Digital implant transfer with coded healing abutments: A site-by-scanner accuracy study

To compare the pose-transfer accuracy of single-implant digital impressions obtained with three intraoral scanners across three predefined site-specific coded healing-abutment configurations.

A partially edentulous maxillary model containing molar, incisor, and premolar implant sites was equipped with 2-mm coded healing abutments (iPhysio, LYRA ETK). A metrological reference was generated with an industrial structured-light scanner. Twenty independently initiated acquisition sessions were performed with each of three IOSs on the same master model: Medit i900, 3Shape TRIOS 6, and Dentsply Sirona Primescan 2. Coded components were replaced with their corresponding CAD-library geometries, generating 60 datasets and 180 site-level observations. Metrological evaluation included translational magnitude, angular deviation, signed mean surface distance, and within-surface distance dispersion. Generalized estimating equations with robust covariance were used to test IOS, site-specific CHA condition, and IOS × condition effects (α=0.05).

The site-specific CHA condition significantly affected translational magnitude (Wald χ²(2)=183.88, p<0.001) and angular deviation (Wald χ²(2)=56.12, p<0.001). Significant IOS × condition interactions were identified for translational magnitude (Wald χ²(4)=120.12, p<0.001), angular deviation (Wald χ²(4)=80.15, p<0.001), signed mean surface distance (Wald χ²(4)=35.87, p<0.001), and within-surface dispersion (Wald χ²(4)=46.55, p<0.001). The marginal IOS effect was not significant for translational magnitude (p=0.427) or angular deviation (p=0.112). Condition-specific comparisons showed different performance patterns among scanners, with no IOS demonstrating uniformly lower pose-transfer deviation across all tested conditions.

Within this in vitro model, coded healing-abutment transfer accuracy differed among the tested site-specific CHA configurations and showed significant IOS × experimental-condition interactions. Because anatomical location and CHA geometry were not independently varied, these effects cannot be attributed exclusively to implant site.

Performance differed across specific scanner-site-component combinations; clinicians should therefore avoid assuming uniform scanner performance across coded healing-abutment configurations until these in vitro findings are clinically validated.