This article is more than 5 years old. Information may no longer be current.

Guidelines for intravitreal injections discourage need for topical antibiotics to prevent endophthalmitis

Issue: May 10, 2015

Add topic to email alerts

Receive an email when new articles are posted on

Please provide your email address to receive an email when new articles are posted on .

Subscribe

Added to email alerts

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

Back to Healio

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Back to Healio

An expert panel that was convened to update guidelines for giving intravitreal injections makes clear that routine use of topical antibiotics before, during or after the procedure is not supported by evidence and is likely unnecessary. Stopping this practice, however, may be difficult for both patients and physicians to accept, if only to maintain the perceived added protection for the patient.

Ten years ago, when the number of intravitreal injections given per year was minimal, a similar panel convened. Those consensus guidelines suggested that physicians “may consider using preinjection topical antibiotics,” with the caveats that limited data supported preinjection use of antibiotics and that antimicrobial resistance was possibly more likely to occur with the antibiotic treatment.

“Intravitreous injections were a relatively new means of administering drugs,” Emmett T. Cunningham Jr., MD, PhD, MPH, who chaired both panels, said. “We wanted to make sure there were consensus guidelines to minimize risk and promote safe and effective injections.”

There is no evidence that routine use of topical antibiotics with intravitreal injections decreases infection risk, according to Emmett T. Cunningham Jr., MD, PhD, MPH.

Source: Brook Bryand Photography

At the time of the first panel, Cunningham was senior vice president of medical strategy at Eyetech Pharmaceuticals, which was developing Macugen (pegaptanib sodium), the first anti-VEGF agent for the treatment of neovascular age-related macular degeneration and the first intravitreal pharmacotherapeutic to be approved by the FDA. Today, a number of such agents are approved, and it is projected that nearly 6 million intravitreal injections will be given in 2016, more than twice the number of cataract procedures, according to Cunningham.

The new guidelines are intended first and foremost to address the safety of injections.

“There has been the realization over the last 10 years that the routine use of topical antibiotics before, immediately after and then for a few days following an intravitreal injection, as prophylaxis to decrease infection risk, is not justified,” Cunningham said. “There is no evidence that their routine use meaningfully decreases infection risk.”

According to OSN Retina/Vitreous Board Member Andrew A. Moshfeghi, MD, MBA, the major difference between the new and old guidelines, both published in Retina, is the clearer stance on the use of topical antibiotics.

“They are no longer recommended and are, in fact, discouraged,” he said. “We have noticed that the rate of endophthalmitis has not been affected by the use of antibiotics but that the rate of resistant periocular flora to the commonly used antibiotics by ophthalmologists has considerably increased. This is not new news, per se, as this has been known for several years now. However, it does represent a major shift in the management of patients receiving intravitreal injections.”

Harry W. Flynn Jr.

Harry W. Flynn Jr., MD, who was co-chair of both the original and new panels, believes the risk of infection has decreased over the past decade due to a combination of better preparation of the medication, improved preparation of the injection site and more careful injection technique.

“[The risk for endophthalmitis following injections] is perhaps the lowest risk of all invasive ocular procedures, including cataract surgery, vitrectomy and glaucoma filtering surgery,” Flynn said, further attributing the low risk to use of a 30-gauge needle and the quickness of the injection, which lasts only 2 to 3 seconds.

“Topical antibiotics administered monthly change the ocular surface flora to allow more resistant and more virulent bacteria to remain, thereby perhaps increasing the risk of post-injection endophthalmitis,” Flynn said. “By eliminating the use of antibiotics, there is also a huge cost savings, as well as a reduction in time spent by patients and pharmacies in filling prescriptions for antibiotics.”

Flynn cited several studies that show that antibiotics do not help decrease the risk of endophthalmitis from intravitreal drug injections, including a 2014 comparative study by Storey and colleagues in Ophthalmology.

Sophie J. Bakri, MD, said that many practitioners will continue to administer topical antibiotics.

“They feel that it might provide some added protection to the patient. But in reality, there is no evidence to support that,” she said.

Julia A. Haller

OSN Retina/Vitreous Board Member and panel member for the current guidelines Julia A. Haller, MD, agrees.

“The updates reflect lack of evidence that antibiotics reduce endophthalmitis and our worry that chronic administration thereof may increase emergence of harder-to-treat organisms,” she said.

