- Email: [email protected]
Ozonated water is inferior to propanol-based hand rubs for disinfecting hands
Journal of Hospital Infection 92 (2016) 340e343 Available online at www.sciencedirect.com
Journal of Hospital Infection journal homepage: www.elsevierhealth.com/journals/jhin
Short report
Ozonated water is inferior to propanol-based hand rubs for disinfecting hands C. Appelgrein a, *, G. Hosgood a, A.L. Dunn b, O. Schaaf c a
College of Veterinary Medicine, Murdoch University, Perth, Australia Adelaide Veterinary Specialists and Referral Centre, Adelaide, South Australia, Australia c University of Adelaide, Roseworthy, South Australia, Australia b
A R T I C L E
I N F O
Article history: Received 29 April 2015 Accepted 6 August 2015 Available online 15 December 2015 Keywords: Ozone Hand disinfection Handwashing
S U M M A R Y
Ozone is a strong oxidizing biocide that has broad-spectrum antimicrobial properties. The aim of the study was to compare the efficacy of ozone to a propanol-based hand rub for hand disinfection. Twenty subjects were enrolled in an in-vivo cross-over trial (prEN 12791). Subjects treated their hands with the reference procedure (propan-1-ol 60%) or with ozone (4 ppm). Post-wash bacterial counts were determined from one hand (immediate effect), and from the other hand that had been gloved for 3 h (delayed effect). The investigation indicated that ozone is inferior to propan-1-ol 60% hand rub for hand asepsis. Crown Copyright ª 2015 Published by Elsevier Ltd on behalf of the Healthcare Infection Society. All rights reserved.
Introduction Many agents available for hand asepsis are alcohol-based and are used with good results but may provoke skin reactions.1 Alternatives to alcohol for hand asepsis that do not cause skin reactions but still provide effective hand asepsis would facilitate workplace compliance. Ozone is a strong oxidizing biocide that has broad-spectrum antimicrobial properties.4 The ozone molecule acts as an oxidant via direct oxidative effects and through release of free * Corresponding author. Address: South Street Campus, 90 South Street, Murdoch, Western Australia, 6150, Australia. Tel.: þ61 04 0886 4493. E-mail address: [email protected] (C. Appelgrein).
radicals when ozone decomposes. It interacts with bacterial cell walls, enzymes and nucleic material and destroys them.2e4 Ozone is used in various workplace settings such as in hospitals, dentistry and extensively in water and food treatment.5e7 This study assessed ozonated water as a potential substitute for a propan-1-ol-based hand rub for hand asepsis, examining specifically the immediate reduction in bacterial loads on the fingertips. It was proposed that ozone was at least as effective in reducing the bacterial load on the fingertips as the alcoholbased rub, such that ozone attained an immediate post-wash bacterial load on the fingertips at least as low as that after application of the alcohol-based rub. We hypothesized that application of ozonated water will result in a colony-forming unit (cfu) count on the fingertips that falls within a predetermined equivalence range for the
http://dx.doi.org/10.1016/j.jhin.2015.08.029 0195-6701/Crown Copyright ª 2015 Published by Elsevier Ltd on behalf of the Healthcare Infection Society. All rights reserved.
C. Appelgrein et al. / Journal of Hospital Infection 92 (2016) 340e343 immediate and delayed post-wash count after application of propan-1-ol 60% hand rub.
Methods Products The following products were used in this study: e propan-1-ol (Chem Supply Pty Ltd, Gillman, South Australia), contains 99.5% propan-1-ol which was diluted to 60% propan-1-ol with sterile saline; e Bigwell Infection Control: ozonated water generator (product no. BTG120235); e Dominant Soft Hands Soap Batch 85488, containing glycerine and lanolin (Dominant Australia Pty Ltd, Brompton, South Australia); e D/E neutralizing broth (product no. MIMOPM133 Thermo Fisher Scientific Australia Pty Ltd); e tryptone soya agar (TSA) (product no. PP 2157 Thermo Fisher Scientific Australia Pty Ltd).
