Making progress with the automation of systematic reviews: principles of the International Collaboration for the Automation of Systematic Reviews (ICASR)

An exploration of available methods and tools to improve the efficiency of systematic review production: a scoping review

BACKGROUND

Systematic reviews (SRs) are time-consuming and labor-intensive to perform. With the growing number of scientific publications, the SR development process becomes even more laborious. This is problematic because timely SR evidence is essential for decision-making in evidence-based healthcare and policymaking. Numerous methods and tools that accelerate SR development have recently emerged. To date, no scoping review has been conducted to provide a comprehensive summary of methods and ready-to-use tools to improve efficiency in SR production.

OBJECTIVE

To present an overview of primary studies that evaluated the use of ready-to-use applications of tools or review methods to improve efficiency in the review process.

METHODS

We conducted a scoping review. An information specialist performed a systematic literature search in four databases, supplemented with citation-based and grey literature searching. We included studies reporting the performance of methods and ready-to-use tools for improving efficiency when producing or updating a SR in the health field. We performed dual, independent title and abstract screening, full-text selection, and data extraction. The results were analyzed descriptively and presented narratively.

RESULTS

We included 103 studies: 51 studies reported on methods, 54 studies on tools, and 2 studies reported on both methods and tools to make SR production more efficient. A total of 72 studies evaluated the validity (n = 69) or usability (n = 3) of one method (n = 33) or tool (n = 39), and 31 studies performed comparative analyses of different methods (n = 15) or tools (n = 16). 20 studies conducted prospective evaluations in real-time workflows. Most studies evaluated methods or tools that aimed at screening titles and abstracts (n = 42) and literature searching (n = 24), while for other steps of the SR process, only a few studies were found. Regarding the outcomes included, most studies reported on validity outcomes (n = 84), while outcomes such as impact on results (n = 23), time-saving (n = 24), usability (n = 13), and cost-saving (n = 3) were less often evaluated.

CONCLUSION

For title and abstract screening and literature searching, various evaluated methods and tools are available that aim at improving the efficiency of SR production. However, only few studies have addressed the influence of these methods and tools in real-world workflows. Few studies exist that evaluate methods or tools supporting the remaining tasks. Additionally, while validity outcomes are frequently reported, there is a lack of evaluation regarding other outcomes.

Automation of duplicate record detection for systematic reviews: Deduplicator

BACKGROUND

To describe the algorithm and investigate the efficacy of a novel systematic review automation tool "the Deduplicator" to remove duplicate records from a multi-database systematic review search.

METHODS

We constructed and tested the efficacy of the Deduplicator tool by using 10 previous Cochrane systematic review search results to compare the Deduplicator's 'balanced' algorithm to a semi-manual EndNote method. Two researchers each performed deduplication on the 10 libraries of search results. For five of those libraries, one researcher used the Deduplicator, while the other performed semi-manual deduplication with EndNote. They then switched methods for the remaining five libraries. In addition to this analysis, comparison between the three different Deduplicator algorithms ('balanced', 'focused' and 'relaxed') was performed on two datasets of previously deduplicated search results.

RESULTS

Before deduplication, the mean library size for the 10 systematic reviews was 1962 records. When using the Deduplicator, the mean time to deduplicate was 5 min per 1000 records compared to 15 min with EndNote. The mean error rate with Deduplicator was 1.8 errors per 1000 records in comparison to 3.1 with EndNote. Evaluation of the different Deduplicator algorithms found that the 'balanced' algorithm had the highest mean F1 score of 0.9647. The 'focused' algorithm had the highest mean accuracy of 0.9798 and the highest recall of 0.9757. The 'relaxed' algorithm had the highest mean precision of 0.9896.

CONCLUSIONS

This demonstrates that using the Deduplicator for duplicate record detection reduces the time taken to deduplicate, while maintaining or improving accuracy compared to using a semi-manual EndNote method. However, further research should be performed comparing more deduplication methods to establish relative performance of the Deduplicator against other deduplication methods.

Accelerating evidence synthesis for safety assessment through ClinicalTrials.gov platform: a feasibility study

BACKGROUND

Standard systematic review can be labor-intensive and time-consuming meaning that it can be difficult to provide timely evidence when there is an urgent public health emergency such as a pandemic. The ClinicalTrials.gov provides a promising way to accelerate evidence production.

METHODS

We conducted a search on PubMed to gather systematic reviews containing a minimum of 5 studies focused on safety aspects derived from randomized controlled trials (RCTs) of pharmacological interventions, aiming to establish a real-world dataset. The registration information of each trial from eligible reviews was further collected and verified. The meta-analytic data were then re-analyzed by using 1) the full meta-analytic data with all trials and 2) emulated rapid data with trials that had been registered and posted results on ClinicalTrials.gov, under the same synthesis methods. The effect estimates of the full meta-analysis and rapid meta-analysis were then compared.

RESULTS

The real-world dataset comprises 558 meta-analyses. Among them, 56 (10.0%) meta-analyses included RCTs that were not registered in ClinicalTrials.gov. For the remaining 502 meta-analyses, the median percentage of RCTs registered within each meta-analysis is 70.1% (interquartile range: 33.3% to 88.9%). Under a 20% bias threshold, rapid meta-analyses conducted through ClinicalTrials.gov achieved accurate point estimates ranging from 77.4% (using the MH model) to 83.1% (using the GLMM model); 91.0% to 95.3% of these analyses accurately predicted the direction of effects.

CONCLUSIONS

Utilizing the ClinicalTrials.gov platform for safety assessment with a minimum of 5 RCTs holds significant potential for accelerating evidence synthesis to support urgent decision-making.

Rapid review methods series: Guidance on the use of supportive software

This paper is part of a series of methodological guidance from the Cochrane Rapid Reviews Methods Group. Rapid reviews (RRs) use modified systematic review methods to accelerate the review process while maintaining systematic, transparent and reproducible methods. This paper guides how to use supportive software for RRs.We strongly encourage the use of supportive software throughout RR production. Specifically, we recommend (1) using collaborative online platforms that enable working in parallel, allow for real-time project management and centralise review details; (2) using automation software to support, but not entirely replace a human reviewer and human judgement and (3) being transparent in reporting the methodology and potential risk for bias due to the use of supportive software.

The effect of pressure injury prevention care bundles on pressure injuries in hospital patients: A complex intervention systematic review and meta-analysis

BACKGROUND

Numerous interventions for pressure injury prevention have been developed, including care bundles.

OBJECTIVE

To systematically review the effectiveness of pressure injury prevention care bundles on pressure injury prevalence, incidence, and hospital-acquired pressure injury rate in hospitalised patients.

DATA SOURCES

The Medical Literature Analysis and Retrieval System Online (via PubMed), the Cumulative Index to Nursing and Allied Health Literature, EMBASE, Scopus, the Cochrane Library and two registries were searched (from 2009 to September 2023).

STUDY ELIGIBILITY CRITERIA

Randomised controlled trials and non-randomised studies with a comparison group published in English after 2008 were included. Studies reporting on the frequency of pressure injuries where the number of patients was not the numerator or denominator, or where the denominator was not reported, and single subgroups of hospitalised patients were excluded. Educational programmes targeting healthcare professionals and bundles targeting specific types of pressure injuries were excluded.

PARTICIPANTS AND INTERVENTIONS

Bundles with ≥3 components directed towards patients and implemented in ≥2 hospital services were included.

STUDY APPRAISAL AND SYNTHESIS METHODS

Screening, data extraction and risk of bias assessments were undertaken independently by two researchers. Random effects meta-analyses were conducted. The certainty of the body of evidence was assessed using Grading of Recommendations, Assessment, Development and Evaluation.

RESULTS

Nine studies (seven non-randomised with historical controls; two randomised) conducted in eight countries were included. There were four to eight bundle components; most were core, and only a few were discretionary. Various strategies were used prior to (six studies), during (five studies) and after (two studies) implementation to embed the bundles. The pooled risk ratio for pressure injury prevalence (five non-randomised studies) was 0.55 (95 % confidence intervals 0.29-1.03), and for hospital-acquired pressure injury rate (five non-randomised studies) it was 0.31 (95 % confidence intervals 0.12-0.83). All non-randomised studies were at high risk of bias, with very low certainty of evidence. In the two randomised studies, the care bundles had non-significant effects on hospital-acquired pressure injury incidence density, but data could not be pooled.

