Except for the anaerobic propionibacteria that are located mainly at the ducts of sebaceous glands, most of these microorganisms reside on the uppermost part of the stratum corneum (6,7), on corneocytes, and are embedded in a mass of lipids and cell detritus of the pars disjuncta (8,9). They multiply in the upper regions of the hair follicle (10). The deeper regions of the skin are, the ducts of eccrine and apocrine glands, not colonized (11). The composition of skin flora has been described in several reviews (12,13,14,15,16,17,18 and 19). Recently, molecular biologic diagnostic tools such as a novel pyrosequencing-based method, allowed characterizing a hitherto unknown diversity of skin bacteria on the palmar surfaces of the hands of young adults (20). According to the findings of Fierer et al. (20) hands harbored, on average, 158 (in the range 46-401) unique species-level bacterial phylotypes, and among the 51 healthy volunteers, they identified a total of 4,742 unique phylotypes across the 102 hands examined. The bacterial diversity was more expressed in females than in males. The bacterial skin flora varies qualitatively and quantitatively by body site, gender, age, health condition, hospitalization, season (6,11,21,22 and 23), handedness, and the time interval between last hand washing and skin sampling (20). Except for areas with large numbers of sebaceous glands where propionibacteria prevail, the main portion of the skin flora is made up of Micrococcaceae such as staphylococcal species (Staphylococcus epidermidis, Staphylococcus
hominis, Staphylococcus capitis, etc.) and micrococci. Also, Staphylococcus aureus may temporarily colonize the skin, especially the perineal region, nose, hands, face, and neck. This occurs more often with children than with adults (3), but healthcare personnel are especially prone to this colonization; the prevalence of colonization with S. aureus in healthcare personnel was reported by Larson et al. (24) to reach 18%. In the intensive care unit (ICU) of a German teaching hospital, Hofmann et al. (25) found S. aureus on 18.4% of nurses’ hands and on 36% of doctors’ hands. These data more likely reflect, however, a state of repeated contamination rather than true long-term carriage. Lipophilic and nonlipophilic corynebacteria are common inhabitants of the skin—the former usually in hairy regions, the latter more in bald regions (6). The antibiotic-resistant Corynebacterium jeikeium may cause therapeutically difficult nosocomial infections in high-risk patients. Although the most common site of isolation is the perineum (23), it may also occur on the hands. Propionibacterium acnes and Propionibacterium granulosum—the latter of which is less often isolated—multiply at sebaceous body sites (11,26), but they can also be found in small numbers on the hands, although most likely as transients (9). Fierer et al. (20), however, identified them as the most abundant bacterial group on the palmar hand surfaces of their undergraduate student-volunteers who had, just prior to sampling, taken an examination. This may be because the students had often and intensely touched their sebum-rich forehead while thinking and, by this, contaminated their hands.

Gram-negative bacteria such as Acinetobacter and Enterobacter species may be isolated mainly from moist skin areas (26,27,28 and 29) but also regularly from the hands, where they may be regarded as residents (30). Larson (31) found that 80% of persons outside the hospital and 21% of hospital personnel persistently carried Acinetobacter species and members of the Klebsiella-Enterobacter group. Males were significantly more likely to be carriers than females, and persons who washed their hands less than eight times per day were more likely to persistently carry the same gram-negative species than those who washed more than eight times. Well known is the report by Casewell and Phillips (32) of a hospital outbreak with Klebsiella colonizing the hands of hospital personnel. Attendants with close patient contact, as in ICUs, were especially likely to carry gram-negative bacteria on their hands (33). A list on the contamination frequency with nosocomial pathogenic species on healthcare workers’ hands and their persistence on hands and inanimate surfaces has been presented in a recent review by Kampf and Kramer (34).

The population density of resident skin bacteria ranges somewhere between 10⁶ colony-forming units (CFU)/cm² on the sebum-rich scalp and 10² to 10⁴/cm² on the forearm (35). Fingertip counts assessed by agar contact methods ranged from 0 to 300 (36). The density remains remarkably stable for any given individual over long periods of time (5,20,36,37 and 38). Only diseases of the skin and agents interfering with the biocenosis, such as antibiotics or disinfectants, may cause long-term alterations (6,14,26,39). The greatest short-term fluctuations (1-2 hours) are seen after intense contact with water (40).

