key: cord-0787366-x7s6ahy6
authors: Curtis, Val; Schmidt, Wolf; Luby, Stephen; Florez, Rocio; Touré, Ousmane; Biran, Adam
title: Hygiene: new hopes, new horizons
date: 2011-03-28
journal: Lancet Infect Dis
DOI: 10.1016/s1473-3099(10)70224-3
sha: 7cc9c36dd4de535aabba7346c12be7b470f6797f
doc_id: 787366
cord_uid: x7s6ahy6

Although promotion of safe hygiene is the single most cost-effective means of preventing infectious disease, investment in hygiene is low both in the health and in the water and sanitation sectors. Evidence shows the benefit of improved hygiene, especially for improved handwashing and safe stool disposal. A growing understanding of what drives hygiene behaviour and creative partnerships are providing fresh approaches to change behaviour. However, some important gaps in our knowledge exist. For example, almost no trials of the effectiveness of interventions to improve food hygiene in developing countries are available. We also need to figure out how best to make safe hygiene practices matters of daily routine that are sustained by social norms on a mass scale. Full and active involvement of the health sector in getting safe hygiene to all homes, schools, and institutions will bring major gains to public health.

Promotion of hygiene might be the single most costeff ective way of reducing the global burden of infectious disease. 1 One might therefore expect hygiene to be the subject of multimillion dollar international initiatives like those for malaria or HIV/AIDS prevention. Perhaps because hygiene does not require clever new technologies or products, or perhaps because it is a domestic and personal issue largely aff ecting women and children, and perhaps because it concerns the neglected diarrhoeal and respiratory diseases (still the two biggest killers of children), hygiene is still very much overlooked in public health. There are signs that the situation is beginning to improve. Governments and funding agencies increasingly accept that hygiene promotion should play a part in health investments across the wider community, not just in health-care settings. Policy makers are also realising that the health benefi ts of increased investment in water and sanitation infrastructure are largely delivered through improvements in personal and domestic hygiene. 2 Original approaches using new insights are modernising the hygiene sector, making it more attractive to investors.

Improved water supplies and sanitation facilities make it easier to practise hygiene, keeping children and adults safe from infection. But even without improved facilities, better hygiene can still make a huge diff erence to health. Although most sanitation and water supply programme implementers seek to improve hygiene alongside hardware, they rarely have the resources and professional support needed to do this eff ectively. Health professionals recognise the need for better hygiene, but too few are actually engaged in programmes to promote it.

In this Review we gather the facts about the importance of hygiene for public health and explore the scale of the problem. We set out what we know about hygiene and assess its promotion in the service of the Millennium Development Goals (MDGs) and beyond. Growing understanding of what shapes hygiene behaviour and creative partnerships are changing the way improvement is being approached. The evidence for giving hygiene a much higher priority is strong, and, to a large extent, we already know what needs to be done. The most important ingredient still missing is the full and active engagement of the health sector in improving global hygiene.

Improvements in hygiene, sanitation, and water can prevent several important infections, in addition to providing other benefi ts. Among these avoidable infections prevention of diarrhoeal diseases is most important. Because the source of infections is human faecal material, the most important hygiene behaviours are clearly those that keep faecal matter out of the domestic environment. Adequate handwashing after contact with faeces is also crucial (after one's own defecation, after handling the faeces of children, or after contact with a faeces-contaminated environment). Other ways of preventing the faecal-oral transmission of infections include keeping water, foods, and surfaces free of faecal contamination and preventing carriage by fl ies. Safe food handling and preparation is also important, especially for children, as is the avoidance of animal faeces and the safe storage and use of water. 3 Other diseases that can be prevented by adequate hygiene include respiratory infections, trachoma, and skin infections. Endoparasites, such as roundworm and hookworm, and ectoparasites including scabies and fl eas, can also be avoided.

