“Iodine Allergy” and the Use of Povidone Iodine for Endophthalmitis Prophylaxis

The antiseptic povidone iodine (PI) is a key component of endophthalmitis prophylaxis. The Council on Drugs first reported on the actions, uses, and dosage of the solution in 1957 under the heading “New and Nonofficial Drugs” in the Journal of the American Medical Association. The authors stated that the complex, effective on skin and mucous membranes, could be expected to produce the same “antimicrobial effects of iodine in general” without the “stinging or irritation usually experienced after topical application of elemental iodine.” ¹ Without substantial alteration and despite the introduction of many new topical antibiotics, PI continues to be essential for reducing the incidence of endophthalmitis after ophthalmic procedures and surgeries. ²

Though the precise mechanism of iodine’s antimicrobial effect remains unclear, it is likely that the oxidation of protein disulfide bonds is critical. ³ Molecular iodine is the formulation responsible for this action, but a complex, unstable mixture of iodine states exists in a pure aqueous solution. To address these issues, iodophores were developed—compounds such as povidone that are used to carry a reservoir of iodine. These substances enhance the medical utility of iodine by releasing more free iodine to replace the ions depleted via the antimicrobial action. ⁴

Despite its availability, 34 years passed before the significant utility of PI was demonstrated in a clinical trial of cataract surgery. In a 1991 prospective surgical study, Speaker and Menikoff demonstrated that PI antisepsis decreased the rate of endophthalmitis from 0.18% to 0.06%, a significant decrease compared with silver protein solution. ⁵ Although few ophthalmologists were employing PI for surgical antisepsis prior to this study, it is now standard practice not only in the operating room but also in intravitreal injection (IVI) protocols. ^{6 -8} Key aspects of PI include short application time for effectiveness, ⁹ efficacy of dilute concentrations, ¹⁰ lack of microbial resistance, ^11,12 ability to penetrate biofilm, ¹³ and safety to the retina. ¹⁴ According to a large survey of US retina specialists, using PI was the defining feature of IVI preparation, ¹⁵ and an expert panel emphasized the use of this antiseptic as “perhaps the single most important step in endophthalmitis prevention.” ¹⁶

Without the use of PI, infections have been reported at higher rates. In 10 studies between 1944 and 1960, the pooled rate of endophthalmitis after cataract surgery was 0.21%. ¹⁷ To decrease this incidence, a cocktail of preoperative antibiotics was touted. A 1974 series of 36 000 consecutive surgical cases reported an infection rate of 0.086%. The most effective prophylaxis regimen involved admitting the patients the day before surgery for preoperative treatment with 3 topical antibiotics (chloramphenicol, polymyxin B sulfate, and erythromycin). Unfortunately, this combination was not effective against all bacterial strains. ¹⁸

Similarly, modern data show that periprocedural topical antibiotics for IVI are more costly and ineffective when compared with PI. ² Though PI was used per protocol in 99.04% of patients in the Comparison of Age-Related Macular Degeneration (AMD) Treatment Trials, the rate of endophthalmitis was 10 times higher among patients who did not receive PI (1/177, 0.565%) compared with those who did (10/18 332, 0.055%). ¹⁹ A recent retrospective review emphasized the increased rate of endophthalmitis in IVIs without PI and the much lower risk when it is used. This study reported that 11.9% of post-IVI endophthalmitis cases occurred when PI was omitted from the preparation. Furthermore, injections lacking PI antisepsis had an estimated infection rate of 9.4%, whereas the study’s overall incidence of endophthalmitis after injection was 0.019%. ²⁰

These findings emphasize that, despite its critical, widely accepted role in endophthalmitis prevention, a clinical dilemma arises when patients endorse an “allergy” that may contraindicate PI application. Reported allergens can range from iodine itself to iodinated contrast to seafood or shellfish. Though guidelines note that true allergy to PI is rare and anaphylaxis after ophthalmic use has never been described, ¹⁶ physicians still hesitate to use the antiseptic in these situations. Although patients may have true allergies to these other substances, no data suggest that using topical PI for endophthalmitis prophylaxis in these patients is contraindicated. Conversely, data do show that avoiding PI because of a shellfish allergy ¹⁹ or a reaction to iodinated contrast ²⁰ unnecessarily increases the risk of postprocedure endophthalmitis. Cases such as these demonstrate the need to clearly address the issue of “iodine allergy.”

Iodine is a necessary element for life, thus an allergy to elemental iodine is not possible. Nevertheless, iodine-containing chemicals can provoke 3 types of reactions on human contact: an irritant contact dermatitis, an allergic contact dermatitis, and a generalized anaphylaxis. ²¹ First, an irritant reaction to iodine is an effect of the body’s innate immune system and, as such, a skin response can be provoked immediately without prior exposure. ²² Compared with tinctures of pure iodine, these types of reactions are less frequent with the lower levels of free iodine encountered with iodophores, but they are still common and depend on the exposure time and concentration. ^23,24

Notably, data have shown that PI effectively reduces bacterial counts in 30 seconds ⁹ and that the fastest in vitro killing time is achieved with 0.1% PI rather than more-concentrated formulations. ¹⁰ The American Academy of Ophthalmology recommends using 5% PI, ²⁵ and recent in vivo studies have shown that 5% PI is as effective as 10% in the setting of IVIs. ²⁶ Using an appropriate concentration of PI for an evidence-based period of time can help reduce its irritant effect.

