Systematic review and meta-analysis protocol: Diagnostic accuracy assessment for soil-transmitted helminth using microscopy and molecular-based methods

Abstract

Soil-transmitted helminth (STH) is a part of NTDs that remains to be eliminated because of its significant global impact. Diagnostics are essential in aiding the elimination of STH infection. Therefore, it is imperative to evaluate the progress of diagnostic techniques, especially molecular-based techniques. Nevertheless, there is currently no comprehensive systematic review and meta-analysis (SRMA) evaluating the performance of molecular-based techniques for diagnosing STH. This protocol, registered on PROSPERO (CRD42024563432), intends to assess the diagnostic accuracy of molecular-based techniques compared to microscopic-based techniques for four common STH infections (A. lumbricoides, T. trichiura, hookworms, and S. stercoralis) across global studies. Two authors will independently conduct the screening, study selection, data extraction, and quality assessment of this SRMA. The obtained studies will be assessed using QUADAS-2 and will be analyzed for publication bias and diagnostic accuracy, including sensitivity, specificity, NPV, PPV, and DOR, and SROC. This developed protocol is expected to be versatile enough to be implemented in the range of studies assessing the diagnostic accuracy of the advance compared to the convention methods with adjustments. The findings of this SRMA are expected to provide a valuable contribution towards achieving the goal of eradicating STH infection by 2030.

Graphical abstract

Specifications table

Subject area:	Medicine and Dentistry
More specific subject area:	Evidence Based Medicine (EBM)
Name of your protocol:	Protocol for conducting systematic review and meta-analysis of diagnostic accuracy for soil-transmitted helminth diagnosis
Reagents/tools:	The analysis in this systematic review and meta-analysis (SRMA) will be performed using R version 4.3.2
Experimental design:	Not applicable
Trial registration:	This systematic review protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the registration number CRD42024563432.
Ethics:	Not applicable
Value of the Protocol:	•Comprehensively evaluate the comparability of molecular-based techniques in comparison to the widely used microscopic-based techniques based on studies published worldwide. •Highlight the importance of choosing a proper diagnostic technique for the diagnosis of different STH species. •Reveal the drawbacks of the current “gold standard” for STH infection by implementing microscopic-based techniques as the reference standard to calculate the diagnostic accuracy of PCR.

Background

Intestinal nematodes, including Ascaris lumbricoides, Trichuris trichiura, hookworms (Ancylostoma duodenale and Necator americanus), and Strongyloides stercoralis, are classified as soil-transmitted helminth (STH) due to the nature of their transmission [1]. STH is classified as one of the neglected tropical diseases (NTDs) and effects approximately 1.5 billion people, accounting for 24 % of the global population [2]. STH diagnosis heavily relies on microscopic-based techniques with Kato-Katz as the reference standard recommended by the World Health Organization (WHO) [3]. Other microscopic-based techniques used for the diagnosis of STH include direct smear, faecal concentration techniques (Baermann concentration and formalin-ether concentration techniques), flotation techniques (McMaster, FLOTAC, and mini-FLOTAC), and culture-based techniques (Harada-Mori filter paper strip culture and agar plate culture) [[3], [4], [5]]. Generally, microscopic-based techniques are known for being simple and inexpensive, however it lacks sensitivity in low-intensity settings and might require a combination of various techniques to detect multiple infections [[5], [6], [7]]. Therefore, a more sensitive diagnostic technique is needed to accurately diagnose STH infection with the goal of elimination [8].

Considering the recent advancements in molecular-based techniques, polymerase chain reaction (PCR) provides an alternative for the diagnosis of STH infections [6,9]. PCR has superior sensitivity and specificity compared to micrsocopic-based techniques. Additionally, it enables the simultaneous detection of multiple infections through the utilizations of multiplex PCR [5]. The presence of STH DNA can be detected even in very low-intensity settings which makes the elimination of STH is feasible [10]. PCR is considered to be relatively more expensive than microscopic-based techniques. Additionally, PCR necessitates specialized equipment and skilled personnels, posing challenges for its implementation in resource-limited settings [4,5]. Despite the existence of several studies comparing the diagnostic performance of PCR with microscopic-based techniques for the detection of STH infection, there has not been a systematic review and meta-analysis (SRMA) that comprehensively examines and summarize all the results for all four STH infections. This protocol serves as a foundation for an upcoming SRMA that will assess the diagnostic accuracy of molecular-based techniques, specifically conventional PCR (cPCR) and quantitative PCR (qPCR), in comparison to the widely used microscopic-based techniques for diagnosing STH infection. Furthermore, the consensus of this designed protocol will be readily adapted for the purpose of diagnostic accuracy assessment. Thus, ensure the adaptability and flexibility of this developed protocol.