Test principle and prerequisites The in-vivo bactericidal efficacy of ozonated water was assessed according to the European test method (prEN 12791) by using 20 veterinary student participants in a cross-over trial.8 Ethical consent was obtained from the University of Adelaide (H-2013-050). The subjects had not used any substances with antibacterial activity or antibacterial soap for one week prior to testing.
341
Propan-1-ol 60%-based hand rub was applied and reapplied in 3 mL portions to ensure that the hands were moist for a period of 3 min and hands were rubbed according to a standard handwashing technique.8 The hands were then allowed to air dry. Ozonated water was allowed to run over the hands for 3 min while using a standard handwashing technique.8 The hands were then dried with sterile paper towels.
Determination of post-wash cfu counts Immediately after disinfection, the subjects rubbed the fingertips of the left hand for 1 min into a Petri dish containing D/E neutralizing broth. This solution was used to calculate the immediate post-wash cfu count. The participants were gowned and the right hand was gloved for 3 h, using standard surgical gloving technique, during which the subjects attended university lectures. After removing the glove, sampling was done identically to that performed previously. This solution was used to calculate the delayed post-wash cfu count. From each sample (immediate and delayed), 1 mL and 0.1 mL of undiluted aliquots and a 0.1 mL aliquot of a 1:10 dilution were seeded into separate TSA Petri dishes. Dishes were incubated at 36 C for 24 h and the cfu were counted 24 h after sampling. To indicate the actual cfu count on the fingertips, the number of cfu/mL in the original sampling liquid was calculated by multiplying the cfu count for each dilution by its dilution factor. There were three plates (with the three dilution factors) per hand totalling six plates per student. For accuracy, only plates with <300 cfu were counted and included in calculations. All pre- and post-wash cfu counts were expressed as log10 counts.
Wash phase Statistical analysis To eliminate transient bacteria and external particles, the subjects washed their hands with 10 mL non-medicated soap and water for 1 min using a standardized handwashing procedure.8 They then rinsed their hands with running tap water and dried them with non-sterile paper towels.
Determination of pre-wash cfu counts The fingertips of the left and right hand, including the thumb, were rubbed for 1 min in two Petri dishes (9 cm diameter) containing 10 mL of tryptic soy broth (TSB). The Petri dishes were labelled indicating the left or right hand. From the sampling fluid for each hand, 1:10 and 1:100 dilutions were made, of which 0.1 mL aliquots of both solutions were spread over tryptic soy agar (TSA) dishes (two plates per hand per participant). Dishes were incubated at 36 C for 24 h and the cfu were counted at 24 h. Only plates with <300 cfu were counted.
Hand asepsis phase Each subject was treated with a reference product (propan1-ol 60%), or ozonated water (4 ppm). A period of at least one week elapsed between each product application to allow the reconstitution of normal skin flora. The ozonized water concentration in the handwash solution produced by the ozone generator was confirmed to be 4 ppm prior to and during testing. The water temperature was maintained at 18 C.
The log10 cfu count was the response of interest and found to be normally distributed using the ShapiroeWilk test with failure to reject the null hypothesis of normality. Thus, the log10 cfu count was summarized as mean [95% confidence interval (CI)]. The pre-wash log10 cfu count for each hand was first tested between products using an unpaired t-test (equal variances) against a two-sided hypothesis with P < 0.05 considered significant, in order to verify that the trials were not different in the initial bacterial load. Failure to find a difference allowed further comparison of absolute log10 cfu counts between products. The log10 cfu count was compared from pre-wash to immediate post-wash for the right hand and pre-wash to delayed post-wash for the left hand, for each product, using a paired ttest. Comparisons between immediate post-wash and delayed post-wash for each product were made using an unpaired ttest. To adjust for multiple comparisons, a restricted P < 0.01 was considered significant. Based on the hypothesis that ozone was at least as effective as the propan-1-ol 60% hand rub, demonstrated by attaining an immediate post-wash log10 cfu count on the fingertips at least as low as that after alcohol-based asepsis, a non-inferiority analysis was performed. For the immediate post-wash log10 cfu count, an equivalence range was established based on the mean 10% of the
342
C. Appelgrein et al. / Journal of Hospital Infection 92 (2016) 340e343
immediate post-wash log10 cfu count for the propan-1-ol 60% hand rub. The mean (95% CI) of the immediate post-wash log10 cfu count for ozone was compared against this equivalence range with equivalence considered inside this range, and inferiority or superiority outside this range.