CONCLUSIONS AND IMPLICATIONS OF KEY FINDINGS

Whilst some studies showed decreases in pressure injuries, this evidence was very low certainty. The potential benefits of adding emerging evidence-based components to bundles should be considered. Future effectiveness studies should include contemporaneous controls and the development of a comprehensive, theory and evidence-informed implementation plan.

SYSTEMATIC REVIEW REGISTRATION NUMBER

PROSPERO CRD42023423058.

TWEETABLE ABSTRACT

Pressure injury prevention care bundles decrease hospital-acquired pressure injuries, but the certainty of this evidence is very low.

Automation tools to support undertaking scoping reviews

OBJECTIVE

This paper describes several automation tools and software that can be considered during evidence synthesis projects and provides guidance for their integration in the conduct of scoping reviews.

STUDY DESIGN AND SETTING

The guidance presented in this work is adapted from the results of a scoping review and consultations with the JBI Scoping Review Methodology group.

RESULTS

This paper describes several reliable, validated automation tools and software that can be used to enhance the conduct of scoping reviews. Developments in the automation of systematic reviews, and more recently scoping reviews, are continuously evolving. We detail several helpful tools in order of the key steps recommended by the JBI's methodological guidance for undertaking scoping reviews including team establishment, protocol development, searching, de-duplication, screening titles and abstracts, data extraction, data charting, and report writing. While we include several reliable tools and software that can be used for the automation of scoping reviews, there are some limitations to the tools mentioned. For example, some are available in English only and their lack of integration with other tools results in limited interoperability.

CONCLUSION

This paper highlighted several useful automation tools and software programs to use in undertaking each step of a scoping review. This guidance has the potential to inform collaborative efforts aiming at the development of evidence informed, integrated automation tools and software packages for enhancing the conduct of high-quality scoping reviews.

Comparing generative and extractive approaches to information extraction from abstracts describing randomized clinical trials

BACKGROUND

Systematic reviews of Randomized Controlled Trials (RCTs) are an important part of the evidence-based medicine paradigm. However, the creation of such systematic reviews by clinical experts is costly as well as time-consuming, and results can get quickly outdated after publication. Most RCTs are structured based on the Patient, Intervention, Comparison, Outcomes (PICO) framework and there exist many approaches which aim to extract PICO elements automatically. The automatic extraction of PICO information from RCTs has the potential to significantly speed up the creation process of systematic reviews and this way also benefit the field of evidence-based medicine.

RESULTS

Previous work has addressed the extraction of PICO elements as the task of identifying relevant text spans or sentences, but without populating a structured representation of a trial. In contrast, in this work, we treat PICO elements as structured templates with slots to do justice to the complex nature of the information they represent. We present two different approaches to extract this structured information from the abstracts of RCTs. The first approach is an extractive approach based on our previous work that is extended to capture full document representations as well as by a clustering step to infer the number of instances of each template type. The second approach is a generative approach based on a seq2seq model that encodes the abstract describing the RCT and uses a decoder to infer a structured representation of a trial including its arms, treatments, endpoints and outcomes. Both approaches are evaluated with different base models on a manually annotated dataset consisting of RCT abstracts on an existing dataset comprising 211 annotated clinical trial abstracts for Type 2 Diabetes and Glaucoma. For both diseases, the extractive approach (with flan-t5-base) reached the best F 1 score, i.e. 0.547 ( ± 0.006 ) for type 2 diabetes and 0.636 ( ± 0.006 ) for glaucoma. Generally, the F 1 scores were higher for glaucoma than for type 2 diabetes and the standard deviation was higher for the generative approach.

CONCLUSION

In our experiments, both approaches show promising performance extracting structured PICO information from RCTs, especially considering that most related work focuses on the far easier task of predicting less structured objects. In our experimental results, the extractive approach performs best in both cases, although the lead is greater for glaucoma than for type 2 diabetes. For future work, it remains to be investigated how the base model size affects the performance of both approaches in comparison. Although the extractive approach currently leaves more room for direct improvements, the generative approach might benefit from larger models.

Production and use of rapid responses during the COVID-19 pandemic in Quebec (Canada): perspectives from evidence synthesis producers and decision makers

BACKGROUND

The COVID-19 pandemic has required evidence to be made available more rapidly than usual, in order to meet the needs of decision makers in a timely manner. These exceptional circumstances have caused significant challenges for organizations and teams responsible for evidence synthesis. They had to adapt to provide rapid responses to support decision-making. This study aimed to document (1) the challenges and adaptations made to produce rapid responses during the pandemic, (2) their perceived usefulness, reported use and factors influencing their use and (3) the methodological adaptations made to produce rapid responses.

METHODS

A qualitative study was conducted in 2021 with eight organizations in the health and social services system in Quebec (Canada), including three institutes with a provincial mandate. Data collection included focus groups (n = 9 groups in 8 organizations with 64 participants), interviews with decision makers (n = 12), and a document analysis of COVID-19 rapid responses (n = 128). A thematic analysis of qualitative data (objectives 1 and 2) and a descriptive analysis of documents (objective 3) were conducted.

RESULTS

The results highlight the teams and organizations' agility to deal with the many challenges encountered during the pandemic (e.g., increased their workloads, adoption of new technological tools or work processes, improved collaboration, development of scientific monitoring, adaptation of evidence synthesis methodologies and products). The challenge of balancing rigor and speed was reported by teams and organizations. When available at the right time, rapid responses have been reported as a useful tool for informing or justifying decisions in a context of uncertainty. Several factors that may influence their use were identified (e.g., clearly identify needs, interactions with producers, perceived rigor and credibility, precise and feasible recommendations). Certain trends in the methodological approaches used to speed up the evidence synthesis process were identified.

CONCLUSIONS

This study documented rapid responses producers' experiences during the COVID-19 pandemic in Quebec, and decision makers who requested, consulted, or used these products. Potential areas of improvements are identified such as reinforce coordination, improve communication loops, clarify guidelines or methodological benchmarks, and enhance utility of rapid response products for decision makers.

A comparison of machine learning methods to find clinical trials for inclusion in new systematic reviews from their PROSPERO registrations prior to searching and screening

Searching for trials is a key task in systematic reviews and a focus of automation. Previous approaches required knowing examples of relevant trials in advance, and most methods are focused on published trial articles. To complement existing tools, we compared methods for finding relevant trial registrations given a International Prospective Register of Systematic Reviews (PROSPERO) entry and where no relevant trials have been screened for inclusion in advance. We compared SciBERT-based (extension of Bidirectional Encoder Representations from Transformers) PICO extraction, MetaMap, and term-based representations using an imperfect dataset mined from 3632 PROSPERO entries connected to a subset of 65,662 trial registrations and 65,834 trial articles known to be included in systematic reviews. Performance was measured by the median rank and recall by rank of trials that were eventually included in the published systematic reviews. When ranking trial registrations relative to PROSPERO entries, 296 trial registrations needed to be screened to identify half of the relevant trials, and the best performing approach used a basic term-based representation. When ranking trial articles relative to PROSPERO entries, 162 trial articles needed to be screened to identify half of the relevant trials, and the best-performing approach used a term-based representation. The results show that MetaMap and term-based representations outperformed approaches that included PICO extraction for this use case. The results suggest that when starting with a PROSPERO entry and where no trials have been screened for inclusion, automated methods can reduce workload, but additional processes are still needed to efficiently identify trial registrations or trial articles that meet the inclusion criteria of a systematic review.

