Hypochlorous Acid 0.01% vs Povidone-Iodine 5% for Ocular Antisepsis

Abstract

Purpose:

An alternative ocular antiseptic is needed for patients who do not tolerate povidone-iodine (PI). The purpose of this study is to compare the antimicrobial effect of hypochlorous acid (HA) 0.01% with PI 5% applied topically to the ocular surface.

Methods:

Swabs of the inferior conjunctiva and posterior lower eyelid margin of 40 patients were taken from both eyes and plated onto blood agar plates. One eye was treated with HA and the other with PI, and swabs were taken after 1-minute exposure. The eye treated with PI was rinsed with sterile saline and another swab was taken. Colony-forming units (CFUs) were recorded after 2 days. Patients rated the level of irritation after treatment in each eye.

Results:

HA and PI both gave significant reduction in CFUs from baseline (P < .001 for HA and P = .002 for PI). The mean reduction in logCFU ± 95% CI was 0.850 ± 0.387 or greater for HA and 0.749 ± 0.385 or greater for PI; this was equivalent to a mean reduction of 7.1-fold or greater or 86% or greater (95% CI, 66%-94%) for HA and 5.6-fold or greater or 82% or greater (95% CI, 57%-93%) for PI. CFUs increased in 17 eyes after saline rinse. PI caused substantial irritation in 31 of the 40 participants, whereas no individuals had any irritation from topical HA.

Conclusions:

Both HA and PI were effective in reducing ocular bacterial load. Unlike PI, HA was not irritating to the eye. Saline rinse after topical PI may increase bacterial counts in some individuals.

Introduction

The most common preparation for disinfection of the ocular surface prior to ophthalmic procedures is irrigation of the ocular surface with povidone-iodine (PI) 5% solution (Betadine 5% Sterile Ophthalmic Prep Solution, Alcon). This has been commonly used since early studies by Apt et al ¹ demonstrated significant reduction in conjunctival bacterial load following topical application of this solution. More than 99% of retina specialists surveyed use PI before intravitreal injections. ² Although its bactericidal efficacy is well established by in vitro and in vivo studies, PI is not without adverse effects and potential complications. Patients often experience ocular and periocular irritation and can develop corneal epithelial defects or keratitis. ³ Studies have shown deterioration in tear film parameters and an increase in the Schein dry eye questionnaire score after application of topical PI, indicating more symptoms of dry eye and discomfort. ^4,5 In our clinical experience, topical PI is often a major source of discomfort for patients receiving intravitreal injections. It can be a deterrent for patients, especially those who need frequent intravitreal treatments.

Although patients cannot have a true allergy to elemental iodine, some patients may have a hypersensitivity reaction to other components of PI. Therefore, an alternative antiseptic is needed that is just as effective as PI for disinfection of the ocular surface but that is painless and has less corneal toxicity than PI; an alternative is especially needed for patients who are allergic to or who do not tolerate PI. Others have recognized the need for an alternative, and studies have been conducted that show that lower concentrations of PI or chlorhexidine could be possible alternatives to 5% PI. ^{6 -11}

Hypochlorous acid (HA) 0.01% (Avenova, NovaBay Pharmaceuticals) has potent in vitro bactericidal activity and has minimal in vitro cytotoxic effects. ¹² Rani et al ¹² demonstrated that HA 0.01% had the highest bactericidal effect (10 000-fold reduction in Staphylococcus aureus after 1-minute exposure) compared with several other skin disinfectants and that it was nontoxic to fibroblasts in cell culture at this concentration. Also, in vitro time-kill studies with PI 5% and HA 0.01% showed that HA reached greater bactericidal effect at 1 minute against staphylococcal isolates from endophthalmitis samples. ¹³ Other studies have shown the effectiveness of HA in killing fungi and a wide range of other microorganisms. ^{14 -16} HA has been widely used as a wound care agent and has been shown to significantly decrease bacterial load and improve wound healing at concentrations that are not cytotoxic. ¹⁷

HA is approved by the US Food and Drug Administration for treatment of blepharitis. It is not approved for direct application to the eye. However, animal studies have demonstrated no eye irritation from topical application of HA to the eye, ¹⁸ and the package insert for Avenova states that it is “nontoxic, nonsensitizing, and nonirritating to the skin and eyes.” Therefore, based on its bactericidal effect in vitro and its apparent safety with topical application, HA could be a reasonable alternative to PI for preinjection antisepsis. This study was designed to assess the effectiveness of HA as an ocular antiseptic in vivo.

