Proptosis—Correlation and Agreement between Hertel Exophthalmometry and Computed Tomography

To determine the correlation and agreement between Hertel exophthalmometry and computed tomography (CT) of the orbits in measuring proptosis.80 patients (40 normal and 40 with proptosis) were recruited for this study. Exophthalmometry were performed on all of them using the Hertel exophthalmometer and CT. Values and correlations between the modalities were analysed with unpaired t-tests, intraclass correlation coefficients (ICC), and Pearson correlation coefficients (PCC). The Bland-Altman method was used to analyse the agreement between the two modalities.Hertel exophthalmometer and CT measurements did not differ significantly (p > 0.05), although exophthalmometry measurements of the normal (14.5 ± 2.2 mm) and proptosis groups (20.5 ± 3.9 mm) were higher than those obtained from CT (13.9 ± 2.4 mm and 20.0 ± 3.7 mm, respectively). ICC for both the Hertel exophthalmometer and CT measurements were both 0.99, indicating high intra-observer reliability and reproducibility. PCC between Hertel exophthalmometer and CT measurements in both normal and proptosis groups were strongly correlated (r = 0.96 and 0.93, respectively, p = 0.01). The 95% limits of agreement (LOA) between Hertel exophthalmometer and CT measurements for the normal and proptosis groups were -0.70 to 1.78 mm and -2.36 to 3.33 mm, respectively.Although Hertel exophthalmometer and CT measurements are similar and strongly correlated, they do not agree well with each other in the presence of proptosis. The measurements should not be used interchangeably as the differences between them may lead to errors in clinical interpretation.