The normal microbial skin flora fulfills the important function of colonization resistance, thereby preventing colonization with other and potentially more pathogenic microorganisms. The influencing factors are the presence of free fatty acids liberated from skin lipids by bacterial metabolism, the presence of bacteriocins and other antibiotic-like bacterial secretions, and the low water content of the stratum corneum (3,17,41,42). The pH value and osmotic conditions are less important (21). Unless introduced into body tissue by trauma or in the presence of foreign bodies such as catheters or implants, the pathogenic potential of the resident flora is usually regarded as low (43,44). Resident flora is difficult to remove by mechanical means. Washing hands with soap and water reduces the release of skin bacteria every 5 minutes by only 50% (5,45,46,47).

Members of this group are characterized by their inability to multiply on the skin. They occur as skin contaminants. Among them, microorganisms with high pathogenic potential may also be found. Usually transient flora does not survive for very long, but sporadically multiply on the skin surface (34). Besides the above-mentioned factor of colonization resistance, the inhospitable physicochemical environment may be another reason for the failure of transient flora to survive. Medical personnel, however, should never rely on this. In contrast to natural microbial skin flora, transient flora is easily removed by mechanical means such as hand washing. If hands are washed for 1 minute with soap and water, the reduction of bacterial release was measured to be two to three orders of magnitude (48,49,50,51,52 and 53,54). Even rubbing hands with water alone is effective (52).

This group includes the etiologic agents of actual infections such as abscesses, panaritium, paronychia, and infected eczema on the hands. They are of proven pathogenicity.
S. aureus and β-hemolytic streptococci are the species most often encountered.

Although in German-speaking countries the term “hand washing” is exclusively reserved for the use of unmedicated soap and water (with or without a brush), in other parts of the world, it also implies the application of antiseptic soaps (disinfectant-detergents). In this chapter, the term is applied sensu strictu to washing hands with unmedicated detergent and water.

The objective of hand washing is to remove dirt (consisting of extraneous substances, sweat, skin lipids, epithelial debris, etc.) and loosely adhering microbial skin flora, which will include most of the transient but only a small part of the resident flora. In fields of application where the microbiologic aspect dominates, the aim is, of course, to reduce microbial release from hands to an extent that may be considered safe for the intended purpose. In the medical field, this purpose is usually to prevent hand-borne infection.

The efficacy of a hand wash depends on the time taken and the technique. Unfortunately, this period is usually rather (too?) short in normal hospital work. The average duration was reported by several authors to be between 8 and 20 seconds (24,69,70). This period of time, however, does not include the additional time needed to approach and return from the washplace. Therefore, the complete process takes considerably longer. In fact, it has been measured to take 40 to 80 seconds (64). Table 91-3 indicates how effectively the release of transient bacteria from artificially contaminated hands can be reduced by hand washing. The greatest reduction is achieved within the first 30 seconds; it ranges between 0.6 and 1.1 log after 15 seconds and between 1.8 and 2.8 log at the end of 30 seconds. Extending the washing time to 1 minute results in reductions of 2.7 to 3.0 log. A further prolongation of the procedure is not worth the effort, because after 2 minutes the reduction increases negligibly to only 3.3 log and after 4 minutes to only 3.7 log.

Although in most instances these reductions are probably sufficient to prevent infection-generating transmission of pathogens (51,53,72,74,75 and 76), this is not always the case. Semmelweis, for instance, observed that normal hand washing did not always prevent the spread of fatal infection. Eleven parturient women died of puerperal fever after having been examined immediately after contact with a patient suffering from a foully discharging medullary carcinoma (77) by attendants who, in between, had washed their hands with only soap and water. After this experience, Semmelweis extended his order to disinfect hands in a solution of chlorinated lime from before entering the delivery or patient room to using it before each vaginal examination (77,78,79 and 80). It is important to understand that some procedures of hand disinfection are significantly more efficient in reducing the bacterial release from hands than hand washing with soap and water.

