non-identifiable. Latent class models for assessing diagnostic accuracy of two diagnostic tests in the absence of a gold standard are not identifiable because the number of unknown parameters exceeds the degrees of freedom ; in this case, multiple unique parameter estimates can be produced from the same model, and a single set of estimates cannot be identified. Within a Bayesian framework, when informative prior distributions are defined for at least two parameters, the model is identifiable . Consistent with prior studies, we used a highly informative prior for specificity of each diagnostic test to allow for identifiability [22,28,57,58] since prior studies have reported high specificity values for double-slide Kato-Katz [59,60] and qPCR [25,61] (S5 Table). For A. lumbricoides, we chose to specify a non-informative prior for double-slide Kato-Katz sensitivity and specificity and a less informative prior for qPCR sensitivity because of discordant test results (summarized below). Since the model’s parameter estimates are sensitive to the assumed prior distributions , we conducted a sensitivity analysis using alternative, more informative priors for A. lumbricoides (S6 Table). To estimate parameters, we used Monte Carlo simulation with 4 chains and 100,000 iterations, recording every 10th result. We computed the mean and 2.5 and 97.5 percentiles of the posterior distribution using Gibbs sampling in WinBUGS version 14 .
We analyzed stool samples from 2,799 children using both double-slide Kato-Katz and qPCR. 51% of children were female, and 66% of children’s caregivers reported that they had received deworming in the past 6 months (S1 Fig). Concordance between individual Kato-Katz technicians and expert parasitologists in the classification of STH positive/negative status was close to 100% for all technicians and STH (S2 Fig). For A. lumbricoides and T. trichiura, kappa statistics comparing individual technicians and expert parasitologists were high, and the kappa statistic for hookworm was high when the expert parasitologist performed Kato-Katz on the same day as the original count (S7 Table). The difference in mean eggs per slide between individual Kato-Katz technicians and experts was < 60 for A. lumbricoides, < 15 for hookworm, and < 20 for T. trichiura (S3 Fig). We excluded the qPCR results of one sample from statistical analyses for all species because the internal amplification control was > 3 standard deviations from the mean upon initial testing and following re-extraction.
The observed prevalence was higher for qPCR than double-slide Kato-Katz for hookworm and T. trichiura but lower for A. lumbricoides (Table 1, Fig 1). The observed prevalence of A. lumbricoides was 37.0% (95% CI 34.7%, 39.3%) using Kato-Katz and 23.3% (95% CI 20.9%, 25.7%) using qPCR. The observed prevalence of hookworm was 7.5% (95% CI 6.3%, 8.7%) using Kato-Katz. Using qPCR the observed prevalence of Necator americanus was 18.4% (95% CI 16.5%, 20.3%), and the observed prevalence of Ancylostoma ceylanicum was 3.8% (95% CI 2.9%, 4.6%); only one out of 2,799 samples tested positive for Ancylostoma duodenale. The observed prevalence of any hookworm species using qPCR was 21.4% (95% CI 19.4%, 23.3%). For T. trichiura, the observed prevalence was 7.0% (95% 5.6%, 8.3%) by Kato-Katz and 12.3% (95% CI 10.3%, 14.2%) by qPCR. Using Kato-Katz, 7% of children were infected with more than one STH, and using qPCR it was 13% (Fig 2). The observed prevalence of moderate intensity infection using Kato-Katz was 11% for A. lumbricoides, less than 1% for hookworm, and 5% for T. trichiura. The observed prevalence of heavy intensity infection using Kato-Katz was <1% for A. lumbricoides, 1% for hookworm, and 0% for T. trichiura. Using Kato-Katz, the geometric mean of epg was 5.14 for A. lumbricoides (134 among positives), 0.43 for hookworm (125 among positives), and 0.40 for T. trichiura (124 among positives).
Prevalence was estimated from stool samples collected from children aged 2–12 years in rural Bangladesh (N = 2,799).
Soil-transmitted helminth ova or DNA were detected in stool samples collected from children aged 2–12 years in rural Bangladesh using Kato-Katz or multi-parallel qPCR (N = 2,799).
Concordance between double-slide Kato-Katz and qPCR was higher for hookworm and T. trichiura than for A. lumbricoides. The p-value for permutation tests assessing whether the observed prevalence estimated by the two detection methods differed was < 0.001 for each STH (Table 2). 6–14% of samples classified as negative for the different species by Kato-Katz were positive using qPCR. 26% of samples classified as positive for A. lumbricoides by Kato-Katz were negative using qPCR. All samples determined to be Kato-Katz positive for A. lumbricoides but negative by qPCR analysis targeting the non-coding repeat also failed to produce a positive result when tested with the ITS-targeting assay. The majority of epg values among A. lumbricoides Kato-Katz-positive, qPCR-negative samples were < 100, although a small number of samples had epg between 1,000 and 50,000 (S4 Fig). The probability that an individual Kato-Katz technician classified a stool sample as A. lumbricoides positive among those classified as negative by qPCR ranged from 20% to 31%, and four out of six technicians were significantly more likely to misclassify a sample than the technician with the lowest misclassification rate (S5 Fig). There was no evidence that the proportion of samples classified as positive for A. lumbricoides by Kato-Katz and negative by qPCR (false positives assuming qPCR was the gold standard) followed a trend by date of Kato-Katz or date of DNA extraction (S6 Fig, S7 Fig); patterns by date were similar for false negatives (S8 Fig, S9 Fig).
