Abstract
Purpose: :
To investigate the inter-observer variability and accuracy of Hertel exophthalmometry and analyze whether it is a learned clinical skill improved through repetition and experience.
Methods: :
In this comparative study, simultaneous bilateral Hertel exophthalmometry was performed on four different patients in consecutive fashion by 36 ophthalmologists stratified by training level experience (residents, oculoplastics fellows, and neuro-ophthalmology and oculoplastics attendings). The patients had no history of orbital disease and were of different genders and ethnic backgrounds. Each observer was blinded and three measurements were taken for each patient using a Marco Prism Exophthalmometer (Marco Instrument Co, Jacksonville, Florida, USA). Additionally, each resident completed a brief survey evaluating their formal education and exposure to exophthalmometry.
Results: :
The PGY-2 residents demonstrated the greatest variation for globe position and intercanthal distance, respectively (PGY-2: 6mm and 20mm; PGY-3: 5mm and 13mm; PGY-4: 3.5mm and 8mm). The subspecialty attendings (gold standard) showed the least variation for globe position at 2mm. There was no statistical difference for accuracy of measurements for globe position between each level of training (ANOVA analysis p = 0.336). The difference in accuracy for intercanthal distance between the PGY-2 residents and all other groups was statistically significant (PGY-3: p = 0.032; PGY-4: p = 0.008; Attg: p = 0.011). The difference in accuracies for intercanthal distance between PGY-3 and PGY-4 and each with the attendings were not statistically significant. Finally, 59% of residents had formal clinical training, 6% had didactic training and 76% felt it was a useful clinical skill.
Keywords: neuro-ophthalmology: diagnosis • orbit