Exploring the 10-year incidence of endophthalmitis at a tertiary care academic referral practice.

One of the most severe complications of cataract surgery is postoperative endophthalmitis. Currently, most clinicians attempt to minimize the risk of this rare but serious intraocular infection with the use of prophylactic topical antibiotics. Within the past decade, there has been a growing practice of using intracameral antibiotic injections postoperatively, as this approach has been reported to significantly reduce the risk of infection. This is a retrospective chart review of patients who underwent cataract surgery between January 1, 2007, and December 31, 2017, at the University of California, Irvine (UCI), and subsequently developed postoperative endophthalmitis. This protocol was approved by the UCI Institutional Review Board. All cases of sterile endophthalmitis were excluded. The current prophylactic protocol at the UCI is to use fourthgeneration fluoroquinolone antibiotic drops 3 days preoperatively, on the day of surgery, and 1 week postoperatively. The overall incidence of postoperative endophthalmitis over this 10-year period at our institution was 3.51 cases per 10 000 surgeries performed (0.035%). A total of 4 patients developed endophthalmitis after cataract surgery at the UCI, out of a total of 11 401 surgical cases performed, with demographics and surgery-specific information shown in Table 1. All 4 patients had standard small-incision phacoemulsification cataract surgery with monofocal intraocular lenses and had no intraoperative complications. The preoperative prophylactic regimen included gatifloxacin drops for 2 patients and moxifloxacin drops for the other 2 patients. Postoperative course for each patient is summarized in Table 2. All patients received treatment on the same day of diagnosis, including vitreous tap and intravitreal injection of ceftazidime/vancomycin (patients 1 and 2), intravitreal dexamethasone (patient 1), anterior chamber washout with partial anterior vitrectomy (patient 3), and a pars plana vitrectomy with intravitreal injection of ceftazidime/vancomycin (patient 4). All patients were promptly treated on the same day of diagnosis, with 3 of 4 patients receiving a vitreous biopsy followed by intravitreal administration of antibiotics of vancomycin and ceftazidime, whereas 1 patient had a partial anterior vitrectomy. The visual acuities (VAs) at diagnosis of endophthalmitis, upon resolution, and best visual outcomes based on previous macular and optic nerve potential for each patient are listed in Table 2. Patient 1’s vision was complicated by macular degeneration, patient 2’s by diabetic retinopathy, patient 3’s by macular degeneration and hypertensive retinopathy, and patient 4’s by vitreous hemorrhage, retinal scarring, and hypertensive retinopathy. These pre-existing ocular pathologies likely limited the final visual outcomes to 20/200, 20/70, 20/80, and counting fingers, respectively; however, these final outcomes are comparable with each patient’s best potential VAs (Table 2). In addition, we found that the final VA of each patient all improved compared with the VA at diagnosis, and either returned to the baseline VA recorded on postoperative day 1 or even improved, highlighting the importance of rapid diagnosis and treatment. Worldwide, post–cataract endophthalmitis incidence rates have ranged from 0.020% to 0.42%. The incidence of post–cataract endophthalmitis at our single tertiary care academic referral practice was 0.035% from 2007 to 2017. This low incidence rate is relatively comparable with that of centers using intracameral injections, such as in Sweden (0.045%) and India (0.02%), as well as centers that did not use intracameral injections, such as in Iran (0.023%) and Canada (0.043%). The use of intracameral antibiotics is also not without risk. Intracameral cefuroxime can cause retinal infarction, whereas intracameral vancomycin can precipitate retinal necrosis and retinal vasculitis. There is also a risk of the miscalculation of dosage or contamination during preparation because the commercial preparation of intracameral antibiotics is not available in all countries, which can lead to intracameral cefuroxime–induced anterior and posterior chamber inflammation. Factors leading to endophthalmitis can vary significantly based on the locale, operating room, and hospital culture, and patient populations. Our low incidence rate at a tertiary university hospital practice in a developed country suggests that a change in the protocol from perioperative antibiotic drops to the use of intracameral antibiotics may not be indicated. The use of perioperative antibiotic drops in addition to perioperative Betadine at our practice likely contributes to the low incidence rate. Future multicenter studies of tertiary academic centers in the United States and worldwide will improve statistical power to further investigate the incidence rates of endophthalmitis and clarify whether the benefit-to-risk ratio of adding intracameral antibiotics is warranted in various settings.