Using Bayesian latent class models, the estimated true prevalence was 27% (95% Bayesian Credible Interval (BCI) 20%, 37%) for A. lumbricoides, 20% (95% BCI 17%, 24%) for hookworm, and 11% (95% BCI 8%, 14%) for T. trichiura. For double-slide Kato-Katz, the sensitivity was 49% (95% BCI 34%, 64%) for A. lumbricoides, 32% (95% BCI 22%, 41%) for hookworm, and 52% (95% BCI 33%, 71%) for T. trichiura, and the specificity was 68% (95% BCI 61%, 77%) for A. lumbricoides, 99% (95% BCI 96%, 100%) for hookworm, and 98% (95% BCI 96%, 100%) for T. trichiura (Table 3). For qPCR, the sensitivity was 79% (95% BCI 61%, 99%) for A. lumbricoides, 93% (95% BCI 82%, 100%) for hookworm, and 90% (95% BCI 81%, 100%) for T. trichiura, and the specificity was 97% (95% BCI 95%, 100%) for all three STH. The sensitivity analysis for A. lumbricoides using alternative more informative priors produced similar estimates for Kato-Katz and higher sensitivity for qPCR; when more informative priors were used for both Kato-Katz and qPCR, the estimated sensitivity of qPCR was 90% (95% BCI 80%, 99%) (S8 Table). Pooling both methods as the gold standard, the sensitivity and specificity of Kato-Katz was similar to those estimated from the Bayesian model, and for qPCR the sensitivity of hookworm and T. trichiura was higher but the specificity was lower than estimated by the Bayesian model.
The distribution of Cq values for A. lumbricoides was bimodal, and to a lesser extent the same pattern was present for N. americanus and T. trichiura (S10 Fig). For each STH, there was a negative, monotonic relationship between qPCR Cq values and epg estimated using double-slide Kato-Katz (Fig 3). The Kendall’s tau comparing these two quantities was -0.442 for A. lumbricoides, -0.346 for N. americanus, -0.266 for A. ceylanicum, and -0.248 for T. trichiura (for each, the p-value was < 0.001); we compared N. americanus and A. ceylanicum to any hookworm species detected by Kato-Katz. We also examined the distribution of Cq values by Kato-Katz infection intensity status for A. lumbricoides (very few children had moderate-to-heavy intensity hookworm or T. trichiura infections using Kato-Katz) (Fig 4). The distribution of Cq values was higher for children who were Kato-Katz negative and children with light intensity infections compared to those with moderate-to-heavy infections. The median Cq value was 9 (range: 4–34) among children with moderate to heavy A. lumbricoides infection, 14 (range: 8–38) among those with light intensity infection, and 27 (range: 9–40) among those who were Kato-Katz negative (t-test p-value < 0.001 for all comparisons).
Soil-transmitted helminth ova or DNA were detected by Kato-Katz and qPCR in stool samples collected from children aged 2–12 years in rural Bangladesh (N = 2,799). The black solid line is the LOESS smoother for values that were classified as positive using both tests. Gray points indicate results that were negative by qPCR but positive by Kato-Katz; they are gray because Kato-Katz cannot differentiate between hookworm species. The Kendall’s tau comparing eggs per gram and Cq values was -0.442 for A. lumbricoides, -0.346 for N. americanus, -0.266 for A. ceylanicum, and -0.248 for T. trichiura (for each, the p-value was < 0.001).
Light intensity infections were defined as < 5,000 epg for A. lumbricoides, < 1,000 epg for T. trichiura, and < 2,000 epg for hookworm; moderate intensity infections were defined as 5,000 ≤ epg < 50,000 for A. lumbricoides, 1,000 ≤ epg < 10,000 for T. trichiura, and 2,000 ≤ epg < 4,000 for hookworm; heavy intensity infections were defined as ≥ 50,000 epg for A. lumbricoides, ≥ 10,000 epg for T. trichiura, and ≥ 4,000 epg for hookworm per the World Health Organization definition.
In the sequencing analyses of 10 samples, we detected Enterobius vermicularis in one sample that was double-slide Kato-Katz negative and qPCR positive, one sample that was Kato-Katz positive and qPCR negative, and one sample that was negative using both methods (S9 Table). In addition, we identified Giardia intestinalis in 3 samples that were Kato-Katz positive and qPCR negative and one sample that was negative using both methods; we detected Dientamoeba fragilis in 3 samples that were Kato-Katz positive and qPCR negative, one sample that was Kato-Katz negative and qPCR positive, and one sample that was negative using both methods. Analysis using the QIIME2 pipeline failed to identify any reads mapping to A. lumbricoides in all seven Kato-Katz positive, qPCR negative samples. A. lumbricoides was detected in one of the two samples that were Kato-Katz negative and qPCR positive. However, this analysis failed to identify A. lumbricoides in the second sample. Because of this discrepancy, the A. lumbricoides-derived consensus sequence of the reads from the A. lumbricoides-positive sample was compared, using BLAST, to all other reads from the entire sample set. This resulted in the identification of a small number of A. lumbricoides-derived reads (n = 5, identity ≥ 97%) within the second Kato-Katz negative, qPCR positive sample, while no A. lumbricoides reads were found within the pool of reads mapping to any of the other samples. As only a single read within this Kato-Katz negative, qPCR positive sample had 100% identity with the A. lumbricoides-derived consensus sequence used as the query sequence in the BLAST analysis, QIIME2’s failure to generate an OTU from these reads follows logically, as it is likely that the 99% similarity threshold was not met for any two reads mapping to A. lumbricoides from this sample. However, the presence of A. lumbricoides-derived reads provides further support for the existence of A. lumbricoides DNA within the sample and speaks to the sensitivity of the qPCR assays used during the analysis.
This study compared the performance of double-slide Kato-Katz and qPCR for detecting STH infections in a large sample of 2,800 children in a setting in rural Bangladesh with biannual mass deworming administration. Consistent with prior studies in low STH infection intensity settings, using qPCR led to substantial increases in hookworm detection and moderate increases in T. trichiura detection compared to Kato-Katz [23,25,27,28,51]. Poorer performance of Kato-Katz, particularly for hookworm, is likely due to rapid disintegration of hookworm ova following stool collection [51,64]. However, unexpectedly, we found that one third of samples classified as A. lumbricoides negative by qPCR were classified as positive by Kato-Katz, and the specificity of Kato-Katz was lower for A. lumbricoides than for hookworm or T. trichiura. This discordance in classification occurred primarily among samples with epg between 10 and 100, but surprisingly, a small number of samples with A. lumbricoides epg between 1,000 and 50,000 classified as negative by qPCR (S4 Fig). Experimental testing with two qPCR assays utilizing independent targets as well as sequencing analysis of a subset of samples confirmed that samples classified as A. lumbricoides negative by qPCR contained no genomic targets for A. lumbricoides. These unexpected findings call into question the common assumption that Kato-Katz yields few false positives [22,28,57,58].