Although highly sophisticated washrooms with fully automated functions have been shown to be even counterproductive rather than motivating healthcare personnel to adhere to hand washing rules (81), certain requirements for washrooms must be fulfilled for minimal compliance. Wash basins should be conveniently located; no overflow or plug is necessary because hands should be washed only under running water. A mixer tap helps to provide water of comfortable temperature that, under the best conditions, is controlled thermostatically. Operating the water flow without using hands (elbow, knee, foot, sensor) may be desirable in certain critical areas. Suitable dispensers for soap, disinfectant (rub is better than detergent), hand lotion, and one-way towels are accepted requirements. There must also be a container furnished with a liner for used towels. If liquid soap is used, dispensers must either be easily removable and heat resistant for thermal reprocessing or they should be equipped with disposable bags. Liquid soap dispensed from refillable containers should be bacteriostatic to prevent microbial growth; topping of these containers is to be strictly prohibited.

An appropriate hand washing technique includes adjusting the water flow and the temperature (both activities can be accelerated by suitable technical devices), wetting hands, taking soap, rubbing hands to produce a lather without splashing, and performing wash movements that include rubbing palm to palm, rotational rubbing with clasped fingers of right hand in palm of left hand and vice versa, moving right palm over left dorsum and vice versa, palm to palm with fingers interlaced, backs of fingers to opposing palm with fingers interlocked, and rotational rubbing of right thumb clasped in left palm and vice versa. Each movement is to be repeated four times. This technique was proposed by Ayliffe et al. (71) as a standard technique when testing antiseptic hand washes. It could also represent a routine hand wash technique. Some authors demonstrated, however, that, after appropriate instruction, allowing each individual his or her own “responsible application” resulted in a better coverage of all hand surfaces (82); and this is a prerequisite for a correct technique. Finally, hands are rinsed with fingertips up, and the water is cautiously shaken off. As concluded from the results of laboratorybased in vivo tests, the whole procedure should take not less than 30 seconds, a goal nearly impossible to attain during patient care. The subungual spaces harbor, by far, the main part of the bacterial hand flora (83). The importance of this observation for the transmission of nosocomial infections by medical personnel is unknown, but Tanner et al. (84) found that, at least in presurgical hand preparation, nailbrushes and nail picks do not decrease bacterial numbers and are, therefore, unnecessary. After washing, the hands are dried with a disposable towel (paper or textile). Unless the water flow is discontinued by an automatic device, the water should be turned off by using the same towel rather than by the freshly washed hands (73). The towel is then discarded into the appropriate container, and a hand lotion should be applied onto the hands. This latter step is extremely important to prevent chapping. Electric hand dryers are useless in hospitals because, with them, hand drying takes too long and because they lack the friction of towels to remove the remaining soap from the skin.

No matter how well and detailed hand washing techniques may be described, Larson and Kretzer (85) are probably right in suggesting that a subject of much greater concern is how to motivate personnel to wash their hands in the first place, because hand washing practices still remain suboptimal.

The objective of a hygienic hand rub is to reduce the release of transient pathogens with maximum efficacy and speed, so that hands can be rendered safe after known or suspected contamination. This should be done in a way that avoids microbial dispersal into the environment. A sustained effect is not required. The fate of the resident skin flora is disregarded in this procedure.

The technique of hygienic hand rubs includes rubbing small portions of 3 to 5 mL of a fast-acting antiseptic preparation onto both hands. This can be a very convenient way of treating hands after known or suspected contamination, because dispensers for hand rubs can easily be made available wherever necessary; for instance, they may be placed in the vicinity of every patient bed in high-risk areas. All areas of the hands must be covered by the disinfectant, but this is often not done (70).

