Effectiveness of Biomarker-Guided Implant Risk (BGIR) vs. IDRA and PiRA Scores in Predicting the Onset of Peri-Implant Diseases: A Long-Term Retrospective Study

Objectives: The long-term prognostic value of the biomarker-guided implant risk (BGIR) score was compared with two other risk models, the IDRA (implant disease risk assessment) and the PiRA (peri-implantitis risk assessment), in predicting the onset of implant diseases in patients attending a private practice and enrolled in a supportive care program (SCP). Methods: From a pool of private implant patients subjected to at least 10 years of SCP, 80 partially edentulous subjects were recruited and retrospectively evaluated. The BGIR, the IDRA and the PiRA scores were calculated during SCP and compared to each other. Prognostic scores of BGIR, IDRA and PiRA were estimated based on the area under the curve for receiver operating characteristic (ROC), including 95% confidence intervals. High-risk BGIR, IDRA and PiRA subgroups versus low/moderate BGIR, IDRA and PiRA subgroups were compared at the patient level using Cox regression analyses. Results: The mean implant follow-up was 10.5 years. Peri-implantitis incidence was 17.5%. The BGIR, IDRA and PiRA high-risk percentages were 27.5%,21.2% and 25%, respectively. The areas under the ROC curve for BGIR, IDRA and PiRA were 0.836, 0.758, and 0.769, respectively, for peri-implantitis events. BGIR, IDRA and PiRA high-risk scores were found to be significantly associated with the onset of peri-implantitis, with hazard ratios (HR) of 4.7, 3.6 and 3.8, respectively. Compared to IDRA and PiRA, BGIR showed higher specificity and sensitivity. However, comparisons between the area under the, curves (AUC) of the three independent ROC curves showed no statistical significance. Conclusions: Compared to IDRA and PiRA, calculation of the BGIR score during follow-up seems to have a greater ability to predict the onset of peri-implant diseases in daily clinical practice.