Premium
The risk of bias of animal experiments in implant dentistry: a methodological study
Author(s) -
Faggion Clovis Mariano,
Diaz Karla Tatiana,
Aranda Luisiana,
Gabel Frank,
Listl Stefan,
Alarcón Marco Antonio
Publication year - 2017
Publication title -
clinical oral implants research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.407
H-Index - 161
eISSN - 1600-0501
pISSN - 0905-7161
DOI - 10.1111/clr.12852
Subject(s) - dentistry , medicine , confidence interval , randomized controlled trial , logistic regression , medline , meta analysis , scopus , univariate , implant , multivariate statistics , surgery , mathematics , biology , statistics , biochemistry
Objectives To evaluate the risk of bias ( ROB ) in reports of randomised controlled trials ( RCT s) of animal experiments published in implant dentistry, and to explore the association between animal experiment characteristics and ROB . Material and methods We searched the MEDLINE (via PubMed), SCOPUS and Sci ELO databases from 2010 to March 2015 for reports of RCT s of animal experiments published in implant dentistry. We evaluated independently and in duplicate the ROB of these experiments by the use of a tool specifically developed to evaluate ROB in animal studies, the SYRCLE 's tool. ROB was judged as low, high or unclear (when there was not enough information to judge ROB ). We used univariate and multivariate logistic regression analyses to evaluate the association of specific study characteristics and extent of ROB . Results We initially selected 850 publications and 161 reports of animal experiments were included. For a total of 1449 entries (records), 486 (34%) were rated as low ROB . High ROB was attributed to 80 (6%) of entries, and 883 (60%) entries were rated as unclear ROB . The characteristics “impact factor” ( IF ), reporting of standard error ( SE ) and reporting of confidence interval ( CI ) were significantly associated with low ROB in some SYRCLE domains. Conclusions A substantial number of items with unclear ROB were observed in this sample of animal experiments in implant dentistry. Furthermore, the present findings suggest that implant dentistry animal experiments published in journals with higher IF and better report of measures of precision; that is, CI and SE may have lower ROB than those not having these characteristics.