The primary stability of the implant was assessed during surgery. Implant success and survival were evaluated in the group of implants restored with abutments [5, 29] at both placements of the provisional and definitive prostheses and at each follow-up visit thereafter. Implants were deemed successful in accordance with the criteria for implant success laid down by Albrektsson et al. [30]. Implants were successful if there was less than 0.2 mm bone loss annually after the first year of loading, if they were clinically immobile, if there was no peri-implant radiolucency, and if there was no persistent and/or irreversible pain, infection, neuropathies, or paresthesia. During the course of the study, the criterion of bone loss by Albrektsson et al. was scrutinized, and the scientific relevance was not considered to be suitable anymore [30, 31]; therefore, implant success was assessed post hoc, according to Buser et al. [29], that is, there was no persistent and/or irreversible signs or symptoms such as pain, infection, neuropathies, or paresthesia, no peri-implant infection with suppuration, no mobility, and no continuous radiolucency around the implant. Radiological evaluation for radiolucency and bone loss was measurable only with available evaluable radiographs; radiographs for some patients were missing or not evaluable. In the case that no complications were reported by the clinician, and the patients reported being satisfied according to the set criteria, then radiographs were not necessary and the implant was deemed successful.

The patients rated their satisfaction regarding the ability to chew, to taste, their comfort, appearance and fit of restoration, and general satisfaction on a categorical scale (very unsatisfied, unsatisfied, middle, satisfied, very satisfied) via a questionnaire at each visit beginning from loading [17, 22].

Safety

Adverse events were recorded on an adverse event form and reported as non-treatment associated or treatment-associated events.

Statistical methods

A minimum of 200 patients were planned to be included in the study. Analyses were performed on the per protocol population. In addition, to assess the correlation of implant success with anatomical and surgical parameters, analyses of the subgroups “platform matching” and “platform switching” (based on abutment type) were performed. Implant success and survival rates were calculated using a life table analysis 1 year after baseline and yearly thereafter. To test for significant differences for repeated measurements, the Wilcoxon signed-rank test was used, and to assess for significant differences between the subgroups, the Kruskal-Wallis test was used. p values of less than p < 0.05 were deemed significant. Changes in the crestal bone were quantitatively evaluated through the standardized measurements of the radiographs. Any non-standardized radiographs allowed for qualitative analysis only. Standardized measurements on radiographs for calculating bone level changes were done with the freely available software ImageJ 1.50i (https://imagej.nih.gov/ij/). Descriptive statistics were performed with IBM SPSS Statistics for Windows V24.0 (IBM Corp., Armonk, NY, USA).