The antimicrobial spectrum necessary for hygienic hand rubs depends on the intended use. Commonly, the antimicrobial spectrum required includes only bacterial and fungal pathogens. Sporicidal activity is, if at all, only
needed in certain situations such as in Clostridium difficile outbreaks. But there are hardly any chemicals that are sufficiently strong and fast acting as well as skin-tolerable at the same time, so the mechanical action of a hand wash is usually used for spore reduction. “Hand washing with soap and water showed the greatest efficacy in removing C. difficile and should be performed preferentially over the use of alcohol-based hand rubs when contact with this pathogen is suspected or likely,” concluded Oughton et al. (86) from the results of their laboratory-based experiments with artificially contaminated hands of volunteers. However, as the usually encountered nosocomial pathogens can still be around, it is recommended to first use an alcohol-based hand rub before washing hands. Activity against mycobacteria is required only at certain places such as in tuberculosis hospitals, wards for acquired immunodeficiency syndrome patients, and in pathology and microbiology laboratories. The antituberculous effect must be proven and stated on the label. Virucidal activity is not a general requirement and is only justified in special situations. Furthermore, it should only be claimed if the (proven) antiviral spectrum of a product also includes enteroviruses such as polio or hepatitis A virus together with an acceptable exposure time.

There is only a small range of possible agents for hand rubs, such as alcohols in high concentration, used alone or mixed with other antiseptics; aqueous solutions containing halogens such as chlorine or iodine; chlorhexidine; quaternary ammonium compounds; phenolics; triclosan; aldehydes; metallo-organic compounds; and oxidizing agents such as peracetic acid. Except for the alcohols, aqueous solutions of chlorine, povidone-iodine, and chlorhexidine, the other agents are usually used solely as adjuncts to alcohols (quaternary ammonium and ampholytic compounds, phenolics); are contained in antiseptic detergents (phenol derivatives, povidone-iodine, chlorhexidine, triclosan); or are not used at all because of poor efficacy, allergenicity, irritant or toxic potential, or ecologic considerations (aldehydes, metallo-organic, and peracetic acid). There is no doubt that alcohols are much more comfortable to rub onto the skin than aqueous solutions because of specific features such as excellent spreading and quick evaporation.

Table 91-4 summarizes examples of results from evaluations of commonly used active agents for their antibacterial efficacy, which was assessed in standardized tests simulating practical conditions on artificially contaminated hands of volunteers (9,45,52,54,71,72,87,98,102,103,104,105,106,107,108 and 109). As demonstrated, the alcohols n-propanol, isopropanol, and ethanol, and the halogen releasers sodium tosylchloramide and povidone-iodine appear superior to aqueous solutions of chlorhexidine diacetate, chlorocresol, and hydrogen peroxide. Among the results with the alcohols, there is a clear positive association between the extent of bacterial reduction and the concentration used. If mean log reductions obtained with the three alcohols are compared with each other at equal concentrations, n-propanol is the most effective and ethanol is the least effective alcohol. The efficacy of aqueous solutions of sodium tosylchloramide and povidone-iodine compares well with that of isopropanol at a concentration of 60% v/v.

Tuberculocidal activity has been demonstrated for the alcohols mentioned (88,89 and 90,92,93,99), although with prolonged exposure (89). Several recommendations suggest disinfection times of 1 to 5 minutes with 70% ethanol, 60% to 70% isopropanol, or 50% to 70% n-propanol (87,91,94). The halogen-based preparations are also regarded as active (60,94).

