Diagnostic Accuracy and Reliability of Visual Assessment of Peri- Implant Keratinized Mucosa Width and Mucosal...

This cross-sectional study aimed to evaluate the diagnostic accuracy, reliability, and reproducibility of visual digital image-based assessments of peri-implant keratinized mucosa width (KMW) and mucosal thickness (MT) compared with direct clinical measurements. Calibrated evaluators assessed KMW and MT, standardized intraoral photographs, and intraoral scans (IOS) obtained from adult patients with non-molar implant-supported prostheses. Diagnostic accuracy was determined using sensitivity, specificity, predictive values, diagnostic odds ratios (DOR), and receiver operating characteristic (ROC) curve analysis. Inter- and intra-observer agreement were assessed using Fleiss' kappa, weighted kappa, and intraclass correlation coefficients (ICC). Based on sample size calculation, a total of 60 patients with 60 non-molar implant-supported prostheses were included. Twenty calibrated evaluators with different professional backgrounds assessed 140 digital images, with 20 images repeated. This resulted in 2800 KMW and 2800 MT evaluations. Overall, a limited ability to distinguish inadequate from adequate KMW and thin from thick MT was shown when assessing digital images. Photographs tended to be associated with slightly higher sensitivity and specificity than IOS, with overall sensitivity ranging from 31.3% to 56.9% and specificity from 45.6% to 63.7% across modalities. ROC analysis confirmed poor discriminatory capacity for both imaging modalities, with area under the curve values ranging from 0.427 to 0.633. Inter-observer agreement was generally fair, whereas intra-observer reproducibility remained poor across evaluator groups and imaging modalities. In conclusion, visual assessment of peri-implant KMW and MT based on digital imaging demonstrated limited diagnostic accuracy and reproducibility compared with direct clinical measurements.