Take-Home Message: A pearl for preventing endophthalmitis following intravitreal injections: avoid eyelid contact with the injection site after the last application of povidone iodine.
By Lynda Charters; Reviewed by Joshua D. Levinson, MD
Endophthalmitis has a small risk of developing, but when it does, the ocular consequences are mighty. To lower that risk further, the effects of different aseptic protocols were evaluated to determine the incidence of endophthalmitis following intravitreal injections.
Investigators found that preventing contact between the eyelid and the injection site after the final application of povidone iodine resulted in a substantial decrease in the incidence rate of the infection.
Application of povidone iodine has been cited as the most important step in preventing endophthalmitis, Joshua Levinson, MD, noted. With this in mind, he and his colleagues, Richard Garfinkel, MD, and Daniel Berinstein MD, set out to analyze the aseptic protocols used among the physicians at their institution.
“The ultimate goal was to improve safety after injections for our patients,” said Dr. Levinson, a fellow at Georgetown University Hospital, Washington Hospital Center, and The Retina Group of Washington.
In this retrospective, case-control series, patients were prepped before the intravitreal injection that began with application of topical tetracaine, and 5% povidone iodine to the eyelids and lashes and to flush the conjunctiva and fornices. The eye was then closed and covered with a sterile patch until the time to perform the intravitreal injection.
A survey of the physicians identified four distinct routes regarding reapplication of povidone iodine, Dr. Levinson reported. These practices were the absence of reapplication of povidone iodine, reapplication of povidone iodine without positioning of a lid speculum, reapplication of povidone iodine before placement of the lid speculum, and reapplication of povidone iodine after placement of the lid speculum.
Besides povidone iodine, multivariate risk factor analysis also considered how the use of gloves, use of a caliper to mark the injection site, and the medications injected might have impacted the findings.
A total of 37,646 injections of anti-vascular endothelial growth factor (anti-VEGF) drugs and steroids performed by 27 retina specialists in 2016 were analyzed. “Of these injections, we identified 33 infections of presumed infectious endophthalmitis after the injections, for an overall prevalence of 0.088%,” Dr. Levinson said.
The results indicated that gloves and calipers and the class of medications were not significant predictors for the development of endophthalmitis.
The heart of the study involved the reapplication of povidone iodine. When investigators analyzed the protocol that did not involve reapplication of povidone iodine after the patients were prepped with povidone iodine, the incidence of endophthalmitis was 0.124%, about 1 case in 800 patients. When physicians reapplied povidone iodine without use of a lid speculum, the incidence of endophthalmitis was 0.110%, about 1 case in 900 patients.
Dr. Levinson found it noteworthy that among the physicians who reapplied povidone iodine before placement of the lid speculum had an incidence rate of endophthalmitis that was almost identical to those who did not reapply povidone iodine.
In the analysis of physicians who re-applied povidone iodine after placement of the lid speculum, Dr. Levinson reported a “drastic reduction” in the incidence of endophthalmitis to 0.017%, which is about 1 case in 6,000 patients (P = 0.004).
When the investigators took the analysis a step further looking at the lid speculum as a factor, that was separate from povidone iodine. “When we looked at the physicians who performed the injections using the lid speculum, we found they had about half the incidence of endophthalmitis (0.065%), compared with those who did not insert a lid speculum (0.131%), which was a significant difference,” Dr. Levinson said.
He added that this finding was in contrast to most published reports that the use of a lid speculum is unnecessary.
When the investigators looked at the physicians who did not reapply povidone iodine, the lid speculum no longer had a protective effect. “This suggests a synergistic effect between isolating the lids from the conjunctiva and reapplication of povidone iodine,” Dr. Levinson emphasized.
He also noted that alternative techniques to isolate the eyelids, such as bimanual lid retraction, were not studied, but they may be acceptable alternatives as long as the eyelid is prevented from contacting the injection site.
The study conclusion was that application of additional povidone iodine after placement of the lid speculum significantly decreased the incidence of endophthalmitis after intravitreal injections.
“We recommend that whatever the aseptic technique used, the eyelid should not be allowed to be in contact with the injection site after final application of povidone iodine, Dr. Levinson concluded. “Since we started this study, an additional 16,000 injections have been administered after we shared out findings with our colleagues, and we noticed a 37% decrease in the incidence of endophthalmitis.”
Joshua D. Levinson, MD
This article was adapted from a presentation that Dr. Levinson delivered at the 2017 American Society of Refractive Surgery. Dr. Levinson has no financial interest in any aspect of this report.