RESUMO
The aim was to investigate the impact of cone-beam computed tomography (CBCT) on implant planning and on prediction of final implant size. Consecutive patients referred for implant treatment were submitted to clinical examination, panoramic (PAN) radiography and a CBCT exam. Initial planning of implant length and width was assessed based on clinical and PAN exams, and final planning, on CBCT exam to complement diagnosis. The actual dimensions of the implants placed during surgery were compared with those obtained during initial and final planning, using the McNemmar test (p < 0.05). The final sample comprised 95 implants in 27 patients, distributed over the maxilla and mandible. Agreement in implant length was 50.5% between initial and final planning, and correct prediction of the actual implant length was 40.0% and 69.5%, using PAN and CBCT exams, respectively. Agreement in implant width assessment ranged from 69.5% to 73.7%. A paired comparison of the frequency of changes between initial or final planning and implant placement (McNemmar test) showed greater frequency of changes in initial planning for implant length (p < 0.001), but not for implant width (p = 0.850). The frequency of changes was not influenced by implant location at any stage of implant planning (chi-square test, p > 0.05). It was concluded that CBCT improves the ability of predicting the actual implant length and reduces inaccuracy in surgical dental implant planning.