Description of protocol

Protocol registration

For the purpose of transparency in the conduct of this SRMA, a protocol summary was registered on the International Prospective Register of Systematic Reviews (PROSPERO) under the registry number CRD42024563432. The SRMA will be performed in accordance with the Cochrane Handbook for Systematic Reviews of Diagnostic Accuracy (version 2) [11] and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [12]. Fig. 1 provides a concise summary of the methodological framework.

Fig. 1.

Methodological framework of the SRMA.

Research question

The research question for this SRMA will be formulated based on PICO (patient, intervention, comparison, and outcome). The PICO framework utilized in this review will consist of the following components:

• •Patient: Patients with or without symptoms of STH infection
• •Intervention: Molecular-based techniques
• •Comparison: Microscopic-based techniques
• •Outcome: Sensitivity, specificity, positive and negative predictive values, and diagnostic odds ratio (DOR)

Search strategy

A systematic literature search for peer-reviewed studies will be conducted using various databases, including Pubmed, ProQuest, Ovid, and EBSCO. The search strategy will be limited to studies published between January 2003 and July 2023. In addition, preprint databases, namely medRxiv and bioRxiv, will also be used to obtain grey literature to minimize publication bias. The search strategy will use keywords and Boolean operators, notably “AND” and “OR”. The keywords used in the search strategy are provided in Supplementary Material 1.

Eligibility criteria

Diagnostic accuracy studies written in English evaluating the performance of molecular-based techniques (cPCR and qPCR) in comparison to microscopic-based techniques (direct smear, Kato-Katz, formalin-ether concentration technique, modified Ritchie concentration technique, Baermann concentration, spontaneous tube sedimentation technique, sodium nitrate flotation, zinc sulphate flotation, Willis Flotation, Harada-Mori paper strip culture, agar plate culture, and coproculture) for the diagnosis of STH infection in human faecal samples were included in this study.

Study selection and data extraction

Studies obtained from the search strategy will be uploaded to Zotero reference manager for the removal of duplicates. Two authors, VW and SJK, will independently screen the obtained studies for potentially eligible studies based on the title, abstract, and lastly, based on the full text. Any discrepancies found during the screening process will be resolved by a third author, MMMK, who will make the final decision. The result of the screening process will be displayed in a PRISMA diagram (Fig. 2).

Fig. 2.

PRISMA diagram summarizing the screening process.

Two authors, VW and SJK, will independently extract selected studies to collect data regarding the first author, year published, study design, study population characteristics, source and sample size of faecal specimens, species of helminths, and the diagnostic techniques used in each study. The prevalence of each STH infection from each study will be determined by calculating the ratio of the number of infections identified by microscopic or molecular-based techniques to the total number of samples. Meanwhile, the numbers of true positive (TP), false positive (FP), false negative (FN), and true negative (TN) will be either obtained or calculated based on the data provided from each selected study. Contact will be made with the corresponding authors of studies without sufficient data to seek the provision of complete data. Given that the diagnostic accuracy of molecular-based techniques can be influenced by various factors, including primer designs, if the forthcoming analyses display yield unsatisfactory results, the primer sequences employed in those studies will be assessed using the basic local alignment search tool (BLAST) provided by national center for biotechnology information (NCBI) to determine the specificity of the primers used in the respective studies. All gathered information will be exported to Microsoft Excel version 2007.

Quality assessment/assessment of risk of bias

The risk of bias and applicability of each study will be assessed using Quality Assessment of Diagnostic Accuracy Studies Version 2 (QUADAS-2) as recommended by the Cochrane Handbook for Systematic Reviews of Diagnostic Accuracy [13]. Two authors, VW and SJK, will independently assess the four key domains of QUADAS 2: patient selection, index test, reference standard, as well as flow and timing. MMMK, a third author, will resolve any disputes that arise during the assessment process. Studies that compare PCR to two or more techniques based on microscopic will be assessed separately according to the specific microscopy technique used.

Data analysis

The data analysis in this review is modified from our previously published study on the diagnostic accuracy of rapid antigen tests for the diagnosis of Coronavirus disease 2019 (COVID-19) [14]. Data analyses consisting of publication bias, diagnostic accuracy, heterogeneity, and sub-group analyses will be conducted using R version 4.3.2. These analyses will be performed using the ‘meta’ and ‘mada’ packages [15,16].

Publication bias

The presence of potential publication bias in the selected studies will be assessed using funnel plots, as well as Egger's regression and Begg's correlation test to evaluate asymmetry. A p-value of <0.05 signifying publication bias. The assessment of publication bias will be evaluated based on the value of the DOR of selected studies.

