BACKGROUND: The aim of this study is to evaluate the interobserver agreement, using computed tomography, regarding the local and regional extension of mouth and oropharynx squamous cell carcinoma. METHOD: From 1991 through 2004, thirty patients, with available CT examinations, were examined at the Head and Neck Surgery, Otorhinolaryngology and Imaging Departments of Heliópolis Hospital, São Paulo, Brazil. They were included in this retrospective non-randomized study. We performed biopsies and histopathologic examinations in all cases. There were 25 males and 5 females, with ages ranging from 25 to 77 years. Two radiologists evaluated all cases individually, without previous clinical staging data, analyzing local and regional sites of lesion dissemination. No patients had received any previous treatment up to the moment of CT examination. RESULTS: We calculated the kappa index by assessing agreement between the two observers, and the significance level was assumed 0.05. Agreement was excellent (Kappa: 0.81 - 1.00) in three analyzed subsets (middle line, masticator space and high jugular lymph node chains). It was good (Kappa: 0.61 - 0.80) in ten subsets (floor of the mouth and oral tongue, mandibular branches, retromolar trigone, submandibular space, tonsillar fossa and submentonian, submandibular, middle jugular, low jugular and spinal lymph node chains). It was moderate (Kappa: 0.41 - 0.60) in two subsets (mandibular body and base of the tongue), and in one subset the analysis was not possible using this coefficient where the Kappa index was not feasible. CONCLUSION: The interobserver agreement was of significant statistical value, predominating from good to excellent.
Carcinoma; squamous cell; Mouth neoplasms; Oropharynx; Tomography
