Clinical and microbiological predictors of peri-implantitis risk in patients with chronic periodontitis

Relevance. Dental implantation is currently regarded as the most effective approach to restoring the continuity of the dentition. However, individuals with a history of chronic periodontitis face a significantly increased risk of inflammatory complications, particularly peri-implant mucositis and peri-implantitis. Developing effective prevention and treatment strategies requires accurate, timely risk assessment to identify at-risk patients. Objective. To develop a risk classification model for peri-implantitis in patients with chronic periodontitis based on clinical and microbiological markers, using a decision tree algorithm. Materials and methods. The study included 177 patients with chronic periodontitis, evenly divided into three groups (n = 59 each): without dental implants, with implants and no clinical signs of peri-implantitis, and with implants and diagnosed peri-implantitis. Clinical examination, microbiological analysis, and PCR testing of subgingival plaque from periodontal pockets were conducted to identify clinical and microbiological markers associated with peri-implantitis risk. Based on these parameters, a machine learning model was developed using the CART (Classification and Regression Tree) algorithm with the Gini index as the splitting criterion. Results. The most prominent clinical signs observed in patients with peri-implantitis were dentin hypersensitivity, halitosis, exudate, and gingival margin overgrowth. Among the identified microorganisms, the most predictive for peri-implantitis risk were R. mucilaginosa , S. mitis , R. dentocariosa , A. odontolyticus , S. australis , P. gingivalis , F. nucleatum and A. actinomycetemcomitans . The classification model demonstrated high discriminative ability, with an accuracy of 0.75, F1–score of 0.7243, and a ROC–AUC of 0.74. Conclusion. A comprehensive clinical and microbiological evaluation, supported by statistical analysis and machine learning methods, provides an effective approach for predicting the risk of peri-implantitis in patients with chronic periodontitis. The findings highlight the importance of integrating microbiological diagnostics into standard clinical protocols to enable early identification of high-risk individuals and to guide personalized preventive and therapeutic strategies.