Superior
Propan-1-ol 60% Equivalence range (mean ± 10% (mean))
Inferior
Ozone 95% confidence interval
Results All subjects completed the study. One participant developed a reaction to the propan-1-ol 60% hand-rub product (1/20, 5%; 95% CI: 0.2‒26.4%) and post-wash cfu counts were not obtained for that participant. There was no significant difference in the pre-wash counts for propan-1-ol 60% hand rub or ozone for either hand (P ¼ 0.741 right hand; P ¼ 0.547 left hand); therefore further comparisons are for absolute log10 cfu counts (Table I). The immediate post-wash and delayed post-wash log10 cfu counts were significantly lower than the pre-wash count for both propan-1-ol 60% hand rub (P < 0.001, respectively) and ozone (P < 0.001, respectively). The delayed post-wash log10 cfu count was significantly higher than the immediate postwash count for both propan-1-ol 60% hand rub (P < 0.001) and ozone (P < 0.001). The log10 cfu counts for immediate postwash and delayed post-wash were significantly lower for propan-1-ol 60% hand rub than ozone (P < 0.001, respectively) (Table I). The equivalence range for immediate post-wash log10 cfu counts for propan-1-ol 60% hand rub was established as 1.93 10% (1.71e2.15). The 95% CI of the mean immediate post-wash log10 cfu count for ozone was 2.99e3.67, well above the equivalence range (Figure 1). Thus, ozone was considered inferior to the propan-1-ol 60% rub in reducing the immediate post-wash log10 cfu count.
Discussion The investigation indicated that ozone is inferior to propan1-ol 60% hand rub for hand asepsis. Ozone left a higher residual bacterial load on the fingertips immediately after handwashing, and, although it showed a mild sustained effect, resulting in delayed counts higher than that obtained after disinfection with alcohol, ozone cannot be recommended as a substitute for alcohol as a hand disinfectant. To be a substitute for a product already accepted, the substitute must be at least as effective, and must offer an advantage. The propan-1-ol 60% hand rub did cause a reaction in one student, which is a known disadvantage of the product and the premise for the study. No reactions were seen with ozone. Traditional hypothesis testing could not prove equivalency, thus a non-inferiority analysis was employed. Failure to find a
1.71
1.93
2.15 2.99 Log10 cfu count
3.30
3.67
Figure 1. Equivalence comparison of the immediate post-wash log10 cfu count for propan-1-ol 60% and ozone comparing cfu counts after 24 h incubation. Note that the cfu counts after ozone fall well outside the equivalence range and ozone is considered inferior to propan-1-ol 60%.
