Corneal abrasions from trauma or epithelial defects created during ophthalmic surgery occur commonly. In this review of nine RCTs in which a total of 556 participants with corneal abrasions are analyzed, we compare the safety and efficacy of topical ophthalmic anesthetic with or without NSAID versus placebo or NSAID. The evidence is of very low certainty for any difference between topical anesthetic and placebo in pain control, healing, or outcomes suggestive of anesthetic abuse. The main reasons for this level of certainty are small sample sizes and an overall high risk of bias. Use of anesthetic eye drops without close monitoring or attention to the amount dispensed (in one trial, 40 mL was prescribed for outpatient use)38 creates a potential for significant ocular complications. Such sequelae may not be observed in trials with short follow-up or with high participant drop-out, characteristics of several trials in this review.
Given the number of abrasions evaluated in emergency departments (EDs) each year, investigators should plan for larger sample sizes and a longer duration of follow-up, perhaps having a follow-up ophthalmic examination to confirm lack of recurrent erosion or contact lens-related keratitis.1,41,42 These two conditions also present with disruption of the corneal epithelium but require ophthalmic care. Because most abrasions heal within a few days with just a topical antibiotic, high participant attrition in any such trial is anticipated, yet the dangers of anesthetic use increase with lack of monitoring, longer duration of use, higher concentration, or more frequent dosing than prescribed.43,44 For a self-limited condition, the risk of blindness from corneal melt, infection, or perforation—all findings of topical anesthetic abuse—must be weighed against any possible benefit.
The current review cannot address whether the numerous case reports of anesthetic complications describe outliers since most trials do not assess safety outcomes beyond one to two weeks. The case literature suggests certain patient populations may be at risk for topical anesthetic abuse.4–21 These characteristics include a history of psychiatric illness, depression, dry eye, or drug abuse; certain occupations like manual labor that might prioritize return to work over close follow-up or recovery; or a medical or veterinary background that provides access to topical anesthetics. There are ethical concerns with enrolling participants at higher risk for anesthetic abuse.
In any future trial, sample size assumptions (effect size, power) should be justified, and efficacy outcomes should include a more complete understanding and documentation of pain. Baseline and follow-up pain levels, function, and quality of life as well as complete demographic data would aid interpretation of applicability. Some trials record the baseline pain level in participants; others do not. A high proportion of participant drop-out is noted in one trial of post-traumatic abrasions.23 This same trial employs a mixed-model approach for the multiple, repeated self-reported pain measurements from individuals that are “adjusted for pain on arrival, and allowing for individuals with different pain levels and pain reducing at different rates.”23 No other included trial utilizes this method. Additionally, reliance on self-reported pain intensity following surgery can result in an overestimation of the needed intervention.45,46 Lessons gained from the current opioid epidemic include the need for better understanding of pain, use of other treatment modalities, and focus on returning patients to normal function.45,46 Given that topical anesthetic use also can be habit-forming, the same considerations may apply to pain control in corneal abrasion.
A smaller systematic review of topical anesthetics for traumatic abrasions that includes 2 trials in this current review similarly concludes that topical anesthetic use is “currently not supported by evidence.”47 One summary (not a systematic review) of interventions for post-PRK pain recommends against the long-term use of topical anesthetics.48 Authors of another summary (not a systematic review) state topical anesthetics do not delay epithelial healing.49 A third highlights the need for careful monitoring.50 The small review47 excludes a trial we had included,34 and all three exclude two non-English language trials that we had included.35,37 The American Academy of Ophthalmology’s 2022 Refractive Surgery Preferred Practice Pattern® states “small quantities of dilute topical anesthetic are sometimes used but warrant close supervision.”51 Treatment guidelines and point-of-care resources generally cite case reports and series when recommending against the use of topical anesthetics for pain control.43,44,52 In the United Kingdom, authors of the Clinical Knowledge Summary by National Institute for Health and Care Excellence (NICE) state that topical anesthetic can be used to aid examination but “repeat doses should be avoided as they can cause corneal epithelium toxicity and impair healing.”53
A few years ago, the American Academy of Ophthalmology invited the American College of Emergency Physicians to issue joint guidelines regarding use of topical anesthetics for corneal abrasions because of concerns of potential harm. A joint working group from both organizations convened and reviewed the peer-reviewed literature. However, sidestepping concerns of the ophthalmologists regarding clinical recommendations, the emergency physicians of the group recently published guidelines without any ophthalmologist authors.27 The guidelines endorse dispensing 1.5 to 2 mL topical anesthetics (q30 minutes for 24 hours) for corneal abrasions in the emergency department setting.27 This amount (60–80 drops) actually would last 30–40 hours, although the guidelines state patients would be asked to discard drops after 24 hours. Guideline authors question the emphasis our Cochrane Systematic Review places on RCTs, question its relevance because it is “written solely” by ophthalmologists (not true), and claim an “existing practice pattern of ophthalmologists prescribing topical anesthetics for photorefractive keratectomy postoperative analgesia either for initial use or for breakthrough pain” (not accurate). Topical anesthetic use is exceedingly rare after PRK, and at least three papers the authors cite to support their use ironically describe complications or increased postoperative pain from anesthetic use.
Only four trials in the current review report specific adverse events; most do not report events beyond one week. This short follow-up may lead to an inaccurately low estimate of adverse event rates. Topical anesthetic abuse is usually described in the literature at time points well beyond those of the trials in this review. In Turkey, where topical anesthetics were sold over the counter until 2012, one study found that median duration of use was 28 days (range, 10–112 days; 7 patients) before amniotic membrane transplantation was needed for corneal melt.54 In a case series of patients with toxic keratopathy from topical anesthetic abuse, epithelial healing took a median of 17 days (range, 6–50 days; 19 patients).21
The time from trauma to participant randomization is longer than for surgical participants by 24 to 36 hours. Post-trauma trials vary in whether patients with rust rings are included, which has implications for potential risk of bias and applicability of findings. Abrasion size in post-trauma trials likely varies as well, whereas the epithelial defect is more uniformly created in surgical procedures. Therefore, the baseline risk for slow epithelial healing and adverse events such as infection may differ systematically by setting. Because of different baseline risks, combining post-trauma and post-surgical subgroups when examining for complications does not make clinical sense and therefore is not done.
In conclusion, there is insufficient evidence that topical anesthetics are more safe or effective than placebo for pain or healing of corneal abrasions of either surgical or traumatic origin. Trials with larger sample sizes, means to minimize attrition of trial participants, efficacy outcomes that reflect a better understanding and measurement of pain, and differentiation of abrasions (e.g., uncomplicated abrasions vs. ones with retained foreign material [or “rust rings”]) would reduce bias and increase certainty of evidence. Future trials and systematic reviews should include core outcome sets, comprised of multistakeholder consensus of measurable, reported standardized outcomes to improve consistency across studies and facilitate comparison of interventions.55–59 Given that most abrasions are self-limited, trial participants should receive appropriate quantities of topical anesthetic to reduce the risk of complication and vision loss. Lessons learned from the opioid epidemic can apply to analgesia for corneal abrasion.