There are several potential explanations for the unexpected discrepancy between Kato-Katz and qPCR for A. lumbricoides. First, poor quality control of Kato-Katz could have caused the discrepancy. The quality control assessment found minimal differences between classification of A. lumbricoides between individual laboratory technicians and expert parasitologists (S2 Fig, S3 Fig, S8 Table), suggesting that Kato-Katz was performed with typical quality. However, the expert parasitologist that performed quality assurance also conducted Kato-Katz training; thus, if the expert mistook regularly misclassified substances in stool as A. lumbricoides, misclassification may have been passed on to technicians during training and would not have been caught during quality assurance. Second, the discrepancy could have resulted from mislabeling of stool samples. While we cannot rule out this possibility, we consider this highly unlikely; if stool samples were mislabeled, we would expect to see discrepancies for hookworm and T. trichiura as well. There was no association between the false positive rate and false negative rate (using qPCR as the gold standard) by date of Kato-Katz or date of DNA extraction, which suggests that mislabeling or other laboratory protocol violations were unlikely since both may have caused an unexpected increase in false positives/negatives at a specific point in time (S6 Fig, S7 Fig, S8 Fig, S9 Fig). Third, it is possible that improper sample storage could have resulted in STH DNA degradation that led to our discrepant findings for A. lumbricoides, however, if that was the case we would expect to also see discrepancies for other STH and in particular for hookworm, which is more fragile and is known to degrade more rapidly than A. lumbricoides. Finally, it is possible that the discrepancy was due to insufficient DNA extraction, but if this was the cause we would expect to see similarly poor performance for all qPCR assays relative to Kato-Katz results. However, qPCR had higher sensitivity than Kato-Katz for hookworm and T. trichiura. Taken together, these findings suggest that there is an alternative explanation for the A. lumbricoides discrepancy.
Misclassification of STH ova has been previously reported: Capillaria philippinensis or Capillaria hepatica ova can be mistaken for A. lumbricoides or T. trichiura ova [65,66] and Trichostrongylus spp. or Schistosoma ova can resemble hookworm ova [67,68]. Though we did identify other enteropathogens (Giardia intestinalis and Dientamoeba fragilis) when sequencing a subset of samples, in stool, these parasites do not share morphological features with most stages of A. lumbricoides . We hypothesize that the substance mistaken for A. lumbricoides in Kato-Katz was most likely plant material, pollen grains, or fungal spores, all of which can resemble certain stages of A. lumbricoides . Some of the sequenced stool samples contained DNA of flowering plants of the division Magnoliophyta, which produce pollen, and fungi of the genera Saccharomyces and Candida, which are commonly found in healthy intestinal microbiota  (S9 Table). We cannot definitively identify whether a different organism or substance was mistaken for A. lumbricoides because the original slides used in Kato-Katz were not preserved. Nevertheless, our findings raise concerns about the quality of the Kato-Katz procedures and underscore the advantage of using molecular methods over copromicroscopy since molecular methods do not rely on visualization of stool contents.
Overall, we found that sensitivity was lower for both double-slide Kato-Katz and qPCR than in prior studies [22,28]. A challenge in quantifying the accuracy of STH diagnostics is the lack of a gold standard measure of infection . Though we used Bayesian latent class analysis, which assumes no gold standard, our estimates of sensitivity and specificity from this method should be interpreted with caution. In studies evaluating fewer than four diagnostic tests, parameters are not identifiable, and as a result parameter estimates are highly dependent upon the assumed prior distributions . Our latent class models used informative priors for the specificity of each diagnostic, which we assumed was between 0.95 and 1.00 for hookworm and T. trichiura; for these STH there was agreement between results from the Bayesian latent class analysis and an analysis pooling both methods as the reference, which suggests that our assumption was appropriate. However, because of the apparent false positives for A. lumbricoides using Kato-Katz, our prior distribution for the specificity of Kato-Katz ranged from 0 to 1. Using this model with less informative priors, our estimate of sensitivity of qPCR for detecting A. lumbricoides using Bayesian latent class analysis was lower than a previous estimate for multi-parallel qPCR using the same method . Using less informative priors for A. lumbricoides may have reduced the identifiability of our model and contributed to differing sensitivity and specificity estimates in this study. However, our sensitivity analysis using more informative priors for A. lumbricoides produced estimates that were similar to our primary analysis and/or closer to previously published estimates (S6 Table, S8 Table). Our model also assumed that the sensitivity and specificity of each diagnostic was constant for each study subject . It is possible that sensitivity and specificity varied due to variation in STH prevalence and infection intensity, differing laboratory technicians, time from stool collection to Kato-Katz or stool archiving, or other factors; however, we expect such heterogeneity to be relatively low because our study used standardized procedures and was in a relatively small geographic area.
