Directly Placed Restorative Materials

For restoring cavitated dental lesions, whether carious or not, a large number of material combinations are available. We aimed to systematically review and synthesize data of comparative dental restorative trials. A systematic review was performed. Randomized controlled trials published between 2005 and 2015 were included that compared the survival of ≥2 restorative and/or adhesive materials (i.e., no need for restorative reintervention). Pairwise and Bayesian network meta-analyses were performed, with separate evaluations for cervical cavitated lesions and load-bearing posterior cavitated lesions in permanent and primary teeth. A total of 11,070 restorations (5,330 cervical, 5,740 load bearing) had been placed in 3,633 patients in the included trials. Thirty-six trials investigated restoration of cervical lesions (all in permanent teeth) and 36 of load-bearing lesions (8 in primary and 28 in permanent teeth). Resin-modified glass ionomer cements had the highest chance of survival in cervical cavitated lesions; composites or compomers placed via 2-step self-etch and 3-step etch-and-rinse adhesives were ranked next. Restorations placed with 2-step etch-and-rinse or 1-step self-etch adhesives performed worst. For load-bearing restorations, conventional composites had the highest probability of survival, while siloranes were found least suitable. Ambiguity remains regarding which adhesive strategy to use in load-bearing cavitated lesions. Most studies showed high risk of bias, and several comparisons were prone for publication bias. If prioritized for survival, resin-modified glass ionomer cements might be recommended to restore cervical lesions. For load-bearing ones, conventional or bulk fill composites seem most suitable. The available evidence is quantitatively and qualitatively insufficient for further recommendations, especially with regard to adhesive strategies in posterior load-bearing situations. Moreover, different material classifications might yield different findings on the same materials. Future trials should aim for sufficient power, longer follow-up times, and high internal validity to prove or refute differences between certain material combinations. An agreed material classification for future syntheses is desirable.