Diagnostic accuracy

Assessing the diagnostic accuracy is essential for evaluating the performance of molecular-based techniques in differentiating between infected and non-infected patients. Molecular-based techniques as the index test will be evaluated in comparison to microscopic-based techniques as the reference standard. The evaluation will focus on the following parameters: sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and DOR. In the context of STH infection, sensitivity refers to the probability of the index test to correctly identify infected patients (true positive) and not classifying infected patients as healthy (false negative). Specificity refers to the probability of the index test to correctly identify healthy patients (true negative) and not classifying healthy patients as infected (false positive). Meanwhile, PPV and NPV represent the probability of having or not having the infection, respectively, based on the result of index test. PPV refers to the probability of the index test to correctly determine infected patients (true positive) and not incorrectly classifying healthy patients as infected (false positive) based on a positive result from the index test. On the other hand, NPV refers to the probability of the index test to correctly determine healthy patients (true negative) and not classifying infected patients as healthy (false negative) based on a negative result from the index test.

In order to prevent confusion between sensitivity and specificity with PPV and NPV, sensitivity and specificity indicate the agreement in correctly identifying patients with or without an infection between the results of the index test and the results of the reference standard. On the other hand, PPV and NPV represent the probability of correctly identifying patients with or without an infection based solely on the results of the index test. Finally, DOR represents the odds of the index test of having positive results in infected patients relative to the odds of having positive results in healthy patients. This SRMA will examine the DOR because it is not dependent on the prevalence of infection, unlike PPV and NPV which are affected by prevalence. DOR can serve as a single indicator of a diagnostic technique's performance, facilitating the comparison between molecular-based techniques and microscopic-based techniques. In short, sensitivity and specificity indicate the probability of the index test to accurately identify patients who are truly infected (sensitivity) or patients who are truly healthy (specificity). On the other hand, PPV and NPV indicate the probability of actually having the infection (PPV) or not having the infection (NPV) based on the index test results. Lastly, DOR represents the odds of positive results in infected patients relative to the odds of positive results in healthy patients. The diagnostic accuracy parameters can be obtained from a 2 × 2 contingency table depicted in Fig. 3, together with the corresponding mathematical formulas for each parameter [17,18].

Fig. 3.

Contingency table for deriving the parameters of diagnostic accuracy.

The Sensitivity, specificity, PPV, NPV, and DOR of cPCR/qPCR in comparison to microscopic-based techniques will be calculated and compared using a random effects model with an inverse variance method and a 95 % confidence interval. Sensitivity, PPV, and DOR will not be analysed for studies with zero true positives. In contrast, for studies with either zero false positives or zero false negatives, a modified Haldane-Anscombe zero-cell correction method will be applied to analyse the value of the DOR [19]. The significant of effect size differences will be assessed using Cochran's Q test and p-value. Bivariate models of summary receiver operating characteristics (SROC) will also be made to summarize the overall results from selected studies. The performance of PCR for the diagnosis of STH infection will be measured according to the value of the area under the curve (AUC) and will be interpreted as poor, fair, good, and excellent for AUC values between 0.6 and 0.7, 0.7 and 0.8, 0.8 and 0.9, and above 0.9, respectively [20].

Heterogeneity

Heterogeneity between selected studies will be evaluated using I² statistics and interpreted according to the Cochrane Handbook for Systematic Reviews of Interventions. Heterogeneity between studies will be classified as insignificant, moderate, substantial, and considerable when I² values are below 40 %, between 30 %−60 %, between 50 %−90 %, and above 75 %, respectively [21].

Subgroup analysis

A sub-group analysis will be conducted based on the microscopic-based techniques used in each study to evaluate the overall diagnostic accuracy of cPCR/qPCR compared to each respective microscopic-based technique and to minimize heterogeneity across studies. Additionally, a sub-group analysis based on the endemicity of the specimen's source will also be made to compare the performance of molecular-based techniques in different infection intensities.

Protocol validation

Not applicable.

Limitations

The methods detailed in this protocol have certain limitations. The results from the methods described above will not be able to directly conclude if molecular-based techniques have a higher diagnostic accuracy compared to microscopic-based techniques, since microscopic-based techniques itself will be applied as the reference standard. Hence, the findings of this review may demostrate the comparability of molecular-based techniques to microscopic-based techniques.

CRediT authorship contribution statement

Vincent Wijaya: Writing – original draft, Methodology. Samuel Johnson Kurniawan: Writing – original draft, Methodology. Maria Mardalena Martini Kaisar: Conceptualization, Visualization, Writing – review & editing, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.