Assessing Antibiotic Use With Intravitreal Injections

Are they a line of defense against infection, or do they increase the risk of resistance?

Assessing Antibiotic Use With Intravitreal Injections

Are they a line of defense against infection, or do they increase the risk of resistance?

By Samantha Stahl, Assistant Editor

As the number of intravitreal injections performed hits a new peak each year due to an increasingly aging population base, so too does use of topical antibiotics. While the efficacy of antibiotics has been hotly debated, many retinal physicians continue to use them before and after injections. For patients who receive regular injections, spending three or more days a month exposed to antibiotics for years on end is a prime set-up for microbial drug resistance.

A growing number of physicians are convinced that antibiotics do little to prevent endophthalmitis. So why do they still bother? Entirely out of fear of getting sued.

This past August it was reported that 16 people in Florida and Tennessee experienced severe eye infections after receiving contaminated injections of Avastin. While the incidents were linked to a compounding pharmacy in Florida and a VA facility in Nashville, the situation is every retina specialist's worst fear. No one wants to risk a lawsuit because he or she didn't use antibiotics as a line of defense.

For years, there has been little peer-reviewed evidence to confirm the suspicion that prophylactic antibiotics aren't worth the effort in these patients. Fortunately, new studies provide just the proof that leery physicians are looking for and offer strategies for how to prevent endophthalmitis—without the drops.


“We're all floundering as a specialty about the good use of antibiotics. Is there a risk for antibiotic resistance? Probably, but the problem is that there hasn't been a good study to prove it,” says David Dyer, MD, of Kansas City.

Dr. Dyer isn't alone in wanting this kind of research. Stephen Kim, MD, of Vanderbilt University School of Medicine in Nashville, recently published a study1 in the Archives of Ophthalmology demonstrating a link between resistance and antibiotic usage after intraocular injections. The longitudinal, randomized controlled study included 24 patients receiving unilateral injections for choroidal neovascularization. All participants received four monthly injections in one eye and one of four antibiotic eye drops (ofloxacin 0.3%, gatifloxacin 0.3%, moxifloxacin hydrochloride 0.5% or azithromycin 1%) both on the day of and for four days after each injection. Each patient got cultures before and after the injection in both eyes and was followed for a year.

When a culture tested positive, the strain was identified and tested for susceptibility to 16 different antibiotics. In untreated eyes, resistance to ofloxacin, levofloxacin, gatifloxacin and moxifloxacin was 59%, 56%, 20% and 26%, respectively. Just as suspected, resistance in fluroquinolone-treated eyes was much higher: 82%, 79%, 42% and 65%, respectively.

Even more troublesome, Dr. Kim says, was that patients were not only resistant to the antibiotic they were administered, but they became resistant to other antibiotics as well. “That's particularly worrisome because we are all concerned about bacteria that are resistant to multiple antibiotic classes.”

One of the most significant findings of the study concerns nasal flora. Dr. Kim obtained nasal cultures on the treated side and found that topically applied antibiotics were altering nasopharyngeal bacterial flora as well.

“If you develop flora in the nasopharynx that become resistant, that's a very serious problem,” he says. “These flora are a known cause of many life-threatening infections such as pneumonia and skin infections. Emerging antimicrobial resistance is one of the greatest fears in medicine.”

Dr. Kim hopes that the study's results will give physicians the confidence to scale back on their use of antibiotics. “We have to have more rational use of antibiotics and modify our behavior. We need to use less powerful fluoroquinolones—and use them less often.” He also hopes that the study trickles down to family physicians and optometrists, who are equally guilty of over-prescribing antibiotics and contributing to the risk of antimicrobial resistance.

In a recent article2 from the American Journal of Ophthalmology, retina specialists Charles Wykoff, Harry Flynn, Jr. and Philip Rosenfeld make a striking point:Assuming the average cost of topical antibiotics given after intravitreal injection ranges from $8 to $90, if “all patients were given post-intravitreal injection antibiotics, in 2009 the cost to the United States Healthcare system attributable to these prescriptions to Part B Medicare recipients alone was between approximately $10,000,000 and $114,000,000.” That's a pretty large chunk of change for a medication that might be doing more harm than good. One cited study suggests that topical moxifloxacin 0.5% achieved no additional reduction of conjunctival bacteria beyond what 5% povidone iodine accomplishes alone.3 The authors add that if endophthalmitis does develop, the use of pre-injection antibiotics may increase the risk of resistance of the causative organism; thus, preinjection antibiotics are not advised. Post-injection use is just as debatable, with recent data showing that topically applied fluoroquinolones don't reach sufficient therapeutic levels in the vitreous cavity.4

The article also points to the Ophthalmic Mutual Insurance Company, a malpractice insurance provider, which received no claims or lawsuits from 2006 through the first quarter of 2011 related to intravitreal injection endophthalmitis prophylaxis—a figure that should offer comfort to those who prescribe drops because of legal fears. Signs are slowly pointing towards a more conservative approach to antibiotics.


Not everyone is convinced that antibiotics are being overused in ophthalmology. Michael Tolentino, MD, of Winter Haven, FL, insists that it is “naïve to think that in community use of antibiotics we are developing these ‘super bugs'.” Instead, he believes opportunistic infections are picked up when patients are hospitalized and given systemic antibiotics—not from anything given in conjunction to ophthalmic procedures.