Second, the presence of povidone or other iodophores makes possible an allergic contact dermatitis to iodine-containing substances. As opposed to an irritant reaction, an allergic reaction to the iodophore is mediated by T lymphocytes. As these cells are a part of the adaptive immune system, an initial exposure for sensitization as well as a subsequent reexposure is necessary to produce an allergic response. ²⁷ True allergy to PI is rare but can be difficult to differentiate from an irritant reaction.

In dermatology, patch testing with 10% aqueous PI is commercially available, but even this gold-standard diagnostic test commonly has high false-positive rates, sometimes higher than true-positive rates. ²⁸ To parse this, a study employed patch testing with PI—the gold standard for allergic contact dermatitis—followed by a repeated open application test to obviate confounding irritant effects. Despite being performed in a susceptible atopic population, only 0.4% of the patients manifested a true allergy to PI. ²⁹ This rarity was powerfully demonstrated in the review of endophthalmitis after IVI in patients with self-reported PI allergies. Although those patients reported an allergy to PI or iodine, each later tolerated PI prophylaxis without adverse effect. ²⁰

Third, anaphylaxis is a serious, systemic B-lymphocyte immunoglobulin E (IgE)-mediated life-threatening reaction. There have been rare case reports of urticarial or anaphylactic reactions to PI or povidone, and only a few have involved severe anaphylactic reactions. ^{30 -32} In many of these cases, the volume used or the concentration of PI is not applicable to ophthalmology. For instance, one case involved an intraarticular injection of a povidone-containing substance and another diffuse application of PI to the vaginal mucosa. ^33,34 Importantly, an anaphylactic reaction to PI has not been reported in the ophthalmic literature.

Allergies to many other substances are often erroneously viewed as increasing the risk of allergy to PI. One commonly cited contraindication to any iodine-containing substance is a history of allergy to seafood or shellfish. Though of unclear origin, this apocryphal link has had remarkable durability. A 1973 study examining adverse reactions to iodinated contrast media delineated a subgroup of patients with “seafood allergy” apart from other “food allergies” without explanation, but it reported no increased risk in this group. ³⁵ A study of more than 100 000 contrast injections identified no increased risk among patients reporting a seafood allergy, ³⁶ and a 2010 systematic review reported no other studies that even estimated the risk for cross-reactivity, much less a heightened threat. ³⁷

Another group of patients in which PI is erroneously avoided is that with prior adverse reactions to iodinated contrast media. These hypersensitivities have been shown to be specific to the particular chemical, without predicting reactions to other iodinated contrast agents. ³⁸ As with seafood or shellfish allergies, allergy to iodinated contrast media does not indicate increased risk when using PI. A 2018 review of radiology literature emphasized that patients with a history of “iodine allergy” did not receive clinically indicated scans, and the authors pointedly concluded, “The term iodine allergy should never be used in the context of radiocontrast-associated adverse reactions because it leads to poorer clinical outcomes.” ³⁹

In the highly unusual cases in which PI is relatively contraindicated, alternative antisepsis must be considered. General surgery literature suggests that chlorhexidine may be an effective alternative. A meta-analysis from 2010 found a significant reduction in surgical site infection with chlorhexidine compared with PI. ⁴⁰ However, the applicability of data from general surgery cases to ophthalmic surgeries or IVIs is unclear. The hesitation to use chlorhexidine in ophthalmic settings stems from data showing corneal endothelial ⁴¹ and epithelial toxicity ⁴² from formulations used on the skin. In contrast to PI, allergic dermatitis is very common with chlorhexidine, having been well described since the 1960s. ⁴³ Furthermore, as opposed to the “exceptional” cases of IgE-mediated reactions to PI, anaphylaxis to chlorhexidine is described as “well documented.” ²³ Nevertheless, dilute, aqueous chlorhexidine has been used in IVIs with success. In more than 40 000 IVIs with 0.05% or 0.1% chlorhexidine gluconate as the antiseptic, Merani et al reported a rate of postprocedural endophthalmitis of 0.0074%. ⁴⁴ The authors emphasized that more-concentrated preparation or formulations with alcohol or detergents might not be safe, but they reported no toxic effects and one case of suspected allergy. Thus, in patients in whom PI cannot be used safely, aqueous chlorhexidine appears to be a clinically effective alternative for achieving antisepsis.

PI is an essential component of reducing the risk of endophthalmitis after ophthalmic procedures. As such, clinical scenarios that justify withholding PI are rare. Given the infrequency of a true allergy to PI, physicians should critically consider any reported reactions to the antiseptic itself. Though an irritant effect is possible, this alone is not a contraindication. Furthermore, a reported allergy to iodine, seafood, shellfish, or contrast is not a contraindication because none of these is associated with a dangerously increased risk of allergic response to PI. Just as PI reduces the rate of endophthalmitis, data strongly indicate that omitting this antiseptic substantially increases the incidence of intraocular infection. Therefore, PI continues to be the foundation of endophthalmitis prophylaxis in ophthalmology.