Role of confocal microscopy in the diagnosis of fungal and acanthamoeba keratitis.

Abstract

To investigate the role of confocal microscopy as a diagnostic modality in microbial keratitis and to determine inter- and intraobserver variation in the analysis and interpretation of confocal microscopy findings. Prospective, double masked, nonrandomized, observational clinical trial. We included 146 consecutive patients with clinically suspected microbial keratitis. Confocal microscopy and microbiology evaluation of study participants. Sensitivity, specificity, and positive and negative predictive values of confocal microscopy in diagnosing fungal and Acanthamoeba keratitis compared with microbiologic evaluation, as well as the intra- and interobserver variation in interpretation of confocal scans. We included 148 cases of infiltrative keratitis. Of the 103 microbiologically proven cases of Acanthamoeba or fungal keratitis, the confocal microscope was able to identify fungal filaments or Acanthamoeba cysts in 91 cases with either fungal or Acanthamoeba keratitis with a sensitivity of 88.3% (95% confidence interval [CI], 82.2-94.5) and a specificity of 91.1% (95% CI, 82.8-99.4). The interobserver agreement in interpreting the scans was good (kappa = 0.6; phi = 0.617). The intraobserver agreement was kappa = 0.795 and phi = 0.807. The confocal microscope seems to be an accurate and reliable diagnostic modality in the etiologic diagnosis of fungal and Acanthamoeba keratitis.