Pages Navigation Menu

Meet 'em, greet 'em, treat 'em and street 'em

SGEM#145: Topical Anesthetics for ED Patients with Corneal Abrasions

SGEM#145: Topical Anesthetics for ED Patients with Corneal Abrasions

Podcast Link: SGEM145
Date: January 20th, 2016

Guest Host: Dr. Chris Carpenter is from Washington University, Deputy Editor of Academic Emergency Medicine and faculty member of Emergency Medical Abstracts.

Guest Skeptics: Dr. Kara Otterness from Stony Brook University Hospital on Long Island.  Kara is an assistant clinical professor of Emergency Medicine at Stony Brook.

Dr. Salim Rezaie is from University of Texas Health Science Center at San Antonio. Salim is an associate clinical professor of Emergency Medicine/Internal Medicine at the University of Texas Health Science Center at San Antonio, Texas. He is the creator/founder of the REBEL EM Blog and REBEL Cast.

Case: A 23-year-old healthy woman presents with right eye pain. She states she felt a foreign body sensation yesterday and then overnight, it became painful. Her visual acuity is 20/20 bilaterally and she doesn’t wear corrective lenses. Instillation of tetracaine drops results in complete resolution of the pain. On your slit lamp examination, you see a small corneal abrasion outside of the visual axis with no evidence of ulceration and no foreign body. You update the patient’s tetanus, prescribe her an antibiotic drop and prepare to discharge her when she approaches you and states, “the pain is starting to come back. Can I have that bottle of medicine you used before to take home with me?”

Background: Corneal abrasions account for approximately 10% of eye related visits to the Emergency Department, making them one of the most common eye related presentations (Verma and Kahn). The cornea is highly innervated, and even small abrasions can cause significant pain. Pain control is one of the fundamental goals of emergency medical care. The first documented use of topical ophthalmologic anesthetics was in 1818. A cocaine derivative was employed to effectively block nerve conduction in the superficial cornea and conjunctiva (Rosenwasser).

However, a number of proposed dangers limit the use of topical anesthetic agents for the treatment of corneal abrasion associated pain. These dangers include delayed healing secondary to mitosis inhibition and decreased corneal sensation. The latter issue is of concern because of the potential for the abrasion to progress to an ulcer without the patient noticing. Additionally, these agents may have direct toxicity to corneal epithelium with prolonged use, leading to increased corneal thickness, opacification, stromal infiltration, and epithelial defects. The fear of these complications has led to the pervasive teaching that topical anesthetics should never be used for outpatient management of corneal abrasions. This is reflected in the condemnation of their use in major Emergency Medicine textbooks including Rosen’s and Tintinalli’s.

Based on this, we have been reluctant to give these agents to patients in spite of the fact that we know they’ll improve pain levels.

Clinical Question: Can emergency department patients with simple corneal abrasions be safely discharged home with a prescription for a topical anesthetic drop?

Reference: Swaminathan A, Otterness K, Milne K, Rezaie S. The safety of topical anesthetics in the treatment of corneal abrasions: A review. J Emerg Med 2015

  • Population: Adult patients with a corneal abrasion
    • Excluded: Animal studies, case reports, case series and non-English
  • Intervention: Topical anesthetic drops (proparacaine, tetracaine or bupivacaine)
  • Comparison: Placebo
  • Outcome: Pain control and adverse events

Some discussion on the process:

  • Dr. Salim Rezaie

    Dr. Salim Rezaie

    Salim: This was a systematic review looking at all articles in humans investigating the use of topical anesthetics in corneal abrasions. We found 36 citations originally. We only included studies that were prospective, human trials using either proparacaine or tetracaine for the treatment of corneal abrasions. This limitation yielded two citations. After reviewing all the citations in the articles we found, we identified four additional studies that were relevant and so, six studies in total were included in the review.