Methods

Participants in this single-center study were 40 healthy individuals with no signs or symptoms of periocular or ocular infection and who were not taking any ocular medications. Each participant was assigned a unique, sequential subject number, and samples were identified and processed using this subject number. A pretreatment swab of the inferior conjunctiva and posterior lower eyelid margin was taken from each eye using a sterile calcium alginate swab. Each swab was plated directly onto blood agar plates (tryptic-soy agar with 5% sheep erythrocytes).

A drop of proparacaine 0.5% was instilled in both eyes. Then, in even-numbered individuals, the right eye was flushed with 0.5-mL HA 0.01% and the left eye was flushed with 0.5- mL PI 5%. No additional periocular skin, eyelid, or eyelash preparation was performed. Odd-numbered individuals received PI in the right eye and HA in the left eye. After 1-minute exposure, posttreatment swabs were taken from the conjunctiva and posterior lower lid margins of both eyes and plated onto blood agar plates. To avoid any bias between treatments from slight differences in exposure time, the right eye was always swabbed first. Thus, after treatment, in half the patients the eye treated with HA was swabbed first and in the other half, the eye treated with PI was swabbed first.

The eye that received PI was flushed with 1 to 3 mL of sterile saline. A final swab was performed on that eye after the saline rinse. HA does not need to be rinsed from the eye, and therefore, the eye that received HA was not rinsed and not swabbed a third time. All swabs were taken using a standard protocol of passing the swab gently over the inferior palpebral conjunctiva and the posterior aspect of the lower lid, taking care to avoid the eyelashes.

Soon after completing the treatments and sampling, patients were asked whether they were experiencing irritation in either eye. Any individual who had irritation was asked which eye had more irritation. They were then asked to rate the level of irritation as slightly, moderately, or much more irritated as compared with the other eye.

Within 1 hour after collection, the inoculated plates were transported to an independent testing laboratory (Midwest Laboratories, Omaha, NE) for processing and incubation. Colony-forming units (CFUs) were recorded after 48 hours of incubation. To eliminate bias in performing colony counts, laboratory personnel had no knowledge of the coding system used to identify the samples.

Because of the wide range of CFUs expected, the logarithm of the CFUs was used in statistical analysis to achieve a better approximation of a normal distribution. Only eyes that had more than 1 CFU on pretreatment swabs were included in the analysis. Posttreatment swabs with no growth were assigned a logCFU value of 0. A 95% CI was calculated for the log reduction achieved by each treatment using the t distribution.

Results

A total of 40 patients (80 eyes) completed the study. There were 32 women and 8 men. Participants ranged in age from 23 to 78 years. There was a trend toward increased baseline bacterial load with increasing age. Of the 10 individuals aged 30 years and younger, 12 of 20 eyes (60%) had no growth from baseline swabs. In participants older than 30, only 20 of 60 eyes (33%) had no growth from baseline swabs. Figure 1 shows the effect of age on total pretreatment bacterial counts from both eyes. There was no noticeable difference between male and female participants in pretreatment counts; however, there were too few male patients in the study to make a valid comparison.

Figure 1.

The effect of age on total bacterial count from both eyes of each of the 40 patients at baseline. In patients aged 30 years and younger, 12 of 20 eyes (60%) had no growth at baseline. In patients older than age 30 years, 20 of 60 eyes (33%) had no growth at baseline. CFU indicates, colony-forming units.

HA and PI both significantly reduced the bacterial load (Figure 2). Only the eyes that had more than 1 CFU at baseline were included in the analysis: 21 for HA and 17 for PI. Because counts were reduced to 0 in several eyes from both groups, the percentage reduction was actually greater than or equal to the calculated means. The reduction of logCFU ± 95% CI was 0.850 ± 0.387 or greater for HA (P < .001) and 0.749 ± 0.385 or greater for PI (P = .002) (paired t test; see Figure 2). This was a mean reduction of 7.1-fold or greater or 86% or greater (95% CI, 66%-94%) for HA and 5.6-fold or greater or 82% or greater (95% CI, 57%-93%) for PI. There was no significant difference between HA and PI as demonstrated by the overlapping CIs (unpaired t test, P = .61).

