Purpose
To compare the diagnostic utility of the Nidek and Heidelberg systems for acanthamoeba keratitis.
Methods
The Nidek Confoscan4 and the Heidelberg HRT3 Rostock confocal microscopes were used in a consecutive series of patients suspected of having acanthamoeba keratitis. Images from both microscopes were presented in a random order to two masked graders. Examiners graded each series of images for the presence or absence of acanthamoeba cysts, and also documented whether their evaluation was made with a low or high degree of certainty. Graders discussed discrepancies and agreed on a single grade and level of certainty for all scans.
Results
18 total scans from 13 eyes were evaluated, including 5 eyes that were smear- or culture-positive and had examinations performed on both microscopes, and 8 control eyes that were smear-and/or and culture-negative, but exhibited suspicious keratitis clinically. Five control eyes were tested on the Nidek only and 3 control eyes were tested on the Heidelberg only. For the Heidelberg camera, all 5 smear- or culture-positive cases were correctly interpreted by expert examiners (sensitivity=100%, 95%CI 47.8 to 100%), and all 3 negative cases were correctly interpreted as negative (specificity=100%, 95%CI 29.2 to 100%). For the Nidek, 3 of 5 smear- or culture-positive cases were correctly identified by expert examiners as positive (sensitivity 60%, 95%CI 14.7 to 94.7%), and 5 out of 5 negative cases were correctly interpreted as negative (specificity=100%, 95%CI 47.8 to 100%). Of the five positive cases for each camera, 2 cases were identified as “high certainty” on each machine. There was no difference between the two microscopes among the 5 eyes that had both scans performed (P=0.50, McNemar’s test).
Conclusions
Both the Nidek and Heidelberg confocal microscopes are useful in diagnosis of acanthamoeba keratitis. Although not statistically significant in this small study, the Heidelberg microscope appeared to be more effective in identifying acanthamoeba cysts than the Nidek. Neither microscope produced false positive tests. These results could be dependent on the graders from this specific study and should be confirmed in future studies.