Purpose: To determine if the recent outbreaks of Fusarium or Acanthamoeba keratitis could be identified from data obtained from a single center, the F.I. Proctor Foundation at the University of California, San Francisco.  
Methods: We conducted a retrospective analysis of the culture results from clinically diagnosed non-herpetic corneal ulcers seen from 1987 through May 2007 at the F.I. Proctor Foundation.   For statistical detection of outbreaks, we used the Maximum Excess Events Test (MEET, with 10,000 Monte Carlo replications), which detects clustering within years and between years.  To control for changing referral patterns, we used the total number of cultures as a denominator and replicated the analyses using the number of positive cultures for any organism as a denominator. 
Results: The average numbers of patients undergoing corneal cultures was 67 per year, ranging from 27 to 110.  The total number of isolates identified as bacteria, fungi or Acanthamoeba was 26 per year, ranging from 15 to 44.  Using total number of cultures as a denominator we found evidence of temporal clustering of Fusarium cases (P=0.0007), with post-hoc analysis identifying an epidemic in 2006 (P=0.001).  We also found evidence of temporal clustering of Acanthamoeba case counts (P=0.02).  Two epidemics were identified, the first from 1987-1988 and the second from 2005 to May 2007 (P=0.006).  Similar results were obtained when the analyses were carried out using the number of positive cultures rather than the number of cultures taken.
Conclusion: Consistent with the recent international epidemic of Fusarium keratitis, our analysis identified an epidemic of Fusarium keratitis at our institution in 2006.  Likewise, we confirmed epidemics of Acanthamoeba keratitis at our institution in 1987-1988 and 2005-2007.  These epidemics could not be explained by a change in our diagnostic techniques or a change in the number of cultures performed.  It is possible for a single center to detect an outbreak, but institutions need to be wary of changing referral patterns and diagnostic criteria. The formation of a multi-center network to monitor for early detection of new epidemics of infectious ocular diseases may be in order.
 
Financial Disclosure: N; Grant support: NIH K23EY017897, Research to Prevent Blindness

OMIG Main Page | 2007 Abstracts | < Previous| Next >