Overcoming the challenges of using automated technologies for public health evidence synthesis

Many organisations struggle to keep pace with public health evidence due to the volume of published literature and length of time it takes to conduct literature reviews. New technologies that help automate parts of the evidence synthesis process can help conduct reviews more quickly and efficiently to better provide up-to-date evidence for public health decision making. To date, automated approaches have seldom been used in public health due to significant barriers to their adoption. In this Perspective, we reflect on the findings of a study exploring experiences of adopting automated technologies to conduct evidence reviews within the public health sector. The study, funded by the European Centre for Disease Prevention and Control, consisted of a literature review and qualitative data collection from public health organisations and researchers in the field. We specifically focus on outlining the challenges associated with the adoption of automated approaches and potential solutions and actions that can be taken to mitigate these. We explore these in relation to actions that can be taken by tool developers (e.g. improving tool performance and transparency), public health organisations (e.g. developing staff skills, encouraging collaboration) and funding bodies/the wider research system (e.g. researchers, funding bodies, academic publishers and scholarly journals).

A mapping exercise using automated techniques to develop a search strategy to identify systematic review tools

The Systematic Review Toolbox aims provide a web-based catalogue of tools that support various tasks within the systematic review and wider evidence synthesis process. Identifying publications surrounding specific systematic review tools is currently challenging, leading to a high screening burden for few eligible records. We aimed to develop a search strategy that could be regularly and automatically run to identify eligible records for the SR Toolbox, thus reducing time on task and burden for those involved. We undertook a mapping exercise to identify the PubMed IDs of papers indexed within the SR Toolbox. We then used the Yale MeSH Analyser and Visualisation of Similarities (VOS) Viewer text-mining software to identify the most commonly used MeSH terms and text words within the eligible records. These MeSH terms and text words were combined using Boolean Operators into a search strategy for Ovid MEDLINE. Prior to the mapping exercise and search strategy development, 81 software tools and 55 'Other' tools were included within the SR Toolbox. Since implementation of the search strategy, 146 tools have been added. There has been an increase in tools added to the toolbox since the search was developed and its corresponding auto-alert in MEDLINE was originally set up. Developing a search strategy based on a mapping exercise is an effective way of identifying new tools to support the systematic review process. Further research could be conducted to help prioritise records for screening to reduce reviewer burden further and to adapt the strategy for disciplines beyond healthcare.

An Automated Literature Review Tool (LiteRev) for Streamlining and Accelerating Research Using Natural Language Processing and Machine Learning: Descriptive Performance Evaluation Study

BACKGROUND

Literature reviews (LRs) identify, evaluate, and synthesize relevant papers to a particular research question to advance understanding and support decision-making. However, LRs, especially traditional systematic reviews, are slow, resource-intensive, and become outdated quickly.

OBJECTIVE

LiteRev is an advanced and enhanced version of an existing automation tool designed to assist researchers in conducting LRs through the implementation of cutting-edge technologies such as natural language processing and machine learning techniques. In this paper, we present a comprehensive explanation of LiteRev's capabilities, its methodology, and an evaluation of its accuracy and efficiency to a manual LR, highlighting the benefits of using LiteRev.

METHODS

Based on the user's query, LiteRev performs an automated search on a wide range of open-access databases and retrieves relevant metadata on the resulting papers, including abstracts or full texts when available. These abstracts (or full texts) are text processed and represented as a term frequency-inverse document frequency matrix. Using dimensionality reduction (pairwise controlled manifold approximation) and clustering (hierarchical density-based spatial clustering of applications with noise) techniques, the corpus is divided into different topics described by a list of the most important keywords. The user can then select one or several topics of interest, enter additional keywords to refine its search, or provide key papers to the research question. Based on these inputs, LiteRev performs a k-nearest neighbor (k-NN) search and suggests a list of potentially interesting papers. By tagging the relevant ones, the user triggers new k-NN searches until no additional paper is suggested for screening. To assess the performance of LiteRev, we ran it in parallel to a manual LR on the burden and care for acute and early HIV infection in sub-Saharan Africa. We assessed the performance of LiteRev using true and false predictive values, recall, and work saved over sampling.

RESULTS

LiteRev extracted, processed, and transformed text into a term frequency-inverse document frequency matrix of 631 unique papers from PubMed. The topic modeling module identified 16 topics and highlighted 2 topics of interest to the research question. Based on 18 key papers, the k-NNs module suggested 193 papers for screening out of 613 papers in total (31.5% of the whole corpus) and correctly identified 64 relevant papers out of the 87 papers found by the manual abstract screening (recall rate of 73.6%). Compared to the manual full text screening, LiteRev identified 42 relevant papers out of the 48 papers found manually (recall rate of 87.5%). This represents a total work saved over sampling of 56%.

CONCLUSIONS

We presented the features and functionalities of LiteRev, an automation tool that uses natural language processing and machine learning methods to streamline and accelerate LRs and support researchers in getting quick and in-depth overviews on any topic of interest.

The use of artificial intelligence for automating or semi-automating biomedical literature analyses: a scoping review

OBJECTIVE

Evidence-based medicine (EBM) is a decision-making process based on the conscious and judicious use of the best available scientific evidence. However, the exponential increase in the amount of information currently available likely exceeds the capacity of human-only analysis. In this context, artificial intelligence (AI) and its branches such as machine learning (ML) can be used to facilitate human efforts in analyzing the literature to foster EBM. The present scoping review aimed to examine the use of AI in the automation of biomedical literature survey and analysis with a view to establishing the state-of-the-art and identifying knowledge gaps.

MATERIALS AND METHODS

Comprehensive searches of the main databases were performed for articles published up to June 2022 and studies were selected according to inclusion and exclusion criteria. Data were extracted from the included articles and the findings categorized.

RESULTS

The total number of records retrieved from the databases was 12,145, of which 273 were included in the review. Classification of the studies according to the use of AI in evaluating the biomedical literature revealed three main application groups, namely assembly of scientific evidence (n=127; 47%), mining the biomedical literature (n=112; 41%) and quality analysis (n=34; 12%). Most studies addressed the preparation of systematic reviews, while articles focusing on the development of guidelines and evidence synthesis were the least frequent. The biggest knowledge gap was identified within the quality analysis group, particularly regarding methods and tools that assess the strength of recommendation and consistency of evidence.

CONCLUSION

Our review shows that, despite significant progress in the automation of biomedical literature surveys and analyses in recent years, intense research is needed to fill knowledge gaps on more difficult aspects of ML, deep learning and natural language processing, and to consolidate the use of automation by end-users (biomedical researchers and healthcare professionals).

Limits in the search date for rapid reviews of diagnostic test accuracy studies

Limiting the search date is a common approach utilised in therapeutic/interventional rapid reviews. Yet the accuracy of pooled estimates is unknown when applied to rapid reviews of diagnostic test accuracy studies. Data from all systematic reviews of diagnostic test accuracy studies published in the Cochrane Database of Systematic Reviews, until February 2022 were collected. Meta-analyses with at least five studies were included to emulate rapid reviews by limiting the search to the recent 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35 and 40 years. The magnitude of the pooled area under the curve (AUC), sensitivity and specificity for the full meta-analysis and the rapid reviews were compared. A total of 846 diagnostic meta-analyses were included. When the search date was limited to the recent 10 and 15 years, more than 75% and 80% of meta-analyses presented less than 5% difference between the pooled AUC, sensitivity and specificity of the full meta-analysis and the rapid review. There was little gain in the precision of the pooled estimates when the emulated rapid reviews included more than 15 years in the search. Rapid reviews restricted by search date are a valid and reliable approach for diagnostic test accuracy studies. Robust evidence can be achieved by restricting the search date to the recent 10-15 years. Future studies need to examine the reduction in workload and time to finish the rapid reviews under different search date limits.

Reducing Literature Screening Workload With Machine Learning

Abstract. In our era of accelerated accumulation of knowledge, the manual screening of literature for eligibility is increasingly becoming too labor-intensive for summarizing the current state of knowledge in a timely manner. Recent advances in machine learning and natural language processing promise to reduce the screening workload by automatically detecting unseen references with a high probability of inclusion. As a variety of tools have been developed, the current review provides an overview of their characteristics and performance. A systematic search in various databases yielded 488 eligible reports, revealing 15 tools for screening automation that differed in methodology, features, and accessibility. For the review on the performance of screening tools, 21 studies could be included. In comparison to sampling records randomly, active screening with prioritization approximately halves the screening workload. However, a comparison of tools under equal or at least similar conditions is needed to derive clear recommendations.

