Open AccessExtensive literature search, selection for relevance and data extraction of studies related to the toxicity of PCDD /Fs and DL ‐ PCB s in humans
Author(s) -
Vedrine Max La,
Hanlon James,
Bevan Ruth,
Floyd Pete,
Brown Terry,
Matthies Franziska
Publication year - 2018
Publication title -
efsa supporting publications
Language(s) - English
Resource type - Journals
ISSN - 2397-8325
DOI - 10.2903/sp.efsa.2018.en-1136
Subject(s) - selection (genetic algorithm) , relevance (law) , data extraction , extraction (chemistry) , information retrieval , toxicology , biology , computer science , chemistry , medline , chromatography , artificial intelligence , political science , biochemistry , law
To enable the hazard identification and characterisation in the risk assessment for humans related to the seventeen 2,3,7,8‐substituted dioxins (PCCDs) and furans (PCDFs) and the twelve dioxin‐like polychlorinated biphenyls (DL‐PCBs), EFSA outsourced an extensive literature search (ELS), followed by selection for relevance and extraction of relevant data for consideration in the risk assessment. Two tailored search strategies for Web of Science (WoS) and PubMed for identifying relevant human studies were developed in discussion with EFSA and used to carry out two ELSs. The outcome of the ELSs were exported into EndNote files, with a total of 4,549 studies identified in WoS and a total of 3,677 studies identified in PubMed. The EndNote files were combined and duplicates were removed, which left 6,699 studies in total. The combined EndNote file was imported into DistillerSR ® , the duplication detection tool in DistillerSR ® was used and additional 598 duplicates were identified and moved to quarantine in DistillerSR ® . Level 1 and Level 2 relevance templates were created in DistillerSR ® using the eligibility criteria (inclusion/exclusion criteria) provided by EFSA in the Technical Specifications and these were discussed with EFSA. Following the discussions and a relevance pilot test the remaining 6,101 studies were checked for relevance. When the selection for relevance had been complete, 257 studies proceeded to Level 3 data extraction. The data extraction templates were created in DistillerSR ® using the criteria provided by EFSA. Following discussions with EFSA and a data extraction pilot test, the project team performed the data extraction on these studies with the relevant information added to the data extraction forms in DistillerSR