Public health practitioners commonly use information from four sources when weighing up the risk of infectious disease. First, they can assess the biological likelihood that a particular practice will place individuals at risk of infection. Second, they can use risk mapping-for example, modelling of the transfer of microbes between surfaces and hosts in homes and hospitals 4 or use of the hazard analysis critical control points method for assessing risk in food preparation. 5 However, these approaches depend on access to good estimates of environmental contamination, which are largely unavailable for developing countries. Third, health practitioners can use correlations between recorded practices and disease incidence from observational studies. These data are more readily available, but can be misleading. 6 Hygiene behaviour is commonly associated with socioeconomic factors, such as wealth, education, Review access to water, and modern lifestyle attitudes, 7 all of which infl uence the risk of infectious disease. Such strong socioeconomic confounding is diffi cult, if not impossible, to address analytically. 8 The fourth source of information for public health policy making is randomised controlled trials (RCTs), which control for confounding. However, very few RCTs of hygiene promotion programmes have been undertaken in developing countries, and those that have been done have several methodological fl aws. For example, the masking of participants to the intervention is diffi cult, and as a result, mothers who are grateful for an intervention may be less likely to report disease in their children, leading to infl ated eff ect sizes. 9 Bias is thus a serious issue in unblinded studies on diarrhoea. 10 Given these caveats, what can we say about the prevention of diarrhoeal disease through hygiene? Table 1 draws together our assessment of the available evidence, from reviews and other key papers, concerning the four sources of information: biological plausibility, risk modelling, observational studies, and RCTs.

The best studied hygiene practice in developing countries is that of handwashing. Evidence from all four types of source is consistent, with RCTs of handwashing interventions showing reductions in diarrhoea of around 30%, and of 43-47% if soap is used. 11, 12 Handwashing can also reduce other infections; one review suggested it could reduce respiratory infection by 16%, 13 and a more recent cluster-randomised trial in Pakistan reported a reduction in acute lower respiratory tract infections of 50%. 14 Handwashing also reduces neonatal mortality, 15 trachoma, 16 and parasitic worm infections. 17 Face or whole-body washing are less well researched but might help to control skin infections and trachoma. 18 An unclean face is associated with increased risk of trachoma, 19, 20 and a randomised trial suggested that face washing reduces the risk of severe trachoma infection substantially. 21 Handwashing with soap mitigated the severe acute respiratory syndrome epidemic 22 and is one of the key practices recommended to counter possible infl uenza pandemics. 23 Although food-borne infection is the main route of transmission of gastrointestinal infections in developed countries, their contribution to the burden of diarrhoea in low-income settings is unclear. Hot climates, poor storage facilities, and faecal contamination of the environment all make food-borne infection more likely. Therefore, food-borne infections are likely to play a major part in diarrhoeal disease transmission in low-income settings. 24 Microbiological studies have shown the ability of many pathogens to grow quickly in food, especially in hot climates. 24, 25 Contaminated weaning food, in particular, has been suggested as a major contributor to diarrhoea in low-income settings, 26 although observational studies gave inconclusive results. 24 Most of what we assume about food-borne infections in low-income settings is based on expert opinion and biological plausibility, rather than fi eld data (eg, WHO's manual Five Keys to Safer Food). 27 Several trials have assessed the eff ect of promoting exclusive breastfeeding on food-borne infections, with equivocal results. 28, 29 Food hygiene interventions have rarely been systematically tested. In one of the very few intervention studies of improving childhood feeding practices, which included some food-hygiene education, Bhandari and colleagues 30 found little eff ect on the nutritional status of children in rural India. The results of a recent trial done in Mali suggested that the microbiological safety of weaning food could be signifi cantly improved with hazard-control principles in homes. 31 Other routes of infection that could be removed by better hygiene are related to contact with child 7,32,33 and animal faeces. 34 A meta-analysis of observational studies of hygiene practices associated with child faeces found that failure to remove child faeces and unhygienic handling practices were associated with a 23% increased risk of diarrhoea. 35 Other observational studies have reported that animals kept in shared outdoor living spaces (compounds) increase the risk of diarrhoea by over 50%. [36] [37] [38] However, no studies that we know of have quantifi ed the risks associated with the use of cow dung for fuel or in house maintenance. So far no intervention trials have aimed to reduce animal faecal contamination in domestic spaces. Neither are there any reports of trials of improving the disposal of child faeces by use of potties, nappies, or child-friendly toilets.

