Adesão às recomendações CONSORT de ensaios clínicos randomizados sobre longevidade de restaurações atraumáticas: uma revisão sistemática.

Abstract

The present systematic review of the literature assessed the conformity of randomized controlled trials on the longevity of atraumatic restorations (ART) with the CONSORT recommendations and the risk of bias. After developing a specific strategy, a broad search in the literature was carried out in different databases to find randomized clinical screenings that assessed the longevity of atraumatic restorations compared to other restorative procedures. A specific tool for evaluating items belonging to the CONSORT recommendations was applied, with scores ranging from 0 (without description of the item), 1 (poor / inadequate description) and 2 (adequate / complete description), in a total of 32 points. The risk of bias analysis was performed using the Cochrane tool for risk assessment, classifying each domain as low risk, undefined risk or high risk. he results were presented in the form of descriptive statistics, with the number of published studies classified according to periodical, follow-up period and country of the main author, with the respective adherence averages to the CONSORT recommendations obtained. The comparison of these averages in relation to each of these classifications was performed with ANOVA per rank (α = 0.05); the correlations between the impact factor of the journals and the CONSORT average and the impact factor and the risk of bias were analyzed using Pearson's correlation (α = 0.05). After searching the literature, 2167 studies were identified. From all of them, 33 randomized clinical trials met the criteria and were included in the systematic review. The overall average of CONSORT adherence of these studies was 21.3 ± 6.1 points, which represents 66% of the maximum possible score. The items considered most critical and with the greatest number of missing or flaws in the description were protocol, flowchart and allocation concealment, since approximately 70% of the studies obtained a score of 0. The correct description of the sample calculation was also missing in more than 50% of the articles. Differences in the means of adhering to CONSORT were observed in the analysis according to the periodical, country of the main author and follow-up period (p <0.001). 15.15% of the studies were judged to be at low risk of bias; 75.75% were classified as unclear risk; and 15.15% as a high risk of bias. It is concluded that, although the CONSORT guidelines have been included in the instructions for authors of different journals, compliance is far from being fully achieved. The results indicate that the adherence of randomized clinical trials on the longevity of atraumatic restorations to the CONSORT recommendations should be encouraged, which will also improve the refinement of the methodology by the authors, which will allow a reduction in the risk of bias in randomized clinical trials.