difference with traditional testing against a null hypothesis cannot be interpreted as being the same and the power of the analysis will always be questioned.9 Equivalency testing established a range of effect within which the investigator was willing to accept as sufficient for substitution. A range of mean 10% was used in this study, to establish a stringent interval. In this study, there was clear inadequacy of ozone, with the immediate post-wash cfu count not only failing equivalency, but also well within the range considered inferior. The equivalence range would have to extend to þ50% even to approach the lower confidence limit of the mean estimate for ozone. In addition, although ozone did reduce the immediate post-wash cfu count, unlike propan-1-ol 60% hand rub, there was only a marginal sustained reduction in cfu counts after 3 h. Based on this study, ozone cannot be considered a suitable alternative for alcohol-based rubs for hand asepsis and other alternates should be sought.9 Aqueous ozone has been evaluated in a previous handwash study comparing an ozone concentration of 4 ppm for 40 s followed by 1.5 mL 0.2% benzalkonium/83% ethanol hand rub with povidoneeiodine hand scrub. The ozone/benzalkonium rub group achieved a reduction factor in cfu of 1.05 log10 0.86 (mean SD), whereas povidoneeiodine achieved a reduction factor of 1.3 0.57. The differences in counts were not statistically different and the treatment groups were ruled equivalent.9 However, no significant difference does not imply equivalence in this case and the effect of ozone is highly confounded by subsequent applications of products known to be effective in hand asepsis. Possible reasons for the failure of ozonated water to match a propan-1-ol 60% hand rub in this study may be related to the water source. Temperature, pH and organic debris influence the antibacterial efficacy of ozone.4 Whereas we evaluated the
Table I Mean (95% confidence interval) counts for colony-forming units (log10 cfu) before and after hand asepsis with propan-1-ol 60% (alcohol) hand rub and ozonated water (ozone) Agent
Ozone Alcohol
Pre-wash (right hand)*
4.4 (4.18e4.64) 4.36 (4.2e4.52)a
a
Immediate post-wash b
3.33 (2.99e3.67) 1.93 (1.71e2.15)c
* No difference between agents for pre-wash (P > 0.05). aee Means with the same superscript are not significantly different (t-test P < 0.01).
Pre-wash (left hand)*
Delayed post-wash
4.51 (4.30e4.70)a 4.4 (4.16e4.66)a
4.17 (4.01e4.34)d 2.15 (2.03e2.29)e
C. Appelgrein et al. / Journal of Hospital Infection 92 (2016) 340e343 temperature of the water and the subjects cleaned their hands of gross contamination, the pH value of the water supply was not determined. Ozone is more stable at low pH values, but the degree of microbial inactivation remains virtually unchanged for pH values in the range of 5.7e10.1.10 The presence of ozone-consuming compounds, such as organic material, may compete with micro-organisms for ozone, thus limiting its efficacy. Although the water was filtered prior to passing through the ozone machine, the presence of organic debris in the water may have played a role. Double-distilled water may prove to be more effective. In conclusion, ozone failed to reduce bacterial load on hands to a level considered equivalent to propan-1-ol 60% hand rub, both immediately and after 3 h in surgical gloves, and thus cannot be considered suitable as a substitute.
Acknowledgements The authors acknowledge laboratory availability provided by S. Pyecroft. Conflict of interest statement None declared. Funding sources Greencross veterinary group (Brisbane, Queensland, Australia) funded the consumables required. The ozone machine was donated by Mint Devices (Potts Point, NSW, Australia).
343
References 1. Humes D, Lobo DN. Asepsis, antisepsis and skin preparation for surgery. Found Years 2006;2:13e14. 2. Khadre MA, Yousef AE, Kim JG. Microbiological aspects of ozone applications in food: a review. J Food Sci 2001;66:1242e1253. 3. Fan L, Song JH. Interaction of ozone and negative air ions to control microorganisms. J Appl Microbiol 2002;93:144e148. 4. Scott DB, Lesher EC. Effect of ozone on survival and permeability of Escherichia coli. J Bacteriol 1963;85:567e576. 5. Berrington AW, Pedler SJ. Investigation of gaseous ozone for MRSA decontamination of hospital side-rooms. J Hosp Infect 1998;40:61e65. 6. Sharma M, Hudson J. Ozone gas is an effective and practical antibacterial agent. Am J Infect Control 2008;36:559e563. 7. Hudson JB, Sharma M, Petric M. Inactivation of norovirus by ozone gas in conditions relevant to healthcare. J Hosp Infect 2007;66:40e45. 8. European Committee for Standardization CEN. prEN 12791. Chemical disinfection and antiseptics. Surgical hand disinfection. Test method and requirements (phase 2, step 2). 2005; 6010. 9. Greene C, Morland L. Noninferiority and equivalence designs: issues and implications for mental health research. J Trauma Stress 2008;21:433e439. 10. Kampf G, Ostermeyer C, Heeg P, Paulson D. Evaluation of two methods of determining the efficacies of two alcohol-based hand rubs for surgical hand antisepsis. Appl Environ Microbiol 2006;72:3856e3861.
