- Email: [email protected]
Handwashing with soap or alcoholic solutions? A randomized clinical trial of its effectiveness
Handwashing with soap or alcoholic solutions? A randomized clinical trial of its effectiveness Magda Zaragoza, RN Montserrat Sallés, RN Julià Gomez, MD Jose M. Bayas, MD, PhD Antoni Trilla, MD, PhD, MSc Barcelona, Spain
Background: The effectiveness of an alcoholic solution compared with the standard hygienic handwashing procedure during regular work in clinical wards and intensive care units of a large public university hospital in Barcelona was assessed. Methods: A prospective, randomized clinical trial with crossover design, paired data, and blind evaluation was done. Eligible health care workers (HCWs) included permanent and temporary HCWs of wards and intensive care units. From each category, a random sample of persons was selected. HCWs were randomly assigned to regular handwashing (liquid soap and water) or handwashing with the alcoholic solution by using a crossover design. The number of colony-forming units on agar plates from hands printing in 3 different samples was counted. Results: A total of 47 HCWs were included. The average reduction in the number of colony-forming units from samples before handwashing to samples after handwashing was 49.6% for soap and water and 88.2% for the alcoholic solution. When both methods were compared, the average number of colony-forming units recovered after the procedure showed a statistically significant difference in favor of the alcoholic solution (P < .001). The alcoholic solution was well tolerated by HCWs. Overall acceptance rate was classified as “good” by 72% of HCWs after 2 weeks use. Of all HCWs included, 9.3% stated that the use of the alcoholic solution worsened minor pre-existing skin conditions. Conclusions: Although the regular use of hygienic soap and water handwashing procedures is the gold standard, the use of alcoholic solutions is effective and safe and deserves more attention, especially in situations in which the handwashing compliance rate is hampered by architectural problems (lack of sinks) or nursing work overload. (AJIC Am J Infect Control 1999;27:258-61)
Handwashing is still considered the most important and effective infection control measure to prevent transmission of nosocomial infections.1 An essential component of any infection control program is the handwashing procedure, before and after all direct contact with patients, biologic fluids, devices, or medical equipment at risk of being contaminated with nosocomial microorganisms. However, several studies performed in many US and European hospitals regularly
From the Hospital Hygiene and Central Sterilization Unit, Microbiology Laboratory, Preventive Medicine Department, and the Hospital Epidemiology and Health Services Research Unit, Hospital Clinic—University of Barcelona, Biomedical Research Institute August Pi I Sunyer (IDIBAPS). Partially supported by a research grant from Beiersdorf SA (Spain). Reprint requests: Magda Zaragoza, RN, Hospital Hygiene and Central Sterilization Service, Hospital Clinic—University of Barcelona, Villarroel 170 (Esc 4-1), 08036-Barcelona, Spain. Copyright © 1999 by the Association for Professionals in Infection Control and Epidemiology, Inc. 0196-6553/99/$8.00 + 0
258
17/46/97622
and repeatedly show that handwashing compliance rate is lower than 50%.2-4 Handwashing with soap and water significantly reduces the burden of microorganisms of the hands. Other agents such as alcoholic solutions with or without other compounds (quaternary ammonia) mixed with products for preventing skin dryness are increasingly used in Europe. Health care workers (HCWs) usually complain about difficulties for washing their hands properly, including several procedure-related problems (skin dryness, minor hand lesions) and other issues (increased workload, interruptions of regular care procedures, performance of tasks without “visible dirtiness”), and because of some important structural problems, mainly the inappropriate number and location of handwashing sinks in most wards and intensive care units (ICUs). Other reasons such as unawareness of the importance of handwashing procedures and personal comfort are sometimes argued. The aim of this study was to assess the effectiveness (reduction of bacterial microflora of hands) of a commercially available alcoholic solution when compared with the standard hygienic handwashing procedure during regular work in clinical wards and ICUs of a large public university hospital in Barcelona (Catalonia, Spain).