  • Kara Otterness

    Kara Otterness

    Kara: All of the articles we included in the systematic review are online on the SGEM blog. Table 1 represents the two emergency department based studies that were relevant to our question. Table 2 includes the four studies from the ophthalmology literature. We thought that this distinction was important because the ophthomology papers were actually looking at the use of topical proparacaine or tetracaine in patients who had undergone PRK or photorefractive keratectomy. This procedure produces an injury to the cornea similar to a corneal abrasion but it’s done under surgical circumstances – so the corneal defects are not contaminated. However, if we believe that the topical anesthetics are directly toxic to the eye or that they inhibit wound healing, the PRK data is still relevant to the discussion.

  • Salim: As part of doing this study, we looked at the original articles that were used as the basis for not giving topical anesthetics. These were all case reports or case series where the patients didn’t have ocular evaluation by a physician prior to starting the drops, used the drops too frequently and/or for a prolonged period of time and often used concentrations that were far too strong. Interestingly, in two of the biggest case series, and by big, we’re talking five patients and nine patients, four of the subjects were doctors.
  • Kara: As opposed to these case series, the six articles we included in our review had patients administer the drops at appropriate intervals and for shorter durations. In the ED study by Ball et al, they also used a dilute solution (1/10th of the standard concentration).
  • Salim: We would have loved to do a meta-analysis of the articles but there was far too high of heterogeneity for us to even attempt.

Author’s Conclusions: “Limited available data suggests that the use of dilute topical ophthalmologic proparacaine or tetracaine for a short duration of time is effective, though their safety for outpatient use is inconclusive.”

Quality Checklist for Therapeutic Systematic Reviews: 

  1. checklistThe clinical question is sensible and answerable. Yes. The question is discrete and it’s a common emergency department presentation so we should be able to get good data on it.
  2. The search for studies was detailed and exhaustive.Yes. We searched PubMed and EMBASE for citations. Additionally, we looked through all the references of the articles we located. We also used a medical librarian to help with our search.
    • Chris: A significant number of clinical studies are never published for one reason or another so (for a thorough systematic review search strategy), it is also important to hand-search research abstracts presented at EM meetings (SAEM, ACEP) and ophthalmology meetings. Another method to discover unpublished research (that won’t show up on PUBMED or EMBASE) is to search Did you do so for this study?
    • Kara: We did not search through abstracts presented at EM meetings or search the database
  3. The primary studies were of high methodological quality. Yes. In particular, the two ED studies by Ball and Waldman were excellent from a methodology standpoint.
    • Chris: A variety of quality assessment instruments exist for systematic reviews of interventional studies (examples include the Cochrane instrument and the Jadad scale). It is important for SR authors to use a validated instrument to assess individual study’s risk of bias to ensure readers that a reliable and valid assessment of quality assessment occurred before lumping the results together. Did your author group use one of these instruments? If not, why not and how can you be confident that the included studies were “high quality”?
    • Kara: This could potentially be a limitation of our systematic review, in that we did not use the Cochrane instrument or the Jadad scale to ensure the quality of the studies included.
  4. The assessment of studies were reproducible.Yes. We detailed our process in the materials and methods.
    • Chris: I don’t see anything in the methods detailing how individual study quality assessment was conducted (by whom, using what instrument, with what inter-rater reliability assessment).
    • Kara: There was no formal inter-rater reliability assessment performed, but instead what we did was each review the search list individually and compared lists to find the six prospective, human trials on the topic.
  5. The outcomes were clinically relevant. Yes. The topical anesthetics provided analgesia and adverse outcomes were evaluated.
  6. There was low statistical heterogeneity for the primary outcomes. No. The studies were highly heterogeneous and as a result, we were not able to do a meta-analysis.
  7. The treatment effect was large enough and precise enough to be clinically significant. Unsure. Since we didn’t do a meta-analysis, we don’t have a single outcome number. In the individual trials, though, topical anesthetics were found to be effective for analgesia. However, from a safety perspective, no patients had ulcerations, corneal opacifications or significantly delayed healing.

