Improving the conduct of systematic reviews: a process mining perspective

Pham, Ba’; Bagheri, Ebrahim; Pourmasoumi, Asef; Robson, Reid; Hwee, Jeremiah; Isaranuwatchai, Wanrudee; Darvesh, Nazia; Page, Matthew J.; Tricco, Andrea C

doi:10.1016/j.jclinepi.2018.06.011

Cited by 25 publications

(22 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2 In cataloging the work process, study selection, including abstract screening, is the most labor-intensive and time-consuming step, accounting for an estimated 26% of project work. 3 As the literature on cancer genetics grows, it is imperative to develop an effective procedure that can reduce the literature review burden without compromising comprehensiveness.…”

Section: Introductionmentioning

confidence: 99%

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Deng

Yin

Bao

et al. 2019

JCO Clinical Cancer Informatics

View full text Add to dashboard Cite

PURPOSE Quantifying the risk of cancer associated with pathogenic mutations in germline cancer susceptibility genes—that is, penetrance—enables the personalization of preventive management strategies. Conducting a meta-analysis is the best way to obtain robust risk estimates. We have previously developed a natural language processing (NLP) –based abstract classifier which classifies abstracts as relevant to penetrance, prevalence of mutations, both, or neither. In this work, we evaluate the performance of this NLP-based procedure. MATERIALS AND METHODS We compared the semiautomated NLP-based procedure, which involves automated abstract classification and text mining, followed by human review of identified studies, with the traditional procedure that requires human review of all studies. Ten high-quality gene–cancer penetrance meta-analyses spanning 16 gene–cancer associations were used as the gold standard by which to evaluate the performance of our procedure. For each meta-analysis, we evaluated the number of abstracts that required human review (workload) and the ability to identify the studies that were included by the authors in their quantitative analysis (coverage). RESULTS Compared with the traditional procedure, the semiautomated NLP-based procedure led to a lower workload across all 10 meta-analyses, with an overall 84% reduction (2,774 abstracts v 16,941 abstracts) in the amount of human review required. Overall coverage was 93%—we are able to identify 132 of 142 studies—before reviewing references of identified studies. Reasons for the 10 missed studies included blank and poorly written abstracts. After reviewing references, nine of the previously missed studies were identified and coverage improved to 99% (141 of 142 studies). CONCLUSION We demonstrated that an NLP-based procedure can significantly reduce the review workload without compromising the ability to identify relevant studies. NLP algorithms have promising potential for reducing human efforts in the literature review process.

show abstract

Section: Introductionmentioning

confidence: 99%

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Deng

Yin

Bao

et al. 2019

JCO Clinical Cancer Informatics

View full text Add to dashboard Cite

show abstract

“…In order to obtain information on the methods and resources used by these agencies, we will develop an online questionnaire containing information about activities, methods and resources organized by review development stages [25]. The online survey will include the following issues:General issues for the development of rapid reviews of diagnostic tests (including time for development, intended audience, commissioner of these reviews, focus of diagnostic reviews).Composition of the review team for conducting of rapid reviews of diagnostic tests.Methodological shortcuts used (including abbreviated search strategies, number of authors involved in the screening, selection of references and collection of data).Methods for performing the synthesis of study results (including use of the GRADE system).Issues regarding the preparation of the review report and their ending.…”

Section: Methodsmentioning

confidence: 99%

Challenges of rapid reviews for diagnostic test accuracy questions: a protocol for an international survey and expert consultation

et al. 2019

Self Cite

View full text Add to dashboard Cite

Background Assessment of diagnostic tests, broadly defined as any element that aids in the collection of additional information for further clarification of a patient’s health status, has increasingly become a critical issue in health policy and decision-making. Diagnostic evidence, including the accuracy of a medical test for a target condition, is commonly appraised using standard systematic review methodology. Owing to the considerable time and resources required to conduct these, rapid reviews have emerged as a pragmatic alternative by tailoring methods according to the decision maker’s circumstances. However, it is not known if streamlining methodological aspects has an impact on the validity of evidence synthesis. Furthermore, due to the particular nature and complexity of the appraisal of diagnostic accuracy, there is need for detailed guidance on how to conduct rapid reviews of diagnostic tests. In this study, we aim to identify the methods currently used by rapid review developers to synthesize evidence on diagnostic test accuracy, as well as to analyze potential shortcomings and challenges related to these methods. Methods We will carry out a two-fold approach: (1) an international survey of professionals working in organizations that develop rapid reviews of diagnostic tests, in terms of the methods and resources used by these agencies when conducting rapid reviews, and (2) semi-structured interviews with senior-level individuals to further explore and validate the findings from the survey and to identify challenges in conducting rapid reviews. We will use STATA 15.0 for quantitative analyses and framework analysis for qualitative analyses. We will ensure protection of data during all stages. Discussion The main result of this research will be a map of methods and resources currently used for conducting rapid reviews of diagnostic test accuracy, as well as methodological shortcomings and potential solutions in diagnostic knowledge synthesis that require further research.

show abstract

“…Novel methods are needed if systematic review production is to keep pace with the publication of new evidence from trials, and if existing systematic reviews are to be kept up-to-date [2][3][4]. Title-abstract screening is one of the more timeconsuming steps in the production of a systematic review [5]. Often, two independent reviewers will screen all potentially eligible records to identify the (relatively) few that are relevant.…”

Section: Introductionmentioning

confidence: 99%

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

Gates

DaRosa

et al. 2020

Syst Rev

View full text Add to dashboard Cite

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.

show abstract

Improving the conduct of systematic reviews: a process mining perspective

Cited by 25 publications

References 33 publications

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Validation of a Semiautomated Natural Language Processing–Based Procedure for Meta-Analysis of Cancer Susceptibility Gene Penetrance

Challenges of rapid reviews for diagnostic test accuracy questions: a protocol for an international survey and expert consultation

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

Contact Info

Product

Resources

About