When should we declare an endophthalmitis outbreak? A Poisson model of post-cataract surgery endophthalmitis outbreak projections

Purpose: To create an accessible, online tool that determines the likelihood of an endophthalmitis outbreak. Setting: Surgical centers that perform cataract surgeries. Design: Mathematical modelling study. Methods: The global risk of acute post-operative endophthalmitis (POE), defined as occurring within 6 weeks following standalone cataract surgery, was determined from a systematic review. An online tool was then created using the Poisson model to determine the likelihood of a POE outbreak. The model used the number of cataract surgeries performed, time, and number of POE cases observed. Significance was defined as p<0.05, p<0.01, or p<0.001. Results: From the 25 included studies, the mean standard risk of POE following standalone cataract surgery was 0.0692% (range: 0.0189-0.102%). The model demonstrated that for a center that performs 1,000 cataract surgeries in a month, the probability of at least 1 POE cases occurring in a month is not statistically significant (p=0.461, p>0.05), and therefore outbreak is unlikely. In contrast, the probability of at least 3 POE cases occurring in a month at this center is statistically significant (p<0.05), and thus, has moderate evidence for outbreak with 95% confidence. Conclusions: The Poisson model represents an evidence-based way to determine whether the number of POE cases are within expectations for the number of cataract operations performed over a period. This online tool provides ophthalmologists and public health agencies helpful data to guide POE outbreak declaration decisions.