AJIC
Zaragoza et al 259
Volume 27, Number 3
PERSONS AND METHODS The study was a prospective, randomized clinical trial, with a crossover design, paired data, and blind evaluation. It was performed at the University of Barcelona Hospital Clinic, an 850-bed tertiary care referral public center located in Barcelona (Catalonia, Spain), during the period from February to March 1997. The Hospital Clinic consists of 12 medical and 8 surgical wards (average number of beds: 24 per ward), and 6 ICUs (average number of beds: 8 per ICU). The study was conducted in 4 randomly selected wards (2 medical and 2 surgical) and in 3 randomly selected ICUs (1 surgical, 1 medicalsurgical, 1 medical). The selection was made by use of a random numbers table applied to the units and ICUs list. The architecture of a regular ward includes a 2-block design of 10 to 12 beds located at the right and left wings of a central nursing station where the HCWs’ handwashing sinks (n = 4) are located. The architecture of the ICUs is less uniform, but the average ratio of handwashing sinks per bed is 1 sink per 2.3 beds, usually located at a central position within the ICU. Eligible HCWs included all permanent and temporary HCWs (faculty, house staff physicians, nurses, and other HCWs) of each ward and ICU. From each category, a random sample of persons was selected. A sample size of 45 was estimated from a population of 175 HCWs (with a 95% confidence level for detecting a 15% difference in each group). All HCWs were asked to volunteer for the study. When an HCW refused to participate, an alternative random selection HCW was included. The selection was made by use of a random numbers table applied to the list of HCWs attending the early morning meeting of each ward and ICU. HCWs of each ward and ICU were thereafter randomly assigned to regular handwashing (liquid soap: water, sodium lauryl sulfate, sodium chloride, glycol distearate, hydrolyzed collagen, cocamide dea, citric acid, hydroxyethilcellulose, with a pH of 5.5, and average 3 mL to 5 mL per pump, standard procedure for hygienic hospital handwashing) or handwashing with the experimental agent (Sterillium, Bode Chemie Hamburg, Germany: liquid 45% 2-propanol, 30% 1-propanol, 0.2% metronium ethyl-sulfate for an average 3 mL to 5 mL per pump) for a total of 15 days by using a crossover design. The protocol for alcoholic solution use includes directions for handwashing (soap and water) before the use of the alcoholic solution whenever there is visible dirtiness. All HCWs were instructed in the use of the alcoholic solution by personal training at the bedside (research nurse), and a written protocol was available at each unit. Three different samples were obtained from each HCW in every experiment. Samples were taken by
printing of both hands onto a blood-agar plate. Sample timing was as follows: sample 1 (T1) was obtained before any handwashing procedure on the first day of the study (baseline sample); sample 2 (T2) was obtained immediately after the handwashing procedure and before any patient care was undertaken; finally, sample 3 (T3) was obtained 10 to 30 minutes after T2, while the HCW performed regular tasks in the ward or ICU. Every HCW included was informed not to change his or her common handwashing procedure and rate during the study period. No special attempt to monitor compliance was made. Agar plates (blood-agar 5%, bioMerieux, France) were incubated overnight at 37°C. The number of colony forming units (CFU) was counted, and the species identification of isolated microorganisms was performed by using standard procedures. 5 The microbiologist responsible for the CFU count was blinded regarding the handwashing method and particular sequence of each sample. Agar plates were used instead of the more sensitive glove-juice method because of their cost and availability. A single research nurse (M.Z.) was responsible for the whole sampling procedure in all wards and ICUs. She stayed in the ward or ICU without any interference in the regular handwashing procedures of HCWs. When the study was finished, a multiple choice questionnaire (available from the authors by request) was given to all HCWs and included a subjective assessment of each handwashing method after use of the product for 2 weeks. Statistical analysis was performed by using the paired t test, with an α level of significance of .05. RESULTS A total of 50 HCWs were included in the study. Of these, 7 were excluded from the final analysis because they were only available for one of the procedures evaluated. The final analysis included paired data for 43 HCWs, each one with 3 samples (T1, T2, and T3) with both handwashing procedures. Microorganisms isolated were mostly coagulase-negative staphylococci, Staphylococcus aureus, Enterobacteriaceae, Corynebacterium spp and Micrococcus spp. All of them are considered normal skin flora. CFU counts for Streptococcus viridans and Bacillus spp, found in few samples, were very low. No differences were seen regarding the species of microorganisms isolated between both handwashing groups. No special microorganism outbreak was observed while the study was in progress in any of the wards and ICUs. The endemic rate of nosocomial methicillin-resistant S aureus infections in our institution was stable throughout the study period.