Unsupervised title and abstract screening for systematic review: a retrospective case-study using topic modelling methodology

BACKGROUND

The importance of systematic reviews in collating and summarising available research output on a particular topic cannot be over-emphasized. However, initial screening of retrieved literature is significantly time and labour intensive. Attempts at automating parts of the systematic review process have been made with varying degree of success partly due to being domain-specific, requiring vendor-specific software or manually labelled training data. Our primary objective was to develop statistical methodology for performing automated title and abstract screening for systematic reviews. Secondary objectives included (1) to retrospectively apply the automated screening methodology to previously manually screened systematic reviews and (2) to characterize the performance of the automated screening methodology scoring algorithm in a simulation study.

METHODS

We implemented a Latent Dirichlet Allocation-based topic model to derive representative topics from the retrieved documents' title and abstract. The second step involves defining a score threshold for classifying the documents as relevant for full-text review or not. The score is derived based on a set of search keywords (often the database retrieval search terms). Two systematic review studies were retrospectively used to illustrate the methodology.

RESULTS

In one case study (helminth dataset), [Formula: see text] sensitivity compared to manual title and abstract screening was achieved. This is against a false positive rate of [Formula: see text]. For the second case study (Wilson disease dataset), a sensitivity of [Formula: see text] and specificity of [Formula: see text] were achieved.

CONCLUSIONS

Unsupervised title and abstract screening has the potential to reduce the workload involved in conducting systematic review. While sensitivity of the methodology on the tested data is low, approximately [Formula: see text] specificity was achieved. Users ought to keep in mind that potentially low sensitivity might occur. One approach to mitigate this might be to incorporate additional targeted search keywords such as the indexing databases terms into the search term copora. Moreover, automated screening can be used as an additional screener to the manual screeners.

Combining globally search for a regular expression and print matching lines with bibliographic monitoring of genomic database improves diagnosis

Introduction: Exome sequencing has a diagnostic yield ranging from 25% to 70% in rare diseases and regularly implicates genes in novel disorders. Retrospective data reanalysis has demonstrated strong efficacy in improving diagnosis, but poses organizational difficulties for clinical laboratories. Patients and methods: We applied a reanalysis strategy based on intensive prospective bibliographic monitoring along with direct application of the GREP command-line tool (to "globally search for a regular expression and print matching lines") in a large ES database. For 18 months, we submitted the same five keywords of interest [(intellectual disability, (neuro)developmental delay, and (neuro)developmental disorder)] to PubMed on a daily basis to identify recently published novel disease-gene associations or new phenotypes in genes already implicated in human pathology. We used the Linux GREP tool and an in-house script to collect all variants of these genes from our 5,459 exome database. Results: After GREP queries and variant filtration, we identified 128 genes of interest and collected 56 candidate variants from 53 individuals. We confirmed causal diagnosis for 19/128 genes (15%) in 21 individuals and identified variants of unknown significance for 19/128 genes (15%) in 23 individuals. Altogether, GREP queries for only 128 genes over a period of 18 months permitted a causal diagnosis to be established in 21/2875 undiagnosed affected probands (0.7%). Conclusion: The GREP query strategy is efficient and less tedious than complete periodic reanalysis. It is an interesting reanalysis strategy to improve diagnosis.

Machine learning computational tools to assist the performance of systematic reviews: A mapping review

BACKGROUND

Within evidence-based practice (EBP), systematic reviews (SR) are considered the highest level of evidence in that they summarize the best available research and describe the progress in a determined field. Due its methodology, SR require significant time and resources to be performed; they also require repetitive steps that may introduce biases and human errors. Machine learning (ML) algorithms therefore present a promising alternative and a potential game changer to speed up and automate the SR process. This review aims to map the current availability of computational tools that use ML techniques to assist in the performance of SR, and to support authors in the selection of the right software for the performance of evidence synthesis.

METHODS

The mapping review was based on comprehensive searches in electronic databases and software repositories to obtain relevant literature and records, followed by screening for eligibility based on titles, abstracts, and full text by two reviewers. The data extraction consisted of listing and extracting the name and basic characteristics of the included tools, for example a tool's applicability to the various SR stages, pricing options, open-source availability, and type of software. These tools were classified and graphically represented to facilitate the description of our findings.

RESULTS

A total of 9653 studies and 585 records were obtained from the structured searches performed on selected bibliometric databases and software repositories respectively. After screening, a total of 119 descriptions from publications and records allowed us to identify 63 tools that assist the SR process using ML techniques.

CONCLUSIONS

This review provides a high-quality map of currently available ML software to assist the performance of SR. ML algorithms are arguably one of the best techniques at present for the automation of SR. The most promising tools were easily accessible and included a high number of user-friendly features permitting the automation of SR and other kinds of evidence synthesis reviews.

Automating document classification with distant supervision to increase the efficiency of systematic reviews: A case study on identifying studies with HIV impacts on female sex workers

There remains a limited understanding of the HIV prevention and treatment needs among female sex workers in many parts of the world. Systematic reviews of existing literature can help fill this gap; however, well-done systematic reviews are time-demanding and labor-intensive. Here, we propose an automatic document classification approach to a systematic review to significantly reduce the effort in reviewing documents and optimizing empiric decision making. We first describe a manual document classification procedure that is used to curate a pertinent training dataset and then propose three classifiers: a keyword-guided method, a cluster analysis-based method, and a random forest approach that utilizes a large set of feature tokens. This approach is used to identify documents studying female sex workers that contain content relevant to either HIV or experienced violence. We compare the performance of the three classifiers by cross-validation in terms of area under the curve of the receiver operating characteristic and precision and recall plot, and found random forest approach reduces the amount of manual reading for our example by 80%; in sensitivity analysis, we found that even trained with only 10% of data, the classifier can still avoid reading 75% of future documents (68% of total) while retaining 80% of relevant documents. In sum, the automated procedure of document classification presented here could improve both the precision and efficiency of systematic reviews and facilitate live reviews, where reviews are updated regularly. We expect to obtain a reasonable classifier by taking 20% of retrieved documents as training samples. The proposed classifier could also be used for more meaningfully assembling literature in other research areas and for rapid documents screening with a tight schedule, such as COVID-related work during the crisis.

Lessons learnt: Undertaking rapid reviews on public health and social measures during a global pandemic

Public health and social measures (PHSM) have been central to the COVID‐19 response. Consequently, there has been much pressure on decision‐makers to make evidence‐informed decisions and on researchers to synthesize the evidence regarding these measures. This article describes our experiences, responses and lessons learnt regarding key challenges when planning and conducting rapid reviews of PHSM during the COVID‐19 pandemic. Stakeholder consultations and scoping reviews to obtain an overview of the evidence inform the scope of reviews that are policy‐relevant and feasible. Multiple complementary reviews serve to examine the benefits and harms of PHSM across different populations and contexts. Conceiving reviews of effectiveness as adaptable living reviews helps to respond to evolving evidence needs and an expanding evidence base. An appropriately skilled review team and good planning, coordination and communication ensures smooth and rigorous processes and efficient use of resources. Scientific rigor, the practical implications of PHSM‐related complexity and likely time savings should be carefully weighed in deciding on methodological shortcuts. Making the best possible use of modeling studies represents a particular challenge, and methods should be carefully chosen, piloted and implemented. Our experience raises questions regarding the nature of rapid reviews and regarding how different types of evidence should be considered in making decisions about PHSM during a global pandemic. We highlight the need for readily available protocols for conducting studies on the effectiveness, unintended consequences and implementation of PHSM in a timely manner, as well as the need for rapid review standards tailored to “rapid” versus “emergency” mode reviewing.

