The aim of this retrospective cohort study was to evaluate the impact of intraoperative perforations of the Schneiderian membrane during sinus floor elevation on the stability of the augmented area and its influence on osseointegration after implant insertion. Therefore, we could re-assess a patient cohort of originally 34 patients with 41 perforations and compare their outcome with a control group of patients with sinus floor elevation but without membrane tear, offering a long-term perspective over a range of 1.8 years in the control group to 2.7 years in the study group.

Although the height of the alveolar crest differed significantly, the total postoperative height of augmented and residual bone was on a comparable level allowing for sufficient elevation of the sinus floor to insert dental implants independently of a perforation. This fact remains reassuring as the latter represents the mostly common complication with a reported variation of 10 to 44% [5, 6, 8, 12, 17,18,19,20,21,22,23].

In this context, the surgical experience and the tissue quality, e.g., scarring due to previous procedures or local inflammation, might lead to a high probability of Schneiderian membrane perforation [16, 22, 24, 25] with anatomical variations, such as sinus septa or thin, vulnerable membrane textures [4, 16, 22, 24, 25]. Some studies have suggested contraindication of sinus floor elevation in patients with anatomical variations such as septa or mucosal swelling [8]. As our results did not indicate any negative impact of membrane tear on the augmented sinus, the results of Nolan et al., including a total of 359 sinus floor augmentations, indicated that graft failure occurred in 6.7% of all procedures and was significantly correlated (p = 0.0028) with membrane perforation. Compared to our evaluation, where procedures were, without exception, performed by attending physicians, the perforation rate was twice as high (21 vs. 41%). Surgical experience was ruled out as an influencing factor, as membrane perforations were equally distributed in cases treated by attending physicians compared to those performed by residents [6]. Another reason for the different findings might be due to the differences between the numbers of compared membrane tears (39 perforations vs. 150 perforations), as both evaluations were performed retrospectively.