Journal of Hospital Infection journal homepage: www.elsevierhealth.com/journals/jhin
Short report
Ozonated water is inferior to propanol-based hand rubs for disinfecting hands C. Appelgrein a, *, G. Hosgood a, A.L. Dunn b, O. Schaaf c a
College of Veterinary Medicine, Murdoch University, Perth, Australia Adelaide Veterinary Specialists and Referral Centre, Adelaide, South Australia, Australia c University of Adelaide, Roseworthy, South Australia, Australia b
A R T I C L E
I N F O
Article history: Received 29 April 2015 Accepted 6 August 2015 Available online 15 December 2015 Keywords: Ozone Hand disinfection Handwashing
S U M M A R Y
Ozone is a strong oxidizing biocide that has broad-spectrum antimicrobial properties. The aim of the study was to compare the efficacy of ozone to a propanol-based hand rub for hand disinfection. Twenty subjects were enrolled in an in-vivo cross-over trial (prEN 12791). Subjects treated their hands with the reference procedure (propan-1-ol 60%) or with ozone (4 ppm). Post-wash bacterial counts were determined from one hand (immediate effect), and from the other hand that had been gloved for 3 h (delayed effect). The investigation indicated that ozone is inferior to propan-1-ol 60% hand rub for hand asepsis. Crown Copyright ª 2015 Published by Elsevier Ltd on behalf of the Healthcare Infection Society. All rights reserved.
Introduction Many agents available for hand asepsis are alcohol-based and are used with good results but may provoke skin reactions.1 Alternatives to alcohol for hand asepsis that do not cause skin reactions but still provide effective hand asepsis would facilitate workplace compliance. Ozone is a strong oxidizing biocide that has broad-spectrum antimicrobial properties.4 The ozone molecule acts as an oxidant via direct oxidative effects and through release of free * Corresponding author. Address: South Street Campus, 90 South Street, Murdoch, Western Australia, 6150, Australia. Tel.: þ61 04 0886 4493. E-mail address: [email protected] (C. Appelgrein).
radicals when ozone decomposes. It interacts with bacterial cell walls, enzymes and nucleic material and destroys them.2e4 Ozone is used in various workplace settings such as in hospitals, dentistry and extensively in water and food treatment.5e7 This study assessed ozonated water as a potential substitute for a propan-1-ol-based hand rub for hand asepsis, examining specifically the immediate reduction in bacterial loads on the fingertips. It was proposed that ozone was at least as effective in reducing the bacterial load on the fingertips as the alcoholbased rub, such that ozone attained an immediate post-wash bacterial load on the fingertips at least as low as that after application of the alcohol-based rub. We hypothesized that application of ozonated water will result in a colony-forming unit (cfu) count on the fingertips that falls within a predetermined equivalence range for the
http://dx.doi.org/10.1016/j.jhin.2015.08.029 0195-6701/Crown Copyright ª 2015 Published by Elsevier Ltd on behalf of the Healthcare Infection Society. All rights reserved.
C. Appelgrein et al. / Journal of Hospital Infection 92 (2016) 340e343 immediate and delayed post-wash count after application of propan-1-ol 60% hand rub.
Methods Products The following products were used in this study: e propan-1-ol (Chem Supply Pty Ltd, Gillman, South Australia), contains 99.5% propan-1-ol which was diluted to 60% propan-1-ol with sterile saline; e Bigwell Infection Control: ozonated water generator (product no. BTG120235); e Dominant Soft Hands Soap Batch 85488, containing glycerine and lanolin (Dominant Australia Pty Ltd, Brompton, South Australia); e D/E neutralizing broth (product no. MIMOPM133 Thermo Fisher Scientific Australia Pty Ltd); e tryptone soya agar (TSA) (product no. PP 2157 Thermo Fisher Scientific Australia Pty Ltd).
