An initial screening yielded a total of 3551 publications of which 26 potentially relevant articles were selected after an evaluation of their titles and abstracts. Full text of these articles was obtained and evaluated thoroughly. Of these, eight articles [26,27,28,29,30,31,32,33] (Table 2) fulfilled the inclusion criteria and subsequently were included in the qualitative analysis (Fig. 2). Reasons for exclusion are displayed in supplementary (Table 3).

The NOS was used to appraise the quality of included studies for a proper understanding of nonrandomized studies. Because no nonrandomized controlled trials were found in the screening process, the 17 included studies were analyzed with NOS. The level of agreement between the reviewers regarding study inclusion calculated using Cohen kappa statistic interagreement rate of 0.8 was reached. A mean NOS score of 5.1 ± 1.4 was obtained after discussing the disagreements between the examinees (GMR and BE) and third consultant (FL-A).

Eight studies [26,27,28,29,30,31, 33] (Table 2) provide information on the survival rate, which consisted of a global sample of 493 implants, of which a subtotal of 340 reported the degree of penetration (Fig. 3), with a mean follow-up of 52.7 months. The implant survival between the different authors range from 70 to 100%, being the weighted mean survival rate 95.6% with an IC 95% [88.4100] (Fig. 4a). A re-estimation is proposed excluding Brånemark et al. [26] from the meta-analysis, due to the year of publication which is before 2000, implant surface, and the surgical technique that can lead to these low survival rates of 70%. The seven articles [27,28,29,30,31,32,33] of the last 10 years remain in the calculations, being the weighted survival rate of 99.3% with an IC 95% [98.4100]. With regard to the influence of the penetration level, it was categorized into two levels with a cutoff point of 4 mm (Fig. 4b). It was analyzed if there were any differences in survival according to this aspect. For this, six studies contribute with 267 implants [27,28,29,30,31,32,33]. The estimated survival rates were 99.5% CI [98.2100] in implant penetrating ≤ 4 mm and 98.5% CI [96.6100] in implant penetrating > 4 mm (Fig. 3). There were no statistical significant differences in survival according to the degree of penetration (p = 0.403) (Fig. 4b).