eAppendix. Clinical Site (Number of Participants with at Least 1 Injection): Investigator Names
Customize your JAMA Network experience by selecting one or more topics from the list below.
Bhavsar AR, Glassman AR, Stockdale CR, Jampol LM, for the Diabetic Retinopathy Clinical Research Network. Elimination of Topical Antibiotics for Intravitreous Injections and the Importance of Using Povidone-Iodine: Update From the Diabetic Retinopathy Clinical Research Network. JAMA Ophthalmol. 2016;134(10):1181–1183. doi:10.1001/jamaophthalmol.2016.2741
This report provides updated endophthalmitis rates for eyes receiving intravitreous injections with and without povidone-iodine and rates with and without topical antibiotics from Diabetic Retinopathy Clinical Research Network clinical trials.
Among 8 Diabetic Retinopathy Clinical Research Network clinical trials conducted from 2006 to 2015, 28 786 intravitreous injections were administered (3123 eyes), and 20 617 of those (2264 eyes) were administered between 2012 and 2015. Eleven cases of endophthalmitis occurred; 4 occurred between 2012 and 2015. Thirteen injections in 3 eyes from 2 participants were administered without povidone-iodine; both participants developed endophthalmitis in 1 eye. Of the remaining 28 773 injections (3120 eyes) performed with povidone-iodine, 9 cases of endophthalmitis occurred: 6 cases (0.05% of 11 565 injections) in eyes receiving topical antibiotics and 3 cases (0.02% of 17 208 injections) in eyes not receiving topical antibiotics (P = .17).
Conclusions and Relevance
While only a small number of eyes did not receive povidone-iodine just prior to an intravitreous injection, this report provides further evidence regarding the risk of endophthalmitis when povidone-iodine is not used before intravitreous injections. Exclusion of topical antibiotics was not associated with a higher risk of endophthalmitis. Continued use of povidone-iodine and consideration to eliminate topical antibiotics from injection procedures seems warranted.
Previously, the Diabetic Retinopathy Clinical Research Network (DRCR.net) reported a low rate of endophthalmitis achieved via a protocol requiring topical povidone-iodine, a sterile eyelid speculum, and topical anesthetic but not requiring topical antibiotics.1,2 These results included 8027 injections (1066 eyes) from single-use vials across 4 DRCR.net randomized trials through 2011.3-6 Seven cases of endophthalmitis had occurred, including 6 (0.13%; n = 4697) with the use of topical antibiotics, compared with 1 (0.03%; n = 3333) without the use of topical antibiotics (P = .25).
As of December 31, 2015, a total of 28 786 intravitreous injections had been administered to 3123 eyes across 8 DRCR.net trials including 20 617 injections in 2264 eyes from 2012 to 2015.3-8 The studies adhered to the tenets of the Declaration of Helsinki. The protocol and Health Insurance Portability and Accountability Act–compliant informed consent forms were approved by multiple institutional review boards. The injections were performed by 379 different investigators at 136 clinical sites in the United States and Canada. The medications injected in a volume of 0.05 mL included aflibercept, 2.0 mg (n = 6632), bevacizumab, 1.25 mg, repackaged into glass vials (n = 4431), bevacizumab from unaltered vials containing 4 mL of bevacizumab (n = 208), ranibizumab, 0.3 mg or 0.5 mg (n = 17 186), or saline (n = 329). A total of 11 cases of endophthalmitis had occurred (0.04% of injections; 95% CI, 0.02% to 0.07%; 0.35% of eyes; 95% CI, 0.18% to 0.63%). This equates to approximately 1 case in every 2700 injections and 1 case of endophthalmitis among every 285 eyes. A higher percentage of cases of endophthalmitis occurred prior to 2012, when 7 cases occurred in the 8162 injections (0.09%; 95% CI 0.03% to 0.18%; approximately 1 in 1100) compared with 4 cases from 2012 to 2015 (0.02% of injections; 95% CI, 0.01% to 0.05%; approximately 1 in 5100; P value = 0.02, Fisher exact test).
The standardized DRCR.net intravitreous injection protocol requires the application of topical anesthetic, the use of a sterile eyelid speculum, and the application of topical povidone-iodine to the conjunctiva. Each injection is required to be drawn up from a single-use vial by the investigator at the time of the injection. The protocol does not require (but does allow) topical antibiotics prior to, on the day of, or after the injection. Furthermore, the protocol does not require the use of sterile gloves, a sterile drape, a face mask, or specific avoidance of talking by the patient or treating ophthalmologist. Data on the use of povidone-iodine and antibiotics were collected in real time. Diagnosis of endophthalmitis was based on investigator’s judgment.
Per DRCR.net protocol, topical povidone-iodine is required to be present for at least 30 seconds over the injection site prior to the injection. For eyes in which povidone-iodine cannot be used (eg, allergy), study injections were to be withheld. Therefore, omission of povidone-iodine constitutes a protocol violation. Among 3123 eyes injected, 3 were injected without povidone-iodine in 2 participants for a total of 13 injections. Both participants who were not pretreated with povidone-iodine developed endophthalmitis in 1 eye during their clinical course (1 participant also received an injection without povidone-iodine in the fellow eye but did not develop endophthalmitis in the fellow eye) (Table). These 2 events constitute a 15% risk of endophthalmitis among 13 injections (95% CI, 2% to 45%), with 66% of the eyes and 100% of the participants developing endophthalmitis during the treatment course.