AJIC
260 Zaragoza et al
June 1999
Table 1. Number of CFU in both handwashing groups, according to ward and sampling timing Water and soap CFU (mean and SD)* T1
Medical wards (HCWs, n = 13) Surgical wards (HCWs, n = 8) ICUs (HCWs, n = 22) Total (HCWs, N = 43)
87 76 82 82
± ± ± ±
78 64 80 75
T2
36 56 40 42
± ± ± ±
Alcoholic solution CFU (mean and SD)
T3
21 65 35 39
62 71 85 76
± ± ± ±
T1
32 56 60 52
71 73 78 75
± ± ± ±
T2
19 34 49 39
9 14 7 9
± ± ± ±
T3
11 13 10 11
56 55 65 61
± ± ± ±
30 53 59 50
Sampling timing: T1: before handwashing; T2: after handwashing; T3: 10 to 30 minutes after T2. *CFU (mean and SD): mean CFU ± standard deviation of the mean.
Table 2. Overall results: Number of CFU before and after handwashing (water and soap vs alcoholic solution) Water and soap handwashing
Number of HCWs CFU mean ± standard deviation Percentage reduction in CFU count
Alcoholic solution handwashing
T1
T2
T1
T2
43 82 ± 75*
43 42 ± 39† 49.6 (P = .002)
43 75 ± 39*
43 9 ± 11† 88.2 (P < .0001)
T1: Before handwashing; T2: after handwashing. *Comparison between both groups in T1 do not reach statistical significance (P = .562) †Comparison between both groups in T2 showed a statistically significant difference (P < .0001).
Table 1 includes the overall results for all wards. Comparisons between the mean number of CFU before and after handwashing with soap and water and with the alcoholic solution did not show any statistically significant difference. Comparisons of the mean number of CFU before and after handwashing between medical wards versus surgical wards, medical wards versus ICUs, and surgical wards versus ICUs did not show any statistically significant differences. In 13 samples (30%) from the soap and water handwashing group, the mean T2 number of CFU found (after handwashing) was greater than the mean T1 number of CFU (before handwashing). This result was not the case when the alcoholic solution group was analyzed, and none of the T2 samples showed higher mean numbers of CFU than the T1 samples in this group. In 33% of the T2 samples (after handwashing) from the alcoholic solution group, no growth of any single microorganism was found. Table 2 shows the overall results for each group of handwashing procedures. The baseline mean number of CFU was similar between both groups (nonsignificant statistical differences). The average reduction in the number of CFU from T1 samples (before handwashing) to T2 samples (after handwashing) was 49.6% for soap and water and 88.2% for the alcoholic solution. When both handwashing methods were compared, the average number of CFU recovered after the handwashing procedure (T2) showed a statistically significant difference in favor of the alcoholic solution (P < .0001). The lasting effect (measured by the CFU after handwashing and 10 to 30 minutes of regular patient care)
also was assessed (T3 samples). No statistically significant difference between both groups was found (mean number of CFU ± standard deviation: soap group: 76 ± 52 CFU; alcoholic solution group: 61 ± 50 CFU; P = .172). When T1 samples (baseline) were compared with T3 samples (after handwashing and 10 to 30 minutes of regular patient care), no statistically significant differences were found within each group. The alcoholic solution was well tolerated by the HCWs. Overall acceptance rate was classified as “good” by 72% of HCWs after use of the product for 2 weeks and “fair” by an additional 18% of HCWs. Of all HCWs studied, 9.3% stated that use of the alcoholic solution worsened minor pre-existing skin conditions, such as hand dryness or erythema. In any case, the use of the alcoholic solution had to be discontinued as a result of these minor problems. DISCUSSION The reduction of the CFU of microflora on hands after hygienic soap and water standard handwashing procedure is well known. In this study, differences in the mean number of CFU before and after both handwashing procedures were highly significant, with a 50% average reduction in the number of CFU. The reduction achieved was greater when the alcoholic solution was used. The lasting effect of pure alcohol is negligible. The mixture used in the alcoholic solution includes 2 different types of alcohol, a quaternary ammonium, and refatting agents. No differences were found when both procedures were evaluated. However, as with any negative result, the statistical power of the test should be
AJIC
Zaragoza et al 261
Volume 27, Number 3
