Hand hygiene: An evidence-based review for surgeons

International Journal of Surgery (2006) 4, 53e65 www.int-journal-surgery.com REVIEW Hand hygiene: An evidence-based review for surgeons C.R. Nicolay Academic Surgical Unit, 10th Floor QEQM Wing, St Mary s Hospital, Praed St, London W2 1NY, UK KEYWORDS Hand hygiene; Handwashing; Infection; Alcohol; Handrub Abstract This review of the literature discusses the scientific evidence behind using different hand hygiene agents on the surgical ward, and in theatre for preoperative disinfection. It considers the mechanism of action of the agents and their effectiveness against different pathogens, as well as possible future agents, and how they are tested. It addresses problems such as the poor compliance with hand hygiene guidelines by healthcare workers (especially doctors) and investigates what can be done to improve compliance. Finally, it demonstrates the reduction in hospital acquired infection (HAI) rate that can be achieved by improving hand hygiene compliance, and shows that the savings associated with this easily outweigh the cost. ª 2005 Surgical Associates Ltd. Published by Elsevier Ltd. All rights reserved. Introduction It has been clear since the 1840s when both Oliver Holmes 1 and Ignaz Semmelweis 2 independently reported the contagious nature of puerperal fever, that healthcare workers should cleanse their hands before contact with patients. Semmelweis insisted that doctors who had performed autopsies must wash their hands before delivering a baby, and in one step he reduced mortality from streptococcal sepsis by nearly 90%. Sixteen decades later, many studies have confirmed that by healthcare workers E-mail address: christopher.nicolay@doctors.org.uk simply washing their hands between contact with patients, the hospital acquired infection rate can be reduced. 3 However, adherence by doctors to hand hygiene guidelines is often unacceptably poor at below 50%, 4e9 and is usually poorer than nurses and other healthcare workers. Approximately 10% of hospital inpatients are suffering from an infection acquired following their admission, 10,11 and Department of Health guidance suggests that approximately 30% of these HAIs could be avoided by better application of existing knowledge and realistic infection control practices. 12 The 70% of HAIs that are not preventable demonstrates the scale of endogenous infection compared with cross-infection. Hospital 1743-9191/$ - see front matter ª 2005 Surgical Associates Ltd. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijsu.2005.06.002

54 C.R. Nicolay acquired infections with antibiotic-resistant bacteria such as methicillin-resistant Staphylococcus aureus (MRSA) are not uncommon and can be difficult and expensive to treat. According to a study funded by the Department of Health, 13 adult inpatients in common specialties who develop a hospital acquired infection remain in hospital 2.5 times longer, incur hospital costs nearly 3 times higher, and incur higher general practitioner, district nurse, and hospital costs after discharge from hospital than uninfected patients. Even after adjustment for various factors including age, diagnosis and the number of comorbidities, patients with an HAI are 7 times more likely to die in hospital than uninfected patients. It has been estimated that 5000 deaths a year in the UK are primarily attributable to HAI (1% of all deaths), and in a further 15,000 deaths, HAI may be a significant contributor. The estimated mean additional hospital cost of an HAI in a surgical patient is about 4000. An extrapolation of figures to NHS hospitals in England suggests a total additional annual cost of HAI in adult patients of all specialties to be nearly 1000 million. A brief summary of part of the National Institute of Clinical Excellence (NICE) evidence-based clinical guidelines for the prevention of hospital acquired infection 14 is as follows: 1. Hands must be decontaminated immediately before each and every episode of direct patient contact or care and after any activity or contact that potentially results in hands becoming contaminated. 2. Hands that are visibly soiled or potentially grossly contaminated with dirt or organic material, must be washed with liquid soap and water. 3. Hands must be decontaminated, preferably with an alcohol-based handrub unless hands are visibly soiled, between caring for different patients and between different care activities for the same patient. 4. An effective handwashing technique involves three stages: preparation, washing and rinsing, and drying. Preparation requires wetting hands under tepid running water before applying liquid soap or an antimicrobial preparation. The handwash solution must come into contact with all of the surfaces of the hand. The hands must be rubbed together vigorously for a minimum of 10e15 s, paying particular attention to the tips of the fingers, the thumbs and the areas between the fingers. Hands should be rinsed thoroughly before drying with good quality paper towels. With hospital hygiene and the superbug MRSA often in the media, there is a vast amount of literature published on this broad topic, and much research forming the basis of hospital hand hygiene guidelines worldwide. John Boyce, MD and Didier Pittet, MD published a comprehensive 48- page report on the subject 15 and this review summarises the parts of the literature particularly relevant to the field of surgery, together with more recent work published. Normal bacterial skin flora It is important to appreciate normal bacterial skin flora, before considering the options for hand hygiene. Normal human skin is colonised by bacteria, but to different extents depending on the site, for example: 100! 10 4 CFU (colony forming units) cm ÿ2 on the scalp, 50! 10 4 CFU cm ÿ2 in the axillae, 4! 10 4 CFU cm ÿ2 on the abdomen, and 1! 10 4 CFU cm ÿ2 on the forearm. 16 The two major groups of bacteria on the skin are those that normally reside on it (resident flora), and contaminants (transient flora). Transient flora such as S. aureus, gram-negative bacteria and yeast inhabit the superficial layers of the skin and are those acquired by doctors during patient contact, or by contact with contaminated surfaces. These are the bacteria that are most responsible for hospital acquired infections and are the ones that are more easily removed by hand hygiene. Resident flora such as coagulase-negative staphylococci (e.g. Staphylococcus epidermidis) and Corynebacteria are attached to deeper layers of the skin and are less easy to remove by hand hygiene and also less likely to cause HAIs. However, low virulence resident flora can cause huge complications for patients, for example in orthopaedic wound infections. Review of hand hygiene agents Plain soap The term soap is a class name for the sodium and potassium salts of stearic acid and other fatty acids, and they are available in different forms, commonly bar or liquid. They have little if any antibacterial activity but may remove loose transient flora. It has been shown that a 1 min handwash with plain soap reduces the artificial bacterial contamination of hands by a log 10 reduction factor of only 2.8, compared with a value of 4.6 for the alcohol propan-2-ol (60% v/v). 17 In a small study,

