Objective: To determine the characteristics of infectious corneal ulcers at
the time of presentation to the cornea specialist associated with a favora
ble response to medical therapy versus a poor outcome manifested by the nee
d for penetrating keratoplasty for therapy or visual rehabilitation.
Design: Retrospective, case-control study.
Participants: A total of 162 patient records were reviewed, including the s
tudy group of 30 patients and the control group of 132 patients.
Intervention: A retrospective review of all cases of microbial keratitis pr
esenting to the Cornea Service between January 1, 1989 and December 31, 199
5 was conducted. The cases were divided into two groups. The study group co
nsisted of patients with microbial keratitis who failed medical therapy and
required penetrating keratoplasty. The control group included patients wit
h infectious ulcers who responded to medical therapy alone.
Main Outcome Measures: The influence of demographics, medical and ocular hi
story, delay in presentation to the primary ophthalmologist or the corneal
specialist, topical medications, and contact lens usage were compared. Visu
al acuity and ulcer characteristics were recorded. The statistical signific
ance was evaluated by the chi-square test for independence and multiple log
istic regression.
Results: Older age (P=0.001), delay in referral to the corneal specialist (
P<0.03), and treatment with topical steroids prior to presentation (P<0.000
1) were statistically significant factors associated with the need for pene
trating keratoplasty. Steroid use and the delay in referral were correlated
. A past history of ocular surgery (P=0.01), poor visual acuity at presenta
tion (P<0.001), and ulcer characteristics, including central location (P<0.
0001), large size (P<0.0001), presence of perforation or descemetocele (P<0
.0001), limbal involvement (P<0.0001), and hypopyon (P=0.05), were all asso
ciated with the need for penetrating keratoplasty.
Conclusions: Older age, delay in referral to the corneal specialist, topica
l steroid treatment, past ocular surgery, poor vision at presentation, larg
e size, and central location of the ulcer are risk factors for poor outcome
of microbial keratitis, as indicated by the need for penetrating keratopla
sty.