taken into account. The small sample size (N = 43) analyzed could be partially responsible, and larger sample sized studies are needed. A surprising result was observed in the soap and water handwashing group, where 30% of T2 samples (after the procedure) showed higher mean CFU counts than paired T1 samples (before the procedure). As mentioned in the “Methods” section, a research nurse acted as an independent observer for the handwashing procedure and did not interfere with any HCW throughout the study. The most common observed break in the standard procedure was manually turning off the faucet, touching it without paper towel protection. Most handwashing sinks in wards of European hospitals do not have automatic systems for turning on and off faucets, which is a fairly common risk for breaking the standard procedure.6-8 There are few well-designed studies showing the relationship with different handwashing methods and a further reduction in the rate of nosocomial infections. In 1991 Doebbeling et al9 reported a study comparing a chlorhexidine solution with 60% isopropyl alcohol and nonmedicated soap. In spite of the strict study conditions, the handwashing compliance rate (number of real handwashing episodes/number of handwashing opportunities) was less than 50%, and the chlorhexidine solution was found to reduce more effectively the rate of nosocomial infections.9 In a case-control study, Freeman10 demonstrated a 26% risk reduction of acquiring a nosocomial infection (urinary tract infections and pneumonia) when the patient was admitted to a bed with an adjacent sink. However, the existence of available sinks at bedside is not as common as expected in modern hospitals. A recent European study of ICU nosocomial infection control shows that only 34% of all European ICUs had 1 sink per bed, and the figure for Spanish ICUs was even lower (12%).11 The lack of availability of sinks in hospitals is a major limitation for improving the compliance rate of handwashing by HCWs. Many different approaches for improving handwashing rates have been tested (educational programs, reminder cards, posters, electronic devices for sink operation, bar-code lecture devices on sinks) but their success, if any, has been limited to the period in which the program is enforced, and handwashing compliance rates usually return to basal levels thereafter.6-8 The current situation is deemed to be bad enough for assuming noncompliance with handwashing procedures as the bottom line in any infection control program.2 An interesting approach, to educate the patients instead of the HCWs, also has been advocated,1 but its limitations in the ICU setting are important.12 Taking into account all previous experiences, an alternative approach to improve handwashing among HCWs could be to provide them with some sort of “indi-
vidual portable sinks.” This implies that in every nursing station and medical records cart, an alcoholic solution dispenser device must be readily available. If the alcoholic solution is an effective handwashing method, as it seems to be according to the results of our study and others,6 it can be safely used as an option to regular handwashing in special situations (ICU setting, work overload, difficult access to ward sinks). The use of these solutions has several other advantages: it reduces the risk of breaking the regular handwashing procedure (by avoiding touching the faucet directly with hands after handwashing), it is well tolerated by the HCWs, and it is easier and faster to use.6 The acquisition cost of the alcoholic solution is higher than the cost of soap and water, but formal cost-effectiveness studies are needed (including other factors such as nursing time savings and reduction in the rate of nosocomial infections) before making a final decision. Although we strongly support the regular use of hygienic soap and water handwashing procedures as the gold standard, we believe that the use of alcoholic solutions deserves more attention, especially in situations in which the compliance rate is hampered by architectural problems or nursing work overload, a common situation nowadays in many hospitals. References 1. Jarvis WR. Handwashing—the Semmelweis lesson forgotten? Lancet 1994;344:1311-2. 2. Goldmann D, Larson E. Hand-washing and nosocomial infections. N Engl J Med 1992;327:120-2. 3. Graham M. Frequency and duration of hand washing in an intensive care unit. AJIC Am J Infect Control 1990;18:77-81. 4. Ehrenkranz NJ. Bland soap handwashing or hand antisepsis? The pressing need for clarity. Infect Control Hosp Epidemiol 1992;13:299-301. 5. Murray PR. Manual of clinical microbiology. Washington (DC): ASM Press; 1995. 6. Voss A, Widmer AF. No time for handwashing? Handwashing versus alcoholic rub: can we afford 100% compliance. Infect Control Hosp Epidemiol 1997;18:205-8. 7. Larson EL, Bryan JL, Adler LM. A multifaceted approach to changing handwashing behavior. AJIC Am J Infect Control 1997;25:3-10. 8. Larson EL. APIC Guideline for handwashing and hand antisepsis in health care settings. AJIC Am J Infect Control 1995;23:251-69. 9. Doebbelling BN, Stanley GL, Sheetz CT, Pfaller MA, Houston AK, Annis L, et al. Comparative efficacy of alternative hand-washing agents in reducing nosocomial infections in Intensive Care Units. N Engl J Med 1992;327:88-93. 10. Freeman J. Prevention of nosocomial infections by location of sinks for hand washing adjacent adjacent to the bedside [abstract #60]. Proceedings of the 33rd International Conference on Antimicrobial Agents and Chemotherapy; 1993; New Orleans, La. 11. Vincent JL, Bihari DJ, Suter PM, Bruining HA, White J, Nicolas MH, et al. The prevalence of nosocomial infection in intensive care units in Europe. Results of the European prevalence of infection in intensive care (EPIC) study. JAMA 1995;274:639-44. 12. Trilla A, Vila J, Zaragoza M, Sallés M. Acinetobacter infections, handwashing and the intensive care unit. Lancet 1994;345:123.