Key Results: Our systematic review found that topical anesthetics provided good pain control with no adverse outcomes if used for 48 hours and having physician follow up.

Table 1

Table 1

Table 2

Table 2

Screen Shot 2015-04-25 at 3.11.12 PM

Dr. Chris Carpenter

Dr. Chris Carpenter

Chris: So we already touched on the issue of heterogeneity but let’s talk a little bit more about whether the treatment effect was large enough, even if you were unable to perform the planned meta-analysis.

  • Kara: I think to address this we should focus in on the two ED based studies. In the study by Ball et al, the authors found a statistically significant difference in pain relief from the topical anesthetic in comparison to placebo. Patients were either given dilute proparacaine (0.05%) and instructed to use it as often as they like for seven days or they were given a placebo drop. They were also given Tylenol with codeine for breakthrough pain. They used a Visual Analog Scale (or VAS) to rate pain, and found a mean improvement of 3.9 in the proparacaine group and only 0.6 in the placebo group.
  • Salim: The second study was from Waldman et al. In this study, patients were given either 1% tetracaine or saline drops and told they could use the drops every 30 minutes for 24 hours. They were given paracetamol for breakthrough pain. They found no significant difference in VAS score between the two groups although the patients receiving tetracaine rated their drops as more effective in relieving pain.

Chris: I find it very surprising that they didn’t find a difference in pain scores because I see these patients every shift. We put in the drop and PRESTO – the pain is gone – at least after the initial severe stinging sensation.

  • Kara: Agreed. Our experience says that this works but in this one study, they didn’t show it. The only explanation I thought of was simply that 24 hours out from the diagnosis of the injury, perhaps the pain isn’t that bad from the corneal abrasion and so the drops don’t help that much.
  • Salim: A couple of other things we should mention about the studies. There were a number of important exclusions…some of the really important ones were: previous eye surgery, contact lens use, chemical contamination, gross contamination from a foreign body and inability to follow up in 48 hours.
    • Exclusions: Injury > 36hrs before presentation, < 18 years of age, previous eye surgery or cataracts, wear contact lenses, injury to both eyes, suffering from infectious or chemical conjunctivitis, grossly contaminated foreign bodies, suffering from an ocular infection, current herpes keratitis, allergy to tetracaine, injury requiring urgent ophthalmologic evaluation (i.e. penetrating eye injuries, large corneal abrasions, or injuries causing a disruption of vision), and unable to attend follow up at 48 hours
  • Kara: The need for follow up is critical here. You don’t want to send these patients out on their own with no one reassessing the eye to ensure they aren’t developing an ulcer.

Chris: You appropriately conclude that “Although the small randomized controlled trials are promising, they do not prove safety.” In general randomized controlled trials are not designed or powered to prove safety. Years ago, Jerry Hoffman from Emergency Medical Abstracts actually tried to find Ophthalmology researchers as co-investigators for an NIH trial to definitively assess the efficacy and safety of topical anesthetics for corneal abrasions. Did you calculate how large a study would be to definitively evaluate whether this practice is safe or not?

  • Kara: We did not do a sample size calculation as part of our study

Chris: I was pretty bored today, it was a snow day in St. Louis, and so I actually did a power calculation. The way I did that was I looked at the Ball et al and the Waldman et al study. Between those two studies you had about 200 patients and none of them identified a single adverse outcome. So at a minimum we need to have 200 patients to get an adverse outcome like a corneal ulcer. As a fraction, 1/200 is 0.05%. Assuming that only ¼ of patients who do not get anesthetic eye drops develop a corneal ulcer, that is 1/800 of 0.0125%. Using Even’s Awesome A/B Tools and a two-sided alpha of 0.05 and a beta of 20% or 90% power you would need over 700,000 patients enrolled in a controlled study to definitively prove safety for these drops. Nobody is ever going to do that trial.

So here’s the real question: would you give a patient with a simple corneal abrasion a prescription for topical tetracaine or proparacaine to go home with for analgesia?

