Predictive Factors of Dental Implant Failure: A Retrospective Study Using Decision Tree Regression

Area of Science:

• Dental Implantology
• Biomaterials Science
• Oral Surgery

Background:

• Dental implants are widely used for tooth replacement.
• Assessing single-unit implant failure rates and identifying risk factors is crucial for improving long-term success.
• This study focuses on a seven-year period (2015-2021) with a minimum two-year follow-up.

Purpose of the Study:

• To evaluate the failure rate of single-unit dental implants.
• To identify significant risk factors contributing to early and late implant failure using statistical analysis.
• To analyze the duration of implant failure in relation to identified risk factors.

Main Methods:

• A retrospective study of 224 patients who received single-unit dental implants between 2014 and 2021.
• Analysis of clinical records and radiographs to assess implant failures.
• Application of machine learning decision tree regression and Kaplan-Meier survival analyses to identify risk factors.

Main Results:

• Smoking and peri-implantitis were the principal contributors to implant failure (p=0.001).
• Late implant failure (LIF) duration averaged 16.87 months, while early implant failure (EIF) averaged 5.71 months.
• Bruxism, peri-implantitis, diabetes mellitus, and osseointegration complications were significant predictors of failure.

Conclusions:

• Age, sex, surgical procedure type, sinus lift, and grafting were not significantly associated with implant failure.
• Bruxism, peri-implantitis, lack of osseointegration, smoking, and type 2 diabetes mellitus are significant predictors of dental implant failure.
• These findings emphasize the importance of managing modifiable risk factors like smoking and bruxism to enhance dental implant survival rates.