Infection Control and Injection Safety: a Work in Progress

As surgeons gain experience and anti-infectives gain strength, guideline updates should be expected.

Infection Control and Injection Safety: a Work in Progress
As surgeons gain experience and anti-infectives gain strength, guideline updates should be expected.

Years ago, retinal specialists used intravitreal injections mainly to combat retina surgery-related eye infections, such as endophthalmitis. Today, intravitreal injections have become an increasingly popular therapeutic tool for the treatment of age-related macular degeneration (AMD). This evolution comes with an ironic twist: the tool that was mainly used to fight infection is now the center of ongoing debate regarding how best to use it to avoid infection.

Pegaptanib sodium (Macugen, Eyetech/Pfizer), the most recent injectable to enter the AMD treatment arena, earned high enough marks from the Food and Drug Administration to attain the "safe and effective" seal of approval, but suggested protocol rather than strict mandates and a relatively high endophthalmitis rate of 1.3% (the risk for 1 year of therapy, or a total of 9 injections) brought the issue of safe usage and infection control to the forefront.

Eyetech revised its pegaptanib-usage protocol while the clinical trials were still ongoing in May 2003. An amendment of the injection protocol required the use of a sterile preparation and drape similar to that used for routine intraocular surgery, and use of either pre-injection topical ophthalmic antibiotic drops for 3 days prior to the injection or a 10 mL povidone-iodine flush immediately prior to injection.

Christopher Ta, MD, of Stanford University, reviewed the available literature and reported that povidone-iodine be used prior to intravitreous injection and that topical antibiotics might also be used prior to and after injections. Dr. Ta also recommended that clinicians follow aseptic guidelines including the use of an eyelid speculum.1


William F. Mieler, MD, professor and chairman of the Department of Ophthalmology and Visual Science at the University of Chicago, says the formula for safer injections and infection control is the use of a sterile technique including a povidone-iodine lid scrub, use of a speculum, and avoidance of an anterior chamber paracentesis.

"The anterior chamber paracentesis is not necessary and is somewhat risky," says Dr. Mieler, who also opts for use of a topical fourth-generation fluoroquinolone antibiotic. "We generally place 1 or 2 drops of a new fourth-generation fluoroquinolone on the eye 15 to 20 minutes prior to the injection even though the need is still not fully proven," he says. "After we do the injection, we monitor the IOP and once we're sure there is no problem, we send the patient home on the same bottle of antibiotic, and have them use a drop 3 or 4 times a day for about 4 days," explains Dr. Mieler. "We generally have the patient return for follow-up within 3 to 7 days following the injection, and instruct them to call us immediately if they encounter any significant pain, discharge from the eye, or change in their vision," he adds.

Jennifer I. Lim, MD, associate professor of ophthalmology at the Doheny Retina Institute of the Doheny Eye Institute Los Angeles, uses povidone-iodine before and after placing the sterile drape to decrease the risk of endophthalmitis in AMD eyes receiving therapeutic injections.

"In the past, injections were usually done to treat infected eyes, so we didn't sterilize the eye prior to obtaining the cultures from the aqueous and vitreous," she says. Her injection technique, too, is geared toward infection prevention. "I like to inject in the superotemporal quadrant vs. inferotemporal quadrant because pooling of tears is less superiorly and this may equate to a reduced risk of infection," says Dr. Lim.

Joel Corwin, MD, founder of Miramer Eye Specialists Group in Ventura County, Calif, says combining the use of preoperative antibiotics with the use of a sterile technique dramatically reduces the number of incidents of gram-positive cultures prior to injection.

"Many physicians argue that the incidence of endophthalmitis is extremely low, regardless of the technique used to perform the injection. I believe, however, that the occurrence of endophthalmitis is still too frequent. For this reason, I believe it is prudent to combine preop and postop antibiotics with sterile technique, because neglecting these vital steps can result in unsafe injections. Following this protocol ensures that every action has been taken to prevent the occurrence of endophthalmitis," says Dr. Corwin.

He uses a fourth-generation fluoroquinolone to further prophylax and prevent endophthalmitis after pegaptanib injection. "I use gatifloxacin (Zymar, Allergan) because it contains preservative, rather than moxifloxacin (Vigamox, Alcon), which is self-preserved," he explains. "This is important, because when patients are receiving injections every 6 weeks, according to protocol, the bottle of antibiotics stays open longer than with a 1-time procedure like cataract surgery. Patients might touch the tip of the bottle to their eyes, lids or foreign surfaces, increasing the risk for contamination and infection. The preservative in gatifloxacin provides protection against bottle contamination, while moxifloxacin doesn't. Once moxifloxacin has been opened it only lasts 1 injection cycle. Gatifloxacin can be used for several injections," he adds.


