Comparison of digital and silicone impressions for single-tooth implants and two- and three-unit implants for a free-end edentulous saddle

Abstract Background The use of intraoral scanners (IOS) has facilitated the use of digital workflows for the fabrication of implant-supported prostheses not only for single missing teeth, but also for multiple missing teeth. However, the clinical application of IOS and computer-aided design/manufacturing (CAD/CAM) in implant-supported prosthodontics remains unclear. This study aimed to compare the accuracy of digital and silicone impressions for single-tooth implants for bounded edentulous spaces and two-unit and three-unit implant-supported fixed dental prostheses for free-end edentulous spaces. Methods This study enrolled 30 patients (n = 10 for each of the three groups) with an average age of 61.9 years. Conventional silicone-based and digital IOS-based impressions were made for all patients, and the implant superstructures were fabricated. We measured the scan-body misfit and compared the accuracy of the impressions for single-unit, two-unit, and three-unit implant prostheses with a bounded edentulous space by superimposing the standard triangulated language (STL) data obtained from IOS over the STL data of the plaster model used for final prosthesis fabrication. The scan bodies of the superimposed single-molar implant, two-unit implant prosthesis without teeth on the mesial side, two-unit implant prosthesis without teeth on the distal side, three-unit implant prosthesis without teeth on the mesial side, and three-unit implant prosthesis without teeth on the distal side were designated as A, B1, B2, C1, and C2, respectively. The misfit for each scan body was calculated and the accuracies were compared using the Tukey–Kramer method. Results The average scan-body misfit for conditions A, B1, B2, C1, and C2 was 40.5 ± 18.9, 45.4 ± 13.4, 56.5 ± 9.6, 50.7 ± 14.9, and 80.3 ± 12.4 μm, respectively. Significant differences were observed between the accuracies of A and B2, A and C2, and C1 and C2 (P < 0.001). Conclusions IOS and CAD/CAM can find clinical applications for implant-supported prostheses of up to three units for a bounded edentulous saddle. The use of IOS could render implant treatment easier, benefiting both the surgeons and patients. Prosthesis maladjustment may lead to peri-implantitis and prosthetic fracture. Therefore, further validation of the accuracy of IOS impressions is required in patients with multiple missing teeth in long-span implant prostheses.