The influence of selective enamel etch and self-etch mode of universal adhesives' application on clinical behavior of composite restorations placed on non-carious cervical lesions: A systematic review and meta-analysis.

BACKGROUND: To answer the PICOS question: "Is the risk of retention loss equal for SEE and SE approach when universal adhesives and composite restorations are indicated for restoring NCCLs?" METHODS: Web of Science, PubMed, Cochrane Central Register of Controlled Trials, Scopus, Scientific Electronic Library Online, LILACS, Google Scholar and OpenGrey were searched. Only randomized controlled clinical trials in which NCCLs were restored with composites and universal adhesives applied in SEE or SE mode were included. The articles were assessed for the risk of bias, after which meta-analyses were run (fixed-effects model was applied; heterogeneity was explored using Cochran Q test and I2 statistics; alpha = 0.05) and the certainty of evidence was assessed by the GRADE tool. RESULTS: Fifteen articles were included in qualitative, while 7 articles were included in quantitative analysis. Seven studies were judged as "low" risk of bias, while 8 were considered as "unclear" risk of bias. Statistically significant difference for retention were observed at 6- and 18/24 months (p = 0.05; OR=0.42, 95% CI [0.18, 0.99]; very low certainty of evidence and p = 0.007; OR=0.31, 95% CI [0.13, 0.72]; low certainty of evidence, respectively), favoring SEE approach. No other significant differences in clinical outcomes were observed between SEE and SE approach (very low certainty of evidence). SIGNIFICANCE: When restoring NCCLs, clinicians might consider applying universal adhesives in SEE mode since it could lead to more predictable retention compared to SE approach up to 2 years of follow-up.