Interrater reliability of emergency medicine residents in measurement of optic nerve sheath diameter with computed tomography

Introduction It is known that optic nerve sheath diameter enlargement relates with increased intracranial pressure; also worsened clinical outcome. Due to differences between measurements among raters, recently there are some studies which evaluates interrater reliability with ultrasonographical techniques. We were unable to find a study which aims interrater reliability as primary outcome in measurement of optic nerve sheath diameter with computed brain tomography in Medline search. In this study our primary aim is evaluating interrater reliability in measurement of optic nerve sheath diameter with computed brain tomography, our secondary aim is determining the factors that affects interrater reliability and define reliability to the measurement of faculty of radiology which is gold standard. Materials and Methods Measurements of 10 enlarged optic nerve sheath diameter (≥5.5 mm) and 10 normal optic nerve sheath diameter (<5.5 mm) on brain computed tomography had been performed by one faculty of radiology and 12 residents of emergency medicine (six seniors, six juniors) from axial and sagittal planes of the left eye, behind 3 mm of optic disc. Interrater reliability of measurements of post graduate residents and their accuracy to the gold standard had been defined. measurements were made either sagittal or axial computed tomography plain, with normal and enlarged optic nerve sheath diameter. Interrater reliability is calculated using intraclass coefficients correlations (ICC), reliability on categorical variable is analyzed with kappa (ĸ). Mean age of post-grad residents and relevance of measurements to normal distribution had been defined by Kolmogorov-Smirnov test. The data that relates to normal distribution are shown by mean, and unrelated data are shown by median and IQR. All analysis were made in 95% confidence interval and p value <0.05 is statistically significant. Results We found that the interrater reliability is 'good' (ICC 0.798; 95 CI 0.777-0.818) among emergency medicine physicians in optic nerve sheath diameter measurements. Interrater reliability is 'good' (ICC 0.824; 95% CI 0.781-0.858) among senior physicians and 'moderate' (ICC 0.748; 95% CI 0.687-0.797) among junior residents. Interrater reliability among residents is; 0.812 (95% CI 0.7833-0838) on axial plane, 0.783 (95% CI 0.750-0.813) on sagittal plane. Interrater reliability between faculty of radiology and residents was found 'good' (ICC 0.891; 95% CI 0.865-0.911), accuracy is 'good' (ĸ:0.688; p<0.001). Conclusion Optic nerve sheath diameter measurements of post-grad residents on computed brain tomography highly reliable in itself, and they are well compatible with the gold standard measurements. Emergency medicine physicians can reliably measure optic nerve sheath diamter on computed brain tomography images.