However, there are also a few clinical studies in which reduced survival rates for implants inserted in grafted areas were found. Differences in numbers of implants analysed, surgical techniques, indications and/or graft materials may account for these inconsistent results and further studies might be needed.

In the retrospective analysis shown here, the comparison of different augmentation procedures using the log-rank test revealed the highest implant survival for bone condensing followed by lateral ridge augmentation. The lowest rankings were found for sinus floor augmentation and no augmentation. In pairwise comparisons of Kaplan-Meier implant survival curves to non-augmented sites, a significantly higher implant survival was found for lateral bone augmentation and sinus floor elevation. However, all procedures provided a high implant survival of more than 94%. This indicates that under daily practice all these augmentation procedures may provide clinically acceptable results. Recent reviews have also reported a high implant survival of more than 90% for sinus augmentation, lateral ridge augmentation, for bone splitting as well as for three-dimensional augmentations using titanium mesh. When augmentation procedures were compared to each other, the authors were not able to draw a clear conclusion on the superiority of a certain augmentation procedure or grafting protocol.

In our analysis, membranes were used in 36.6% of the implant sites. Small defects were treated with either a bone substitute or bone particles without an additional membrane. The membrane which was used in almost 75% of the cases was a native collagen membrane. Various other studies have reported successful results using this membrane in bone augmentation as well as a low complication rate. Although in our analysis, the membrane was associated with a high absolute survival rate of 96.24% and the best result in the log-rank test, it is still not possible to draw clear conclusions on the superiority of any membrane, when implant survival is the only parameter under consideration.