Test principle and prerequisites The in-vivo bactericidal efficacy of ozonated water was assessed according to the European test method (prEN 12791) by using 20 veterinary student participants in a cross-over trial.8 Ethical consent was obtained from the University of Adelaide (H-2013-050). The subjects had not used any substances with antibacterial activity or antibacterial soap for one week prior to testing.
341
Propan-1-ol 60%-based hand rub was applied and reapplied in 3 mL portions to ensure that the hands were moist for a period of 3 min and hands were rubbed according to a standard handwashing technique.8 The hands were then allowed to air dry. Ozonated water was allowed to run over the hands for 3 min while using a standard handwashing technique.8 The hands were then dried with sterile paper towels.
Determination of post-wash cfu counts Immediately after disinfection, the subjects rubbed the fingertips of the left hand for 1 min into a Petri dish containing D/E neutralizing broth. This solution was used to calculate the immediate post-wash cfu count. The participants were gowned and the right hand was gloved for 3 h, using standard surgical gloving technique, during which the subjects attended university lectures. After removing the glove, sampling was done identically to that performed previously. This solution was used to calculate the delayed post-wash cfu count. From each sample (immediate and delayed), 1 mL and 0.1 mL of undiluted aliquots and a 0.1 mL aliquot of a 1:10 dilution were seeded into separate TSA Petri dishes. Dishes were incubated at 36 C for 24 h and the cfu were counted 24 h after sampling. To indicate the actual cfu count on the fingertips, the number of cfu/mL in the original sampling liquid was calculated by multiplying the cfu count for each dilution by its dilution factor. There were three plates (with the three dilution factors) per hand totalling six plates per student. For accuracy, only plates with <300 cfu were counted and included in calculations. All pre- and post-wash cfu counts were expressed as log10 counts.
Wash phase Statistical analysis To eliminate transient bacteria and external particles, the subjects washed their hands with 10 mL non-medicated soap and water for 1 min using a standardized handwashing procedure.8 They then rinsed their hands with running tap water and dried them with non-sterile paper towels.
Determination of pre-wash cfu counts The fingertips of the left and right hand, including the thumb, were rubbed for 1 min in two Petri dishes (9 cm diameter) containing 10 mL of tryptic soy broth (TSB). The Petri dishes were labelled indicating the left or right hand. From the sampling fluid for each hand, 1:10 and 1:100 dilutions were made, of which 0.1 mL aliquots of both solutions were spread over tryptic soy agar (TSA) dishes (two plates per hand per participant). Dishes were incubated at 36 C for 24 h and the cfu were counted at 24 h. Only plates with <300 cfu were counted.
Hand asepsis phase Each subject was treated with a reference product (propan1-ol 60%), or ozonated water (4 ppm). A period of at least one week elapsed between each product application to allow the reconstitution of normal skin flora. The ozonized water concentration in the handwash solution produced by the ozone generator was confirmed to be 4 ppm prior to and during testing. The water temperature was maintained at 18 C.
The log10 cfu count was the response of interest and found to be normally distributed using the ShapiroeWilk test with failure to reject the null hypothesis of normality. Thus, the log10 cfu count was summarized as mean [95% confidence interval (CI)]. The pre-wash log10 cfu count for each hand was first tested between products using an unpaired t-test (equal variances) against a two-sided hypothesis with P < 0.05 considered significant, in order to verify that the trials were not different in the initial bacterial load. Failure to find a difference allowed further comparison of absolute log10 cfu counts between products. The log10 cfu count was compared from pre-wash to immediate post-wash for the right hand and pre-wash to delayed post-wash for the left hand, for each product, using a paired ttest. Comparisons between immediate post-wash and delayed post-wash for each product were made using an unpaired ttest. To adjust for multiple comparisons, a restricted P < 0.01 was considered significant. Based on the hypothesis that ozone was at least as effective as the propan-1-ol 60% hand rub, demonstrated by attaining an immediate post-wash log10 cfu count on the fingertips at least as low as that after alcohol-based asepsis, a non-inferiority analysis was performed. For the immediate post-wash log10 cfu count, an equivalence range was established based on the mean 10% of the
342
C. Appelgrein et al. / Journal of Hospital Infection 92 (2016) 340e343
immediate post-wash log10 cfu count for the propan-1-ol 60% hand rub. The mean (95% CI) of the immediate post-wash log10 cfu count for ozone was compared against this equivalence range with equivalence considered inside this range, and inferiority or superiority outside this range.
