Eligibility criteria

Articles were included in this systematic review if they met the following inclusion criteria: human prospective or retrospective studies, reporting outcomes of implant placed perforating the sinus floor with implant burs, and without regenerative procedure (lateral sinus lift or transalveolar technique) and graft material. The intrusion into the sinus cavity can occur during drilling or implant placement, with and without punch out Schneiderian membrane. Only studies with at least 6 months of follow-up were included in the qualitative assessment.

The following articles were excluded: case report and case series; animal studies; systematic reviews; in vitro studies; human studies using grafting materials, lateral access sinus lift, and transalveolar technique; and follow-up less than 6 months.

Risk of bias

Two reviewers (GMR and BE) designed and assessed the proposal for the present project to make sure that the STROBE statement and PRISMA guideline were followed. STROBE stands for an international, collaborative initiative of epidemiologists, methodologists, statisticians, researchers, and journal editors involved in the conduction and dissemination of observational studies, and consists of a 22-item checklist to be fulfilled in a systematic review.

Qualitative assessment

The quality of the selected randomized controlled trials (RCTs) was established from the randomized clinical trial checklist of the Cochrane Center and CONSORT (Consolidated Standards of Reporting Trials) statement, which provided guidelines for the following parameters: (1) sequence generation, (2) allocation concealment method, (3) masking of the examiner, (4) address of incomplete outcome data, and (5) free of selective outcome reporting. The Newcastle-Ottawa scale (NOS) was used to assess the risk of bias of non-randomized studies. This was performed by two investigators (GMR and BE). Cohen ́s kappa coefficient was used to assess inter-rater agreement.

Statistical analysis

The R 3.0.2 software package was used to perform the meta-analysis. The primary variable was implant survival rate. The secondary variable was the relationship between the degree of penetration and clinical and radiological complication. The analysis was performed using the methodology described below. The pooled weighted mean (WM) and the 95% confidence interval (CI) of each variable were estimated using a computer program (Comprehensive Meta-analysis version 2, Biostat). Random effects meta-analyses of the selected studies were applied to account for potential bias arising from methodology.

Study of heterogeneity

Heterogeneity was assessed based on calculation of the I² statistic (percentage variability of estimated effect that can be attributed to the heterogeneity of the effects) and the null statistic test. Galbraith graphs displayed the degree of heterogeneity. In studies where great heterogeneity was detected, a sensitivity analysis was performed to determine its source. Funnel plots and the Egger test were used to assess risk of bias of the accepted statistical significance level of 5% (p = 0.05).