The virucidal activity of alcohols is generally good with enveloped viruses (54,95,96), including the human immunodeficiency virus (HIV). An exception is the rabies virus, which is reported to be ethanol-resistant (97). Naked viruses, such as enteroviruses, are inactivated only by high concentrations of alcohols (100), the most effective of which is reported to be ethanol (96,101). Laboratory in vivo tests have shown that the effectiveness of alcohols against some difficult viruses such as entero- and rotavirus is significantly better than that of hand washing with unmedicated soap (101,110,111,112,113,114,115,116,117 and 118). Absolute ethanol reduced, for instance, the viral release from the hands by 3.2 log, 80% ethanol (v/v) by 2.2 log, and absolute n-propanol by 2.4 log (110). In contrast, individual hand washing for 10 to 55 seconds caused a reduction of only 1 log. Testing a commercial preparation containing a high-alcohol concentration, Schürmann and Eggers (100) concluded that this rub was effective against enteroviruses only under favorable environmental conditions such as high temperature, large disinfectant-to-virus volume ratio, and low protein load. In another study, the reduction in the release of human rotavirus strains from the hands by 70% (v/v) ethanol or isopropanol was approximately 100 times that of the reduction attainable with tap water or liquid soap (111). A reduction of >3 log by a 60% ethanol preparation was demonstrated in vivo with the nonenveloped rota-, adeno-, and rhinoviruses (112). Over the last several years, another nonenveloped virus, the norovirus (the former Norwalk-like virus), belonging to the family of caliciviruses, has been recognized as an important cause for epidemic and sporadic food-, water-, and airborne diarrheal disease. As, at present, the human norovirus cannot be grown in cell culture systems, related animal noroviruses have been and are used as surrogate viruses to evaluate the virucidal efficacy of antiseptic agents, especially alcohols. In fingertip experiments according to ASTM International E-1838-96 (113), Gehrke et al. (114) found that a 70% (v/v) concentration of each of three tested alcohol species was more effective than their 90% counterpart. Ethanol turned out to be the most efficacious, followed by 1-propanol and 2-propanol, with the respective viral reductions being 3.78, 3.58, and 2.15 log.

In other fingerpad experiments, a combination of ethanol with 10% 1-propanol, 5.9% 1,2-propandiol, 5.7% 1,3-butandiol, and 0.7% phosphoric acid proved active at a much lower concentration of 55% against polio type 1 with a log reduction of 3.04 within 30 seconds, whereas with 2-propanol only 1.32 log were achieved. Within the same exposure time, feline calicivirus was reduced by 2.8 log (115). (In quantitative suspension tests, with and without protein load, this formula reduced infectivity titers of seven enveloped and four nonenveloped viruses by >103 log within 30 seconds. Only ethanol at a concentration as high as 95% exerted a comparable activity.) Similarly, in another in vivo study using the then-amended fingerpad test method E 1838-02 of ASTM International (119), it was also a combination of ethanol 70% (v/v), in this case, with polyquaternium-37 and citric acid that succeeded in
reducing the release of murine norovirus, which is nowadays regarded as a more relevant surrogate virus (116), by 2.84 log within 30 seconds as compared to only 0.91 log achieved with pure ethanol 75% v/v (117). (When tested in suspension, this test product reduced the infectivity of the nonenveloped viruses: human rotavirus, polio type 1, feline calicivirus, and murine norovirus by >3 log after a 30-second exposure.)

For years, HBV was thought to be extremely resistant to the action of chemical disinfectants. Dried or liquid human plasma containing high-titer HBV, however, did not cause hepatitis if the sera were treated before inoculation into susceptible chimpanzees with 70% isopropanol for 10 minutes, 80% ethanol for 2 minutes, 0.1% glutaraldehyde for 5 minutes, povidone-iodine with 0.8% available iodine for 10 minutes, or hypochlorite solution with 500 mg/L free chlorine for 10 minutes, whereas the control animals receiving untreated plasma developed the disease (120,121). In another test system—the so-called morphology alteration and disintegration test—the HBV appeared significantly altered and disintegrated after exposure to 82% ethanol (122). Ethanol was reported active against HBV at a concentration as low as 70% when in combination with agents such as hexachlorophene, quaternary ammonium compounds, octenidine, biphenylol, or hydrogen peroxide (121,122,123,124 and 125).

Hepatitis C virus is likely to be inactivated by concentrations of 60% to 70% ethanol (126).