Background: The effectiveness of an alcoholic solution compared with the standard hygienic handwashing procedure during regular work in clinical wards and intensive care units of a large public university hospital in Barcelona was assessed. Methods: A prospective, randomized clinical trial with crossover design, paired data, and blind evaluation was done. Eligible health care workers (HCWs) included permanent and temporary HCWs of wards and intensive care units. From each category, a random sample of persons was selected. HCWs were randomly assigned to regular handwashing (liquid soap and water) or handwashing with the alcoholic solution by using a crossover design. The number of colony-forming units on agar plates from hands printing in 3 different samples was counted. Results: A total of 47 HCWs were included. The average reduction in the number of colony-forming units from samples before handwashing to samples after handwashing was 49.6% for soap and water and 88.2% for the alcoholic solution. When both methods were compared, the average number of colony-forming units recovered after the procedure showed a statistically significant difference in favor of the alcoholic solution (P < .001). The alcoholic solution was well tolerated by HCWs. Overall acceptance rate was classified as “good” by 72% of HCWs after 2 weeks use. Of all HCWs included, 9.3% stated that the use of the alcoholic solution worsened minor pre-existing skin conditions. Conclusions: Although the regular use of hygienic soap and water handwashing procedures is the gold standard, the use of alcoholic solutions is effective and safe and deserves more attention, especially in situations in which the handwashing compliance rate is hampered by architectural problems (lack of sinks) or nursing work overload. (AJIC Am J Infect Control 1999;27:258-61)
Handwashing is still considered the most important and effective infection control measure to prevent transmission of nosocomial infections.1 An essential component of any infection control program is the handwashing procedure, before and after all direct contact with patients, biologic fluids, devices, or medical equipment at risk of being contaminated with nosocomial microorganisms. However, several studies performed in many US and European hospitals regularly
From the Hospital Hygiene and Central Sterilization Unit, Microbiology Laboratory, Preventive Medicine Department, and the Hospital Epidemiology and Health Services Research Unit, Hospital Clinic—University of Barcelona, Biomedical Research Institute August Pi I Sunyer (IDIBAPS). Partially supported by a research grant from Beiersdorf SA (Spain). Reprint requests: Magda Zaragoza, RN, Hospital Hygiene and Central Sterilization Service, Hospital Clinic—University of Barcelona, Villarroel 170 (Esc 4-1), 08036-Barcelona, Spain. Copyright © 1999 by the Association for Professionals in Infection Control and Epidemiology, Inc. 0196-6553/99/$8.00 + 0
258
17/46/97622
and repeatedly show that handwashing compliance rate is lower than 50%.2-4 Handwashing with soap and water significantly reduces the burden of microorganisms of the hands. Other agents such as alcoholic solutions with or without other compounds (quaternary ammonia) mixed with products for preventing skin dryness are increasingly used in Europe. Health care workers (HCWs) usually complain about difficulties for washing their hands properly, including several procedure-related problems (skin dryness, minor hand lesions) and other issues (increased workload, interruptions of regular care procedures, performance of tasks without “visible dirtiness”), and because of some important structural problems, mainly the inappropriate number and location of handwashing sinks in most wards and intensive care units (ICUs). Other reasons such as unawareness of the importance of handwashing procedures and personal comfort are sometimes argued. The aim of this study was to assess the effectiveness (reduction of bacterial microflora of hands) of a commercially available alcoholic solution when compared with the standard hygienic handwashing procedure during regular work in clinical wards and ICUs of a large public university hospital in Barcelona (Catalonia, Spain).
