Abstract
Abstract: :
Purpose: It is known that prevailing weather conditions can influence the pathogenicity of many diseases and there are previous reports of associations between season and postoperative endophthalmitis. After observing that the incidence of postoperative endophthalmitis in Western Australia followed a cyclical pattern, we wondered if underlying weather conditions were associated with the risk of postoperative endophthalmitis. Methods: We conducted a retrospective, population–based cohort study. Using the Western Australian Data Linkage System we identified all cataract and lens–related operations performed in Western Australia from 1980 to 2000, in addition to those patients that developed endophthalmitis postoperatively. Cases of endophthalmitis were further verified by cross–referencing with clinicians and other clinical sources prior to medical record review. Associations with environmental parameters were explored using cross–correlation time series techniques and multivariate logistic regression modelling. Results: Initial modelling of various non–surgical risk factors found a marginal association between season of operation and postoperative endophthalmitis, with winter operations at higher risk compared to spring (OR 1.48, 95% CI 1.00–2.18). Cross–correlation of monthly endophthalmitis incidence with various weather indices found a significant association between total monthly rainfall and endophthalmitis. Further analysis has shown that higher cumulative rainfall over the period 3 days prior to admission, day of admission and 3 days after discharge for surgery increases the risk of subsequent endophthalmitis (OR 1.35 for every 50 mm increase in rainfall, p–value 0.03, 95% CI 1.01–1.81). Conclusions: This is the first report that rainfall around the time of cataract surgery affects the risk of endophthalmitis and may be related to alterations in commensal flora of the conjunctiva in the pre– and peri–operative period.
Keywords: endophthalmitis • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • clinical (human) or epidemiologic studies: risk factor assessment