As is evident from the above, alcoholic rubs are very well suited for hygienic hand disinfection, because their antimicrobial performance is excellent and fast, thus saving time; no wash basin is necessary for their use, and they can be positioned next to any patient bed; furthermore, their application does not cause microbial contamination of nurses’ uniforms. However, one must bear in mind that the antimicrobial efficacy of alcohols is very sensitive to dilution with water and is, therefore, vulnerable to inactivation, especially with the small volumes of 3 to 6 mL, which, for hygienic hand rubs, are distributed all over both hands. If, for instance, 60% (v/v) isopropanol is rubbed onto wet hands in two portions, each of 3 mL, for 30 seconds, the mean log bacterial reduction achieved was measured to be 3.7, as opposed to 4.3 with dry hands (107,127). Although being not as comfortable, an aqueous solution of povidone-iodine may be used as an alternative hand rub, if necessary, for any reason.

Mainly in North America, there is now a trend toward gel formulations. Comparative tests between liquid and gel formulations of alcohol-based hand rubs revealed, however, that the bactericidal efficacy of gels is significantly lower than that of rinses. Kramer et al. (128) compared the efficacy of 10 commercial gels and four rinses using the test method of the European standard EN 1500. No single gel met the requirements within 30 seconds of application, whereas all rinses did. From a report by Kampf et al. (129), it appears, however, that gels with a very high alcohol content can meet the requirement of EN 1500. A new gel containing 85% (by weight) ethanol proved to be bactericidal in suspension (when tested according to prEN 12054) and on volunteers’ hands (EN 1500). Furthermore, in suspension tests, the gel was shown to be fungicidal (EN 1275), tuberculocidal (test according to the German Society of Hygiene and Microbiology with Mycobacterium terrae as a surrogate test bacterium for Mycobacterium tuberculosis), and virucidal (defined as a ≥4 log reduction, within different exposure times) for orthopox and herpes simplex 1 and 2 viruses (15 seconds); rotavirus and HIV (30 seconds); and adeno- (2 minutes), polio- (3 minutes), and papovavirus (15 minutes).

Also, in a prospective clinical trial, where immediately before and after direct contact with a patient pre- and postcontaminations were assessed, it was demonstrated that, on the hands of healthcare workers, equally good respective bacterial reductions of 1.28 and 1.29 log were achieved with both a rinse and a gel—the former containing a mixture of a high concentration of n-propanol (30% w/w) and isopropanol (45% w/w), the latter containing 85% (w/w) ethanol—whereas with another gel containing a mixture of 53% (w/w) ethanol plus 17% (w/w) isopropanol, a significantly inferior reduction of 0.51 log was seen (130).

Despite the high alcohol concentration, the user acceptability was described as excellent, although it was even better with the gels. Owing to the ease of its performance, the hygienic hand rub not only offers the advantage of being fast and efficacious but also has the potential to improve the compliance of healthcare givers with hand hygiene.

For hygienic hand rubs, only a few clinical correlates with results from disinfectant testing exist. In one controlled study trying to relate the use of alcoholic rubs to infection rates (131), isopropanol was used in a way that cannot be regarded as a real hygienic hand rub, namely, with average volumes of 0.9 mL per application, which is much too small to cover the surface of both hands and to remain there long enough to exert bactericidal effects. The study demonstrates, however, that a hygienic hand wash with chlorhexidine detergent, which was also tested, is clinically more effective than an individual hand wash with soap and water, which was followed by rubbing a bit of alcohol onto the hands; the study also demonstrated that, despite a preceding intensive education program, it seems very difficult to persuade and motivate medical personnel to observe the simplest rules for the most efficient procedure in the prevention of healthcare-associated infections.

Because hand rubs have a high antimicrobial potential, they can also be used in situations where direct contact with dangerous pathogens has occurred, such as after spillage in the microbiology laboratory or after touching infectious lesions. A high-level requirement for efficacy is, therefore, justified. With this perspective, a requirement for the performance of a reference hand rub was formulated by the Austrian and German Microbiological Societies (132,133) and, finally, by the European Committee for Standardization (134), choosing among the best-acting rubs available. From these, 60% (v/v) isopropanol was taken arbitrarily as an active agent to be used in two portions, each of 3 mL, during a total disinfection period of 60 seconds. The requirement of EN 1500 is that the reduction of transient flora assessed with a product for hand rubs shall not be significantly inferior to that with the reference rub, when tested in parallel, with the same volunteers, on the same day, in a crossover fashion.