AJIC
Zaragoza et al 259
Volume 27, Number 3
PERSONS AND METHODS The study was a prospective, randomized clinical trial, with a crossover design, paired data, and blind evaluation. It was performed at the University of Barcelona Hospital Clinic, an 850-bed tertiary care referral public center located in Barcelona (Catalonia, Spain), during the period from February to March 1997. The Hospital Clinic consists of 12 medical and 8 surgical wards (average number of beds: 24 per ward), and 6 ICUs (average number of beds: 8 per ICU). The study was conducted in 4 randomly selected wards (2 medical and 2 surgical) and in 3 randomly selected ICUs (1 surgical, 1 medicalsurgical, 1 medical). The selection was made by use of a random numbers table applied to the units and ICUs list. The architecture of a regular ward includes a 2-block design of 10 to 12 beds located at the right and left wings of a central nursing station where the HCWs’ handwashing sinks (n = 4) are located. The architecture of the ICUs is less uniform, but the average ratio of handwashing sinks per bed is 1 sink per 2.3 beds, usually located at a central position within the ICU. Eligible HCWs included all permanent and temporary HCWs (faculty, house staff physicians, nurses, and other HCWs) of each ward and ICU. From each category, a random sample of persons was selected. A sample size of 45 was estimated from a population of 175 HCWs (with a 95% confidence level for detecting a 15% difference in each group). All HCWs were asked to volunteer for the study. When an HCW refused to participate, an alternative random selection HCW was included. The selection was made by use of a random numbers table applied to the list of HCWs attending the early morning meeting of each ward and ICU. HCWs of each ward and ICU were thereafter randomly assigned to regular handwashing (liquid soap: water, sodium lauryl sulfate, sodium chloride, glycol distearate, hydrolyzed collagen, cocamide dea, citric acid, hydroxyethilcellulose, with a pH of 5.5, and average 3 mL to 5 mL per pump, standard procedure for hygienic hospital handwashing) or handwashing with the experimental agent (Sterillium, Bode Chemie Hamburg, Germany: liquid 45% 2-propanol, 30% 1-propanol, 0.2% metronium ethyl-sulfate for an average 3 mL to 5 mL per pump) for a total of 15 days by using a crossover design. The protocol for alcoholic solution use includes directions for handwashing (soap and water) before the use of the alcoholic solution whenever there is visible dirtiness. All HCWs were instructed in the use of the alcoholic solution by personal training at the bedside (research nurse), and a written protocol was available at each unit. Three different samples were obtained from each HCW in every experiment. Samples were taken by
printing of both hands onto a blood-agar plate. Sample timing was as follows: sample 1 (T1) was obtained before any handwashing procedure on the first day of the study (baseline sample); sample 2 (T2) was obtained immediately after the handwashing procedure and before any patient care was undertaken; finally, sample 3 (T3) was obtained 10 to 30 minutes after T2, while the HCW performed regular tasks in the ward or ICU. Every HCW included was informed not to change his or her common handwashing procedure and rate during the study period. No special attempt to monitor compliance was made. Agar plates (blood-agar 5%, bioMerieux, France) were incubated overnight at 37°C. The number of colony forming units (CFU) was counted, and the species identification of isolated microorganisms was performed by using standard procedures. 5 The microbiologist responsible for the CFU count was blinded regarding the handwashing method and particular sequence of each sample. Agar plates were used instead of the more sensitive glove-juice method because of their cost and availability. A single research nurse (M.Z.) was responsible for the whole sampling procedure in all wards and ICUs. She stayed in the ward or ICU without any interference in the regular handwashing procedures of HCWs. When the study was finished, a multiple choice questionnaire (available from the authors by request) was given to all HCWs and included a subjective assessment of each handwashing method after use of the product for 2 weeks. Statistical analysis was performed by using the paired t test, with an α level of significance of .05. RESULTS A total of 50 HCWs were included in the study. Of these, 7 were excluded from the final analysis because they were only available for one of the procedures evaluated. The final analysis included paired data for 43 HCWs, each one with 3 samples (T1, T2, and T3) with both handwashing procedures. Microorganisms isolated were mostly coagulase-negative staphylococci, Staphylococcus aureus, Enterobacteriaceae, Corynebacterium spp and Micrococcus spp. All of them are considered normal skin flora. CFU counts for Streptococcus viridans and Bacillus spp, found in few samples, were very low. No differences were seen regarding the species of microorganisms isolated between both handwashing groups. No special microorganism outbreak was observed while the study was in progress in any of the wards and ICUs. The endemic rate of nosocomial methicillin-resistant S aureus infections in our institution was stable throughout the study period.
AJIC
260 Zaragoza et al
June 1999
Table 1. Number of CFU in both handwashing groups, according to ward and sampling timing Water and soap CFU (mean and SD)* T1
Medical wards (HCWs, n = 13) Surgical wards (HCWs, n = 8) ICUs (HCWs, n = 22) Total (HCWs, N = 43)
87 76 82 82
± ± ± ±
78 64 80 75
T2
36 56 40 42
± ± ± ±
Alcoholic solution CFU (mean and SD)
T3
21 65 35 39
62 71 85 76
± ± ± ±
T1
32 56 60 52
71 73 78 75
± ± ± ±
T2
19 34 49 39
9 14 7 9
± ± ± ±
T3
11 13 10 11
56 55 65 61
± ± ± ±
30 53 59 50
Sampling timing: T1: before handwashing; T2: after handwashing; T3: 10 to 30 minutes after T2. *CFU (mean and SD): mean CFU ± standard deviation of the mean.
