We read with great interest the article entitled “Is inclusion of Sabouraud dextrose agar essential for the laboratory diagnosis of fungal keratitis?” by Das et al. We wish to congratulate the authors for conducting a comprehensive study on a very relevant issue. The use of a single culture medium in the diagnosis of infective keratitis has been advocated by Waxman et al.; however, authors have not quoted this study. The authors conclude that Sabouraud dextrose agar (SDA) may be excluded as blood agar (BA) and chocolate agar (CA) provide equivalent positive fungal cultures. However, we have the following concerns.
In the present study, the size of the inoculum appears as the main confounding factor. As BA/CA were first inoculated, both received the highest inoculum. This is further supported by author's experiment with controlled inoculum in which both BA and SDA grew fungus on day 2. However, SDA provided better colony characteristics and species identification. The authors did not eliminate this bias regarding the size of the inoculum.
The authors calculated sensitivity, specificity, and kappa coefficient of the detection of fungi on BA, CA, non-nutrient agar (NNA), and brain heart infusion (BHI) with reference to SDA. It appears that some error has crept in inadvertently during calculations. As per our calculations, the actual figures for sensitivity, specificity, and kappa for various media come out to be different than those reported by the authors [Table 1]. The sensitivity of the detection of fungi on BA, CA, NNA, and BHI is 94.1%, 88.2%, 88.2%, and 47.0%, respectively. The specificity of the detection of fungi on BA, CA, NNA, and BHI is 33.3%, 66.8%, 77.9%, and 44.4%, respectively. Contrary to authors′ assertion, the specificity is less than 80% for all and is the lowest for BA (33.3%).
The authors have included only potassium hydroxide (KOH)-positive fungal corneal ulcers. In the clinical situation, fungal keratitis represents 50% of nonviral infective keratitis. The rate of KOH positivity in fungal keratitis is around 30–34% (32%). Thus in clinical practice of 100 cases of corneal ulcers, only 16 cases may be KOH positive. The utility of BA/CA for the detection of fungi in the clinics may differ from that in the study. We calculated the predictive values of the positive test for various media. The values were 72.7%, 83.3%, 88.2%, and 61.5% for BA, CA, NNA, and BHI, respectively. The predictive value of the positive test depends upon sensitivity, specificity, and disease prevalence. These predictive values are 100% for KOH-positive cases and values may be significantly lower in the clinics, where KOH positivity is 16%.
The authors have favored BA/CA over NNA for the culture of fungal pathogens. Although the sensitivity of NNA was 88.2% and the specificity 77.8% was the highest, was there any specific reason for not considering NNA for the growth of fungal pathogens?
1. Das S, Sharma S, Kar S, Sahu SK, Samal B, Mallick A. Is inclusion of Sabouraud dextrose agar essential for the laboratory diagnosis of fungal keratitis? Indian J Ophthalmol. 2010;58:281–6
2. Waxman E, Chechelnitsky M, Mannis MJ, Schwab IR. Single culture media in infectious keratitis Cornea. 1999;18:257–61
3. Daniel WW. Biostatistics: A foundation for analysis in the health sciences Replika Press Pvt, Ltd. 2000 Delhi, India John Wiley and Son's Inc:71–5