STATEMENT OF PROBLEM: Despite the increasing demand for a digital workflow in the fabrication of indirect restorations, information on the accuracy of the resulting definitive casts is limited.

PURPOSE: The purpose of this in vitro study was to compare the accuracy of definitive casts produced with digital scans and conventional impressions.

MATERIAL AND METHODS: Chamfer preparations were made on the maxillary right canine and second molar of a typodont. Subsequently, 9 conventional impressions were made to produce 9 gypsum casts, and 9 digital impressions were made to produce stereolithography additive (SLA) casts from 2 manufacturers: 9 Dreve SLA casts and 9 Scanbiz SLA casts. All casts were then scanned 9 times with an extraoral scanner to produce the reference data set. Trueness was evaluated by superimposing the data sets obtained by scanning the casts with the reference data set. Precision was evaluated by analyzing the deviations among repeated scans. The root mean square (RMS) and percentage of points aligned within the nominal values (±50 μm) of the 3-dimensional analysis were calculated by the software.

RESULTS: Gypsum had the best alignment (within 50 μm) with the reference data set (median 95.3%, IQR 16.7) and the least RMS (median 25.8 μm, IQR 14.6), followed by Dreve and Scanbiz. Differences in RMS were observed between gypsum and the SLA casts (P<.001). Within 50 μm, gypsum was superior to Scanbiz (P<.001). Gypsum casts exhibited the highest precision, showing the best alignment (within 50 μm) and the least RMS, followed by Scanbiz and Dreve.

CONCLUSIONS: This study found that gypsum casts had higher accuracy than SLA casts. Within 50 μm, gypsum casts were better than Scanbiz SLA casts, while gypsum casts and Dreve SLA casts had similar trueness. Significant differences were found among the investigated SLA casts used in the digital workflow.