Hand hygiene: An evidence-based review 55 plain soap handwashing failed to prevent gramnegative bacterial transfer to a catheter in 11 of 12 experiments, 18 and it can even carry the risk of spreading bacterial contamination. 19,42 Over time, soap usage can make the skin s ph more alkaline, and cause significant skin dryness and irritation, as measured by self-assessment, external visual assessment and epidermal water content measurement (skin capacitance). 20 This effect can be reduced by adding emollients to preparations or by using hand creams. The combination of dryness and removing the natural defensive acidic ph of the skin can lead to easier colonisation with potential pathogens. Alcohols Alcohol-based hand hygiene products contain either ethanol, propan-2-ol (isopropanol, isopropyl alcohol), propan-1-ol (n-propanol, n-propyl alcohol), or combinations thereof (Fig. 1). They have antimicrobial activity due to their ability to denature proteins, 21 and concentrations of 60e95% are most effective, with higher concentrations being less potent because water is needed for the denaturation process. 21 In vitro, alcohols have powerful antibacterial activity against gram-positive and gram-negative bacteria including antibiotic-resistant organisms such as MRSA, Mycobacterium tuberculosis and certain fungi. 21e23 Enveloped lipophilic viruses such as HSV, HIV, RSV and influenza virus are also susceptible. 21,24,25 Hepatitis B and hepatitis C viruses are less susceptible but destroyed by 60e70% alcohol. 26 Alcohols have very poor in vitro activity against bacterial spores, non-enveloped viruses and protozoan cysts. In vivo, alcohols effectively reduce bacterial counts on hands 22 and the log 10 reduction factor of test bacteria from artificially contaminated hands averages 4.6 after a 1-min application of 60% propan-2-ol 17 (the European EN 1500 reference standard for alcohol-based handrub products). Although rapidly bactericidal when applied to skin the effect is short-lived and alcohol gels unfortunately lack the important characteristic of residual activity, 31 but re-growth does occur more slowly after use. 27 Despite very poor in vitro activity against non-enveloped viruses, in vivo activity has been shown, with 70% propan-2-ol and 70% ethanol being more effective than both soap and other antiseptic agents in reducing rotavirus titers (a common cause of infantile gastroenteritis) on fingerpads. 28,29 Although alcohol use is not recommended when hands are visibly soiled or potentially grossly contaminated with dirt or organic material, when small amounts of proteinaceous material such as blood is present, ethanol and propan-2-ol may reduce bacterial counts more than plain soap or antiseptic agents. 30 Alcohols have been shown to prevent the transfer of pathogens from healthcare workers. Gramnegative bacilli were transferred from a colonised patient s skin to a piece of catheter material via the hands of nurses in only 2 out of 12 occasions after alcohol handrub, compared with 11 out of 12 occasions after handwashing with soap and water. 18 Alcohol-based products are more effective for standard hand hygiene by healthcare workers than plain soap or antimicrobial soaps. 18,31e42 In nearly all trials, alcohol reduced antibacterial counts on hands more than washing with plain soap or detergents containing povidoneeiodine, 4% chlorhexidine, or triclosan alone. Importantly, looking at antibiotic-resistant organisms, alcohol-based products reduced the number of pathogens recovered from the hands of healthcare workers more effectively than soap and water. 43,44 The efficacy of alcohol-based products is influenced by several factors including the amount of alcohol used, the concentration, the time contact Figure 1 From left to right: ethanol, propan-1-ol and propan-2-ol molecular structures.

56 C.R. Nicolay lasts for, the type of alcohol and whether hands are wet or not. For example, applying 1 ml of alcohol is much less effective than using 3 ml. 45 Frequent use of alcohol-based formulations can cause skin drying unless other substances such as glycerol and emollients, etc. are added. Interestingly, and even more in the favour of alcohol use, alcohol-based products containing emollients cause significantly less skin irritation than soaps or antimicrobial detergents tested. 20,42,46 Chlorhexidine Chlorhexidine gluconate is a cationic bisbiguanide developed in the UK in the 1950s (Fig. 2) and is found in antiseptic solutions in different concentrations e.g. Hibiscrub â and Hibisol â, as well as other products such as dental mouthwashes. It is antibacterial due to its ability to attach to and disrupt cytoplasmic membranes leading to cell content precipitation, 47 but it has a slower immediate activity than alcohols. It has good activity against gram-positive bacteria, but less against gram-negative and fungi. Chlorhexidine in alcohol has been shown to have better activity against MRSA in vitro than both aqueous chlorhexidine and povidoneeiodine. 48,49 It has minimal activity against tubercle bacilli and it is not sporicidal. It also has in vitro activity against enveloped viruses such as HSV, HIV, CMV, RSV and influenza, but less activity against non-enveloped viruses. As chlorhexidine is a cation, its activity is reduced by natural soaps, inorganic anions, nonionic surfactants and hand creams containing anionic emulsifying agents. 50 Aqueous or detergent formulations containing 0.5% or 0.75% chlorhexidine are more effective than plain soap but less effective than 4% chlorhexidine gluconate. Importantly, unlike alcohol, chlorhexidine has substantial residual activity, 31,51,52 and so addition of low concentrations (0.5e1%) to alcohol-based preparations gives better residual activity than alcohol alone. Chlorhexidine is safe to use 53 with minimal skin absorption. Contact with eyes in concentrations greater than 1% must be avoided as it can cause conjunctivitis and severe corneal damage. Its ototoxicity prevents its use in surgery of the inner or middle ear, and contact with brain and meninges should also be avoided. Iodine Iodine has been used as an antiseptic since the 1800s but because of skin irritation and discolouring, it has been mostly replaced by iodophors (iodine, iodide or triiodide together with a high molecular weight polymer carrier) as the active component in antiseptics. Iodine rapidly penetrates the cell wall of bacteria and forms complexes with amino acids and unsaturated fatty acids leading to disrupted protein synthesis and cell membranes. The amount of free molecular iodine present determines the level of antimicrobial activity, for example 10% povidoneeiodine formulations contain 1% available iodine giving a 1 ppm free iodine concentration. 54 The combination with various polymer carriers increases iodine solubility, reduces skin irritation and promotes sustained iodine release. These polymers are often polyvinyl pyrrolidone (povidone) (Fig. 3) and ethoxylated nonionic detergents (poloxamers). Figure 2 Chlorhexidine (free base) molecular structure.

