ABSTRACT Objective The following study aims to highlight the variables that occur during implant surgery which may contribute to peri-implant disease.

Material and methods This retrospective study was first approved by the University of Pennsylvania IRB (#854482). Data was collected from the electronic health record from a total of 309 patients and 1004 implants placed between 2010 and 2023 who were pre-screened as a part of the Arestin study IRB (#815821) and later approved for this study. Data was extracted from clinical notes, radiographs, and periodontal records, with 96 variables assessed per implant using Microsoft Excel. Three researchers conducted a partially-blinded review, and statistical analysis was performed by an investigator using JMP Pro 17.7. Linear regression, ANOVA, Chi-Squared and T-testing were used, with significance set at p < 0.05. Surgical variables analyzed included depth of implant placement, mesio-distal spacing to adjacent teeth or implants, insertion torque, and surgical approach (single vs. two-stage). Additional implant-related and restorative data was recorded, but this study focuses exclusively on surgical factors.

Results Peri-implant disease was present in two-thirds of implants, with one in seven meeting criteria for peri-implantitis. Subcrestal implant placement showed a weak positive correlation with increased probing depths but a negative trend with marginal bone loss, suggesting potential bone preservation. Greater inter-implant distances were associated with reduced distal bone loss, supporting the 3 mm spacing guideline. In contrast, increased distance from adjacent teeth showed a weak association with greater bone loss. A weak positive correlation between insertion torque and bone loss percentage was seen although the results were widely scattered.

Conclusion This retrospective study suggests that subcrestal implant placement and increased inter-implant distance may support crestal bone preservation. Though probing depths may increase with subcrestal placement. The broad scatter of torque values suggests that torque alone has limited predictive power for bone loss, with implants of comparable values exhibiting markedly different bone loss percentages. The findings underscore the importance of surgical planning and the multifactorial nature of peri-implant disease.