Figure 2.

Mean reduction in bacterial counts with 95% CI for hypochlorous acid (HA) 0.01% and povidone-iodine (PI) 5%. Mean log reductions of colony-forming units (CFUs) were calculated from eyes that had more than 1 CFU at baseline. Log reduction of CFU (logCFU after treatment – logCFU before treatment) was determined for each eye, and the mean was calculated from the paired samples. The 95% CIs were calculated using the t distribution, n = 21 for HA and n = 17 for PI. The log reduction values translated to a mean reduction (95% confidence limits) of 86% (66%, 94%) for HA, and 82% (57%, 93%) for PI. Because CFU was reduced to 0 in many eyes, the actual reductions were greater than or equal to the calculated means.

Figures 3 and 4 show the actual pretreatment and posttreatment CFU for each eye with positive baseline cultures for HA and PI, respectively. One participant in the PI group had a slight increase in CFUs from baseline to posttreatment.

Figure 3.

Reduction in colony-forming units (CFU) with hypochlorous acid 0.01% for each patient who had more than 1 CFU at baseline.

Figure 4.

Reduction in colony-forming units (CFU) with povidone-iodine 5% for each patient who had more than 1 CFU at baseline.

After saline rinse of the eye that was treated with PI, 17 of 40 eyes (42%) had an increase in CFUs compared with the posttreatment swab. This included 7 eyes that were not a part of the comparison of HA and PI because they had 0 or only 1 CFU at baseline.

None of the participants reported any irritation in the eye treated with HA. However, 31 of 40 (78%) patients reported moderately more or much more stinging and burning or irritation in the eye treated with PI compared with the eye treated with HA (Figure 5). Some individuals had only mild irritation in the first few minutes after instillation of PI and then reported much more irritation after 20 to 30 minutes, presumably after the anesthetic effect of the proparacaine wore off. The highest level of irritation within the first hour was recorded. There were 4 patients who reported no irritation with either treatment.

Figure 5.

Comparative irritation of hypochlorous acid (HA) 0.01% vs povidone-iodine (PI) 5%. Each patient received HA in one eye and PI in the fellow eye. Participants were asked if they had any burning, stinging, or irritation in either eye; if they did, they were asked how much more irritation they had in the affected eye compared with the fellow eye.

Conclusions

This in vivo study supports other in vitro studies that demonstrate the strong bactericidal effect of HA. Although neither HA nor PI completely sterilize the ocular surface, both show significant reduction in bacterial load. The reduction in bacterial load in this study is comparable to that reported in other studies with PI. ^1,19,20 Apt et al ¹ reported a mean reduction of 91% with 5% PI based on 15 individuals with no CI reported. Friedman et al ¹⁹ did not calculate a percentage reduction, but based on the data reported in that publication, the reduction was between 79% and 96% with 5% PI after 30 to 60 seconds’ exposure. Ferguson et al ²⁰ showed an overall reduction in median CFUs of 60% with 5% PI after 60 seconds, although eyes with higher baseline counts (>1000 CFUs) had up to a 97% reduction in the median CFUs. Our study showed reductions in CFUs with both HA and PI that were comparable to those reported for 5% PI in these previous studies, 82% or greater (95% CI, 57%-93%) with 5% PI and 86% or greater (95% CI, 66%-94%) with HA.

All these studies have a high degree of variability in reduction of CFUs. For example, although overall reduction in median CFU was significant in the study by Ferguson and colleagues, 16 of 100 eyes had an increase in CFUs after PI treatment. ²⁰ In our study, of the 17 eyes treated with PI that had baseline cultures with positive findings, 1 had a slight increase in CFUs compared with baseline. Although there was not a statistically significant difference between HA and PI in our study (P = .61), with a high variance the size of our study did not provide adequate power to conclude that HA is truly equivalent to PI. However, our study did demonstrate that HA and PI both significantly reduce bacterial load.