Hand hygiene: An evidence-based review 57 * Figure 3 Iodine and iodophors have antibacterial action against gram-positive, gram-negative bacteria and mycobacteria, viruses and fungi, but at the concentrations used, they are usually not sporicidal. In vivo studies have shown that they reduce the number of viable organisms recovered from the hands of personnel. 37,41,44 The antibacterial activity is significantly reduced if organic material is present such as blood or sputum. 55 Future applications of povidoneeiodine include MRSA eradication from nasal mucosa and other mucous membranes, and treatment for herpes simplex and Chlamydia infections. 56 Iodophors cause less skin irritation and fewer allergic reactions than iodine but more irritant contact dermatitis than other antiseptics commonly used today. 19 Triclosan Triclosan (2,4,4#-trichloro-2#-hydroxydiphenyl ether) is a nonionic colourless substance developed in the 1960s and introduced into soaps and other products such as washing-up liquid, mouthwash and toothpaste (Fig. 4). Concentrations of 0.2e2% have antimicrobial activity by entering bacterial cells and affecting RNA, fatty acid and protein synthesis. 57 More recent research suggests this is due to Figure 4 N I Povidoneeiodine molecular structure. I O Triclosan molecular structure. n * triclosan binding to the active site of enoylacyl carrier protein reductase, an enzyme that is important for completing cycles of elongation in type II fatty acid synthase systems. 58,59 Triclosan has a broad range of antimicrobial activity but is often bacteriostatic. It has significantly greater activity against gram-positive organisms, particularly MRSA 60 (introducing triclosan preparations eliminated MRSA outbreaks in two independent neonatal units 61,62 ) than gram-negative and has reasonable activity against mycobacteria and Candida spp. In several studies, however, log 10 reductions of bacterial counts on hands after washing with triclosan have been poorer than after chlorhexidine, iodophors or alcohol-based products. 31,63 Future agents Clearly, the search for the optimum hand hygiene agent is still underway. With such factors as skin ph, skin irritation, allergy, effective concentration, interaction with other substances, e.g. organic matter, and human preference such as ease of rinsing, odour, etc., it is a difficult challenge. A recent study looking at the effectiveness of rubbing hands with a 4% w/w hypochlorite solution until the hands are slippery (about 5 min), as recommended by Semmelweis showed that it was 30 times more effective than a 1-min rub with 60% propan-2-ol. 64 It is, however, often irritating to the skin, has a strong odour, and compliance is already poor without introducing a 5-min washing regimen. Future products are likely to be combinations of existing substances at different concentrations or the combination of novel compounds with existing products to optimise their activity, for example improving the residual antimicrobial activity of alcohols. A new preparation that has persistent antimicrobial activity on surfaces and human skin has been created by adding silver-containing polymers to an ethanol carrier e Surfacine â. Microorganisms contacting this intelligent coating accumulate silver until their toxicity threshold is exceeded, and they eventually lyse and detach from the surface. 65 Research must still continue because just as strains of antibiotic-resistant bacteria emerge, resistance to antiseptic agents could become a problem in the future. Strains of S. aureus that demonstrate increased resistance to chlorhexidine and triclosan have already been isolated. The future environmental effects of rinsing thousands of litres of these agents into the water supply also need to be borne in mind.

58 C.R. Nicolay Testing hand hygiene substances e EN 1499 and EN 1500 EN 1499 is the European standard by which antiseptic liquid soaps must demonstrate efficacy under practical conditions compared with the reference, plain soap, tested on Escherichia coli K12 (NTCC 10538). The product should be significantly more effective than the reference soap. EN 1500 is the standard by which products for hygienic hand disinfection such as hand rinses or gels must demonstrate efficacy under practical conditions in comparison with the reference, propan-2-ol (60% v/v) tested on E. coli K12 (NTCC 10538). The product should be significantly more effective than this reference alcohol. 17 These test protocols involve a crossover design with 12 to 15 volunteers (not doctors or nurses) using the hand hygiene product for 1 min. This is despite the fact that numerous observational studies show that doctors and other healthcare workers only spend between 7 and 24 s washing their hands. 15 Indeed, the NICE guidelines themselves state that hands must be rubbed together vigorously for a minimum of 10 to 15 s. 14 Protocols fail to use protein contamination of the hands that can interact with the agents, and also the bacterial skin flora of doctors hands may be very different to that of volunteers. These shortcomings in testing unfortunately mean that there is little data on the efficacy of these products under the conditions in which they will actually be used. This problem was reported in the Lancet in 2002, with no alcohol gel tested meeting EN 1500 requirements within 30 s of application. 17 In fact, the 30-s efficacy of most alcohol gels is closer to a simple handwash with soap than to the reference hand disinfection, i.e. only a log 10 reduction factor of 2.8, rather than 4.6. Sterillium â gel was the only alcohol gel that passed EN 1500 within 30 s. Technique is also an important variable in determining the effectiveness of hand hygiene products (Fig. 5). A study published in 2004 demonstrated that the mean log 10 reduction in healthcare workers using alcohol-based handrub was only 2.0, with 25% achieving less than 1.1. 