We observed a bimodal distribution of Cq values for A. lumbricoides and to a lesser extent for N. americanus and T. trichiura. While we cannot definitively identify the cause of this pattern, we propose two potential hypotheses. First, it is possible that the lower mode of Cq values (near Cq = 12) reflects target sequences from intact eggs, while the higher mode (near Cq = 32) reflects the presence of target sequences from samples with sub-egg quantities (i.e. sub-genome quantities) of DNA (S10 Fig). While the underlying cause was not addressed, similar distributions have been reported in prior studies . Furthermore, an egg spiking study utilizing the same repeat-targeting assay employed in this study and employing the same controls and instrumentation used in this study, determined that on average, DNA extraction from a sample containing a single A. lumbricoides egg gives a Cq value of approximately 24, with the most efficient extractions commonly producing Cq values in the range of 19–22 . This implies that efficiently extracted sub-egg quantities of target DNA should result in Cq values greater than 22. At nearly 250,000 copies per haploid genome, the genomic target of the employed Ascaris assay is more numerous than the repetitive sequences utilized as targets in any of the other assays. As such, its ability to detect trace amounts of genomic DNA is likely greater, providing an explanation for the more pronounced secondary distribution with higher Cq values. Additional research is needed to determine whether individuals with sub-egg levels of STH DNA in their stool experience clinical symptoms of infection and/or contribute to STH transmission.
A second possible explanation for the bimodal Cq distribution is that the lower mode represents fertilized eggs with diploid genomes, and the higher mode represents unfertilized eggs shed from female-only A. lumbricoides infections with haploid genomes. Unfertilized adult female A. lumbricoides worms may still shed large numbers of single-genome eggs. However, fertilized eggs undergo multiple cellular divisions during the process of embryonation, meaning that a single egg may contain many copies of the genome. Typically, under permissive environmental conditions, embryonation occurs over the course of approximately 2–3 weeks following fecal shedding, generally requiring approximately 5 days for eggs to reach the 8-cell stage . When stored at colder temperatures approaching 5°C, such development is arrested . Given that all stool samples collected as part of this study were placed on ice within 18 hours of sample production, it is extremely unlikely that significant embryonic development could have occurred prior to DNA extraction and qPCR testing. Therefore, while fertilized eggs would contain diploid genomes it is unlikely that significant cell division would have occurred. As such, in this study, any effect on Cq value resulting from fertilization status is likely to have been minimal, making this unlikely to account for the presence of the observed bimodal distribution.
Whether Cq values can be used to approximate intensity of STH infection is an open question. The strength of the correlation between epg and Cq values informs whether Cq value thresholds can be defined for STH infection intensity as is done using Kato-Katz . Prior studies have reported moderate-to-high correlations between epg and DNA concentration for all three STH [23–25,28,75], and one study reported that Cq values correlated well with expelled adult worms following anthelminthic treatment . We found moderate correlations between epg and Cq values that were lower than the correlations reported in prior studies [24,25,28]. A. lumbricoides was the only STH in this study with a sufficiently large number of samples classified as moderate-to-heavy intensity using Kato-Katz. The majority of moderate-to-heavy intensity infections had Cq values of 10 or lower (Fig 4). However, because of the misclassification of A. lumbricoides, we interpret this pattern with caution because epg counts may have been overestimated in some samples and underestimated in others. In general, a limitation of using Cq values to approximate infection intensity is that the quantity of cells per egg (and thus the copy number of a target) varies depending on fertilization status, cell number, diminution status, and the STH species , complicating inferences about epg using Cq values. When comparing Cq values to epg estimated from Kato-Katz, another limitation is that there is variability in the number of STH eggs secreted in stool, the number of haploid genomes present per egg, and eggs are not distributed evenly within stool or in stool samples [77,78]. Despite these limitations, Cq values represent the only quantitative output from a qPCR reaction, and as such, they may represent the best (albeit imperfect) measure for approximating infection intensity.
In settings with predominantly low STH infection intensity, highly sensitive diagnostics are needed to detect low intensity infections prior to elimination and to detect resurgent infections. Though a common critique of qPCR is its higher cost relative to Kato-Katz, comparisons of the materials required for each method suggest that costs for multi-parallel qPCR can be lower than those for Kato-Katz . As pointed out by Turner et al., even if qPCR is more costly than Kato-Katz, in an elimination setting, the continued use of low sensitivity diagnostics may impede efforts to determine when STH transmission has been interrupted which may unnecessarily prolong mass deworming administration. The cost of prolonged MDA is likely to exceed that of a more costly diagnostic . Our results support the use of qPCR in elimination settings due to its higher sensitivity and specificity. In addition, importantly, using qPCR instead of Kato-Katz in intervention trials can also reduce misclassification and bias of intervention effect estimates .
Kato-Katz remains a highly cost-effective, feasible diagnostic method appropriate for certain uses cases, such as identifying where to perform MDA, determining the frequency of MDA, and assessing progress towards program goals . Prior to this study, the greatest disadvantage of Kato-Katz was typically considered to be its low sensitivity; here we show that not only was sensitivity lower for Kato-Katz than for qPCR, but also in our setting, Kato-Katz had low specificity for A. lumbricoides. Though we were not able to definitively determine the cause of the apparent A. lumbricoides misclassification using Kato-Katz, regardless of the cause, our results highlight an inherent limitation of Kato-Katz. In addition, our results echo conclusions of other recent studies that qPCR may be more appropriate for use cases such as identifying whether STH transmission has been interrupted and confirming sustained transmission interruption due to its higher sensitivity in low STH intensity settings .
S2 Checklist. STARD Bayesian Latent Class Models Analysis Checklist.
S1 Table. Percent agreement between the original Kato-Katz technician and the senior parasitologist by days elapsed between each assessment.
S2 Table. Target information, sequences, manufacturers, and relevant references for all qPCR primers and probes used during this study.
S3 Table. Results of plasmid control testing for each assay.
For a given plate, in the event that any concentration of control plasmid was 3 or more standard deviations greater than the mean (calculated from all experimental plates), the plate was rerun in its entirety and the results of initial testing were considered invalid.
S4 Table. Primer sequences used for 18S SSU sequencing.
S5 Table. Prior distributions used in Bayesian latent class analysis models.
S6 Table. Alternative prior distributions used in sensitivity analysis for Bayesian latent class analysis models for A. lumbricoides.
S7 Table. Kappa statistics comparing the original classification of a single slide as having any STH ova between individual Kato-Katz technicians and expert parasitologists.