Synthesizing evidence from clinical trials with dynamic interactive argument trees

Background

Evidence-based medicine propagates that medical/clinical decisions are made by taking into account high-quality evidence, most notably in the form of randomized clinical trials. Evidence-based decision-making requires aggregating the evidence available in multiple trials to reach –by means of systematic reviews– a conclusive recommendation on which treatment is best suited for a given patient population. However, it is challenging to produce systematic reviews to keep up with the ever-growing number of published clinical trials. Therefore, new computational approaches are necessary to support the creation of systematic reviews that include the most up-to-date evidence.We propose a method to synthesize the evidence available in clinical trials in an ad-hoc and on-demand manner by automatically arranging such evidence in the form of a hierarchical argument that recommends a therapy as being superior to some other therapy along a number of key dimensions corresponding to the clinical endpoints of interest. The method has also been implemented as a web tool that allows users to explore the effects of excluding different points of evidence, and indicating relative preferences on the endpoints.

Results

Through two use cases, our method was shown to be able to generate conclusions similar to the ones of published systematic reviews. To evaluate our method implemented as a web tool, we carried out a survey and usability analysis with medical professionals. The results show that the tool was perceived as being valuable, acknowledging its potential to inform clinical decision-making and to complement the information from existing medical guidelines.

Conclusions

The method presented is a simple but yet effective argumentation-based method that contributes to support the synthesis of clinical trial evidence. A current limitation of the method is that it relies on a manually populated knowledge base. This problem could be alleviated by deploying natural language processing methods to extract the relevant information from publications.

Oedema as a predictor of the incidence of new pressure injuries in adults in any care setting: A systematic review and meta-analysis

BACKGROUND

Oedema measurement, also termed sub-epidermal moisture measurement is recommended as an adjunct pressure injury prevention intervention in international guidelines because it indicates early tissue damage.

OBJECTIVE

To determine the prognostic value of oedema measurement in predicting future pressure injury in adults in any care setting.

DESIGN

Systematic review and meta-analysis.

SETTING

Participants were recruited from nursing homes or aged care facilities, hospitals, or post-acute facilities.

PARTICIPANTS

Adults.

METHODS

A modified 2-week systematic review was undertaken. Study designs included cohort (prospective and retrospective), case-control, case series if relevant comparisons were reported, randomised controlled trials if the association between oedema measurement and pressure injury was reported, and registry data. Databases searched included: Medical Literature Analysis and Retrieval System Online, The Cumulative Index to Nursing and Allied Health Literature, Excerpta Medica and the Cochrane Library from inception to 13 July 2021 with no language restrictions. Screening, data extraction using Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies - Prognostic Factors (CHARMS-PF) and quality assessment using Quality in Prognostic Factor Studies (QUIPS) were undertaken independently by ≥2 authors and adjudicated by another if required. Meta-analyses and meta-regression were undertaken. The certainty of the evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach.

RESULTS

Six studies (n = 483 total) were included. Two studies were set in nursing homes and four in either hospitals or post-acute facilities. Fives studies were prospective cohorts, and one was a randomised control trial. Two studies were assessed as low risk and four studies as moderate risk of bias. The pooled risk ratio in four studies (n = 388) for the relationship between oedema and pressure injury cumulative incidence was 18.87 (95% CI 2.13-38.29) and for time to pressure injury was 4.08 days (95% CI 1.64-6.52). Using GRADE, the certainty of the body of evidence was low for all outcomes. Meta-regression indicated that age, gender, and sample size were poor predictors for the association between oedema and pressure injury.

CONCLUSIONS

Measuring oedema as a predictor for pressure injury development is showing promise but a stronger body of evidence that takes into consideration other prognostic factors is needed to better understand its benefit.

REGISTRATION

PROSPERO CRD42021267834.

TWEETABLE ABSTRACT

Measuring oedema is a promising strategy to prevent pressure injuries but the certainty of evidence for this claim is low.

Conducting and critically appraising a high-quality systematic review and Meta-analysis pertaining to COVID-19

With constantly emerging new information regarding the epidemiology, pathogenesis, diagnosis and management of Coronavirus Disease 2019 (COVID-19), reviewing literature related to it has become increasingly complicated and resource-intensive. In the setting of this global pandemic, clinical decisions are being guided by the results of multiple pertinent studies; however, it has been observed that these studies are often heterogenous in design and population characteristics and results of initial trials may not be replicated in subsequent studies. The resulting clinical conundrum can be resolved by high-quality systematic review and meta-analysis with a robust and reliable methodology, encapsulating and critically appraising all the available literature relevant to the clinical scenario under scrutiny. It can condense the large volume of scientific information available and can also identify the cause of differences in the degree of effect under consideration across different studies. It can identify optimal diagnostic algorithms, assess efficacy of treatment strategies, and analyze inherent factors influencing the efficacy of treatment for COVID-19. The current review aims to provide a basic guide to plan and conduct a high-quality systematic review and meta-analysis pertaining to COVID-19, describing the main steps and addressing the pitfalls commonly encountered at each step. Knowledge of the basic steps would also allow the reader to critically appraise published systematic review and meta-analysis and the quality of evidence provided therein.

Toxic effects of nanomaterials for health applications: How automation can support a systematic review of the literature?

Systematic reviews of the scientific literature can be an important source of information supporting the daily work of the regulators in their decision making, particularly in areas of innovative technologies where the regulatory experience is still limited. Significant research activities in the field of nanotechnology resulted in a huge number of publications in the last decades. However, even if the published data can provide relevant information, scientific articles are often of diverse quality, and it is nearly impossible to manually process and evaluate such amount of data in a systematic manner. In this feasibility study, we investigated to what extent open-access automation tools can support a systematic review of toxic effects of nanomaterials for health applications reported in the scientific literature. In this study, we used a battery of available tools to perform the initial steps of a systematic review such as targeted searches, data curation and abstract screening. This work was complemented with an in-house developed tool that allowed us to extract specific sections of the articles such as the materials and methods part or the results section where we could perform subsequent text analysis. We ranked the articles according to quality criteria based on the reported nanomaterial characterisation and extracted most frequently described toxic effects induced by different types of nanomaterials. Even if further demonstration of the reliability and applicability of automation tools is necessary, this study demonstrated the potential to leverage information from the scientific literature by using automation systems in a tiered strategy.

Tools to support the automation of systematic reviews: A scoping review

OBJECTIVE

The objectives of this scoping review are to identify the reliability and validity of the available tools, their limitations and any recommendations to further improve the use of these tools.

STUDY DESIGN

A scoping review methodology was followed to map the literature published on the challenges and solutions of conducting evidence synthesis using the JBI scoping review methodology.

RESULTS

A total of 47 publications were included in the review. The current scoping review identified that LitSuggest, Rayyan, Abstractr, BIBOT, R software, RobotAnalyst, DistillerSR, ExaCT and NetMetaXL have potential to be used for the automation of systematic reviews. However, they are not without limitations. The review also identified other studies that employed algorithms that have not yet been developed into user friendly tools. Some of these algorithms showed high validity and reliability but their use is conditional on user knowledge of computer science and algorithms.

CONCLUSION

Abstract screening has reached maturity; data extraction is still an active area. Developing methods to semi-automate different steps of evidence synthesis via machine learning remains an important research direction. Also, it is important to move from the research prototypes currently available to professionally maintained platforms.

Reporting of methods to prepare, pilot and perform data extraction in systematic reviews: analysis of a sample of 152 Cochrane and non-Cochrane reviews

BACKGROUND

Previous research on data extraction methods in systematic reviews has focused on single aspects of the process. We aimed to provide a deeper insight into these methods by analysing a current sample of reviews.

METHODS

We included systematic reviews of health interventions in humans published in English. We analysed 75 Cochrane reviews from May and June 2020 and a random sample of non-Cochrane reviews published in the same period and retrieved from Medline. We linked reviews with protocols and study registrations. We collected information on preparing, piloting, and performing data extraction and on use of software to assist review conduct (automation tools). Data were extracted by one author, with 20% extracted in duplicate. Data were analysed descriptively.