Superior
Propan-1-ol 60% Equivalence range (mean ± 10% (mean))
Inferior
Ozone 95% confidence interval
Results All subjects completed the study. One participant developed a reaction to the propan-1-ol 60% hand-rub product (1/20, 5%; 95% CI: 0.2‒26.4%) and post-wash cfu counts were not obtained for that participant. There was no significant difference in the pre-wash counts for propan-1-ol 60% hand rub or ozone for either hand (P ¼ 0.741 right hand; P ¼ 0.547 left hand); therefore further comparisons are for absolute log10 cfu counts (Table I). The immediate post-wash and delayed post-wash log10 cfu counts were significantly lower than the pre-wash count for both propan-1-ol 60% hand rub (P < 0.001, respectively) and ozone (P < 0.001, respectively). The delayed post-wash log10 cfu count was significantly higher than the immediate postwash count for both propan-1-ol 60% hand rub (P < 0.001) and ozone (P < 0.001). The log10 cfu counts for immediate postwash and delayed post-wash were significantly lower for propan-1-ol 60% hand rub than ozone (P < 0.001, respectively) (Table I). The equivalence range for immediate post-wash log10 cfu counts for propan-1-ol 60% hand rub was established as 1.93 10% (1.71e2.15). The 95% CI of the mean immediate post-wash log10 cfu count for ozone was 2.99e3.67, well above the equivalence range (Figure 1). Thus, ozone was considered inferior to the propan-1-ol 60% rub in reducing the immediate post-wash log10 cfu count.
Discussion The investigation indicated that ozone is inferior to propan1-ol 60% hand rub for hand asepsis. Ozone left a higher residual bacterial load on the fingertips immediately after handwashing, and, although it showed a mild sustained effect, resulting in delayed counts higher than that obtained after disinfection with alcohol, ozone cannot be recommended as a substitute for alcohol as a hand disinfectant. To be a substitute for a product already accepted, the substitute must be at least as effective, and must offer an advantage. The propan-1-ol 60% hand rub did cause a reaction in one student, which is a known disadvantage of the product and the premise for the study. No reactions were seen with ozone. Traditional hypothesis testing could not prove equivalency, thus a non-inferiority analysis was employed. Failure to find a
1.71
1.93
2.15 2.99 Log10 cfu count
3.30
3.67
Figure 1. Equivalence comparison of the immediate post-wash log10 cfu count for propan-1-ol 60% and ozone comparing cfu counts after 24 h incubation. Note that the cfu counts after ozone fall well outside the equivalence range and ozone is considered inferior to propan-1-ol 60%.
