Purpose: The goal of successful implant placement is to maintain a long-term stable marginal peri-implant bone level (MBL). This retrospective study investigated the MBL of reduceddiameter implants (RDI). The implants were analyzed in different indication in the context of various surgical and prosthetic treatment strategies using heterogeneous data from a private practice. Mixed model analysis was used to process the heterogeneous data structure. This method has not been a standard method of data analysis in implantology, and it shows considerable advantages over conventional variance analyses regarding data integration from outpatient practices. Method: 123 patients were treated with 326 implants. 247 implants were diameter-reduced, the remaining 79 implants were standard implants (SDI) as patient related controls. The mean observation time was 24.4 months, and the maximum observation time 76.0 months. The peri-implant bone level of the implants was evaluated, while considering the diameter, as well as the time of implant placement, time of loading, extent of augmentation and localization of the implants. The data were evaluated after restructuring using mixed model analysis. Result: No significant difference was found between the use of RDI or SDI in the analyzed indication. Furthermore, no significant difference was found for the implant placement time, loading time, and the use of two-stage augmentations regarding the stability of the peri-implant bone level. Conclusion: Narrow-diameter implants are a sufficient treatment option in horizontally deficient bone conditions. The use of diameterreduced implants in the posterior region shows promising results; 3.5mm diameter implants may be indicated considering the individual patient situation. The use of mixed model analysis for the evaluation of heterogeneous practice data can lead to a significant increase in the number of retrospective studies and data integration from practices, forming a sound basis for evidence-based dentistry.