Potential systematic error in laboratory experiments on microbial leakage through filled root canals: review of published articles

P>Aim To systematically evaluate whether published studies on microbial leakage through filled root canals in human teeth embedded in a two-chamber system were properly controlled. Specifically, the control for the assumption that leakage should occur through the root canal rather than other routes was investigated. Methodology A systematic search was conducted using Medline, Biosis, Cochrane, Embase, and Web of Science databases. In addition, the reference lists of review articles pertaining to the topic were searched. No language restriction was applied. Two independent reviewers screened titles and abstracts. All articles deemed appropriate by either reviewer were included in the full-text evaluation. In case of disagreement, a referee arbitrated between the reviewers. Results With 93.8% agreement prior to discussion and arbitration, 67 articles were included. On average, the size of the negative control group was 30% (mean) of the n in the experimental groups (minimum = 0.0%, maximum = 100%, SD = 27%). The majority of studies (57 of 67) used inadequate negative controls. The whole root was covered with the sealing material in these specimens, whilst the root tip was left uncovered in the experimental groups. Consequently, leakage between outer root surface and sealing material was not controlled for. The authors of the remaining 10 communications did not state clearly how negative control assessments were performed. Conclusions Experimental investigations should be performed to assess the routes of microbial leakage in two-chamber models.