66 Years of experience was the single most important factor predicting antimicrobial activity, demonstrating the importance of technique and senior staff leading by example and how training should be given before switching policies from handwashing to alcohol handrub. The fact remains though that it is not really known what log 10 reduction in hand bacterial colonisation is needed to prevent the transmission of bacteria responsible for HAIs. The most powerful evidence promoting the use of different agents does not come from laboratory testing and the counting of colony forming units on agar plates, but rather from an observed fall in the HAI rate after their introduction and use on the ward. Preoperative hand disinfection Lister recommended the application of carbolic acid (phenol) to the hands of surgeons in the 1800s and since then the preoperative ritual of washing hands and forearms with an antiseptic has been a common practice, with a presumed 100% compliance (unlike hand hygiene on the ward). No randomised controlled trials have ever taken place to show that surgical infection rates are reduced by using an antiseptic agent rather than plain soap. However, there is evidence that the bacteria on the hands of surgeons can cause wound infections if introduced into the operative field during surgery. A single strain of S. epidermidis caused an outbreak of postoperative wound infections and endocarditis during a 6-month period and was traced back to the hands of a single cardiac surgeon. 67 Also, rapid bacterial growth occurs under surgical gloves if plain soap is used, but growth is slower after using an antiseptic 68 (perhaps gloves internally impregnated with an antibacterial coating would be a cost-effective solution). Reducing the resident skin flora on the surgeon s hands for the duration of the procedure reduces the risk of contamination if gloves become punctured during surgery. When vascular surgeons who normally used an antiseptic began using plain soap instead, at least one outbreak of surgical site infection occurred. 69 Preoperative antiseptic preparations are selected for their ability to reduce the number of bacteria on hands immediately after washing, after wearing gloves for 6 h (persistence) and after multiple applications over five days (cumulative activity). They also need to be non-irritating and have broad-spectrum activity. The most effective agents for reducing bacterial counts immediately after hand disinfection are formulations containing 60e95% alcohol alone or 50e95% when combined with chlorhexidine gluconate. These are followed by, in order of decreasing activity, chlorhexidine gluconate, iodophors, triclosan, and plain soap. 70e74 Persistence, however, is a very important virtue in an antiseptic, and one that alcohols lack, although bacteria appear to grow more slowly on

Hand hygiene: An evidence-based review 59 Figure 5 Effective hand hygiene technique. Based on procedure described by Ayliffe GAJ, Babb JR, Quoraishi AH. J Clin Pathol. 1978;31(10):923e8. Images produced by author. the hands after alcohol use, and counts after wearing gloves for between 1 and 3 h rarely exceed pre-disinfection values. A recent study showed that a 61% ethanol preparation did not achieve adequate persistent activity at 6 h after use. 75 The addition of 0.5% or 1% chlorhexidine gluconate to alcohol preparations, however, produces an antiseptic with persistence that has equalled or exceeded that of 2% or 4% chlorhexidine gluconate, the agent with best persistence. The next greatest persistent antibacterial activity is found with triclosan, then iodophors. 52,74,76,77 Hexachlorophene has not been discussed as it is absorbed into the blood with repeated use, so seldom used for surgical hand hygiene. Duration of preoperative hand disinfection Disinfection for 10 min preoperatively often leads to skin damage, and 5 min of equivalent activity reduces bacterial counts as effectively, or even more effectively. 78,79 Studies have shown that a 2 or 3 min regime reduces bacterial counts to acceptable levels. 77,80,81 Based on the characteristics of the antiseptic substances discussed already, however (immediate activity, persistence and cumulative activity), it would seem sensible that perhaps the actual products used rather than the length of time spent scrubbing would be a more sensible avenue to explore. Alcohols have the optimum immediate antibacterial effect, and chlorhexidine gluconate the optimum persistent effect, so a combination of the two, or a two-stage disinfection regime would seem a better option than washing for endless minutes with povidoneeiodine. Indeed, an initial 1e2 min wash with 4% chlorhexidine gluconate or povidoneeiodine followed by the application of an alcohol-based product has been shown to be as effective as a 5-min regime with an antiseptic. 51,82 Sponge and brush? Some surgical preoperative protocols require surgeons to scrub with a brush, but this can lead to skin damage and result in increased shedding of bacteria from the hands. 83 Several studies indicate that neither a brush nor a sponge is necessary to reduce bacterial counts, especially when alcoholbased products are used. 46,75,76,78,84 Indeed, one study showed that using 1% chlorhexidine with 61% ethanol without scrubbing or using water produced significantly greater microbial reduction on the hands of volunteers than using 4% chlorhexidine with a scrub brush in 2e3-min surgical scrubs. 75 Compliance with hand hygiene guidelines Compliance of healthcare workers to recommended hand hygiene guidelines is notoriously poor,

60 C.R. Nicolay with a baseline ranging from 5% to 81%, mean 40%. 15 However, the definitions and criteria used and the methods of collecting data vary widely between studies and most of them have been carried out in intensive care units (ICUs), which are probably not representative of behaviour elsewhere in the hospital. Improved compliance after different interventions has been reported, but most studies have short follow-up periods so longlasting effects are unknown. This is probably partly because of the high turnover of staff rotating jobs and hospitals in the medical profession. In a large hospital wide survey of hand hygiene practices, 2834 opportunities for hand hygiene were observed, with an average compliance of 48%. 85 Compliance was greatest with nurses and at weekends rather than a weekday. Noncompliance was greatest in intensive care units compared with medical wards, and when intensity of patient care was high. There have been several other observational studies that have determined the risk factors for noncompliance (Table 1). 