difference with traditional testing against a null hypothesis cannot be interpreted as being the same and the power of the analysis will always be questioned.9 Equivalency testing established a range of effect within which the investigator was willing to accept as sufficient for substitution. A range of mean 10% was used in this study, to establish a stringent interval. In this study, there was clear inadequacy of ozone, with the immediate post-wash cfu count not only failing equivalency, but also well within the range considered inferior. The equivalence range would have to extend to þ50% even to approach the lower confidence limit of the mean estimate for ozone. In addition, although ozone did reduce the immediate post-wash cfu count, unlike propan-1-ol 60% hand rub, there was only a marginal sustained reduction in cfu counts after 3 h. Based on this study, ozone cannot be considered a suitable alternative for alcohol-based rubs for hand asepsis and other alternates should be sought.9 Aqueous ozone has been evaluated in a previous handwash study comparing an ozone concentration of 4 ppm for 40 s followed by 1.5 mL 0.2% benzalkonium/83% ethanol hand rub with povidoneeiodine hand scrub. The ozone/benzalkonium rub group achieved a reduction factor in cfu of 1.05 log10 0.86 (mean SD), whereas povidoneeiodine achieved a reduction factor of 1.3 0.57. The differences in counts were not statistically different and the treatment groups were ruled equivalent.9 However, no significant difference does not imply equivalence in this case and the effect of ozone is highly confounded by subsequent applications of products known to be effective in hand asepsis. Possible reasons for the failure of ozonated water to match a propan-1-ol 60% hand rub in this study may be related to the water source. Temperature, pH and organic debris influence the antibacterial efficacy of ozone.4 Whereas we evaluated the
Table I Mean (95% confidence interval) counts for colony-forming units (log10 cfu) before and after hand asepsis with propan-1-ol 60% (alcohol) hand rub and ozonated water (ozone) Agent
Ozone Alcohol
Pre-wash (right hand)*
4.4 (4.18e4.64) 4.36 (4.2e4.52)a
a
Immediate post-wash b
3.33 (2.99e3.67) 1.93 (1.71e2.15)c
* No difference between agents for pre-wash (P > 0.05). aee Means with the same superscript are not significantly different (t-test P < 0.01).
Pre-wash (left hand)*
Delayed post-wash
4.51 (4.30e4.70)a 4.4 (4.16e4.66)a
4.17 (4.01e4.34)d 2.15 (2.03e2.29)e
C. Appelgrein et al. / Journal of Hospital Infection 92 (2016) 340e343 temperature of the water and the subjects cleaned their hands of gross contamination, the pH value of the water supply was not determined. Ozone is more stable at low pH values, but the degree of microbial inactivation remains virtually unchanged for pH values in the range of 5.7e10.1.10 The presence of ozone-consuming compounds, such as organic material, may compete with micro-organisms for ozone, thus limiting its efficacy. Although the water was filtered prior to passing through the ozone machine, the presence of organic debris in the water may have played a role. Double-distilled water may prove to be more effective. In conclusion, ozone failed to reduce bacterial load on hands to a level considered equivalent to propan-1-ol 60% hand rub, both immediately and after 3 h in surgical gloves, and thus cannot be considered suitable as a substitute.
Acknowledgements The authors acknowledge laboratory availability provided by S. Pyecroft. Conflict of interest statement None declared. Funding sources Greencross veterinary group (Brisbane, Queensland, Australia) funded the consumables required. The ozone machine was donated by Mint Devices (Potts Point, NSW, Australia).
343
References 1. Humes D, Lobo DN. Asepsis, antisepsis and skin preparation for surgery. Found Years 2006;2:13e14. 2. Khadre MA, Yousef AE, Kim JG. Microbiological aspects of ozone applications in food: a review. J Food Sci 2001;66:1242e1253. 3. Fan L, Song JH. Interaction of ozone and negative air ions to control microorganisms. J Appl Microbiol 2002;93:144e148. 4. Scott DB, Lesher EC. Effect of ozone on survival and permeability of Escherichia coli. J Bacteriol 1963;85:567e576. 5. Berrington AW, Pedler SJ. Investigation of gaseous ozone for MRSA decontamination of hospital side-rooms. J Hosp Infect 1998;40:61e65. 6. Sharma M, Hudson J. Ozone gas is an effective and practical antibacterial agent. Am J Infect Control 2008;36:559e563. 7. Hudson JB, Sharma M, Petric M. Inactivation of norovirus by ozone gas in conditions relevant to healthcare. J Hosp Infect 2007;66:40e45. 8. European Committee for Standardization CEN. prEN 12791. Chemical disinfection and antiseptics. Surgical hand disinfection. Test method and requirements (phase 2, step 2). 2005; 6010. 9. Greene C, Morland L. Noninferiority and equivalence designs: issues and implications for mental health research. J Trauma Stress 2008;21:433e439. 10. Kampf G, Ostermeyer C, Heeg P, Paulson D. Evaluation of two methods of determining the efficacies of two alcohol-based hand rubs for surgical hand antisepsis. Appl Environ Microbiol 2006;72:3856e3861.