Dear Editor,

We read with great interest the article titled 'Characteristic clinical features as an aid to the diagnosis of suppurative keratitis caused by filamentous fungi' by Thomas and associates1. We would like to congratulate the authors on this attempt to validate the signs of fungal keratitis, which would be helpful to the ophthalmologists of developing nations. We would like to make following comments:

In the light of the fact that in the tropics laboratory facilities are rare and the diagnosis and treatment of cases is based on clinical characteristics, the authors conclusion that "the probability of fungal etiology in a case of microbial keratitis is 63% if one of the 3 clinical signs are present (serrated margin, raised slough and coloration other than yellow) and at least 85% if all of three are present" becomes very important. However, before applying this conclusion in clinical practice we will have to take into account following facts of the study:

In this study the authors excluded cases with mixed infection (1.4% and 5.5% in Ghana and India respectively), acanthamoeba keratitis (0.3% & 0.9%), unconfirmed laboratory diagnosis (49.7% & 31.1%), and small infiltrates (11.7% overall) 2. Since the score is designed for application in community practice, it would have been better to apply the scores to all cases (290) from Ghana rather than a subset of patients.

Since this study aims at ophthalmologists at the primary health care level, including cases that had small infiltrate i.e. less than 4 mm would have been very useful because clinicians at the primary care level are likely to see early cases.

The predictive value of a positive test depends on relative pretest probability or prevalence of the disease in the group of individuals tested. Therefore, the "score" will be useful only in countries with a high prevalence of fungal keratitis. At a prevalence of 30% the positive predictive value of even a high score will not allow a conclusive diagnosis.

Therefore, the statement that the three signs described by the authors allow the diagnosis of fungal keratitis with 85% confidence can be misleading. Rather, this study further highlights that there are no exclusive clinical signs in microbial keratitis to diagnose the etiological agent.

To address this difficulty, addition of simple, cost effective microbiology tests such as microscopic examination of smears using 10% potassium hydroxide or lactophenol cotton blue will help ophthalmologist to be surer of diagnosis and to start initial treatment with more certainty.

References

1 Thomas P A, Leck A K and Myatt M. Characteristic clinical features as an aid to the diagnosis of suppurative keratitis caused by filamentous fungi. Br. J. Ophthalmol 2005;89;1554-1558

2 Leck AK, Thomas PA, Hagan M, et al. Aetiology of suppurative corneal ulcers in Ghana and south India, and epidemiology of fungal keratitis. Br J Ophthalmol 2002;86:1211-15.