Dr. Tolentino usually skips pre-injection antibiotics in favor of using betadine to reduce bacterial counts more rapidly than any antibiotic eye drops. Post-injection, he prefers to prescribe the latest fourth-generation fluoroquinolones alternating with the less expensive antibiotics, such as sulfonamides, to keep ahead of the ocular flora's changing resistance and to manage cost issues. He prescribes the medication for two days, giving the injection tract from the external surface to the vitreous cavity time to adequately close and prevent ingress of any surface bacteria into the eye, and to put the patient's mind at ease.

In the 2011 PAT survey results, only 27.07% of respondents said they used prophylactic topical antibiotic drops before intravitreal anti-VEGF injections, but 61.67% said they use topical antibiotics after injection.

“I think the antibiotics before an injection don't make sense—we are squirting a lot of antibiotics everywhere,” agrees Thomas Friberg, MD, of Pittsburgh. He extends particular credit to W. Sanderson Grizzard, MD, for revolutionizing the use of povidone iodine as a more effective preoperative standard. The only time Dr. Friberg may consider preop antibiotics is in a patient with very poor lid hygiene who needs to have blepharitis cleared prior to giving an injection. Postop, he typically prescribes antibiotics as a matter of habit (either Vigamox or Polytrim), but remains skeptical about efficacy.

“I think a lot of our colleagues use the antibiotics for a feeling of comfort after the patient goes home,” he says. While the real root of endophthalmitis susceptibility has yet to be determined, he believes that postop antibiotics could make sense to change the patients' flora, since their own flora very well may be the cause of the problem. He agrees with the major concern about emerging drug resistance, “but at the same time if no antibiotics were used and the patient gets a severe endophthalmitis, we as physicians feel a bit vulnerable as the lay public might ultimately decide who is to blame. The cogent scientific reasons not to use them might get lost in translation.” This, he says, is especially relevant now, as severe infections related to Avastin injections have received so much press.

Another proponent of post-injection antibiotics is Dr. Dyer, who puts his patients on drops for three days following injection, but like the others, questions whether they're needed. “I honestly don't know if they work,” he admits. “I have a partner who doesn't use them and has never had endophthalmitis. I have another partner that uses them both three days before and three days after. Does it matter? We really don't know.”


If antibiotics use is questionable, then what is the best way to avoid sight-threatening endophthalmitis? Many experts point to the basics: simply refining injection technique. While it sounds easy enough, there are common errors that some believe are the reason endophthalmitis fears continue to concern the community.

“Be meticulous,” says Dr. Friberg. “Don't let the eyelid get anywhere near the needle, and do the injection very quickly after the lid speculum is in place.” He also recommends using the smallest needle possible.

“Expensive antibiotics make surgeons feel that they don't have to have appropriate technique,” says Dr. Tolentino. He insists upon vigilance with ethyl alcohol hand washes, which are virtually impossible to develop resistance to, and paying close attention to the pre-injection environment. “How many times have I seen a patient blow their nose and then use the same tissue to wipe their eye? How often do patients start talking to you once the needle has been uncapped and spittle is introduced into the surrounding air?”

Colin McCannel, MD, and Joanne Wen, MD, of the Jules Stein Eye Institute recently published a study5 in Archives of Ophthalmology about bacterial dispersal while talking to a simulated patient without wearing a surgical face mask. In this simulated intravitreal injection setup, 15 volunteers were positioned over an open blood agar plate and instructed to either read a five-minute script with a face mask, without a face mask or without a face mask and the face turned away from the plate, or to stand in silence for five minutes. Each volunteer was also asked to read a fiveminute script while reclined in an exam chair with an open blood agar plate secured to his or her forehead.

The researchers found significantly greater bacterial growth when a face mask was not used, suggesting that bringing down infection risk could be as simple as donning a mask and cutting the chit-chat once the needle is uncapped.

Dr. McCannel says that although the risk of infection has supposedly been pushed as low as one in 8000, prospective studies like MARINA, ANCHOR, CATT, VIEW and the VISION trials have reported less promising numbers—perhaps as high as one in 100 to one in 200 cases of endophthalmitis per patient per year.

“The risk for infection is higher than what we like to think about,” he says. “It really is worth the effort to minimize risk, but I just don't think pre- or post-injection topical antibiotics are the right way. The focus should be on injection technique.” RP


1. Kim SJ, Hassanain TS. Ophthalmic antibiotics associated with antimicrobial resistance after intraocular injection therapy. Arch Ophthalmol. 2011;129:1180-1188.
2. Wykoff CC, Flynn HW Jr, Rosenfeld PJ. Prophylaxis for endophthalmitis following intravitreal injection: antisepsis and antibiotics. Am J Ophthalmol. 2011 Sept 19. [Epub ahead of print]
3. Halachmi-Eyal O, Lan Y, Keness Y, Moron D. Preoperative topical moxifloxacin 0.5% and povidone-iodine 5.0% versus lone resistance in ocular cultures. Arch Opthalmol. 2011;129:399-402.
4. Costello P, Bakri SJ, Beer PM, et al. Vitreous penetration of topical moxifloxacin and gatifloxacin in humans. Retina. 2006;26:191-195.
5. Wen JC, McCannel CA, Mochon AB, Garner OB. Bacterial dispersal associated with speech in the setting of intravitreous injections. Arch Ophthalmol. 2011 Aug 8. [Epub ahead of print]