Comparison of Glycine Air Polishing and a Rotary Titanium Brush for the Removal of a Surrogate Surface Marker from Implants in Class Ib and IIIb Peri-Implant Defects: An In Vitro Study

Background/Objectives: Debridement of the implant surface is the critical point in the treatment of peri-implantitis, and there is no gold-standard instrument available for that purpose. The aim of the present study was to determine whether the difference in efficacy between glycine air polishing and a rotary titanium brush for the removal of a surface marker depends on peri-implant defect type (class Ib vs. class IIIb), to compare the overall efficacy of the two instruments, and to determine whether titanium brush performance varied with repeated usage. Methods: Forty implants (3.8 × 11.5 mm, double acid-etched) were stained with permanent blue ink as a surrogate for contamination and inserted in 3D-printed resin blocks mimicking Monje class Ib and class IIIb peri-implant bone defects, with 5 mm of the implant surface remaining visible. Each group included ten implants according to the defect type and instrument used (brush or air-polishing); within each brush group, five brushes were each used on one implant when new and on a second implant after sterilization. The implants were instrumented for 30 s. Only the frontal aspect of each implant was imaged; standardized photographs taken before and after the procedure were analyzed by image segmentation using hue–saturation–value color space, and the amount of stain removed was analyzed through two-way analysis of variance. Results: Neither main effect reached statistical significance (defect p = 0.063; instrument p = 0.052), but the interaction defect × instrument did (F(1,36) = 7.01, p = 0.012; partial eta squared 0.163, 95% confidence interval 0.009 to 0.379). Air polishing removed more marker in class IIIb defects (56.34%, standard deviation [SD] 19.43) than in class Ib defects (31.05%, SD 16.49; p = 0.016), and more than the brush within class IIIb defects (30.52%, SD 20.71; p = 0.013); none of these pairwise comparisons remained significant under Bonferroni-corrected non-parametric testing. No significant difference was detected for the brush between defect types (p = 1.00) or between first and second use (paired p = 0.364), although the design could detect only large differences. No group achieved complete removal. Conclusions: In this surrogate-marker model, the relative performance of the two instruments depended on defect morphology. The interaction rests on the combination of a higher value for air polishing and a lower, more variable one for the brush in class IIIb defects, and it is not robust to the omission of the three lowest brush values. Ink removal is not equivalent to biofilm decontamination, and a water-based marker may favor the instrument that delivers more fluid; these findings therefore need verification in a biofilm model before they can inform instrument selection.