S8 Table. Sensitivity analysis using alternative prior distributions in Bayesian latent class analysis models for A. lumbricoides.
S9 Table. Taxonomic assignments for OTUs resulting from sequencing analysis with the Qiime2 pipeline.
Following de-multiplexing by barcode, OTUs underwent taxonomic assignment. For each sample, the number of interlaced read pairs mapping to each taxonomic assignment is provided.
S1 Text. Bayesian latent class model specification.
S2 Fig. Percent agreement between laboratory technicians and expert technicians in STH positive/negative status in individual slides.
S3 Fig. Mean difference in Kato-Katz single slide egg count between laboratory technician and expert counter among samples that were positive by either the laboratory technician or expert.
S4 Fig. Distribution of A. lumbricoides eggs per gram classified by concordance status between Kato-Katz and qPCR.
S5 Fig. Probability that a stool sample was classified as positive for A. lumbricoides using Kato-Katz among those classified as negative by qPCR by individual Kato-Katz technician.
S6 Fig. Cumulative probability that a stool sample was classified as positive for A. lumbricoides using Kato-Katz among those classified as negative by qPCR by date of Kato-Katz.
S7 Fig. Cumulative probability that a stool sample was classified as positive for A. lumbricoides using Kato-Katz among those classified as negative by qPCR by date of DNA extraction.
S8 Fig. Cumulative probability that a stool sample was classified as negative using Kato-Katz among those classified as positive by qPCR by date of Kato-Katz.
S9 Fig. Cumulative probability that a stool sample was classified as negative using Kato-Katz among those classified as positive by qPCR by date of DNA extraction.
The authors would like to acknowledge Brian P. Abrams for his contribution to laboratory analyses.
- 1. GBD 2016 DALYs and HALE Collaborators. Global, regional, and national disability-adjusted life-years (DALYs) for 333 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet. 2017;390: 1260–1344. pmid:28919118
- 2. Bethony J, Brooker S, Albonico M, Geiger SM, Loukas A, Diemert D, et al. Soil-transmitted helminth infections: ascariasis, trichuriasis, and hookworm. The Lancet. 2006;367: 1521–1532.
- 3. Lo NC, Heft-Neal S, Coulibaly JT, Leonard L, Bendavid E, Addiss DG. State of deworming coverage and equity in low-income and middle-income countries using household health surveys: a spatiotemporal cross-sectional study. Lancet Glob Health. 2019;7: e1511–e1520. pmid:31558383
- 4. Bundy DAP, Silva N de Horton S, Jamison DT. Disease Control Priorities, Third Edition (Volume 8): Child and Adolescent Health and Development. World Bank Publications; 2017.
- 5. World Health Organization. Accelerating work to overcome the global impact of neglected tropical diseases: A roadmap for implementation. 2012. Available: http://www.who.int/neglected_diseases/NTD_RoadMap_2012_Fullversion.pdf
- 6. Anderson R, Hollingsworth TD, Truscott J, Brooker S. Optimisation of mass chemotherapy to control soil-transmitted helminth infection. The Lancet. 2012;379: 289–290.
- 7. Anderson R, Truscott J, Hollingsorth TD The coverage and frequency of mass drug administration required to eliminate persistent transmission of soil-transmitted helminths. Philos Trans R Soc B Biol Sci. 2014;369: 20130435. pmid:24821921
- 8. Ásbjörnsdóttir KH, Means AR, Werkman M, Walson JL. Prospects for elimination of soil-transmitted helminths. Oct 2017 [cited 8 Feb 2019]. pmid:28700363
- 9. Ásbjörnsdóttir KH, Ajjampur SSR, Anderson RM, Bailey R, Gardiner I, Halliday KE, et al. Assessing the feasibility of interrupting the transmission of soil-transmitted helminths through mass drug administration: The DeWorm3 cluster randomized trial protocol. PLoS Negl Trop Dis. 2018;12: e0006166. pmid:29346377
- 10. Medley GF, Turner HC, Baggaley RF, Holland C, Hollingsworth TD. Chapter Six—The Role of More Sensitive Helminth Diagnostics in Mass Drug Administration Campaigns: Elimination and Health Impacts. In: Basáñez MG, Anderson RM, editors. Advances in Parasitology. Academic Press; 2016. pp. 343–392. https://doi.org/10.1016/bs.apar.2016.08.005 pmid:27756457
- 11. Kato K, Miura M. Comparative examinations. Jpn J Parasitol. 1954;3.
- 12. Katz N, Chaves A, Pellegrino J. A simple device for quantitative stool thick-smear technique in Schistosomiasis mansoni. Rev Inst Med Trop São Paulo. 1972;14: 397–400. pmid:4675644
- 13. Allen AV, Ridley DS. Further observations on the formol-ether concentration technique for faecal parasites. J Clin Pathol. 1970;23: 545–546. pmid:5529256
- 14. Cringoli G, Rinaldi L, Maurelli MP, Utzinger J. FLOTAC: new multivalent techniques for qualitative and quantitative copromicroscopic diagnosis of parasites in animals and humans. Nat Protoc. 2010;5: 503–515. pmid:20203667
- 15. Barda B, Cajal P, Villagran E, Cimino R, Juarez M, Krolewiecki A, et al. Mini-FLOTAC, Kato-Katz and McMaster: three methods, one goal; highlights from north Argentina. Parasit Vectors. 2014;7: 271. pmid:24929554
- 16. Barda B, Albonico M, Ianniello D, Ame SM, Keiser J, Speich B, et al. How Long Can Stool Samples Be Fixed for an Accurate Diagnosis of Soil-Transmitted Helminth Infection Using Mini-FLOTAC? PLoS Negl Trop Dis. 2015;9: e0003698. pmid:25848772
- 17. Cringoli G, Maurelli MP, Levecke B, Bosco A, Vercruysse J, Utzinger J, et al. The Mini-FLOTAC technique for the diagnosis of helminth and protozoan infections in humans and animals. Nat Protoc. 2017;12: 1723–1732. pmid:28771238
- 18. Cools P, Vlaminck J, Albonico M, Ame S, Ayana M, Antonio BPJ, et al. Diagnostic performance of a single and duplicate Kato-Katz, Mini-FLOTAC, FECPAKG2 and qPCR for the detection and quantification of soil-transmitted helminths in three endemic countries. PLoS Negl Trop Dis. 2019;13: e0007446. pmid:31369558
- 19. Ayana M, Vlaminck J, Cools P, Ame S, Albonico M, Dana D, et al. Modification and optimization of the FECPAKG2 protocol for the detection and quantification of soil-transmitted helminth eggs in human stool. PLoS Negl Trop Dis. 2018;12: e0006655. pmid:30321180
- 20. WHO. Assessing the Epidemiology of Soil-Transmitted Helminths During a Transmission Assessment Survey in the Global Programme for the Elimination of Lymphatic Filariasis. Geneva, Switzerland: WHO Press; 2015.