RESULTS

Of the 152 included reviews, 77 reported use of a standardized extraction form (51%); 42 provided information on the type of form used (28%); 24 on piloting (16%); 58 on what data was collected (38%); 133 on the extraction method (88%); 107 on resolving disagreements (70%); 103 on methods to obtain additional data or information (68%); 52 on procedures to avoid data errors (34%); and 47 on methods to deal with multiple study reports (31%). Items were more frequently reported in Cochrane than non-Cochrane reviews. The data extraction form used was published in 10 reviews (7%). Use of software was rarely reported except for statistical analysis software and use of RevMan and GRADEpro GDT in Cochrane reviews. Covidence was the most frequent automation tool used: 18 reviews used it for study selection (12%) and 9 for data extraction (6%).

CONCLUSIONS

Reporting of data extraction methods in systematic reviews is limited, especially in non-Cochrane reviews. This includes core items of data extraction such as methods used to manage disagreements. Few reviews currently use software to assist data extraction and review conduct. Our results can serve as a baseline to assess the uptake of such tools in future analyses.

Irreproducibility in searches of scientific literature: A comparative analysis

Repeatability is the cornerstone of science, and it is particularly important for systematic reviews. However, little is known on how researchers' choice of database, and search platform influence the repeatability of systematic reviews. Here, we aim to unveil how the computer environment and the location where the search was initiated from influence hit results.We present a comparative analysis of time-synchronized searches at different institutional locations in the world and evaluate the consistency of hits obtained within each of the search terms using different search platforms.We revealed a large variation among search platforms and showed that PubMed and Scopus returned consistent results to identical search strings from different locations. Google Scholar and Web of Science's Core Collection varied substantially both in the number of returned hits and in the list of individual articles depending on the search location and computing environment. Inconsistency in Web of Science results has most likely emerged from the different licensing packages at different institutions.To maintain scientific integrity and consistency, especially in systematic reviews, action is needed from both the scientific community and scientific search platforms to increase search consistency. Researchers are encouraged to report the search location and the databases used for systematic reviews, and database providers should make search algorithms transparent and revise access rules to titles behind paywalls. Additional options for increasing the repeatability and transparency of systematic reviews are storing both search metadata and hit results in open repositories and using Application Programming Interfaces (APIs) to retrieve standardized, machine-readable search metadata.

Artificial intelligence and the conduct of literature reviews

Artificial intelligence (AI) is beginning to transform traditional research practices in many areas. In this context, literature reviews stand out because they operate on large and rapidly growing volumes of documents, that is, partially structured (meta)data, and pervade almost every type of paper published in information systems research or related social science disciplines. To familiarize researchers with some of the recent trends in this area, we outline how AI can expedite individual steps of the literature review process. Considering that the use of AI in this context is in an early stage of development, we propose a comprehensive research agenda for AI-based literature reviews (AILRs) in our field. With this agenda, we would like to encourage design science research and a broader constructive discourse on shaping the future of AILRs in research.

Systematic review automation tools improve efficiency but lack of knowledge impedes their adoption: a survey

OBJECTIVE

We investigated systematic review automation tool use by systematic reviewers, health technology assessors and clinical guideline developerst.

STUDY DESIGN AND SETTING

An online, 16-question survey was distributed across several evidence synthesis, health technology assessment and guideline development organizations. We asked the respondents what tools they use and abandon, how often and when do they use the tools, their perceived time savings and accuracy, and desired new tools. Descriptive statistics were used to report the results.

RESULTS

A total of 253 respondents completed the survey; 89% have used systematic review automation tools - most frequently whilst screening (79%). Respondents' "top 3" tools included: Covidence (45%), RevMan (35%), Rayyan and GRADEPro (both 22%); most commonly abandoned were Rayyan (19%), Covidence (15%), DistillerSR (14%) and RevMan (13%). Tools saved time (80%) and increased accuracy (54%). Respondents taught themselves to how to use the tools (72%); lack of knowledge was the most frequent barrier to tool adoption (51%). New tool development was suggested for the searching and data extraction stages.

CONCLUSION

Automation tools will likely have an increasingly important role in high-quality and timely reviews. Further work is required in training and dissemination of automation tools and ensuring they meet the desirable features of those conducting systematic reviews.

The Impact of Systematic Review Automation Tools on Methodological Quality and Time Taken to Complete Systematic Review Tasks: Case Study

BACKGROUND

Systematic reviews (SRs) are considered the highest level of evidence to answer research questions; however, they are time and resource intensive.

OBJECTIVE

When comparing SR tasks done manually, using standard methods, versus those same SR tasks done using automated tools, (1) what is the difference in time to complete the SR task and (2) what is the impact on the error rate of the SR task?

METHODS

A case study compared specific tasks done during the conduct of an SR on prebiotic, probiotic, and synbiotic supplementation in chronic kidney disease. Two participants (manual team) conducted the SR using current methods, comprising a total of 16 tasks. Another two participants (automation team) conducted the tasks where a systematic review automation (SRA) tool was available, comprising of a total of six tasks. The time taken and error rate of the six tasks that were completed by both teams were compared.

RESULTS

The approximate time for the manual team to produce a draft of the background, methods, and results sections of the SR was 126 hours. For the six tasks in which times were compared, the manual team spent 2493 minutes (42 hours) on the tasks, compared to 708 minutes (12 hours) spent by the automation team. The manual team had a higher error rate in two of the six tasks-regarding Task 5: Run the systematic search, the manual team made eight errors versus three errors made by the automation team; regarding Task 12: Assess the risk of bias, 25 assessments differed from a reference standard for the manual team compared to 20 differences for the automation team. The manual team had a lower error rate in one of the six tasks-regarding Task 6: Deduplicate search results, the manual team removed one unique study and missed zero duplicates versus the automation team who removed two unique studies and missed seven duplicates. Error rates were similar for the two remaining compared tasks-regarding Task 7: Screen the titles and abstracts and Task 9: Screen the full text, zero relevant studies were excluded by both teams. One task could not be compared between groups-Task 8: Find the full text.

CONCLUSIONS

For the majority of SR tasks where an SRA tool was used, the time required to complete that task was reduced for novice researchers while methodological quality was maintained.

Data extraction methods for systematic review (semi)automation: Update of a living systematic review

Background: The reliable and usable (semi)automation of data extraction can support the field of systematic review by reducing the workload required to gather information about the conduct and results of the included studies. This living systematic review examines published approaches for data extraction from reports of clinical studies. Methods: We systematically and continually search PubMed, ACL Anthology, arXiv, OpenAlex via EPPI-Reviewer, and the  dblp computer science bibliography. Full text screening and data extraction are conducted within an open-source living systematic review application created for the purpose of this review. This living review update includes publications up to December 2022 and OpenAlex content up to March 2023. Results: 76 publications are included in this review. Of these, 64 (84%) of the publications addressed extraction of data from abstracts, while 19 (25%) used full texts. A total of 71 (93%) publications developed classifiers for randomised controlled trials. Over 30 entities were extracted, with PICOs (population, intervention, comparator, outcome) being the most frequently extracted. Data are available from 25 (33%), and code from 30 (39%) publications. Six (8%) implemented publicly available tools Conclusions: This living systematic review presents an overview of (semi)automated data-extraction literature of interest to different types of literature review. We identified a broad evidence base of publications describing data extraction for interventional reviews and a small number of publications extracting epidemiological or diagnostic accuracy data. Between review updates, trends for sharing data and code increased strongly: in the base-review, data and code were available for 13 and 19% respectively, these numbers increased to 78 and 87% within the 23 new publications. Compared with the base-review, we observed another research trend, away from straightforward data extraction and towards additionally extracting relations between entities or automatic text summarisation. With this living review we aim to review the literature continually.