Table 2. Overall results: Number of CFU before and after handwashing (water and soap vs alcoholic solution) Water and soap handwashing
Number of HCWs CFU mean ± standard deviation Percentage reduction in CFU count
Alcoholic solution handwashing
T1
T2
T1
T2
43 82 ± 75*
43 42 ± 39† 49.6 (P = .002)
43 75 ± 39*
43 9 ± 11† 88.2 (P < .0001)
T1: Before handwashing; T2: after handwashing. *Comparison between both groups in T1 do not reach statistical significance (P = .562) †Comparison between both groups in T2 showed a statistically significant difference (P < .0001).
Table 1 includes the overall results for all wards. Comparisons between the mean number of CFU before and after handwashing with soap and water and with the alcoholic solution did not show any statistically significant difference. Comparisons of the mean number of CFU before and after handwashing between medical wards versus surgical wards, medical wards versus ICUs, and surgical wards versus ICUs did not show any statistically significant differences. In 13 samples (30%) from the soap and water handwashing group, the mean T2 number of CFU found (after handwashing) was greater than the mean T1 number of CFU (before handwashing). This result was not the case when the alcoholic solution group was analyzed, and none of the T2 samples showed higher mean numbers of CFU than the T1 samples in this group. In 33% of the T2 samples (after handwashing) from the alcoholic solution group, no growth of any single microorganism was found. Table 2 shows the overall results for each group of handwashing procedures. The baseline mean number of CFU was similar between both groups (nonsignificant statistical differences). The average reduction in the number of CFU from T1 samples (before handwashing) to T2 samples (after handwashing) was 49.6% for soap and water and 88.2% for the alcoholic solution. When both handwashing methods were compared, the average number of CFU recovered after the handwashing procedure (T2) showed a statistically significant difference in favor of the alcoholic solution (P < .0001). The lasting effect (measured by the CFU after handwashing and 10 to 30 minutes of regular patient care)
also was assessed (T3 samples). No statistically significant difference between both groups was found (mean number of CFU ± standard deviation: soap group: 76 ± 52 CFU; alcoholic solution group: 61 ± 50 CFU; P = .172). When T1 samples (baseline) were compared with T3 samples (after handwashing and 10 to 30 minutes of regular patient care), no statistically significant differences were found within each group. The alcoholic solution was well tolerated by the HCWs. Overall acceptance rate was classified as “good” by 72% of HCWs after use of the product for 2 weeks and “fair” by an additional 18% of HCWs. Of all HCWs studied, 9.3% stated that use of the alcoholic solution worsened minor pre-existing skin conditions, such as hand dryness or erythema. In any case, the use of the alcoholic solution had to be discontinued as a result of these minor problems. DISCUSSION The reduction of the CFU of microflora on hands after hygienic soap and water standard handwashing procedure is well known. In this study, differences in the mean number of CFU before and after both handwashing procedures were highly significant, with a 50% average reduction in the number of CFU. The reduction achieved was greater when the alcoholic solution was used. The lasting effect of pure alcohol is negligible. The mixture used in the alcoholic solution includes 2 different types of alcohol, a quaternary ammonium, and refatting agents. No differences were found when both procedures were evaluated. However, as with any negative result, the statistical power of the test should be
AJIC
Zaragoza et al 261
Volume 27, Number 3
taken into account. The small sample size (N = 43) analyzed could be partially responsible, and larger sample sized studies are needed. A surprising result was observed in the soap and water handwashing group, where 30% of T2 samples (after the procedure) showed higher mean CFU counts than paired T1 samples (before the procedure). As mentioned in the “Methods” section, a research nurse acted as an independent observer for the handwashing procedure and did not interfere with any HCW throughout the study. The most common observed break in the standard procedure was manually turning off the faucet, touching it without paper towel protection. Most handwashing sinks in wards of European hospitals do not have automatic systems for turning on and off faucets, which is a fairly common risk for breaking the standard procedure.6-8 There are few well-designed studies showing the relationship with different handwashing methods and a further reduction in the rate of nosocomial infections. In 1991 Doebbeling et al9 reported a study comparing a chlorhexidine solution with 60% isopropyl