Inter-observer variability between radiologists reporting on cerebellopontine angle tumours on magnetic resonance imaging

Abstract

Abstract Background: Studies demonstrate the significance of intra- and inter-observer variability when measuring cerebellopontine angle tumours on magnetic resonance imaging, with measured differences as high as 2 mm. Objective: To determine intra- and inter-observer measurement variability of cerebellopontine angle tumours in a specialised institution. Methods: The magnetic resonance imaging maximal diameter of 12 randomly selected cerebellopontine angle tumours were independently measured by 4 neuroradiologists at a tertiary referral centre using a standard definition for maximal tumour diameter. Average deviation and intraclass correlation were subsequently calculated. Results: Inter-observer difference averaged 0.33 ± 0.04 mm (range, 0.0–0.8 mm). Intra-observer measurements were more consistent than inter-observer measurements, with differences averaging 0.17 mm (95 per cent confidence interval = 0.27–0.06, p = 0.002). Inter-observer reliability was 0.99 (95 per cent confidence interval = 0.97–0.99), suggesting high reliability between the readings. Conclusion: The use of a standard definition for maximal tumour volume provided high reliability amongst radiologists' readings. To avoid oversizing tumours, it is recommended that conservative monitoring be conducted by the same institution with thin slice magnetic resonance imaging scans.