- 21. Anderson RM, Schad GA. Hookworm burdens and faecal egg counts: an analysis of the biological basis of variation. Trans R Soc Trop Med Hyg. 1985;79: 812–825. pmid:3832493
- 22. Nikolay B, Brooker SJ, Pullan RL. Sensitivity of diagnostic tests for human soil-transmitted helminth infections: a meta-analysis in the absence of a true gold standard. Int J Parasitol. 2014;44: 765–774. pmid:24992655
- 23. Verweij JJ, Brienen EAT, Ziem J, Yelifari L, Polderman AM, Van Lieshout L. Simultaneous Detection and Quantification of Ancylostoma duodenale, Necator americanus, and Oesophagostomum bifurcum in Fecal Samples Using Multiplex Real-Time PCR. Am J Trop Med Hyg. 2007;77: 685–690. pmid:17978072
- 24. Llewellyn S, Inpankaew T, Nery SV, Gray DJ, Verweij JJ, Clements ACA, et al. Application of a Multiplex Quantitative PCR to Assess Prevalence and Intensity Of Intestinal Parasite Infections in a Controlled Clinical Trial. PLoS Negl Trop Dis. 2016;10: e0004380. pmid:26820626
- 25. Mejia R, Vicuña Y, Broncano N, Sandoval C, Vaca M, Chico M, et al. A Novel, Multi-Parallel, Real-Time Polymerase Chain Reaction Approach for Eight Gastrointestinal Parasites Provides Improved Diagnostic Capabilities to Resource-Limited At-Risk Populations. Am J Trop Med Hyg. 2013;88: 1041–1047. pmid:23509117
- 26. Pilotte N, Papaiakovou M, Grant JR, Bierwert LA, Llewellyn S, McCarthy JS, et al. Improved PCR-Based Detection of Soil Transmitted Helminth Infections Using a Next-Generation Sequencing Approach to Assay Design. PLoS Negl Trop Dis. 2016;10: e0004578. pmid:27027771
- 27. Cimino RO, Jeun R, Juarez M, Cajal PS, Vargas P, Echazú A, et al. Identification of human intestinal parasites affecting an asymptomatic peri-urban Argentinian population using multi-parallel quantitative real-time polymerase chain reaction. Parasit Vectors. 2015;8: 380. pmid:26183074
- 28. Easton AV, Oliveira RG, O’Connell EM, Kepha S, Mwandawiro CS, Njenga SM, et al. Multi-parallel qPCR provides increased sensitivity and diagnostic breadth for gastrointestinal parasites of humans: field-based inferences on the impact of mass deworming. Parasit Vectors. 2016;9: 38. pmid:26813411
- 29. van Mens SP, Aryeetey Y, Yazdanbakhsh M, van Lieshout L, Boakye D, Verweij JJ. Comparison of real-time PCR and Kato smear microscopy for the detection of hookworm infections in three consecutive faecal samples from schoolchildren in Ghana. Trans R Soc Trop Med Hyg. 2013;107: 269–271. pmid:23426113
- 30. Lim MD, Brooker SJ, Jr VYB, Gay-Andrieu F, Gilleard J, Levecke B, et al. Diagnostic tools for soil-transmitted helminths control and elimination programs: A pathway for diagnostic product development. PLoS Negl Trop Dis. 2018;12: e0006213. pmid:29494581
- 31. Easton AV, Oliveira RG, Walker M, O’Connell EM, Njenga SM, Mwandawiro CS, et al. Sources of variability in the measurement of Ascaris lumbricoides infection intensity by Kato-Katz and qPCR. Parasit Vectors. 2017;10: 256. pmid:28545561
- 32. Arnold BF, Null C, Luby SP, Unicomb L, Stewart CP, Dewey KG, et al. Cluster-randomised controlled trials of individual and combined water, sanitation, hygiene and nutritional interventions in rural Bangladesh and Kenya: the WASH Benefits study design and rationale. BMJ Open. 2013;3: e003476. pmid:23996605
- 33. Luby S, Rahman M, Arnold B, Unicomb L, Ashraf S, Winch P, et al. Effects of water quality, sanitation, handwashing and nutritional interventions on diarrhoea and child growth in rural Bangladesh: A cluster randomized trial. Lancet Glob Health. 2018;6: e302–e315. pmid:29396217
- 34. Ercumen A, Benjamin-Chung J, Arnold BF, Lin A, Hubbard AE, Stewart C, et al. Effects of water, sanitation, handwashing and nutritional interventions on soil-transmitted helminth infections in young children: a cluster-randomized controlled trial in rural Bangladesh. PLoS Negl Trop Dis. 2019;13: e0007323. pmid:31050672
- 35. Benjamin-Chung J, Nazneen A, Halder AK, Haque R, Siddique A, Uddin MS, et al. The Interaction of Deworming, Improved Sanitation, and Household Flooring with Soil-Transmitted Helminth Infection in Rural Bangladesh. PLoS Negl Trop Dis. 2015;9: e0004256. pmid:26624994
- 36. Deer DM, Lampel KA, González‐Escalona N. A versatile internal control for use as DNA in real-time PCR and as RNA in real-time reverse transcription PCR assays. Lett Appl Microbiol. 50: 366–372. pmid:20149084
- 37. Papaiakovou M, Pilotte N, Baumer B, Grant J, Asbjornsdottir K, Schaer F, et al. A comparative analysis of preservation techniques for the optimal molecular detection of hookworm DNA in a human fecal specimen. PLoS Negl Trop Dis. 2018;12: e0006130. pmid:29346412
- 38. Pilotte N, Maasch JRMA, Easton AV, Dahlstrom E, Nutman TB, Williams SA. Targeting a highly repeated germline DNA sequence for improved real-time PCR-based detection of Ascaris infection in human stool. PLoS Negl Trop Dis. 2019;13: e0007593. pmid:31329586
- 39. Papaiakovou M, Pilotte N, Grant JR, Traub RJ, Llewellyn S, McCarthy JS, et al. A novel, species-specific, real-time PCR assay for the detection of the emerging zoonotic parasite Ancylostoma ceylanicum in human stool. PLoS Negl Trop Dis. 2017;11: e0005734. pmid:28692668