Household surfaces seem to play a major part in disease transmission, although most evidence is from developed countries, [39] [40] [41] and few intervention studies have tested whether surface cleansing can reduce transmission in any setting. Larson 

with similar products without an antibacterial agent, and found no additional benefi t. However, a small study in a school setting suggested that regular cleaning of desks and other classroom surfaces reduces the risk of gastrointestinal illness. 43 Epidemiological evidence of the health risk associated with solid-waste disposal in low-income settings is scarce. Observational studies have shown a strong link between environmental exposure to solid waste and diarrhoea, 44 perhaps because waste heaps are sometimes used for open defecation and disposal of excreta. In addition to attracting insect vectors and fl ies, waste is associated with Lassa fever infection which is transmitted by rats. 45 In some settings fl y control might reduce diarrhoea risk by around 25%, [46] [47] [48] and lessen the risk of trachoma.

Because there are multiple routes for the transmission of gastroenteric pathogens, many hygiene intervention studies have targeted several behaviours at once. Such an approach can dilute the eff ect of the intervention. For example, Haggerty and co-workers 49 did a large clusterrandomised trial to test the eff ect of promoting four diff erent hygiene behaviours (handwashing after faecal contact, handwashing before food contact, disposal of animal faeces, and disposal of child faeces). No eff ect on diarrhoea was reported in this study, perhaps suggesting that changing four distinct hygiene practices over a short time is unrealistic.

The biological plausibility of most hygiene interventions is high (table 1) ; there is, however, a major shortage of evidence from trials. Trials on this topic can be complex and the results misleading; it is hard to mask participants to the nature of the intervention, which can lead to bias in outcome reporting. One way to improve this situation is to use more objective out come measures, such as health-care seeking, assessments by health-care workers masked to intervention status, or mortality. Future hygiene trials need to be larger to model full-scale programme implementation and more intensive (and therefore costly) than previous trials to objectively assess outcomes. Large, adequately funded trials are urgently needed to assess the eff ects of intervening to improve three key practices in particular: handwashing, safe disposal of child stools, and promotion of food hygiene.

The immediate question is what public health actions should be taken now? Whether an intervention can be recommended for implementation depends not only on the evidence of disease reduction, but also on its scalability, acceptability, and the risk of adverse eff ects. 50 The weight of evidence suggests that hygiene promotion is eff ective in reducing disease, can be promoted both directly and by mass media programmes with relatively low expenditure per person targeted, 1,51 and has few adverse eff ects. Even if the true eff ect on disease in low-income settings is smaller than studies suggest, hygiene improvements will likely have an eff ect on disease control at large scale. Although additional intervention trials using improved outcome measures are urgently needed to confi rm previous fi ndings, hygiene promotion can already be recommended for large-scale implementation.

While surveys such as multiple indicator cluster surveys and demographic and health surveys systematically collect data on key health indicators, only recently have they begun to include data on hygiene practices. One reason for this is that questionnaire-based surveys are inadequate for gathering data about private and morally bound issues such as food and hand hygiene because they overestimate rates of handwashing, for example, by two to three times. 52 Eff orts are continuing to identify indicators of hygiene practice that are both valid and simple to collect. [53] [54] [55] An article 56 published in 2009 collated data about directly observed handwashing in 11 countries, and we identifi ed another survey 57 from Bangladesh in 2008 (table 2) . Handwashing with soap by child carers at key moments, such as after using the toilet, was rare, varying from 3% in Ghana to 42% in Kerala, India. Handwashing with water alone happens on a further 45% of occasions, on average. Handwashing with soap was also rare after cleaning up children and before handling food. If these fi gures are a good guide, less than one in six children in developing countries is protected from disease by handwashing with soap at key moments. This contributes perhaps a million unnecessary deaths to the global toll. 61 Handwashing behaviour is far from ideal in developed countries. In a motorway service station in the south of England, just 65% of women and 31% of men washed their hands with soap after using the toilet facilities, 62 Review dirty nappy. 63 A survey by Judah and co-workers 64 reported that 28% of commuters in fi ve UK cities had bacteria of faecal origin on their hands. If improvement of hygiene practices, such as handwashing with soap, has the potential to be one of the most cost-eff ective ways in which public health can be improved in developing countries, how should we go about it? Though changing behaviour is diffi cult, we know a lot more about hygiene behaviour than we did 10 years ago and promising approaches to changing hygiene on a large scale are emerging.

