To the Editor:

We would like to thank Kato et al. for their interesting article on the use of the mil-kin® for early detection of fungal and acanthamoeba keratitis in two separate clinical cases [1]. We believe that this technology, which boasts no need for fixing and staining corneal scrapings and is smartphone compatible, may prove useful in resource-limited settings, but may be of less clinical value in more conventional medical settings with ready access to laboratory analysis.

Microscopy-based diagnostics for detection of fungal keratitis have modest sensitivity at best, ranging from 60–70% depending on a myriad factors including sample preparation, user experience, and corneal disease burden [2]. This sensitivity would conceivably decline further without the contrast-enhancement of histochemical stains (e.g., gram, acridine orange, calcofluor white, haematoxylin and eosin), which are inexpensive and take little time to perform [2]. Polymerase chain reaction (PCR) of corneal scrapings, moreover, detects fungal organisms with sensitivity in the 90–95% range with a processing time of a few hours [2]. Regardless of the diagnostic method used, the organism often remains unidentified, and the importance of early initiation of treatment cannot be understated. For instance, topical natamycin is typically started for 7–10 days unless there is suspicion for fungemia, and within this time period microbiology cultures will have already resulted [3]. Even in cases where no organism is identified, ophthalmologists will assess the patient’s response to the initial empiric therapy and choose to escalate or maintain treatment, or search for other cases of keratitis, ultimately making early detection of causal organisms unnecessary.

The authors’ cases highlighting the mil-kin® to identify fungi and acanthamoeba on corneal scrapings are impressive proofs-of-concept. We envision that one particular area of use maybe in resource-poor settings where advanced molecular genetic testing and laboratory access is unavailable. However, while the convenience of a portable microscope and smartphone connectivity is attractive, it should ideally add or maintain the accuracy of our existing microscopy-based methods. Formal head-to-head comparison of diagnostic accuracy of the mil-kin® with traditional microscopy is thus warranted. We commend the authors for their innovation.