Haller no longer treats with antibiotics before or after injection; instead, her staple is Betadine (5% povidone-iodine ophthalmic solution, Alcon) on the conjunctiva to sterilize the ocular surface.

Maintaining a sterile field

The use of povidone-iodine is another area of general agreement for the panel.

“Povidone-iodine (5-10%) should be the last agent applied to the intended injection site before injection,” according to the updated guidelines. “If a gel anesthetic is used, povidone-iodine should be applied both before and after application of gel, as retained gel may prevent povidone-iodine from contacting the conjunctival surface of the injection site.”

Povidone-iodine is the best antiseptic agent available in ophthalmology, according to Flynn.

“In general, we prefer to have the povidone-iodine solution in contact with the ocular surface for at least 30 seconds in order to allow maximum elimination of surface bacteria,” he said.

The basic principles remain the same between the two sets of guidelines, according to Bakri, including maintaining a sterile field.

The recent panel found no evidence to support the routine use of a sterile drape but did support use of sterile or nonsterile gloves. Flynn said that he dons gloves as a universal precaution to protect both the patient and himself.

Droplet transmission

Endophthalmitis after an injection is uncommon, about one in 2,000 injections. An important source of these infections is likely the oral flora of people in the procedure room: the patient, the technicians and the injecting physician.

“There are some pilot data to suggest that talking increases the risk of infection with known oropharyngeal bacterial flora,” Moshfeghi said, adding that this is a discussion that has permeated the retina field for several years.

Haller said that she and her colleagues are now extremely aware of the potential for aerosolized contaminants, so she keeps the treatment room quiet.

“I used to chat more to make the patient at ease,” she said, but no longer. Although most of Haller’s medical staff practice silence, some personnel opt for masks instead.

Sophie J. Bakri

“Prior to 2004, the issue of droplet contamination was rarely mentioned,” Bakri said. A few years ago, however, a meta-analysis by Colin McCannel, MD, found that mouth organisms from droplets grow in Petri dishes.

“Many of the organisms that come from the oral cavity are a Streptococcus species, and they tend to be virulent,” Cunningham said. “Therefore, we included the recommendation that people in the room try to minimize that contamination risk — the generation of contaminated aerosolized oral droplets — by not talking during the procedure (immediately before, during and immediately after) or by using masks.”

If the mask option is chosen, every person in the room must wear one.

Eyelid speculum

Cunningham said the new panel devoted considerable time to debating the pros and cons of using an eyelid speculum for eyelid retraction, but that no clear consensus was reached by committee members.

“On the patient himself, the primary source of infection is probably contamination by the lids, which are typically heavily colonized,” he said.

Contamination can occur if the lids touch the needle as it penetrates, or the injection site itself may be contaminated before needle insertion. In the first set of guidelines, a speculum was advocated.

“In the early randomized clinical trials, the cases of endophthalmitis that did occur, although uncommon, were considered protocol violations because the investigator did not use a speculum,” Cunningham said. However, it is possible for experienced injectors to hold the lids open manually during the procedure.

“The lids can often be kept open without a speculum, merely by using one’s gloved fingers to gently hold the lids open because it is done so quickly,” Cunningham said. “It is also much more comfortable for many patients to not have a speculum.”

Prompted mostly by trial and experience, with guiding principles, Haller has transitioned from employing a lid speculum to having a gloved assistant retract the lids.

“This change is much more comfortable and better tolerated by patients, plus it is equally effective at keeping the lashes and lid margins away from the needle,” she said.

“As long as the patient’s eyelashes are away from the eye and the patient does not blink during the injection, I think it is fine not to use an eyelid speculum,” Bakri said, noting that she still personally uses a speculum and will continue to do so. “I feel a speculum is a very good method of keeping the eyelashes away from the eye and keeping the eye open, although there is occasionally some patient discomfort from the speculum. But the injection does not take very long, so the speculum is not in too long.”

Same-day injections

The new guidelines address bilateral same-day intravitreal injections, which were not a common occurrence when the original guidelines were introduced in 2004.

“The main implication is that each eye is to be treated as a completely separate procedure, without using any of the same preparatory materials from one eye on the other,” Moshfeghi said.