On the other side of the debate sits Minneapolis retinal surgeon David F. Williams, MD, MBA, who says a careful sterile technique and use of topical povidone-iodine are the single most important factors to decrease the risk of endophthalmitis associated with invasive ophthalmic procedures.

Dr. Williams does not use antibiotics in association with intravitreal injections for 2 reasons. "The incidence of endophthalmitis following properly done intravitreal injections is extremely low at about 1/1000 and there is no evidence that use of pre-or post-injection topical antibiotics decreases this risk," he says. Dr. Williams says he would make an exception if the patient showed evidence of significant ocular surface disease and/or severe blepharitis. "In these cases, I ask the patient to go through a regimen of lid hygiene, and I would strongly consider both pre- and post-injection antibiotics."

Dr. Williams points out that antibiotics, such as the fourth-generation fluoroquinolones, are also expensive, costing about $50 wholesale, and $75 or more retail. "Just to theoretically decrease the risk of the 1/1000 cases of endophthalmitis would therefore cost somewhere between $50,000 and $75,000 per case of endophthalmitis, with no assurance that using the antibiotics would even be effective," says Dr. Williams.


Terrence P. O'Brien, MD, the Charlotte Breyer Rodgers distinguished professor of Ophthalmology at Bascom Palmer Eye Institute in Miami and former director of ocular infectious diseases and ocular microbiology laboratory at The Wilmer Eye Institute at Johns Hopkins University School of Medicine in Baltimore, notes that patients desperate with sight-threatening diseases of the posterior segment tend to view these therapeutic intravitreal injections as a "magic bullet," sometimes failing to fully appreciate that the route of delivery carries with it the risk of introducing a microorganism into the eye.

"We have some valuable surrogate evidence from cataract surgery, and I think retinal surgeons may be able to adopt some of the lessons learned from cataract surgeons, particularly with respect to proper utilization of antiseptics and antibiotic agents, including the fourth-generation fluoroquinolones, in an effort to better prepare the ocular surface for injections," says Dr. O'Brien. "The best way to reduce ocular surface contamination is through a 2-pronged strategy. The first prong is represented by the use of an antiseptic such as povidone-iodine that has a direct contact mechanism and rapid onset of action. The use of povidone-iodine 5% directly onto the field without irrigation will help keep it in contact with the conjunctival and ocular surface to help reduce contamination. The second prong is the use of a fourth-generation fluoroquinolone in a brief perioperative pulse."

Dr. O'Brien says fourth-generation fluoroquinolones are more effective against a broader spectrum of gram-positive bacteria than earlier antibiotics were. "With the newer generation we have better coverage against Staphylococcus epidermis, Staphylococcus aureus, and even streptococci," he said. "Moreover, with the fourth-generation fluoroquinolones, we have compelling pharmacodynamic kill curve data that says that using gatifloxacin or moxifloxacin the day of [treatment] is perhaps as good as beginning it 2 or 3 days prior," says Dr. O'Brien.

While the updated pegaptanib protocol calls for either pre-injection topical ophthalmic antibiotic drops for 3 days prior to the injection or a 10 mL povidone-iodine flush immediately prior to injection, Dr. O'Brien thinks the combination of both is better than just 1 or the other because those guidelines were based on studies using ofloxacin, an earlier generation of antibiotic. "This is why we have to update those recommendations to be consistent with data achievable with the new generation of fluoroquinolones — gatifloxacin and moxifloxacin," he explains.

How long after treatment the fluoroquinolone should be used remains to be seen, and whether a topically administered drop achieves sufficiently high concentrations in vitreous humor to be protective is still in question, he points out. "We may be fooling ourselves thinking that we can achieve levels in the vitreous that are therapeutic," says Dr. O'Brien. "We're doing studies right now to find out just how many drops are needed to achieve a level that might be protective if there were organisms that were introduced at the time of the injection," he adds. Findings may be reportable by the annual fall meeting of the American Academy of Ophthalmology, according to Dr. O'Brien.


1. Ta C. Minimizing the risk of endophthalmitis following intravitreous injections. Retina. 2004;24(5):699-705.