  • Kara: Yes. In patients with uncomplicated corneal abrasions who meet the inclusion criteria of the Waldman paper, I would give them a topical anesthetic for comfort. Typically what I do is sterilely draw up 1 cc of the tetracaine 0.5% solution into a syringe of 9 cc sterile saline solution in order to dilute it 10:1 so that I now have a 0.05% solution. I then fill the bottle with about 1-2 cc of this solution and give them that to go home with. By doing this, I’ve only given the patient enough for about a day or so, thus limiting the potential for abuse. This technique was reviewed on REBEL EM. Finally, I make sure the patient has a 48 hour follow up either in the ED or, even better, with an ophthalmologist.
  • Salim: I am doing the same as Kara. I send the patients home with a diluted proparacaine with only 48 hours worth of topical anesthetic and re-iteration of 48 hour follow up. I literally tell the patient, if you don’t follow up you may lose your vision, or worse yet your eye. I think it is also important to document this conversation in your note. If you didn’t document it, it didn’t happen.

Comment on author’s conclusion compared to SGEM Conclusion (Kara): Since we’re the authors, we agree with our conclusion. We’d love to see more robust data on the topic and, we may see that soon. As part of our research, we were able to contact Neal Waldman and he let us know that they are currently enrolling patients in a larger, prospective study. We’ll have to keep our eyes out for that one.

Chris: Excellent – the trial that Jerry Hoffman always wanted to conduct. RCTs assessing the efficacy of any intervention are time-consuming and expensive. Do you know who is paying for Dr. Waldman’s ongoing trial?

  • Kara: I do not know who is sponsoring this trial.

Chris: I found on a second trial that is underway by Dr. Robert Bryson at Queens University comparing tetracaine 0.5% vs. normal saline eye drops just registered in June 2015 after your systematic review came out. Unfortunately, they are only looking at about 260 patients and they haven’t started yet.

SGEM Bottom Line: The best evidence that we currently have demonstrates that dilute topical anesthetic drops of either proparacaine or tetracaine are safe for use in ED patients with simple corneal abrasions to provide analgesia. The studies are small but the data contained in them is far superior to the case series published 50 years ago which led to the dogma that using them is dangerous.

Case Resolution (Kara): You discuss the options with the patient for analgesia. She prefers to not take any opiate medications and opts for a short course of dilute topical tetracaine. You schedule her a follow-up appointment with an ophthalmologist in two days.

What do I tell my patients (Kara): You have a corneal abrasion, which is a scratch on the surface of your eye. This scratch can be quite painful. There are various options available to help control your pain once you leave the Emergency Department, including topical eye drops and oral pain medications. The advantage of the eye drops is that their effects are limited to the eye, so they cause less body-wide side effects compared to the oral medications. While we don’t have robust evidence regarding their use and further research is needed, recent studies suggest that they are effective and probably safe, as long as you use them as prescribed and for a short amount of time. Furthermore, it is very important that you follow up with an ophthalmologist within two days for your eye to be re-checked.

  • Salim: Lastly, we want to give a shout out to the other authors who worked on this. Also, thanks to Kevin Read, our medical librarian who was essential in getting this together.

Keener Kontest: Last weeks’ winner was Kristopher Maday from the University of Alabama at Birmingham. He knew methanol and ethylene glycol are two toxic alcohols that can cause an anion gap metabolic acidosis and are highly toxic.

Listen to the podcast for this weeks’ keener question. If you know the answer then send an email to with “keener” in the subject line. The first correct answer will receive a cool skeptical prize.

Additional FOAMed Resource:

Remember to be skeptical of anything you learn, even if you heard it on the Skeptics’ Guide to Emergency Medicine.

Screen Shot 2016-01-10 at 11.13.56 AM

Screen Shot 2016-01-10 at 11.12.16 AM

Screen Shot 2016-01-10 at 11.16.25 AM