Abstract
Purpose::
To measure the operator and inter-visit variability of retinal nerve fiber layer (RNFL) thickness in Stratus Optical Coherence Tomography (StratusOCT).
Methods::
Three operators (‘Expert’, ‘Moderate’, and ‘Novice’) scanned both eyes of fourteen (14) healthy subjects using the StratusOCT version 4.0 (Carl Zeiss Meditec, Inc., Dublin, CA) Fast RNFL (3.4 mm) scan. Subjects were imaged three times by each operator during each of three visits on different days within one week. Subjects were asked to sit back between scans. For each subject, one eye was chosen at random for analysis. Average, Superior, Inferior, Temporal and Nasal RNFL thickness measurements were analyzed using analysis of variance (ANOVA) with interactions of the three components: Subject, Operator, and Visit.
Results::
The interaction between subject, operator and visit was significant for Average, Superior and Nasal RNFL thickness (variance = 1.8, 5.9 and 8.1 µm2, respectively; P < 0.05, ANOVA). The effect of the interaction term was found to increase the test-retest variability (twice the standard deviation) by 0.64 µm, 0.84 µm, and 1.2 µm.
Conclusions::
The significant interaction term probably occurs because some subjects are more difficult to image than others, and this varies from day to day and affects operators differently. Nonetheless, our findings confirm previous studies that operator and inter-visit variability is small for StratusOCT scans [1-3]. Although variance increases when multiple operators are used, the increase is small and not clinically significant. Therefore, different operators can successfully perform follow-up scans on a patient.References:[1] Budenz et al, ARVO 2006 presentation 3368[2] Paunescu et al, IOVS 2004 Vol 45 (6), p 1716[3] Guaqueta et al, ARVO 2006 presentation 3640
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)