Differential care of corneal ulcers in the community based on apparent severity

To describe current practice patterns in treating infectious keratitis.

A questionnaire was designed that asked ophthalmologists to describe the diagnostic equipment accessible to their practice for performing smears and obtaining scrapings for microbial culture and sensitivity testing. The questionnaire also presented two hypothetical cases of patients with infectious keratitis. Bacterial keratitis was relatively early and less severe in the first patient, and it was more advanced and more severe in the second patient. Recipients were asked about their diagnostic and therapeutic approach. The survey was mailed to 300 ophthalmologists in Florida, Illinois, and New York.

One hundred twenty-four completed surveys (45%) were returned. Six surveys were from cornea specialists, who were excluded from the analysis. Only 18 practices (15%) maintained access to Gram stain supplies, and 58 (50%) maintained culture supplies. Whereas 56% of respondents would treat the patient with the less severe bacterial keratitis without obtaining samples for cultures, only 13% would treat the patient with the more severe condition in this manner (P<0.00001). Of the respondents, 82% would treat the patient with the less severe bacterial keratitis with a fluoroquinolone, compared with 62% for the patient with the more severe infection (P=0.002). The mean frequency of fluoroquinolone administration for the patient with more-severe bacterial keratitis was one drop every 0.88 hours, compared with one drop every 1.48 hours for the patient with the less severe infection.

Ophthalmologists appear to treat suspected infectious keratitis differently, depending on perceived severity; they choose different antibiotic regimens; and are more likely to forgo scrapings for Gram staining and cultures for ulcers that appear less severe. The justification for this approach should be established.