Of the 28 773 injections given with povidone-iodine, application included 5% or 10% topical drops on a cotton-tipped applicator placed on the conjunctival surface at the injection site in 13 206 of 28 773 injections (46%), 5% topical drops on the conjunctival surface in 16 507 of 28 773 injections (57%), or a 10% povidone iodine swab (Povidone Iodine Swabstick; Medline Industries, Inc) applied to the conjunctival surface overlying the planned injection site in 4979 of 28 773 injections (17%). Investigators could use more than 1 type of povidone-iodine application. Application of povidone-iodine to the lower fornix or to upper and lower eyelids and eyelashes were optional but were performed in 18 515 of 28 773 and 16 140 of 28 773 injections, respectively (64% and 56%). Although application of povidone-iodine to the conjunctival surface was required, 30 injections (0.10%) occurred with only application to the fornix, eyelids, or eyelashes, which was also a protocol violation. However, none of these eyes developed endophthalmitis.
In the DRCR.net protocols considered here, use of topical antibiotics is at investigator discretion. Percentage of injections where topical antibiotics have been used has decreased over time. In the first 6 years of DRCR.net injections (2006-2011), among the first 8163 injections, 4749 injections (58%) included topical antibiotics. Within the past 4 years, among the 20 610 injections, 6816 (33%) included topical antibiotics. Throughout the total duration, 17 208 injections (60%) were given without any topical antibiotics, and 11 565 injections (40%) were given with preinjection or postinjection topical antibiotics. The 13 injections without povidone-iodine were excluded from these numbers, although topical antibiotics were given for all 13 of those injections. Among injections performed with povidone-iodine, endophthalmitis occurred in 6 cases (0.05% of 11 565 injections, approximately 1 in 2000) following injections with topical antibiotic use, compared with 3 cases (0.02% of 17 208 injections, approximately 1 in 5700) following injections without the use of any topical antibiotics (P = .17, Fisher exact test).
The rates of the 6 cases of endophthalmitis per injection within the first 6 years by antibiotic use were 0.03% (1 case) vs 0.11% (5 cases) without antibiotics and with antibiotics, respectively. In the last 4 years, the rate of endophthalmitis per injection was 0.01% both without antibiotics (2 cases) and with antibiotics (1 case). These results support the previous DRCR.net publications demonstrating that a low rate of endophthalmitis can be achieved without topical antibiotics when the injection procedure includes topical povidone-iodine, a sterile eyelid speculum, and topical anesthetic. These prospectively accumulated results support similar outcomes noted in large retrospective case series and suggest that topical antibiotics before or after intravitreal injections,9-11 as given in the DRCR.net, are not reducing the risk of endophthalmitis, prompting many DRCR.net investigators to no longer use topical antibiotics as part of the injection procedure. Antibiotic cost, potential organism resistance, potential drug allergy, and patient comfort are reasons why antibiotic use should be eliminated in most circumstances.
Although only 3 eyes of 2 participants were given intravitreous injections without topical povidone-iodine, these results demonstrate a substantial and striking increase in the risk of endophthalmitis when povidone-iodine is omitted. When povidone-iodine is used, topical antibiotics do not appear to mitigate the risk of endophthalmitis. These results have led the DRCR.net to continue to require povidone-iodine prior to injections and to change the injection procedure for new protocols to prohibit topical antibiotic use for injections without first consulting the protocol chair.
Corresponding Author: Adam R. Glassman, MS, Jaeb Center for Health Research, 15310 Amberly Dr, Ste 350, Tampa, FL 33647 (firstname.lastname@example.org).
Published Online: August 11, 2016. doi:10.1001/jamaophthalmol.2016.2741
Author Contributions: Mr Glassman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Bhavsar, Glassman.
Critical revision of the manuscript for important intellectual content: Bhavsar, Stockdale, Jampol.
Statistical analysis: Bhavsar, Glassman.
Obtaining funding: Glassman, Jampol.
Administrative, technical, or material support: Bhavsar, Glassman, Stockdale.
Study supervision: Glassman, Jampol.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Mr Glassman and Ms Stockdale report grants from the National Eye Institute, Genentech/Roche and Regeneron. Dr Jampol reports grants from the National Eye Institute.
Funding/Support: This study was supported through cooperative agreements from the National Eye Institute and the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Department of Health and Human Services EY14231, EY14229, and EY018817. Ranibizumab for the studies was provided by Genentech Inc and Aflibercept for the studies was provided by Regeneron.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Group Information: The most recently published list of the Diabetic Retinopathy Clinical Research Network investigators and staff who participated in this study can be found in the eAppendix in the Supplement.
Additional Information: A complete list of all DRCR.net investigator financial disclosures can be found at http://www.drcr.net.
Create a personal account or sign in to: