Is the evaluation of risk of bias in periodontology and implant dentistry comprehensive? A systematic review.

Background/Objective The objective of this study was to assess how authors of systematic reviews ( SRs) with meta-analyses published in periodontology and implant dentistry evaluate risk of bias ( ROB) in primary studies included in these reviews. Material/Methods A literature search for SRs with meta-analyses was performed in PubMed and Cochrane library databases up to July 20th 2014. The reference lists of included articles were screened for further reviews. The standards of evaluating ROB in primary studies were evaluated by using a 14-item checklist based on the Cochrane approach for evaluating ROB. Standards in ROB evaluations in Cochrane and paper-based SRs were compared using the Fisher's exact test. All searches, data extraction and evaluations were performed independently and in duplicate. Results Seventy SRs were included (45 paper-based and 25 Cochrane SRs, respectively). The median percentage of items addressed was 58% (interquartile range 4-100%). Cochrane SRs more frequently included ROB assessments than paper-based reviews in terms of examiner blinding ( p = 0.0026), selective outcome reporting ( p = 0.0207) and other bias ( p = 0.0241). Conclusions The ROB evaluation in primary studies currently included in SRs with meta-analyses in periodontology and implant dentistry is not sufficiently comprehensive. Cochrane SRs have more comprehensive ROB evaluation than paper-based reviews. [ABSTRACT FROM AUTHOR]