9,85e87 Improving compliance with hand hygiene guidelines Since a substantial proportion of HAIs result from cross-infection, and transmission of microorganisms by the hands of healthcare workers is recognized as the main route of spread, 88 it is essential to implement changes to improve compliance. From the list of risk factors above, one common factor is clear e lack of time for hand hygiene. In one study in ICU, it took nurses an average of 62 s to leave a patient s bedside, walk to a sink, wash their hands and return to the patient. 89 If alcoholbased handrub dispensers are placed at each patient s bedside or carried on each person as a small clip-on dispenser, however, a significant amount of time can be saved. In addition to ease and quickness of use, alcohol gels, when used properly, act faster and irritate hands less than plain soap or antiseptic agents. They were also used in the only programme that has shown a sustained improvement in hand hygiene compliance combined with a fall in HAI rate. 90 If a hand hygiene agent is introduced in the future with better residual activity than those available now, it may no longer become necessary to disinfect hands as often and compliance may increase and guidelines may change. Expecting nurses to disinfect their hands 20 times per day for example may be more acceptable and achievable than perhaps 60 or 80 times per day. A small study in the UK 91 looked at changing hand hygiene behaviour by educating patients, and involving them to ask all healthcare workers who were going to have direct contact them Did you wash your hands?, based on the American Partners in Your Care ª programme. 92 This empowers patients with responsibility for their care and increased soap usage and handwashings by an average of 50%. Interestingly, but perhaps not unpredictably, all patients asked nurses but only 35% asked doctors. Of course, for this to be effective patients need to be well enough to be alert and able to understand and speak English. Considering the breadth of factors, both behavioural and institutional interfering with hand hygiene compliance, clearly a solution to this problem will involve intervention at a personal, and at an institutional level. Regular education demonstrating correct hand hygiene technique and Table 1 Risk factors for noncompliance to hand hygiene guidelines Observed risk factors Self-reported risk factors Being a doctor (rather than nurse) Skin irritation Being a nursing assistant (rather than a nurse) Inaccessible supplies Being male Too busy or lack of time Working in ICU Inconveniently located sinks Working on a weekday Insufficient number of sinks Activities with high risk of cross-contamination Low risk of acquiring infection from patient Wearing gloves or gowns High workload and understaffing Higher number of hand hygiene opportunities per hour of patient care Interference with patienteworker relationship Patient needs perceived as priority Wearing gloves Forgetfulness Ignorance of or disagreement with guidelines Lack of scientific evidence for the impact of hand hygiene on HAI rates

Hand hygiene: An evidence-based review 61 the scientific evidence behind the benefits of hand hygiene is important. Encouragement from seniors, and senior staff leading by example on ward rounds would help to promote good hand hygiene behaviour. A system of continually published cycles of audit looking at both hand hygiene compliance and hospital acquired infection rate with feedback to individual wards would create an environment where people are constantly aware of infection control. Impact of improved hand hygiene on HAI rate The lack of scientific evidence showing that improved hand hygiene leads to a reduction in HAI is cited as a reason why some healthcare workers are not compliant with hand hygiene guidelines. Of seven hospital-based studies, most demonstrated a temporal relationship between improved hand hygiene and lower HAI rates (Table 2). In 1977, an ICU-based study 93 showed that the introduction of routine handwashing by staff before moving from one patient to the next was associated with a significant and sustained reduction in the number of patients colonised or infected with Klebsiella spp. A different study in a neonatal ICU (NICU) showed the elimination of endemic MRSA after the introduction of 1% triclosan w/v for hand hygiene, with all other infection control measures remaining in place. 61 Compared with the previous 12 months, fewer antibiotics were prescribed and fewer hospital acquired infections were recorded (p! 0.05). In 2000, a study at the University of Geneva Hospital demonstrated the effectiveness of a hospital wide, longstanding programme to promote hand hygiene. 90 This was achieved with posters, feedback meetings, and bedside and pocket alcohol handrub dispensers, and HAI rates, attack rates of MRSA cross-transmission and alcoholic handrub consumption were measured. Hand hygiene compliance improved from 48% in 1994 to 66% in 1997 (p! 0.001), and although handwashing with soap and water remained constant, frequency of hand disinfection increased (p! 0.001). Hand hygiene improved significantly among nurses and nursing assistants but still remained poor amongst doctors. Alcohol handrub consumption increased from 3.5 Table 2 Association between improved hand hygiene compliance and HAI rate Year Site Intervention Results Follow-up 1977 ICU 93 Introduction of handwashing between patients 1989 ICU 94 6-week trial each of 10% povidoneeiodine vs 4% aqueous chlorhexidine vs soap 1992 ICU 95 Prospective multiple crossover trial comparing chlorhexidine with 60% propan-2-ol and soap Reduction in Klebsiella spp. HAI rate 50% reduction in HAI rate with 10% povidone & 4% aqueous chlorhexidine vs soap. Chlorhexidine use reduced the HAI rate more effectively than 60% propan-2-ol and soap, partly explained by better compliance with chlorhexidine. 1994 NICU 61 Introduction of triclosan 1% w/v MRSA eliminated. A$17,000 saved from reduction in vancomycin use. 1995 Newborn nursery 62 After an MRSA skin infection outbreak, aggressive infection control measures instituted including changing plain soap to 0.3% triclosan. 2000 ICU 96 Intervention with multiple components designed to increase handwashing with plain soap, compared with control hospital. 2000 Hospital-wide 90 Hand hygiene promotion with posters, feedback meetings, and bedside and pocket alcohol dispensers. NICU Z neonatal ICU. MRSA eliminated. Reduction in VRE infection, no change in MRSA incidence. Hand hygiene compliance increased from 48% to 66%, and significant reduction in HAI rate and MRSA incidence. 2 years 18 weeks 8 months 1 year 3.5 years 6 months 5 years

