RESUMO
BACKGROUND: The diagnosis of chronic sinusitis is dependent on the radiographic evidence of sinus disease. METHODS: We evaluated the performance of radiographs and computed tomographic (CT) scans for the examination of the paranasal sinuses of 91 patients of both sexes, ranging in age from 2 to 17 years, who had chronic upper respiratory tract symptoms for at least 3 months. The CT scan findings were categorized as no disease; minimal disease, and mild, moderate, and severe sinusitis. RESULTS: Fifty-eight patients (63%) had chronic sinusitis: CT scan abnormalities were minimal in 17%, mild in 19%, moderate in 21%, and severe in 43%. There was a statistically significant correlation between rhinorrhea (r = 0.25, p = 0.01), cough (r = 0.27, p = 0.009), and the severity of sinus abnormality as determined by CT scan. Clinical presentation in the mild, moderate, and severe sinusitis groups (p < 0.05) was significantly different from that of the no disease group, whereas the minimal disease group had subclinical presentation (p = 0.11). Clinically significant chronic sinusitis often occurred at multiple sites: 44% of patients had pansinusitis, 50% had disease involvement of at least two sinuses, and 6% had disease in a single sinus. When sinus radiographs were compared with CT scans (n = 70 cases), radiographs could not identify minimal disease. For clinically significant sinusitis, sinus radiographs detected disease in 1 of 5 (20%) frontal sinuses, 0 of 12 (0%) sphenoidal sinuses, and 17 of 31 (54%) ethmoidal sinuses. With the minimal criteria of 40% to 50% opacification or fluid level filling of the maxillary antrum, radiographs detected disease in 37 of 49 (75%) cases. The sensitivity and specificity for a Waters view to confirm clinically significant chronic sinusitis without specifying the sites and severity were acceptable at 76% and 81%, respectively. When limited sinus CT scans were compared with full CT evaluation (n = 49 cases), limited studies detected 5 of 5 (100%) frontal, 9 of 11 (82%) sphenoidal, 14 of 19 (73%) ethmoidal, and 39 of 40 (97%) cases of maxillary sinusitis. The overall agreement was 88%. CONCLUSIONS: A single Waters view is an acceptable part of the initial evaluation of pediatric chronic sinusitis; however, a limited CT scan is a better alternative.