alcohol and nonmedicated soap. In spite of the strict study conditions, the handwashing compliance rate (number of real handwashing episodes/number of handwashing opportunities) was less than 50%, and the chlorhexidine solution was found to reduce more effectively the rate of nosocomial infections.9 In a case-control study, Freeman10 demonstrated a 26% risk reduction of acquiring a nosocomial infection (urinary tract infections and pneumonia) when the patient was admitted to a bed with an adjacent sink. However, the existence of available sinks at bedside is not as common as expected in modern hospitals. A recent European study of ICU nosocomial infection control shows that only 34% of all European ICUs had 1 sink per bed, and the figure for Spanish ICUs was even lower (12%).11 The lack of availability of sinks in hospitals is a major limitation for improving the compliance rate of handwashing by HCWs. Many different approaches for improving handwashing rates have been tested (educational programs, reminder cards, posters, electronic devices for sink operation, bar-code lecture devices on sinks) but their success, if any, has been limited to the period in which the program is enforced, and handwashing compliance rates usually return to basal levels thereafter.6-8 The current situation is deemed to be bad enough for assuming noncompliance with handwashing procedures as the bottom line in any infection control program.2 An interesting approach, to educate the patients instead of the HCWs, also has been advocated,1 but its limitations in the ICU setting are important.12 Taking into account all previous experiences, an alternative approach to improve handwashing among HCWs could be to provide them with some sort of “indi-
vidual portable sinks.” This implies that in every nursing station and medical records cart, an alcoholic solution dispenser device must be readily available. If the alcoholic solution is an effective handwashing method, as it seems to be according to the results of our study and others,6 it can be safely used as an option to regular handwashing in special situations (ICU setting, work overload, difficult access to ward sinks). The use of these solutions has several other advantages: it reduces the risk of breaking the regular handwashing procedure (by avoiding touching the faucet directly with hands after handwashing), it is well tolerated by the HCWs, and it is easier and faster to use.6 The acquisition cost of the alcoholic solution is higher than the cost of soap and water, but formal cost-effectiveness studies are needed (including other factors such as nursing time savings and reduction in the rate of nosocomial infections) before making a final decision. Although we strongly support the regular use of hygienic soap and water handwashing procedures as the gold standard, we believe that the use of alcoholic solutions deserves more attention, especially in situations in which the compliance rate is hampered by architectural problems or nursing work overload, a common situation nowadays in many hospitals. References 1. Jarvis WR. Handwashing—the Semmelweis lesson forgotten? Lancet 1994;344:1311-2. 2. Goldmann D, Larson E. Hand-washing and nosocomial infections. N Engl J Med 1992;327:120-2. 3. Graham M. Frequency and duration of hand washing in an intensive care unit. AJIC Am J Infect Control 1990;18:77-81. 4. Ehrenkranz NJ. Bland soap handwashing or hand antisepsis? The pressing need for clarity. Infect Control Hosp Epidemiol 1992;13:299-301. 5. Murray PR. Manual of clinical microbiology. Washington (DC): ASM Press; 1995. 6. Voss A, Widmer AF. No time for handwashing? Handwashing versus alcoholic rub: can we afford 100% compliance. Infect Control Hosp Epidemiol 1997;18:205-8. 7. Larson EL, Bryan JL, Adler LM. A multifaceted approach to changing handwashing behavior. AJIC Am J Infect Control 1997;25:3-10. 8. Larson EL. APIC Guideline for handwashing and hand antisepsis in health care settings. AJIC Am J Infect Control 1995;23:251-69. 9. Doebbelling BN, Stanley GL, Sheetz CT, Pfaller MA, Houston AK, Annis L, et al. Comparative efficacy of alternative hand-washing agents in reducing nosocomial infections in Intensive Care Units. N Engl J Med 1992;327:88-93. 10. Freeman J. Prevention of nosocomial infections by location of sinks for hand washing adjacent adjacent to the bedside [abstract #60]. Proceedings of the 33rd International Conference on Antimicrobial Agents and Chemotherapy; 1993; New Orleans, La. 11. Vincent JL, Bihari DJ, Suter PM, Bruining HA, White J, Nicolas MH, et al. The prevalence of nosocomial infection in intensive care units in Europe. Results of the European prevalence of infection in intensive care (EPIC) study. JAMA 1995;274:639-44. 12. Trilla A, Vila J, Zaragoza M, Sallés M. Acinetobacter infections, handwashing and the intensive care unit. Lancet 1994;345:123.