62 C.R. Nicolay to 15.4 l per 1000 patient-days (p! 0.001), and during the same period HAI prevalence decreased 41% from 16.9% in 1994 to 9.9% in 1998 (p Z 0.04), and MRSA rates decreased 57% from 2.16 to 0.93 episodes per 10,000 patient-days, (p Z 0.001). Even more impressively, annual incidence of MRSA bacteraemia decreased 68% from 0.74 to 0.24 episodes per 10,000 patient-days (p! 0.001). The cost of hand hygiene The financial cost of hand hygiene products used in hospitals is an important consideration, and in a study in 1999 a 450-bed community teaching hospital in America spent US$22,000 on plain soap, 2% chlorhexidine preparations, and alcohol handrub. 97 Compared with liquid soap, the equivalent amount of 2% chlorhexidine gluconate costs about 1.7 times the amount, and alcohol-based handrub almost double the cost. However, the expenditure by hospital trusts on hand hygiene products must be compared against the excess costs arising from hospital acquired infections, and must not be seen as a barrier to promoting optimum hand hygiene. The costs associated with only four or five HAIs may be the same as the entire annual hand hygiene product budget. Indeed, just one severe surgical site infection, lower respiratory tract infection or septicaemia may cost more than this budget! Based on current BNF prices, the cost of 10 days treatment for MRSA is 346.40 with vancomycin (1 g bd iv) or 391.82 with teicoplanin (400 mg od iv), without the additional costs of drug level monitoring, nursing time, cannulae, etc. One study demonstrated a saving of about A$17,000 over just 7 months from using less vancomycin due to the reduction in MRSA incidence. 61 Of course, less use of antibiotics also reduces the chance of antibiotic resistance developing. In the study at the University of Geneva Hospital discussed previously, 90 the cost of the programme was estimated at no more than 155,000, including direct and indirect costs. The authors conservatively assumed that 25% of the observed reduction in HAI was due to the hand hygiene programme, thereby preventing 900 infections. At an average (very conservative) estimate of 1400 per infection, their estimate of overall savings of 1.25 million far outweighs the costs. It is clear that the savings made by reducing the HAI rate by improving hand hygiene compliance far outweigh the cost to hospital trusts of purchasing more effective and more acceptable hand hygiene products, improving hand hygiene education and increasing HAI surveillance. Conclusion Hospital acquired infection affects approximately 10% of inpatients, leads directly to an estimated 5000 deaths a year in the UK and costs the NHS an estimated 1000 million a year. Despite nearly one-third of these infections being caused by cross-infection by healthcare workers, compliance with recommended hand hygiene guidelines is poor, particularly amongst doctors. This review discusses the effectiveness and the properties of different hand hygiene agents available for use on the ward and for preoperative hand disinfection. There are a vast number of studies looking at different hand hygiene agents and their effectiveness, but there is a lack of good research using adequate numbers, control groups, long-term follow-up, and using HAI rate as an outcome rather than hand hygiene frequency or bacterial growth from hands. Hand hygiene products are needed with better residual activity, and testing of these agents is needed using short application times, under the conditions they are actually used in hospitals rather than in mock protocols by volunteers in laboratories. Improved hand hygiene compliance has been shown to significantly reduce the HAI rate. This improvement requires a multifactorial approach including the availability of alcohol handrubs on every person and bedside, education, peer example, and a continually published cycle of audit looking at hand hygiene compliance and accurately measuring the HAI rate. The financial savings associated with fewer HAIs and less antibiotic use easily outweigh the cost of programmes to improve hand hygiene compliance. In the 21st century, 16 decades after the reports of Semmelweis and Holmes, we are facing an everincreasing battle against hospital acquired infection, antibiotic-resistant bacteria and increasing patient numbers from an ageing population. There has never been a more simple and cost-effective intervention to help our patients than hand hygiene. References 1. Holmes OW. The contagiousness of puerperal fever. N Engl Q J Med Surg April 1843;1:503e30. Reprinted in Med Class 1936;1:211e43. 2. Semmelweis IP. Die Aetiologie, derbegriff und die Prophylaxis des Kindbettfiebers. Pest-Wien-Leipzig, 1861. Reprinted with a new introduction by A.F. Guttmacher. New York/London, 1966. Translated into English by F.R. Murphy as The Etiology, the Concept and Prophylaxis of Childbed Fever, in Med Class 1941;5:350e773. 3. Larson EL. APIC Guidelines for handwashing and hand antisepsis in health care settings. Am J Infect Control 1995;23:251e69.