Data extraction methods for systematic review (semi)automation: A living systematic review

Background: The reliable and usable (semi)automation of data extraction can support the field of systematic review by reducing the workload required to gather information about the conduct and results of the included studies. This living systematic review examines published approaches for data extraction from reports of clinical studies. Methods: We systematically and continually search MEDLINE, Institute of Electrical and Electronics Engineers (IEEE), arXiv, and the dblp computer science bibliography databases. Full text screening and data extraction are conducted within an open-source living systematic review application created for the purpose of this review. This iteration of the living review includes publications up to a cut-off date of 22 April 2020. Results: In total, 53 publications are included in this version of our review. Of these, 41 (77%) of the publications addressed extraction of data from abstracts, while 14 (26%) used full texts. A total of 48 (90%) publications developed and evaluated classifiers that used randomised controlled trials as the main target texts. Over 30 entities were extracted, with PICOs (population, intervention, comparator, outcome) being the most frequently extracted. A description of their datasets was provided by 49 publications (94%), but only seven (13%) made the data publicly available. Code was made available by 10 (19%) publications, and five (9%) implemented publicly available tools. Conclusions: This living systematic review presents an overview of (semi)automated data-extraction literature of interest to different types of systematic review. We identified a broad evidence base of publications describing data extraction for interventional reviews and a small number of publications extracting epidemiological or diagnostic accuracy data. The lack of publicly available gold-standard data for evaluation, and lack of application thereof, makes it difficult to draw conclusions on which is the best-performing system for each data extraction target. With this living review we aim to review the literature continually.

The decisions and processes involved in a systematic search strategy: a hierarchical framework

OBJECTIVE

The decisions and processes that may compose a systematic search strategy have not been formally identified and categorized. This study aimed to (1) identify all decisions that could be made and processes that could be used in a systematic search strategy and (2) create a hierarchical framework of those decisions and processes.

METHODS

The literature was searched for documents or guides on conducting a literature search for a systematic review or other evidence synthesis. The decisions or processes for locating studies were extracted from eligible documents and categorized into a structured hierarchical framework. Feedback from experts was sought to revise the framework. The framework was revised iteratively and tested using recently published literature on systematic searching.

RESULTS

Guidance documents were identified from expert organizations and a search of the literature and Internet. Data were extracted from 74 eligible documents to form the initial framework. The framework was revised based on feedback from 9 search experts and further review and testing by the authors. The hierarchical framework consists of 119 decisions or processes sorted into 17 categories and arranged under 5 topics. These topics are "Skill of the searcher," "Selecting information to identify," "Searching the literature electronically," "Other ways to identify studies," and "Updating the systematic review."

CONCLUSIONS

The work identifies and classifies the decisions and processes used in systematic searching. Future work can now focus on assessing and prioritizing research on the best methods for successfully identifying all eligible studies for a systematic review.

Developing a fully automated evidence synthesis tool for identifying, assessing and collating the evidence

Evidence synthesis is a key element of evidence-based medicine. However, it is currently hampered by being labour intensive meaning that many trials are not incorporated into robust evidence syntheses and that many are out of date. To overcome this, a variety of techniques are being explored, including using automation technology. Here, we describe a fully automated evidence synthesis system for intervention studies, one that identifies all the relevant evidence, assesses the evidence for reliability and collates it to estimate the relative effectiveness of an intervention. Techniques used include machine learning, natural language processing and rule-based systems. Results are visualised using modern visualisation techniques. We believe this to be the first, publicly available, automated evidence synthesis system: an evidence mapping tool that synthesises evidence on the fly.

Decoding semi-automated title-abstract screening: findings from a convenience sample of reviews

BACKGROUND

We evaluated the benefits and risks of using the Abstrackr machine learning (ML) tool to semi-automate title-abstract screening and explored whether Abstrackr's predictions varied by review or study-level characteristics.

METHODS

For a convenience sample of 16 reviews for which adequate data were available to address our objectives (11 systematic reviews and 5 rapid reviews), we screened a 200-record training set in Abstrackr and downloaded the relevance (relevant or irrelevant) of the remaining records, as predicted by the tool. We retrospectively simulated the liberal-accelerated screening approach. We estimated the time savings and proportion missed compared with dual independent screening. For reviews with pairwise meta-analyses, we evaluated changes to the pooled effects after removing the missed studies. We explored whether the tool's predictions varied by review and study-level characteristics.

RESULTS

Using the ML-assisted liberal-accelerated approach, we wrongly excluded 0 to 3 (0 to 14%) records that were included in the final reports, but saved a median (IQR) 26 (9, 42) h of screening time. One missed study was included in eight pairwise meta-analyses in one systematic review. The pooled effect for just one of those meta-analyses changed considerably (from MD (95% CI) - 1.53 (- 2.92, - 0.15) to - 1.17 (- 2.70, 0.36)). Of 802 records in the final reports, 87% were correctly predicted as relevant. The correctness of the predictions did not differ by review (systematic or rapid, P = 0.37) or intervention type (simple or complex, P = 0.47). The predictions were more often correct in reviews with multiple (89%) vs. single (83%) research questions (P = 0.01), or that included only trials (95%) vs. multiple designs (86%) (P = 0.003). At the study level, trials (91%), mixed methods (100%), and qualitative (93%) studies were more often correctly predicted as relevant compared with observational studies (79%) or reviews (83%) (P = 0.0006). Studies at high or unclear (88%) vs. low risk of bias (80%) (P = 0.039), and those published more recently (mean (SD) 2008 (7) vs. 2006 (10), P = 0.02) were more often correctly predicted as relevant.

CONCLUSION

Our screening approach saved time and may be suitable in conditions where the limited risk of missing relevant records is acceptable. Several of our findings are paradoxical and require further study to fully understand the tasks to which ML-assisted screening is best suited. The findings should be interpreted in light of the fact that the protocol was prepared for the funder, but not published a priori. Because we used a convenience sample, the findings may be prone to selection bias. The results may not be generalizable to other samples of reviews, ML tools, or screening approaches. The small number of missed studies across reviews with pairwise meta-analyses hindered strong conclusions about the effect of missed studies on the results and conclusions of systematic reviews.

Creating and curating a community of practice: introducing the evidence synthesis Hackathon and a special series in evidence synthesis technology

Evidence synthesis is a vital part of evidence-informed decision-making, but high growth in the volume of research evidence over recent decades has made efficient evidence synthesis increasingly challenging. As the appreciation and need for timely and rigorous evidence synthesis continue to grow, so too will the need for tools and frameworks to conduct reviews of expanding evidence bases in an efficient and time-sensitive manner. Efforts to future-proof evidence synthesis through the development of new evidence synthesis technology (ESTech) have so far been isolated across interested individuals or groups, with no concerted effort to collaborate or build communities of practice in technology production. We established the evidence synthesis Hackathon to stimulate collaboration and the production of Free and Open Source Software and frameworks to support evidence synthesis. Here, we introduce a special series of papers on ESTech, and invite the readers of environmental evidence to submit manuscripts introducing and validating novel tools and frameworks. We hope this collection will help to consolidate ESTech development efforts and we encourage readers to join the ESTech revolution. In order to future-proof evidence synthesis against the evidence avalanche, we must support community enthusiasm for ESTech, reduce redundancy in tool design, collaborate and share capacity in tool production, and reduce inequalities in software accessibility.

Development, testing and use of data extraction forms in systematic reviews: a review of methodological guidance

Background

Data extraction forms link systematic reviews with primary research and provide the foundation for appraising, analysing, summarising and interpreting a body of evidence. This makes their development, pilot testing and use a crucial part of the systematic reviews process. Several studies have shown that data extraction errors are frequent in systematic reviews, especially regarding outcome data.

Methods

We reviewed guidance on the development and pilot testing of data extraction forms and the data extraction process. We reviewed four types of sources: 1) methodological handbooks of systematic review organisations (SRO); 2) textbooks on conducting systematic reviews; 3) method documents from health technology assessment (HTA) agencies and 4) journal articles. HTA documents were retrieved in February 2019 and database searches conducted in December 2019. One author extracted the recommendations and a second author checked them for accuracy. Results are presented descriptively.

