A Randomized Controlled Clinical Trial Comparing Conventional and Computer-Assisted Implant Planning and Placement in Partially Edentulous Patients. Part 4: Accuracy of Implant Placement

The objective of this study was to compare the accuracy of conventional and computer-assisted implant planning and template-guided placement (CAIPP) protocols. Partially edentulous patients (N = 73) were randomly assigned to either a conventional implant planning and freehand placement protocol (control group, n = 26) or one of two different CAIPP protocols (stereolithographic guide [T1, n = 24] or 3D-printed guide [T2, n = 23]). The virtually planned and final implant positions were compared. Differences between the planned and the obtained implant position were evaluated as horizontal, vertical, and angular measurements. Descriptive statistics were calculated for overall deviation values and their fragmented mesiodistal and bucco-oral vectors at each evaluation plane. To study overall accuracy differences between study groups, analysis of variance (ANOVA) was used with Bonferroni post hoc test (Scheffé method). Possible confounding variables were analyzed using multiple linear regression with respect to treatment group. The mesiodistal or bucco-oral distribution of the positioning errors was evaluated with chi-square test. A multiple linear logistic regression was used to identify confounding variables. Inaccuracy at the level of the occlusal plane of the restoration averaged 0.65 ± 0.26 mm in the control group, 0.59 ± 0.4 mm in T1, and 0.76 ± 0.5 mm in T2. At the implant shoulder level, the inaccuracy amounted to 1.25 ± 0.62 mm, 0.97 ± 0.36 mm, and 0.72 ± 0.31 mm in the control group, T1, and T2, respectively. At the implant apex, mean deviations of 2.32 ± 1.24 mm were recorded in the control group, 0.97 ± 0.57 mm in T1, and 1.08 ± 0.57 mm in T2. Mean discrepancies in vertical direction measured 0.28 ± 1.01 mm (control), 0.2 ± 0.65 mm (T1), and -0.1 mm ± 1.0 mm (T2). Angular deviations of 7.36 ± 3.36 degrees (control), 4.23 ± 2.68 degrees (T1), and 3.13 ± 2.12 degrees (T2) were measured. Statistically significant differences were observed between the conventional and the two CAIPP groups for overall deviations at implant shoulder, apex, and implant angulation. CAIPP protocols seemed to provide a higher accuracy and precision compared to conventional freehand protocols. Still, the amount of inaccuracy using guides demands a safety margin. Moreover, intrasurgical verification during drilling and the implant placement procedure should be performed, including clinical parameters that may not be available from cone beam computed tomography data during the planning phase.