Diagnostic accuracy of maxillofacial trauma two-dimensional and three-dimensional computed tomographic scans: comparison of oral surgeons, head and neck surgeons, plastic surgeons, and neuroradiologists.

[BACKGROUND] The authors' objectives were to study differences in diagnostic accuracy between two- and three-dimensional computed tomographic scans and among the specialties of plastic surgery, head and neck surgery, oral surgery, and neuroradiology, since this had not previously been done.

[METHODS] Four groups of subspecialists completed time-proctored tests of 20 maxillofacial trauma scans with zygomatic arch, zygomatic complex, orbital, Le Fort I, II, III, mandibular and panfacial fractures from five institutions (n = 40). Accuracy of diagnosis and indication for surgery, efficiency, and preference were assessed. Comparison between two- and three-dimensional scans, between expert (experienced attending) versus novice (resident/fellow), and among the four subspecialties was performed.

[RESULTS] For two- and three-dimensional scans, two-dimensional was more accurate for orbital floor/medial wall (40 percent and 34 percent) and frontal sinus (26 percent for diagnostic) fractures. Two-dimensional examinations took 2.3 times longer but were preferred (85 percent). Experts and novices had similar accuracy with three-dimensional scanning, but experts were more accurate with the two-dimensional scanning. Experts were 3.3 times faster with two-dimensional scanning but not with three-dimensional scanning. Accuracy of diagnosis among subspecialists was similar, except that oral surgery was less accurate with orbitozygomatic fractures (79 percent versus 90 to 92 percent); neuroradiology was less accurate with indications for surgery (65 percent versus 87 to 93 percent).

[CONCLUSIONS] Differences in diagnostic accuracy exist between two- and three-dimensional maxillofacial scans and between expert and novice readers but not between subspecialties. Combined modalities are preferred.