Results

Our analysis includes recommendations from 25 documents: 4 SRO handbooks, 11 textbooks, 5 HTA method documents and 5 journal articles. Across these sources the most common recommendations on form development are to use customized or adapted standardised extraction forms (14/25); provide detailed instructions on their use (10/25); ensure clear and consistent coding and response options (9/25); plan in advance which data are needed (9/25); obtain additional data if required (8/25); and link multiple reports of the same study (8/25). The most frequent recommendations on piloting extractions forms are that forms should be piloted on a sample of studies (18/25); and that data extractors should be trained in the use of the forms (7/25). The most frequent recommendations on data extraction are that extraction should be conducted by at least two people (17/25); that independent parallel extraction should be used (11/25); and that procedures to resolve disagreements between data extractors should be in place (14/25).

Conclusions

Overall, our results suggest a lack of comprehensiveness of recommendations. This may be particularly problematic for less experienced reviewers. Limitations of our method are the scoping nature of the review and that we did not analyse internal documents of health technology agencies.

Evidence-Based Research Series-Paper 2 : Using an Evidence-Based Research approach before a new study is conducted to ensure value

BACKGROUND AND OBJECTIVES

There is considerable actual and potential waste in research. The aim of this article is to describe how using an evidence-based research approach before conducting a study helps to ensure that the new study truly adds value.

STUDY DESIGN AND SETTING

Evidence-based research is the use of prior research in a systematic and transparent way to inform a new study so that it is answering questions that matter in a valid, efficient, and accessible manner. In this second article of the evidence-based research series, we describe how to apply an evidence-based research approach before starting a new study.

RESULTS

Before a new study is performed, researchers need to provide a solid justification for it using the available scientific knowledge as well as the perspectives of end users. The key method for both is to conduct a systematic review of earlier relevant studies.

CONCLUSION

Describing the ideal process illuminates the challenges and opportunities offered through the suggested evidence-based research approach. A systematic and transparent approach is needed to provide justification for and to optimally design a relevant and necessary new study.

Evidence-Based Research Series-Paper 1: What Evidence-Based Research is and why is it important?

OBJECTIVES

There is considerable actual and potential waste in research. Evidence-based research ensures worthwhile and valuable research. The aim of this series, which this article introduces, is to describe the evidence-based research approach.

STUDY DESIGN AND SETTING

In this first article of a three-article series, we introduce the evidence-based research approach. Evidence-based research is the use of prior research in a systematic and transparent way to inform a new study so that it is answering questions that matter in a valid, efficient, and accessible manner.

RESULTS

We describe evidence-based research and provide an overview of the approach of systematically and transparently using previous research before starting a new study to justify and design the new study (article #2 in series) and-on study completion-place its results in the context with what is already known (article #3 in series).

CONCLUSION

This series introduces evidence-based research as an approach to minimize unnecessary and irrelevant clinical health research that is unscientific, wasteful, and unethical.

Artificial intelligence and automation of systematic reviews in women's health

PURPOSE OF REVIEW

Evidence-based women's healthcare is underpinned by systematic reviews and guidelines. Generating an evidence synthesis to support guidance for clinical practice is a time-consuming and labour-intensive activity that delays transfer of research into practice. Artificial intelligence has the potential to rapidly collate, combine, and update high-quality medical evidence with accuracy and precision, and without bias.

RECENT FINDINGS

This article describes the main fields of artificial intelligence with examples of its application to systematic reviews. These include the capabilities of processing natural language texts, retrieving information, reasoning, and learning. The complementarity and interconnection of the various artificial intelligence techniques can be harnessed to solve difficult problems in automation of reviews. Computer science can advance evidence-based medicine through development, testing, and refinement of artificial intelligence tools to deploy automation, creating 'living' evidence syntheses.

SUMMARY

Groundbreaking, high-quality, and impactful artificial intelligence will accelerate the transfer of individual research studies seamlessly into evidence syntheses for contemporaneously improving the quality of healthcare.

Current methods for development of rapid reviews about diagnostic tests: an international survey

BACKGROUND

Rapid reviews (RRs) have emerged as an efficient alternative to time-consuming systematic reviews-they can help meet the demand for accelerated evidence synthesis to inform decision-making in healthcare. The synthesis of diagnostic evidence has important methodological challenges. Here, we performed an international survey to identify the current practice of producing RRs for diagnostic tests.

METHODS

We developed and administered an online survey inviting institutions that perform RRs of diagnostic tests from all over the world.

RESULTS

All participants (N = 25) reported the implementation of one or more methods to define the scope of the RR; however, only one strategy (defining a structured question) was used by ≥90% of participants. All participants used at least one methodological shortcut including the use of a previous review as a starting point (92%) and the use of limits on the search (96%). Parallelization and automation of review tasks were not extensively used (48 and 20%, respectively).

CONCLUSION

Our survey indicates a greater use of shortcuts and limits for conducting diagnostic test RRs versus the results of a recent scoping review analyzing published RRs. Several shortcuts are used without knowing how their implementation affects the results of the evidence synthesis in the setting of diagnostic test reviews. Thus, a structured evaluation of the challenges and implications of the adoption of these RR methods is warranted.

A focus on cross-purpose tools, automated recognition of study design in multiple disciplines, and evaluation of automation tools: a summary of significant discussions at the fourth meeting of the International Collaboration for Automation of Systematic Reviews (ICASR)

The fourth meeting of the International Collaboration for Automation of Systematic Reviews (ICASR) was held 5-6 November 2019 in The Hague, the Netherlands. ICASR is an interdisciplinary group whose goal is to maximize the use of technology for conducting rapid, accurate, and efficient systematic reviews of scientific evidence. The group seeks to facilitate the development and acceptance of automated techniques for systematic reviews. In 2018, the major themes discussed were the transferability of automation tools (i.e., tools developed for other purposes that might be used by systematic reviewers), the automated recognition of study design in multiple disciplines and applications, and approaches for the evaluation of automation tools.

Data extraction methods for systematic review (semi)automation: A living review protocol

Background: Researchers in evidence-based medicine cannot keep up with the amounts of both old and newly published primary research articles. Support for the early stages of the systematic review process - searching and screening studies for eligibility - is necessary because it is currently impossible to search for relevant research with precision. Better automated data extraction may not only facilitate the stage of review traditionally labelled 'data extraction', but also change earlier phases of the review process by making it possible to identify relevant research. Exponential improvements in computational processing speed and data storage are fostering the development of data mining models and algorithms. This, in combination with quicker pathways to publication, led to a large landscape of tools and methods for data mining and extraction. Objective: To review published methods and tools for data extraction to (semi)automate the systematic reviewing process. Methods: We propose to conduct a living review. With this methodology we aim to do constant evidence surveillance, bi-monthly search updates, as well as review updates every 6 months if new evidence permits it. In a cross-sectional analysis we will extract methodological characteristics and assess the quality of reporting in our included papers. Conclusions: We aim to increase transparency in the reporting and assessment of automation technologies to the benefit of data scientists, systematic reviewers and funders of health research. This living review will help to reduce duplicate efforts by data scientists who develop data mining methods. It will also serve to inform systematic reviewers about possibilities to support their data extraction.

Data extraction methods for systematic review (semi)automation: A living review protocol

Background: Researchers in evidence-based medicine cannot keep up with the amounts of both old and newly published primary research articles. Conducting and updating of systematic reviews is time-consuming. In practice, data extraction is one of the most complex tasks in this process. Exponential improvements in computational processing speed and data storage are fostering the development of data extraction models and algorithms. This, in combination with quicker pathways to publication, led to a large landscape of tools and methods for data extraction tasks. Objective: To review published methods and tools for data extraction to (semi)automate the systematic reviewing process. Methods: We propose to conduct a living review. With this methodology we aim to do monthly search updates, as well as bi-annual review updates if new evidence permits it. In a cross-sectional analysis we will extract methodological characteristics and assess the quality of reporting in our included papers. Conclusions: We aim to increase transparency in the reporting and assessment of machine learning technologies to the benefit of data scientists, systematic reviewers and funders of health research. This living review will help to reduce duplicate efforts by data scientists who develop data extraction methods. It will also serve to inform systematic reviewers about possibilities to support their data extraction.