Bakri, who was principal investigator of a retrospective case series on bilateral simultaneous intravitreal injections in an office setting, prefers to inject each eye on a different day, ideally spacing the injections about 5 days apart because it can take up to 5 days for an infection to manifest. However, for patients who are unable to travel to the office that often, Bakri counsels them about the risk of bilateral injections.

“Theoretically, if an infection happens, it could happen in both eyes,” she said.

To minimize the risk of infection with the same medication in both eyes on the same day, Bakri uses different lot numbers. She may also use a different medication in each eye.

“Each eye is considered a different prep and a different setup, with a separate povidone-iodine prep and a separate eyelid speculum and, obviously, a separate syringe and needle,” Bakri said.

In response to the flurry of outbreaks of contaminated medication in 2010 that caused endophthalmitis in multiple patients, the revised guidelines emphasize the need to scrutinize compounding pharmacies. Avastin (bevacizumab, Genentech) is a drug that may be prepared by a compounding pharmacy and separated into individual syringes from a large vial for intravitreal administration.

“Compounding pharmacies should be licensed by the state and should have regular inspections to guarantee that they are following the USP Chapter 797 (Pharmaceutical Compounding — Sterile Preparations) on how to prepare compounded medication,” Flynn said. “This includes a sterile field and the pharmacist wearing sterile gloves, a gown, a mask and a hat, as well as there being a laminar flow hood to avoid contamination of the medication.”

If a clinician is uncertain or not confident about his compounding pharmacy, “this might be a reason to not use the same medication on the same day,” Flynn said.

Optional pupil dilation

The new guidelines state that dilating the eye is now optional, compared with the older guidelines for which it was highly recommended.

“We found that dilating the eye was of no benefit to injection,” Bakri said, noting that she stopped dilating the eye around 2008. She said that clinicians primarily dilate to look inside the eye to confirm that the optic nerve is perfused after injection.

“In 2004, the majority of the injections performed were actually of higher volume, 0.1 cc, because they were mainly steroid injections. But since 2005, most injections performed have been anti-VEGF therapies, for which the volume injected in most cases is half, 0.05 cc,” she said. “By injecting this smaller volume, the immediate post-injection IOP is lower than when we injected 0.1 cc, and the risk of non-perfusion is lower.” So instead of dilating the eye, the patient can now be checked for formed vision to ensure that the optic nerve is perfused.

Monitoring IOP

Monitoring patients who are at risk for glaucoma is suggested in the guidelines.

Bakri was the lead author of a study that evaluated IOP in eyes receiving monthly injections of Lucentis (ranibizumab, Genentech). The study, which appeared in Ophthalmology in 2014, showed that there is a subset of patients in whom IOP increases in the long term, she said. Hence, the guidelines recommend that IOP be monitored routinely at every eye exam or before every injection in every patient.

“Evidence suggests that there is a subset of patients who may be at risk for developing sustained ocular hypertension and/or glaucoma following a series of injections,” Moshfeghi said. “This is something that retina specialists have become increasingly vigilant about.”

Guiding principles

As a general consideration, according to the updated guidelines, treatment should be “tailored to the individual patient as necessary in the best clinical judgment of the injecting physician.”

Overall, the revised guidelines represent “basic principles that are generally applicable across therapeutic indications and agents,” Flynn said. However, the expert panel selected the term “guidelines” as opposed to “recommendations” because the conclusions were developed through consensus, “but not necessarily based on science,” he said. “There are no absolute proven studies to make any of these steps mandatory.”

At Bascom Palmer Eye Institute, Flynn’s institution, about 14,000 injections were given in 2013, rising to 16,000 in 2014. Such numbers demand adherence to policies and procedures to reduce or eliminate endophthalmitis, he said.

The advent of drugs that can effectively treat blinding retinal diseases has been a game-changer, according to Haller. But habits are hard to change, she said; for instance, many physicians and patients think of antibiotic drops as part of any procedure. Thus, discussion and procedural implementation are necessary.

“However, at this point, most of the big hurdles to adoption of intravitreal injections are behind us, and most retina specialists are very consonant in their approaches,” she said. “And given the pharmacotherapeutic inaccessibility of the back of the eye — the vitreous cavity and retina — via topical or systemic routes, intravitreal injections are here to stay.”

But new technology may affect the wide use of intravitreal injections someday.