Hand hygiene: An evidence-based review 63 4. Steere AC, Mallison GF. Handwashing practices for the prevention of nosocomial infections. Ann Intern Med 1975; 83:683e90. 5. Albert RK, Condie F. Hand-washing patterns in medical intensive-care units. N Engl J Med 1981;304:1465e6. 6. Jarvis WR. Handwashingdthe Semmelweis lesson forgotten? Lancet 1994;344:1311e2. 7. Sproat LJ, Inglis TJ. A multicentre survey of hand hygiene practice in intensive care units. J Hosp Infect 1994;26: 137e48. 8. Thompson BL, Dwyer DM, Ussery XT, Denman S, Vacek P, Schwartz B. Handwashing and glove use in a long-term-care facility. Infect Control Hosp Epidemiol 1997;18:97e103. 9. Pittet D. Improving adherence to hand hygiene practice: a multidisciplinary approach. Emerg Infect Dis 2001;7(2): 234e40. 10. Meers PD, Ayliffe GA, Emmeron AM, Leigh DA, Mayon- White RT, Mackintosh CA, et al. Report on the national survey of infection in hospitals 1980. J Hosp Infect (Suppl) 1981;2:1e11. 11. Emmerson AM, Enstone JE, Griffin M, Kelsey MC, Smyth ET. The second national prevalence survey of infection in hospitals e an overview of results. J Hosp Infect 1996;32: 175e90. 12. Hospital Infection Working Group of the Department of Health & Public Health Laboratory Service. Hospital infection control: guidance on the control of infection in hospitals. London: Department of Health; 1995:84.8. 13. Plowman R, Graves N, Griffin M, Roberts JA, Swan A, Cookson B, et al. The socio-economic burden of hospitalacquired infection. London: Public Health Laboratory Service and The London School of Hygiene and Tropical Medicine; 2000. 14. Infection control. Prevention of healthcare associated infection in primary and community care. NICE clinical guideline 2; June 2003. 15. Centers for Disease Control and Prevention. Guideline for hand hygiene in health-care settings: recommendations of the Healthcare Infection Control Practices Advisory Committee and the HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. MMWR 2002;51(RR-16):1e48. 16. Selwyn S. Microbiology and ecology of human skin. Practitioner 1980;224:1059e62. 17. Kampf G, Ostermeyer C. Efficacy of alcohol-based gels compared with simple hand wash and hygienic hand disinfection. J Hosp Infect 2004;56:S13e5. 18. Ehrenkranz NJ, Alfonso BC. Failure of bland soap handwash to prevent hand transfer of patient bacteria to urethral catheters. Infect Control Hosp Epidemiol 1991;12:654e62. 19. Larson E, Leyden JJ, McGinley KJ, Grove GL, Talbot GH. Physiologic and microbiologic changes in skin related to frequent handwashing. Infect Control 1986;7:59e63. 20. Boyce JM, Kelliher S, Vallande N. Skin irritation and dryness associated with two hand-hygiene regimens: soapand-water handwashing versus hand antisepsis with an alcoholic hand gel. Infect Control Hosp Epidemiol 2000;21: 442e8. 21. Larson EL, Morton HE. Alcohols. [chapter 11]. In: Block SS, editor. Disinfection, sterilization and preservation. 4th ed. Philadelphia, PA: Lea and Febiger; 1991. p. 642e54. 22. Girou E, Loyeau S, Legrand P, Oppein F, Brun-Buisson C. Efficacy of handrubbing with alcohol based solution versus standard handwashing with antiseptic soap: randomised clinical trial. BMJ 2002;325:362. 23. Kampf G, Höfer M, Wendt C. Efficacy of hand disinfectants against vancomycin-resistant enterococci in vitro. J Hosp Infect 1999;42:143e50. 24. Platt J, Bucknall RA. The disinfection of respiratory syncytial virus by isopropanol and a chlorhexidine-detergent handwash. J Hosp Infect 1985;6:89e94. 25. Krilov LR, Harkness SH. Inactivation of respiratory syncytial virus by detergents and disinfectants. Pediatr Infect Dis J 1993;12:582e4. 26. Sattar SA, Tetro J, Springthorpe VS, Giulivi A. Preventing the spread of hepatitis B and C viruses: where are germicides relevant? Am J Infect Control 2001;29:187e97. 27. Lilly HA, Lowbury EJL, Wilkins MD, Zaggy A. Delayed antimicrobial effects of skin disinfection by alcohol. J Hyg (Lond) 1979;82:497e500. 28. Ansari SA, Springthorpe VS, Sattar SA, Tostowaryk W, Wells GA. Comparison of cloth, paper, and warm air drying in eliminating viruses and bacteria from washed hands. Am J Infect Control 1991;19:243e9. 29. Ansari SA, Sattar SA, Springthorpe VS, Wells GA, Tostowaryk W. In vivo protocol for testing efficacy of hand-washing agents against viruses and bacteria: experiments with rotavirus and Escherichia coli. Appl Environ Microbiol 1989;55:3113e8. 30. Larson E, Bobo L. Effective hand degerming in the presence of blood. J Emerg Med 1992;10:7e11. 31. Ayliffe GAJ, Babb JR, Davies JG, Lilly HA. Hand disinfection: a comparison of various agents in laboratory and ward studies. J Hosp Infect 1988;11:226e43. 32. Lilly HA, Lowbury EJL, Wilkins MD. Detergents compared with each other and with antiseptics as skin degerming agents. J Hyg (Lond) 1979;82:89e93. 33. Bartzokas CA, Gibson MF, Graham R, Pinder DC. A comparison of triclosan and chlorhexidine preparations with 60 per cent isopropyl alcohol for hygienic hand disinfection. J Hosp Infect 1983;4:245e55. 34. Rotter ML, Koller W, Wewalka G, Werner HP, Ayliffe GAJ, Babb JR. Evaluation of procedures for hygienic handdisinfection: controlled parallel experiments on the Vienna test model. J Hyg (Lond) 1986;96:27e37. 35. Namura S, Nishijima S, Asada Y. An evaluation of the residual activity of antiseptic handrub lotions: an in use setting study. J Dermatol 1994;21:481e5. 36. Ayliffe GA, Bridges K, Lilly HA, Lowbury EJ, Varney J, Wilkins MD. Comparison of two methods for assessing the removal of total organisms and pathogens from the skin. J Hyg (Lond) 1975;75:259e74. 37. Rotter M, Koller W, Wewalka G. Povidoneeiodine and chlorhexidine gluconate-containing detergents for disinfection of hands. J Hosp Infect 1980;1:149e58. 38. Rotter ML. Hygienic hand disinfection. Infect Control 1984; 1:18e22. 39. Rotter ML, Koller W. Test models for hygienic handrub and hygienic handwash: the effects of two different contamination and sampling techniques. J Hosp Infect 1992;20:163e71. 40. Leyden JJ, McGinley KJ, Kaminer MS, Bakel J, Nishijima S, Grove MJ, et al. Computerized image analysis of full-hand touch plates: a method for quantification of surface bacteria on hands and the effect of antimicrobial agents. J Hosp Infect 1991;18(Suppl. B):13e22. 41. Cardoso CL, Pereira HH, Zequim JC, Guilhermetti M. Effectiveness of hand-cleansing agents for removing Acinetobacter baumannii strain from contaminated hands. Am J Infect Control 1999;27:327e31. 42. Winnefeld M, Richard MA, Drancourt M, Grobb JJ. Skin tolerance and effectiveness of two hand decontamination procedures in everyday hospital use. Br J Dermatol 2000; 143:546e50. 43. Casewell MW, Law MM, Desai N. A laboratory model for testing agents for hygienic hand disinfection: handwashing

