Author: Ali Lamin (United Kingdom)
Co-authors: Ali Lamin, Abha Gupta, Romi Chhabra
Purpose
To determine the incidence of infectious endophthalmitis in patients receiving intravitreal anti-VEGF injections.
Setting/Venue
Single tertiary center, retrospective study, at Manchester Royal Eye Hospital, UK.
Methods
Patients with a diagnosis of proven infectious endophthalmitis secondary to intravitreal injection (IVI) from 2015 to 2020 were identified from Manchester Royal Eye Hospital electronic database. Data was collected retrospectively including anti-VEGF agent (aflibercept vs ranibizumab), injection location, antiseptic solution (povidone-iodine vs chlorhexidine), lens status, and underling ocular disease (wet AMD vs diabetic macular oedema). Patient demographic data was also collected including age, gender and laterality. In our centre we perform lids check at every visit on slit lamp, prior to IVI procedure to ensure no blepharitis present. All injections were performed by an ophthalmic trained injector in an enclosed, dedicated, sterile room using standardised Aseptic Non Touch Technique (ANTT). Injection protocol included skin preparation with povidone iodine 10%, povidone iodine 5% minims on the ocular surface, lid speculum and a single stat dose of topical antibiotic post injection. In patients with presumed iodine allergy/hypersensitivity, chlorhexidine 0.05% was used instead.
Results
Over a 6-year period, 6 cases of infectious endophthalmitis were reported out of 77199 anti-VEGF injections - an incidence rate of 0.007% per injection (approximately one case per 13,000 injections). Causative organism was gram positive in all 6 cases. The incidence of culture positive endophthalmitis in patients where chlorhexidine was used as an alternative to povidone iodine was zero. There was no difference in endophthalmitis rate among the anti-VEGF agents (aflibercept n = 3, ranibizumab n =3). All 6 cases’ injection location was inferior (inferotemporal n =3, inferonasal n=3). Four out of the six cases were pseudophakic. Five cases had wet AMD, while only one case had diabetic macular oedema. Mean age of presentation was 79 years with no gender differences (males n= 3, females n=3).
Conlusions
There was an exceptionally low incidence of infectious endophthalmitis (0.007%) in our tertiary center when compared to the literature (0.015% - 0.025%). This may be explained by our strict adoption of local ANTT guidelines. At our centre, every clinician undergoes competency assessment prior to performing IVI procedures. In addition, nurse injectors are also required to have an annual assessment to ensure that these high standards of ANTT are always maintained despite very high throughput.
Financial Disclosure
None
Comments
-
