A total of 10 qualified clinicians with a minimum of 3 years of general practice experience were invited to participate in the study. The number of participants was similar to previously published studies [12, 19], and was confirmed by sample size calculation. A mean horizontal deviation of 1 mm and an expected standard deviation of 0.75 mm that were reported from earlier studies [13, 19] were used for the calculation. With the assumption of 80% statistical power and a 5% significance level, at least nine clinicians were needed to participate in the study. The clinicians were new to implant dentistry and were undertaking formal implant training at Melbourne Dental School, Melbourne University. As part of their training, they had covered the principles of restorative and surgical implant treatment. Ethics approval was obtained from the University of Melbourne Human Research Ethics Committee (1851406.1).

The different phases of the experiment were summarized in Fig. 1. The maxillary Nissin training model (Nissin Dental Products Inc., Kyoto, Japan) was used to simulate clinical patient presentations. An intact Nissin model was scanned by a laboratory surface scanner (Identica T300, Medit Identica, DT Technologies, Davenport, IA) to generate an ideal virtual maxillary arch for implant planning. The right central incisor, the left first molar, and their associated tissue formers were removed; and the sockets were sealed with wax. Subsequently, the modified model was scanned by the laboratory scanner to produce a virtual model with the missing teeth. The virtual model was converted to a surface tessellation language (STL) format that was used to produce physical polyurethane models by a 3D printer (ProJet, 3510 DP Pro, 3D systems, Rock Hill, SC, USA). For each clinician, 3 polyurethane models were fabricated to allow for implant placement according to the different implant placement protocols. 3D printing was implemented to ensure consistency and similar accuracy of all the produced models. To simulate the clinical scenario, the polyurethane models were fixed on mounting plates compatible with the maxillary compartment of manikin heads. In addition, a standard mandibular Nissin model was attached to the lower compartment of the manikin heads.