Abstract
Purpose:
We examined the incidence of infective endophthalmitis after intravitreal anti-VEGF in the clinical setting in a large series of injections performed in a single centre. We also aimed to analyze the role of diabetic status, indication for treatment, and nurse injectors in postoperative endophthalmitis rates.
Methods:
We analyzed clinical data recorded prospectively within an electronic medical record (EMR, Medisoft®) between 2006 and 2014, including patient age, diabetic status, treatment indication, anti-VEGF drug used and whether a doctor or nurse administered the injection. The unit used predominantly ranibizumab but data regarding bevacizumab and aflibercept were also analyzed. The standard operating procedure included: use of designated rooms, povidone-iodine antisepsis, and either a drape and speculum or the Invitrea device. EMR audit tools and microbiology records of vitreous samples were used to identify cases of endophthalmitis. Statistical analysis of results was performed using SPSS software.
Results:
Presumed infective endophthalmitis occurred in 21 eyes of 21 patients after 31,561 injections of 3,171 eyes in 2,589 individuals, an incidence of 0.067% per injection, 0.66% per eye and 0.81% per patient. It occurred after 18 of 29608 ranibizumab, 2 of 1332 bevacizumab, and 1 of 675 aflibercept treatments. Of the 17 vitreous samples processed, 9 (53%) were culture positive. Indications for treatment included: neovascular AMD (28,172), diabetic retinopathy (2,180) and retinal vein occlusion (1,198). Out of 21 cases of endophthalmitis, 20 occurred in patients treated for nAMD and 1 for diabetic retinopathy. Between the groups of patients who developed endophthalmitis and those that did not, there was no difference in diabetic status (odds ratio 0.94, p = 0.63) or mean age (82.9 vs. 80.0 years). Since nurses began administering injections in 2012, 17,968 anti-VEGfs have been given: 5,692 by nurses and 12,276 by doctors with 6 cases of endophthalmitis in each group (OR 2.16, p = 0.17).
Conclusions:
The incidence of endophthalmitis in this clinical study is comparable to that in the ANCHOR and MARINA trials (0.05% and 0.1% per injection, respectively). Nurses have played an important role in increasing treatment capacity and their endophthalmitis rate was comparable to that of doctors.