64 C.R. Nicolay and chlorhexidine for the removal of Klebsiella. J Hosp Infect 1988;12:163e75. 44. Huang Y, Oie S, Kamiya A. Comparative effectiveness of hand-cleansing agents for removing methicillin-resistant Staphylococcus aureus from experimentally contaminated fingertips. Am J Infect Control 1994;22:224e7. 45. Larson EL, Eke PI, Wilder MP, Laughon BE. Quantity of soap as a variable in handwashing. Infect Control 1987;8:371e5. 46. Larson EL, Aiello AE, Heilman JM, Lyle CT, Cronquist A, Stahl JB, et al. Comparison of different regimens for surgical hand preparation. AORN J 2001;73:412e20. 47. Russell AD. Chlorhexidine: antibacterial action and bacterial resistance. Infection SeptembereOctober 1986;14(5): 212e5. 48. Sakuragi T, Yanagisawa K, Dan K. Bactericidal activity of skin disinfectants on methicillin-resistant Staphylococcus aureus. Anesth Analg 1995;81:555e8. 49. Kampf G, Jarosch R, Rüden H. Limited effectiveness of chlorhexidine based hand disinfectants against methicillinresistant Staphylococcus aureus (MRSA). J Hosp Infect 1998; 38:297e303. 50. Walsh B, Blakemore PH, Drabu YJ. The effect of handcream on the antibacterial activity of chlorhexidine gluconate. J Hosp Infect 1987;9:30e3. 51. Pereira LJ, Lee GM, Wade KJ. An evaluation of five protocols for surgical handwashing in relation to skin condition and microbial counts. J Hosp Infect 1997;36:49e65. 52. Larson EL, Butz AM, Gullette DL, Laughon BA. Alcohol for surgical scrubbing? Infect Control Hosp Epidemiol 1990;11: 139e43. 53. Rosenberg A, Alatary SD, Peterson AF. Safety and efficacy of the antiseptic chlorhexidine gluconate. Surg Gynecol Obstet 1976;143:789e92. 54. Anderson RL. Iodophor antiseptics: intrinsic microbial contamination with resistant bacteria. Infect Control Hosp Epidemiol 1989;10:443e6. 55. Larson EL, APIC Guidelines Committee. APIC guideline for handwashing and hand antisepsis in health care settings. Am J Infect Control 1995;23:251e69. 56. Reimer K, Wichelhaus TA, Schäfer V, Rudolph P, Kramer A, Wutzler P, et al. Antimicrobial effectiveness of povidonee iodine and consequences for new application areas. Dermatology 2002;204(Suppl. 1):114e20. 57. Jones RD, Jampani HB, Newman JL, Lee AS. Triclosan: a review of effectiveness and safety in health care settings. Am J Infect Control 2000;28:184e96. 58. Ward WH, Holdgate GA, Rowsell S, McLean EG, Pauptit RA, Clayton E, et al. Kinetic and structural characteristics of the inhibition of enoyl (acyl carrier protein) reductase by triclosan. Biochemistry 1999;38:12514e25. 59. Heath RJ, Li J, Roland GE. Inhibition of the Staphylococcus aureus NADPH-dependent enoyl-acyl carrier protein reductase by triclosan and hexachlorophene. J Biol Chem 2000;275:4654e9. 60. Faoagali JL, George N, Fong J, Davy J, Dowser M. Comparison of the antibacterial efficacy of 4% chlorhexidine gluconate and 1% triclosan handwash products in an acute clinical ward. Am J Infect Control 1999;27:320e6. 61. Webster J, Faoagali JL, Cartwright D. Elimination of methicillin resistant Staphylococcus aureus from a neonatal intensive care unit after hand washing with triclosan. J Paediatr Child Health 1994;30:59e64. 62. Zafar AB, Butler RC, Reese DJ, Gaydos LA, Mennonna PA. Use of 0.3% triclosan (Bacti-Stat â ) to eradicate an outbreak of methicillin resistant Staphylococcus aureus in a neonatal nursery. Am J Infect Control 1995;23:200e8. 63. Larson E, Mayur K, Laughon BA. Influence of two handwashing frequencies on reduction in colonizing flora with three handwashing products used by health care personnel. Am J Infect Control 1988;17:83e8. 64. Rotter ML. Semmelweis sesquicentennial: a little-noted anniversary of handwashing. Curr Opin Infect Dis 1998;11: 457e60. 65. Rutala WA, Weber DJ. New disinfection and sterilization methods. Emerg Infect Dis MarcheApril 2001;7(2):348e53. 66. Widmer AE, Dangel M. Alcohol-based handrub: evaluation of technique and microbiological efficacy with international infection control professionals. Infect Control Hosp Epidemiol March 2004;25(3):207e9. 67. Boyce JM, Potter-Bynoe G, Opal SM, Dziobek L, Medeiros AA. A common-source outbreak of Staphylococcus epidermidis infections among patients undergoing cardiac surgery. J Infect Dis 1990;161:493e9. 68. Dewar NE, Gravens DL. Effectiveness of septisol antiseptic foam as a surgical scrub agent. Appl Microbiol 1973;26: 544e9. 69. Grinbaum RS, de Mendonça JS, Cardo DM. An outbreak of handscrubbing-related surgical site infections in vascular surgical procedures. Infect Control Hosp Epidemiol 1995; 16:198e202. 70. Babb JR, Davies JG, Ayliffe GAJ. A test procedure for evaluating surgical hand disinfection. J Hosp Infect 1991; 18(Suppl. B):41e9. 71. Aly R, Maibach HI. Comparative antibacterial efficacy of a 2-minute surgical scrub with chlorhexidine gluconate, povidoneeiodine, and chloroxylenol sponge-brushes. Am J Infect Control 1988;16:173e7. 72. Faoagali J, Fong J, George N, Mahoney P, O Rourke V. Comparison of the immediate, residual, and cumulative antibacterial effects of Novaderm R, Novascrub R, Betadine Surgical Scrub, Hibiclens, and liquid soap. Am J Infect Control 1995;23:337e43. 73. Herruzo-Cabrera R, Vizcaino-Alcaide MJ, Fdez-AciZero MJ. Usefulness of an alcohol solution of N-duopropenide for the surgical antisepsis of the hands compared with handwashing with iodineepovidone and chlorhexidine: clinical essay. J Surg Res 2000;94:6e12. 74. Peterson AF, Rosenberg A. Comparative evaluation of surgical scrub preparations. Surg Gynecol Obstet 1978; 146:63e5. 75. Mulberry G, Snyder AT, Heilman J, Pyrek J, Stahl J. Evaluation of a waterless, scrubless chlorhexidine gluconate/ethanol surgical scrub for antimicrobial efficacy. Am J Infect Control 2001;29:377e82. 76. Hobson DW, Woller W, Anderson L, Guthery E. Development and evaluation of a new alcohol-based surgical hand scrub formulation with persistent antimicrobial characteristics and brushless application. Am J Infect Control 1998;26: 507e12. 77. Pereira LJ, Lee GM, Wade KJ. The effect of surgical handwashing routines on the microbial counts of operating room nurses. Am J Infect Control 1990;18:354e64. 78. Galle PC, Homesley HD, Rhyne AL. Reassessment of the surgical scrub. Surg Gynecol Obstet 1978;147:215e8. 79. O Farrell DA, Kenny G, O Sullivan M, Nicholson P, Stephens M, Hone R. Evaluation of the optimal hand-scrub duration prior to total hip arthroplasty. J Hosp Infect 1994; 26:93e8. 80. Hingst V, Juditzki I, Heeg P, Sonntag HG. Evaluation of the efficacy of surgical hand disinfection following a reduced application time of 3 instead of 5 min. J Hosp Infect 1992; 20:79e86.