Purpose: To investigate the effect of implant angulation, non-free-end partial edentulism, and number of scan bodies on the accuracy of digital impressions of multiple implants in partially edentulous arches.

Materials and methods: Four reference models of partially edentulous mandibles with implants (RM1, RM2, RM3, and RM4) representing different intraoral situations were each scanned 10 times by an intraoral scanner. Reference scans were obtained by a laboratory scanner. Test scans were compared with reference scans to obtain the distance deviations (Δd) and angular deviations (Δθ) between scan bodies for trueness assessment. Differences among the repeated test scans of each model were measured and recorded as Δdp and Δθp for precision assessment. The Student t test (α = .05) was used to compare Δd, Δθ, Δdp, and Δθp of different reference models, including RM2 vs RM1 (effect of non-free-end partial edentulism), RM3 vs RM1 (effect of implant angulation), and RM4 vs RM1 (effect of number of scan bodies).

Results: The implant with 17-degree angulation in RM3 showed significantly lower Δd, Δθ, and Δθp compared with the parallel implant in RM1 (Δd: P = .0382, Δθ: P = .0267, Δθp: P = .0417). The RM2 of non-free-end partial edentulism had lower distance and angular deviations than RM1, but without a significant difference. The number of scan bodies had no significant effect on the Δd, Δθ, Δdp, and Δθp of RM4 and RM1.

Conclusion: Angulated implants showed better accuracy of digital impressions in partially edentulous arches compared with parallel implants. Non-free-end partial edentulism was attributed to improved accuracy, while the number of scan bodies showed no effect.