Early and late implant failure of submerged versus non-submerged implant healing: A systematic review, meta-analysis and trial sequential analysis

Introduction: The aim of this systematic review was to analyse current evidenceregarding differences in early and late implant failure as well as in marginal bonelevel (MBL) changes between submerged and non-submerged healed dentalimplants.Methods: PUBMED, SCOPUS, EMBASE and Web of Science databases weresearched for prospective randomized and non-randomized controlled studies addressing direct comparison between submerged and non-submerged implant healing, without performing immediate loading. Early and late implant failure (before orafter 6 months from implant placement, respectively) together with MBL were theinvestigated outcomes. Risk of bias assessment was performed using the CochraneCollaboration Tool for Randomized clinical trials. Meta-analysis was performed andthe power of the meta-analytic findings determined by trial sequential analysis(TSA).Results: Eleven studies met the inclusion criteria and were included in the review.Results of this systematic review revealed a small higher rate (2%) of early implantfailure when a non-submerged healing approach is performed. Late implant failureappears not to be different in submerged or non-submerged healing, but the powerof evidence, as determined by TSA, is not high. If we consider MBL changes at 1 yearfrom implant load, it seems that non-submerged healing may better preserve marginal bone, although with a small effect size (0.13 mm).Conclusions: Implants placed with a non-submerged technique have a higher risk(2%) of early failure. The power of the evidence about the effects on MBL is low, butpresent results seem to favour non-submerged healing, although with a very smalleffect size.