- 40. Liu J, Gratz J, Amour C, Kibiki G, Becker S, Janaki L, et al. A Laboratory-Developed TaqMan Array Card for Simultaneous Detection of 19 Enteropathogens. J Clin Microbiol. 2013;51: 472–480. pmid:23175269
- 41. Caraguel CGB, Stryhn H, Gagné N, Dohoo IR, Hammell KL. Selection of a Cutoff Value for Real-Time Polymerase Chain Reaction Results to Fit a Diagnostic Purpose: Analytical and Epidemiologic Approaches: J Vet Diagn Invest. 2011 [cited 24 Apr 2019]. pmid:21217022
- 42. Schmaedick MA, Koppel AL, Pilotte N, Torres M, Williams SA, Dobson SL, et al. Molecular Xenomonitoring Using Mosquitoes to Map Lymphatic Filariasis after Mass Drug Administration in American Samoa. PLoS Negl Trop Dis. 2014;8: e3087. pmid:25122037
- 43. WHO. Prevention and control of schistosomiasis and soil-transmitted helminthiasis. World Health Organ Tech Rep Ser. 2002;912: i–vi, 1–57, back cover. pmid:12592987
- 44. Amaral-Zettler LA, McCliment EA, Ducklow HW, Huse SM. A Method for Studying Protistan Diversity Using Massively Parallel Sequencing of V9 Hypervariable Regions of Small-Subunit Ribosomal RNA Genes. PLOS ONE. 2009;4: e6372. pmid:19633714
- 45. Tanaka R, Hino A, Tsai IJ, Palomares-Rius JE, Yoshida A, Ogura Y, et al. Assessment of Helminth Biodiversity in Wild Rats Using 18S rDNA Based Metagenomics. PLOS ONE. 2014;9: e110769. pmid:25340824
- 46. Hino A, Maruyama H, Kikuchi T. A novel method to assess the biodiversity of parasites using 18S rDNA Illumina sequencing; parasitome analysis method. Parasitol Int. 2016;65: 572–575. pmid:26784367
- 47. Vestheim H, Jarman SN. Blocking primers to enhance PCR amplification of rare sequences in mixed samples–a case study on prey DNA in Antarctic krill stomachs. Front Zool. 2008;5: 12. pmid:18638418
- 48. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet C, Al-Ghalith GA, et al. QIIME 2: Reproducible, interactive, scalable, and extensible microbiome data science. PeerJ Inc.; 2018 Dec. Report No.: e27295v2.
- 49. Freedman DA. On The So-Called “Huber Sandwich Estimator” and “Robust Standard Errors.” Am Stat. 2006;60: 299–302.
- 50. Imbens GW, Rubin DB. Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction. New York, NY, USA: Cambridge University Press; 2015.
- 51. O’Connell EM, Nutman TB. Molecular Diagnostics for Soil-Transmitted Helminths. Am J Trop Med Hyg. 2016;95: 508–513. pmid:27481053
- 52. Dendukuri N, Joseph L. Bayesian Approaches to Modeling the Conditional Dependence Between Multiple Diagnostic Tests. Biometrics. 2001;57: 158–167. pmid:11252592
- 53. Branscum AJ, Gardner IA, Johnson WO. Estimation of diagnostic-test sensitivity and specificity through Bayesian modeling. Prev Vet Med. 2005;68: 145–163. pmid:15820113
- 54. Assefa LM, Crellen T, Kepha S, Kihara JH, Njenga SM, Pullan RL, et al. Diagnostic Accuracy and Cost-Effectiveness of Alternative Methods for Detection of Soil-Transmitted Helminths in a Post-Treatment Setting in Western Kenya. PLoS Negl Trop Dis. 2014;8: e2843. pmid:24810593
- 55. Walter SD, Irwig LM. Estimation of test error rates, disease prevalence and relative risk from misclassified data: a review. J Clin Epidemiol. 1988;41: 923–937. pmid:3054000
- 56. Joseph L, Gyorkos TW, Coupal L. Bayesian Estimation of Disease Prevalence and the Parameters of Diagnostic Tests in the Absence of a Gold Standard. Am J Epidemiol. 1995;141: 263–272. pmid:7840100
- 57. Knopp S, Speich B, Hattendorf J, Rinaldi L, Mohammed KA, Khamis IS, et al. Diagnostic Accuracy of Kato-Katz and FLOTAC for Assessing Anthelmintic Drug Efficacy. PLoS Negl Trop Dis. 2011;5: e1036. pmid:21532740
- 58. Levecke B, Behnke JM, Ajjampur SSR, Albonico M, Ame SM, Charlier J, et al. A Comparison of the Sensitivity and Fecal Egg Counts of the McMaster Egg Counting and Kato-Katz Thick Smear Methods for Soil-Transmitted Helminths. PLoS Negl Trop Dis. 2011;5: e1201. pmid:21695104
- 59. Knopp S, Mgeni AF, Khamis IS, Steinmann P, Stothard JR, Rollinson D, et al. Diagnosis of Soil-Transmitted Helminths in the Era of Preventive Chemotherapy: Effect of Multiple Stool Sampling and Use of Different Diagnostic Techniques. PLoS Negl Trop Dis. 2008;2: e331. pmid:18982057
- 60. Booth M, Vounatsou P, N’Goran EK, Tanner M, Utzinger J. The influence of sampling effort and the performance of the Kato-Katz technique in diagnosing Schistosoma mansoni and hookworm co-infections in rural Côte d’Ivoire. Parasitology. 2003;127: 525–531. pmid:14700188
- 61. Verweij JJ. Application of PCR-based methods for diagnosis of intestinal parasitic infections in the clinical laboratory. Parasitology. 2014;141: 1863–1872. pmid:24780241
- 62. van Smeden M, Naaktgeboren CA, Reitsma JB, Moons KGM, de Groot JAH. Latent Class Models in Diagnostic Studies When There is No Reference Standard—A Systematic Review. Am J Epidemiol. 2014;179: 423–431. pmid:24272278
- 63. Lunn D, Thomas A, Best N, Spiegelhalter D. WinBUGS—a Bayesian modelling framework: concepts, structure, and extensibility. Stat Comput. 2000;10: 325–337.