Our study results contrasted with the results of a study by Kanclerz et al. ²¹ They compared rates of conjunctival cultures with positive findings in eyes treated with PI vs eyes treated with HA. They did not report the source or the concentration of the HA that was used. In their study, 10 of 51 eyes had positive culture results after conjunctival lavage with HA compared with 0 of 59 with PI. Their study design did not allow calculation of a percentage reduction in CFUs because findings from swabs were counted only as positive or negative regardless of the number of bacteria collected on the swab. While that study raised concern about the relative effectiveness of HA compared with PI, our study found significant reduction in CFUs with both HA and PI. Our data provided more quantitative information and demonstrated a significant in vivo bactericidal effect of HA.

A limitation of our study is the low baseline bacterial count in many eyes. While that is not uncommon for swabs from healthy individuals, the high rate of pretreatment swabs that were negative for bacteria contributed to the lack of sufficient power to determine whether HA is significantly more or less effective than PI in reducing bacterial load. Both HA and PI may be more effective than our results showed since the possible magnitude of the reduction in CFUs was limited by low pretreatment counts in many participants.

The trend toward higher rates of negative bacterial counts in younger individuals was noted incidentally during the study and was not part of the original design. Therefore, statistical testing comparing different age groups was not performed because these tests would be of equivocal validity. However, the high rate of negative culture results from patients younger than 30 years may be useful information for planning future studies. For a similar type of study, it may be beneficial to selectively recruit older individuals.

A wide variety of techniques have been used in previous studies to culture the ocular surface. In preliminary studies at our laboratory, there was a high rate of negative culture results when only the inferior conjunctival fornix was sampled. Therefore, the posterior margin of the lower eyelid was included in sampling for this study since those swabs were more likely to have positive culture findings and since bacteria present there could be a source of infection. Calcium alginate swabs were used because in preliminary trials, cotton-tipped applicators gave low rates of positive culture results, and ESwabs (Copan Diagnostics Inc) caused irritation in patients when used to swab the conjunctiva. Initial comparisons of various culture techniques demonstrated that highest counts were achieved by plating swabs immediately onto blood agar plates rather than extracting bacteria from swabs in various diluents and plating the diluent onto various growth media. Some studies have incubated swabs in liquid culture media for as long as 3 to 6 hours before plating onto agar plates. ^8,20 While incubation in broth prior to plating may give higher counts, it is also likely to alter the results because of variable growth rates in the culture medium that will affect the number of CFUs once the sample is plated.

A standard protocol was used for collections of swabs. Despite this standardization, a limitation of this and similar studies is that investigators cannot be completely masked to treatments since PI has a distinct color and HA is a clear liquid. However, laboratory personnel were masked for processing of samples and for colony counts.

It is well known from clinical practice that HA is not irritating to the eye. It is also well known from clinical practice that PI causes substantial irritation for many patients. The reported levels of irritation in this study confirm the obvious difference between PI and HA regarding patients’ comfort.

The increase in bacterial counts in many participants after saline rinse of PI from the eye was an important discovery. This suggests that rinsing PI from the eye with saline after a preoperative preparation could flush bacteria back into the eye and increase the risk of infection. Patients being prepared for surgery will generally also have the eyelids, lashes, and periocular skin swabbed with PI. No periocular preparation was performed in this study, and it is possible that the observed increases in bacterial load would have been less likely to occur in eyes that also had periocular treatment. However, scrubbing of lids and lashes is not recommended for intravitreal injections, ²² and the observed increase after saline rinse in many eyes is concerning. Fam et al ²³ showed improvement in patients’ comfort using a rinse with HA after application of PI drops compared with PI without a rinse. If HA also has a substantial bactericidal effect, as shown in our study, then it may be better to use HA rather than saline to rinse PI from the eye. A study comparing HA with saline for rinsing PI from the eye will be necessary to determine whether HA is better for maintaining a reduced bacterial load.

The combination of clinical experience, previous in vitro data, and the in vivo data from this study supports the potential use of HA as a reasonable alternative for an ocular antiseptic in patients who do not tolerate PI. A larger study would be needed to determine the equivalency of HA to PI for reducing bacterial load on the ocular surface, and a multicenter randomized clinical trial would be necessary to determine equivalency for the prevention of postinjection endophthalmitis.