Methods: Seven sectional two-implant (BIOMET-3i 5.0x15.0mm) master models were fabricated with inter-implant angles: Control - 0°; Mesial-Distal (MD) - 10°, 20°, 30°; and Buccal-Lingual (BL) - 10°, 20°, 30°. The master models incorporated gauge blocks that defined the local coordinate reference system. The centroids of implants (Before) or implant analogs (After) were measured with a Coordinate Measuring Machine to compute translational and angular distortions. EN two-piece healing abutments were attached for the EN test-models; direct impression copings were used in the DI test-models; and impressions made with polyether and custom-trays. The EN test-models were sent to BIOMET-3i for implant analog placement. 35 test-models were fabricated per technique (n=5). Two-way ANOVA (p<0.05) for TECHNIQUE and ANGLE was performed.
Results: Global distortion, dR, ranged from 13.0±1.0μm (DIMD20) to 107.0±21.0μm (ENBL30); ABSdXZ angular distortion ranged from 0.050±0.024° (DIMD10) to 1.190±0.197° (ENBL20); and ABSdYZ ranged from 0.046±0.013° (DIMD10) to 0.993±0.332° (ENBL20). The largest negative and largest positive mean dX and dY were from the EN system. EN test-groups with the largest negative and largest positive mean dY were significantly different from all DI test-groups. The largest dR and ABS dXZ and dYZ angular distortions were from the EN system and were significantly larger than all the DI test-groups. TECHNIQUE had significant effect on the dY, dR, as well as ABSdXZ and ABSdYZ. ANGLE had significant effect on dX and dY.
Conclusions: Distortions were found in both EN and DI impression techniques. However, results suggested that the Encode® System was less accurate than the DI technique.