Technology-assisted title and abstract screening for systematic reviews: a retrospective evaluation of the Abstrackr machine learning tool

Abstract

BackgroundMachine learning tools can expedite systematic review (SR) processes by semi-automating citation screening. Abstrackr semi-automates citation screening by predicting relevant records. We evaluated its performance for four screening projects.MethodsWe used a convenience sample of screening projects completed at the Alberta Research Centre for Health Evidence, Edmonton, Canada: three SRs and one descriptive analysis for which we had used SR screening methods. The projects were heterogeneous with respect to search yield (median 9328; range 5243 to 47,385 records; interquartile range (IQR) 15,688 records), topic (Antipsychotics, Bronchiolitis, Diabetes, Child Health SRs), and screening complexity. We uploaded the records to Abstrackr and screened until it made predictions about the relevance of the remaining records. Across three trials for each project, we compared the predictions to human reviewer decisions and calculated the sensitivity, specificity, precision, false negative rate, proportion missed, and workload savings.ResultsAbstrackr’s sensitivity was > 0.75 for all projects and the mean specificity ranged from 0.69 to 0.90 with the exception of Child Health SRs, for which it was 0.19. The precision (proportion of records correctly predicted as relevant) varied by screening task (median 26.6%; range 14.8 to 64.7%; IQR 29.7%). The median false negative rate (proportion of records incorrectly predicted as irrelevant) was 12.6% (range 3.5 to 21.2%; IQR 12.3%). The workload savings were often large (median 67.2%, range 9.5 to 88.4%; IQR 23.9%). The proportion missed (proportion of records predicted as irrelevant that were included in the final report, out of the total number predicted as irrelevant) was 0.1% for all SRs and 6.4% for the descriptive analysis. This equated to 4.2% (range 0 to 12.2%; IQR 7.8%) of the records in the final reports.ConclusionsAbstrackr’s reliability and the workload savings varied by screening task. Workload savings came at the expense of potentially missing relevant records. How this might affect the results and conclusions of SRs needs to be evaluated. Studies evaluating Abstrackr as the second reviewer in a pair would be of interest to determine if concerns for reliability would diminish. Further evaluations of Abstrackr’s performance and usability will inform its refinement and practical utility.Electronic supplementary materialThe online version of this article (10.1186/s13643-018-0707-8) contains supplementary material, which is available to authorized users.

Machine learning for screening prioritization in systematic reviews: comparative performance of Abstrackr and EPPI-Reviewer

“…Specifically, for large reports, potential reductions in screening burden were 4 to 49% (Abstrackr) and 9 to 60% (EPPI-Reviewer). Our findings are consistent with other studies which have found Abstrackr to offer potential gains [7,19,21]. However, to our knowledge, our study is the first to assess EPPI-Reviewer.…”

“…Both tools are web-based and facilitate citation screening through text-mining tools and machine learning techniques; they employ "active learning," in which ongoing feedback from reviewers is used dynamically to improve predictive accuracy [14,17]. Several studies have incorporated Abstrackr into workflow (e.g., as a second screener) and reported a range of improvements in efficiency [19]. For instance, three studies using Abstrackr for screening citations for new reviews (i.e., not updates) reported significant, but variable workload reductions ranging from 9 to 57% [7,13,17].…”

Absence of evidence for the conservation outcomes of systematic conservation planning around the globe: a systematic map

“…Automated duplicate checking, and web scraping also led to significant time savings. Despite progress in the use of machine learning to help automate abstract screening [70,71], we concluded this technology is not sufficiently developed to have been applicable without further testing. Innovative methods are much needed, particularly as the size and scope of maps and reviews continue to increase [72,73].…”

Decoding semi-automated title-abstract screening: a retrospective exploration of the review, study, and publication characteristics associated with accurate relevance predictions

“…Because our evaluation was retrospective, we estimated time savings based on a screening rate of two records per minute. Although ambitious, this rate allowed for more conservative estimates of time savings and for comparisons to previous studies that have used the same rate [10,15,16].…”

“…Previously published studies undertaken at our evidence synthesis centre [10,15,16] have addressed the bene ts (workload and estimated time savings) and risks (omitting relevant studies) of various MLassisted screening approaches in systematic and rapid reviews. We have also explored the usability of some available ML tools [10].…”