Andrew A. Moshfeghi

“Although intravitreal anti-VEGF injections have proven to be extremely efficacious for our patients with AMD, RVO-related macular edema, diabetic macular edema and a host of additional indications over the last decade, the long-term need for these injections at such a high frequency has become a burden to patients and physicians alike, not to mention the indirect costs to the medical system at large,” Moshfeghi said. “We all look forward to the day when we have an efficacious, sustained-release drug delivery device, either in the form of an implant, gene therapy or other novel delivery system.”

“We look forward to the advent of sustained-release implants that will minimize the number of injections given to patients over the long term,” Bakri said. “This will make the procedure much more convenient for the patient. A sustained-release implant may also improve efficacy by increasing compliance.”

“We are waiting for an anti-VEGF implant,” Bakri said, that can last for 3 years. “Ten years ago, we predicted 10 years, but we are still waiting.” – by Bob Kronemyer

How does patient positioning affect the intravitreal injection procedure?

Sit patients upright

David R.P. Almeida

We do all of our injections with the patient sitting upright and facing forward. We do not tip back or lean forward patients. We do not really position patients in any specific way.

We then inject subconjunctival lidocaine in the superotemporal quadrants in at least 90% of patients. Next, we inject by pointing the needle hub directly posterior toward the optic nerve at approximately 3.5 mm to 4 mm posterior to the limbus.

Our internal data reflect an infection rate that is lower than published rates. We feel our technique is a safe and straightforward method to inject patients. However, we do not think that the position of the patient has a significant impact on infection.

We use a lid speculum, which keeps the lashes away from the conjunctiva and the injection needle. This is key to preventing infection. Moreover, by not positioning the patient in any particular way, especially supine or 45° or forward, we find this minimizes any manipulation on the patient’s eyelids and eyelashes, which are known sources of ocular flora that could potentially be related to infectious endophthalmitis. You want to minimize contact with the conjunctiva, wound site, needle hub and actual needle tip.

We use a Betadine prep over the site of injection for 30 seconds. We believe this is an important contributor to a low infection rate. In addition, we use a mask and gloves for all injections. We do not use pre- or post-injection antibiotics.

We perform our injections in a specialized injection clinic that has procedure suite chairs. This allows patients to be secure and not move during the injection. I believe this helps reduce the likelihood of infection because it maintains a controlled, predictable environment that is stereotypical and repeated for all injections.

By not having a procedure suite chair or a chair that can secure your patient in the desired position, there may be unintended movements of the patient, eye, eyelids or eyelashes, which might affect your ability to prevent infection.

David R.P. Almeida, MD, MBA, PhD, is a vitreoretinal disease and surgery fellow at the University of Iowa, Iowa City. Disclosure: Almeida reports no relevant financial disclosures.

Recline patients 45°

Judy E. Kim

I usually recline patients about 45° in the exam chair for intravitreal injections. Although some have speculated that full recline position may increase the endophthalmitis rate, the reason why I do not fully recline patients is not necessarily for endophthalmitis prevention, but more for patient comfort and my convenience. Many of my older patients do not find full recline to be comfortable, due to difficulty with breathing, dizziness or back problems. With about 45° recline of patients, rather than full recline or vertical positions, I find that I do not have to move the chair up and down as much for my height, while being able to achieve good exposure and angle for injections. Partial recline also allows for stable positioning of the patient’s head against the head rest, stable positioning of my hand in order to anticipate potential patient movement, and the lid speculum staying in place.

Compared with vertical position, there is also less likelihood of povidone-iodine running down onto the patient’s face or shirt. Because some studies have found that injections performed in the operating room, where the patients are fully reclined, may have lower endophthalmitis rates than those performed in the office, I believe that factors other than the position of the patient impact risk for infection.

In my opinion, important aspects of infection control during intravitreal injection include aseptic technique during drawing up the drug; lids and lashes away from the injection site, which is usually accomplished with the use of a lid speculum; either wearing a mask or not talking while facing the patient; having the patient not talk, especially after the last povidone-iodine application; and use of povidone-iodine in all patients during the injection preparation and additional povidone-iodine as the last step prior to injection.

Finally, I do not use topical antibiotics pre- or post-injections. With the patients needing to receive repeated injections, chronic use of topical antibiotics may lead to development of drug-resistant organisms and may actually increase the risk for endophthalmitis.

Judy E. Kim, MD, is an OSN Retina/Vitreous Board Member. Disclosure: Kim reports no relevant financial disclosures.