Agreement Between Manual and Digital Mesiodistal Tooth Measurements on Digitized Plaster Models: A Tooth-Level Analysis.

Background: Reliable mesiodistal tooth-width measurements are essential for orthodontic diagnosis and treatment planning. Digital models obtained by scanning plaster casts may facilitate orthodontic workflows; however, the extent of agreement between manual caliper measurements and digital measurements on digitized plaster models remains clinically relevant, particularly at the individual-tooth level. Objective: To evaluate the agreement, repeatability, and inter-method differences between manual and digital mesiodistal tooth-width measurements from the canine to the first molar on digitized plaster models. Methods: Plaster models from 30 orthodontic patients were analyzed. The models were digitized using a 3Shape TRIOS 4 intraoral scanner. Mesiodistal widths of teeth 13-16, 23-26, 33-36, and 43-46 were measured using digital calipers on plaster models and using 3Shape Unite platform version 1.7.33.2 (3Shape A/S, Copenhagen, Denmark) on the corresponding digital models. Manual and digital measurements were each performed twice at 24 h intervals. For both methods, the mean of the two repeated measurements was used for inter-method comparison. Manual and digital intra-observer repeatability was assessed using the intraclass correlation coefficient (ICC). Inter-method differences (ΔMD = mean manual - digital) were analyzed using Bland-Altman analysis. Differences exceeding 0.3 mm were considered to exceed the predefined threshold of potential clinical relevance. Results: A total of 480 teeth were analyzed. Manual and digital intra-observer repeatability were excellent (ICC = 0.985 and ICC = 0.976, respectively). Manual measurements were on average 0.150 mm higher than digital measurements (SD = 0.407 mm; median = 0.145 mm; IQR = 0.29 mm), with 95% limits of agreement ranging from -0.647 mm to 0.947 mm. At the single-tooth level, 212 out of 480 measurements (44.2%) exceeded the predefined 0.3 mm threshold. First molars and selected premolars showed the highest discrepancy rates, whereas canines generally showed lower discrepancy rates. Conclusions: Manual and digital mesiodistal tooth-width measurements performed on digitized plaster models showed good agreement under the tested conditions. However, individual tooth-level differences exceeding the predefined 0.3 mm threshold were frequent, particularly for posterior teeth. Therefore, digital measurements obtained from digitized plaster models should be interpreted with caution in clinical situations requiring high precision at the level of individual tooth dimensions.