Test-retest reliability of computed tomography in the assessment of chronic rhinosinusitis

Abstract

Objective: Computed tomography (CT) of the paranasal sinuses has emerged as the standard test for the assessment of chronic rhinosinusitis, as evidenced by the emergence of several CT-based staging systems. Despite its central role in the diagnosis and treatment planning for chronic rhinosinusitis, the sinus CT represents a 'snapshot in time.' This study was conducted to determine the reliability of the CT scan for chronic sinus disease: are the CT findings in chronic rhinosinusitis stable over time? Methods: A prospective series of patients scheduled for endoscopic sinus surgery was evaluated. A total of 45 patients received two CT scans: an initial scan obtained during routine diagnostic evaluation, and a second scan performed for use as part of an image-guided sinus surgery protocol. No surgical intervention occurred between scans. The patients' scans were staged according to the Lund system by a blinded observer. The correlation between scans for each patient was determined using the matched pairs t test and the Pearson correlation coefficient. Results: The mean time interval between scans was 122.6 days (range, 5-364 d). The average Lund scores for the initial and second scans were 13.56 and 13.27, respectively. The Lund score for 5 patients remained the same, increased in 22 patients, and decreased in 18 patients. Overall, 75.6% of patients' second scans were within ±2 points of the first scan Lund score. The mean change in score between scans of - 0.29 was not significant (P = .606, paired samples t test). The Pearson correlation coefficient between scans was 0.796 (P < .0001). Conclusions: CT scan assessment of chronic rhinosinusitis is a reliable test. The CT findings in patients with chronic rhinosinusitis remain consistent over time.