- 64. Dacombe RJ, Crampin AC, Floyd S, Randall A, Ndhlovu R, Bickle Q, et al. Time delays between patient and laboratory selectively affect accuracy of helminth diagnosis. Trans R Soc Trop Med Hyg. 2007;101: 140–145. pmid:16824566
- 65. Walden J. Parasitic diseases. Other roundworms. Trichuris, hookworm, and Strongyloides. Prim Care. 1991;18: 53–74. pmid:2011642
- 66. Fischer K, Gankpala A, Gankpala L, Bolay FK, Curtis KC, Weil GJ, et al. Capillaria Ova and Diagnosis of Trichuris trichiura Infection in Humans by Kato-Katz Smear, Liberia. Emerg Infect Dis. 2018;24: 1551–1554. pmid:30016245
- 67. Ralph A, O’Sullivan MVN, Sangster NC, Walker JC. Abdominal pain and eosinophilia in suburban goat keepers—trichostrongylosis. Med J Aust. 2006;184: 467–469. pmid:16646749
- 68. Tarafder MR, Carabin H, Joseph L, Balolong E, Olveda R, McGarvey ST. Estimating the sensitivity and specificity of Kato-Katz stool examination technique for detection of hookworms, Ascaris lumbricoides and Trichuris trichiura infections in humans in the absence of a ‘gold standard.’ Int J Parasitol. 2010;40: 399–404. pmid:19772859
- 69. Garcia LS. Diagnostic Medical Parasitology. American Society for Microbiology Press; 2006.
- 70. Colmer-Hamood JA. Fecal Microscopy Artifacts Mimicking Ova and Parasites. Lab Med. 2001;32: 80–84.
- 71. Lloyd-Price J, Abu-Ali G, Huttenhower C. The healthy human microbiome. Genome Med. 2016;8. https://doi.org/10.1186/s13073-016-0307-y
- 72. Bermingham ML, Handel IG, Glass EJ, Woolliams JA, Bronsvoort BM de C, McBride SH, et al. Hui and Walter’s latent-class model extended to estimate diagnostic test properties from surveillance data: a latent model for latent data. Sci Rep. 2015;5: 11861. pmid:26148538
- 73. Kim M-K, Pyo K-H, Hwang Y-S, Park KH, Hwang IG, Chai J-Y, et al. Effect of Temperature on Embryonation of Ascaris suum Eggs in an Environmental Chamber. Korean J Parasitol. 2012;50: 239–242. pmid:22949753
- 74. Anderson RM, May RM. Helminth Infections of Humans: Mathematical Models, Population Dynamics, and Control. Advances in parasitology; 1985.
- 75. Knopp S, Salim N, Schindler T, Voules DAK, Rothen J, Lweno O, et al. Diagnostic Accuracy of Kato–Katz, FLOTAC, Baermann, and PCR Methods for the Detection of Light-Intensity Hookworm and Strongyloides stercoralis Infections in Tanzania. Am J Trop Med Hyg. 2014;90: 535–545. pmid:24445211
- 76. Papaiakovou M, Gasser RB, Littlewood DTJ. Quantitative PCR-Based Diagnosis of Soil-Transmitted Helminth Infections: Faecal or Fickle? Trends Parasitol. 2019;35: 491–500. pmid:31126720
- 77. Speich B, Ali SM, Ame SM, Albonico M, Utzinger J, Keiser J. Quality control in the diagnosis of Trichuris trichiura and Ascaris lumbricoides using the Kato-Katz technique: experience from three randomised controlled trials. Parasit Vectors. 2015;8: 82. pmid:25652120
- 78. Sinniah B. Daily egg production of Ascaris lumbricoides: the distribution of eggs in the faeces and the variability of egg counts. Parasitology. 1982;84: 167–175. pmid:7063252
- 79. Turner HC, Bettis AA, Dunn JC, Whitton JM, Hollingsworth TD, Fleming FM, et al. Economic Considerations for Moving beyond the Kato-Katz Technique for Diagnosing Intestinal Parasites As We Move Towards Elimination. Trends Parasitol. 2017;33: 435–443. pmid:28187989
- 80. Ercumen A, Benjamin-Chung J, Arnold BF, Lin A, Hubbard AE, Stewart C, et al. Effects of water, sanitation, handwashing and nutritional interventions on soil-transmitted helminth infections in young children: a cluster-randomized controlled trial in rural Bangladesh. bioRxiv. 2019; 512509.