Risky hygiene behaviours persist around the world because of a web of factors that can be hard to shift. Poor environmental conditions, such as lack of water, sanitation, and drainage, have a role. Other obstacles include the absence of hard surfaces that can easily be kept clean, unavailability of cleaning materials such as soap and surface cleansers, and limited access to hygiene aids, such as potties or nappies. Local social structures and cultural norms, as well as individual psychological factors, also help to keep present practices locked in place. For behaviour to change one, or several, of these factors will have to be addressed, but to do so will require a better understanding of them. Several formative research studies that aimed to provide an understanding to enable the design of eff ective handwash programmes have investigated the behavioural aspects of hygiene. [65] [66] [67] A review of 11 studies done in Africa, Asia, and Latin America concluded that, although there are local diff erences, common patterns exist. Three kinds of hygiene behaviour were identifi ed: habitual, motivated, and planned. 56 Hygiene habits were learnt at an early age, but soap use was rarely taught by parents or schools. Key motivations for handwashing were disgust of contamination on hands and to do what everyone else was perceived to be doing (social norms). Other motivations included comfort and nurturance (the desire to care for one's children). Planned handwashing, with the aim of preventing disease, took place rarely. Mothers did not fi nd the threat of diarrhoeal disease particularly relevant and found the connection between handwashing and possible diarrhoea in children tenuous. Mothers did, however, plan to teach their children good manners, and they also planned to economise by ensuring that soap was not wasted. Aunger and colleagues 68 observed that habit was the most powerful determinant of handwashing in Kenya, followed by several motives including disgust, and social norms, and cognitive plans to save money. An investigation into nurses' handwashing in Australia saw evidence of a similar distinction between planned and motivated or habitual handwashing behaviour. 69 A study of routine behaviour and hygiene in rural India suggested that some handwashing behaviours are deeply embedded in daily routines and hence highly habitual, whereas others are motivated by the transient disgust or discomfort of having dirty hands. 70 The psychological factors determining hygiene are related to factors in the environment. For example, when local social norms are the source of poor handwashing habits people commonly practise what they perceive everyone else to be doing, which reinforces the norm of not using soap. Lack of water and a perception that soap is too expensive for handwashing could also constrain handwashing (though this might be post-rationalisation because one review suggested that soap was present in 97% of all households in a review, but it was used mainly for clothes, body, and dish washing). 56 Finally, fear of epidemics such as cholera or severe acute respiratory syndrome (but not of endemic diarrhoeas that cause far more deaths) could also lead to improved hygiene. 56 Although an understanding of the determinants of handwashing behaviour is helpful, how such insight can be used in behaviour change programmes is not always obvious. The fi gure summarises the psychological and environmental factors that are likely to determine hygiene behaviour. 56 If much handwashing is habitual, then the cues that trigger these habits need to be found and the habits established at an early age. Environmental changes that make handwashing easier and cheaper, such as the introduction of simple water-saving technologies (eg, socalled tippy taps), could be helpful (although how their use could become widespread is unclear), as could information suggesting that handwashing with soap is a desirable social norm where it is not one.

Some of these hypotheses have already been tested. An experimental study 71 in Australia recently reported that promoting disgust led to increased handwashing in a public toilet, as it did in a service station intervention in the UK. 62 A national handwash programme in Ghana that used disgust and nurture to motivate handwashing increased self-reported handwashing before eating by 41% and after defecation by 13%. 51 Disgust was also used humorously in an urban social marketing programme in Burkina Faso. The project increased observed handwashing with soap by mothers from 1% to 17% after using the toilet 

and from 13% to 31% after cleaning up a child (panel 1). 72 A norms-based message, "Is the person next to you washing with soap?", worked best to encourage handwashing in a motorway service station in the south of England. 62 Other promising approaches, such as trying to establish hygiene habits in schools, are thought to be eff ective. Unpublished evidence from Kenya, Peru, and Uganda suggests that working through schools might have a double advantage: children take up what they are taught and might also take messages home, hence infl uencing their families.

The standard approach to hygiene promotion, whether through schools, clinics, or health outreach programmes, has, until recently, been educational. However, knowledge about possible long-term health eff ects does not necessarily translate into practice. There is little proof that such educational approaches are eff ective, either in developing, 73, 74 or developed countries. 75 In the past two decades an approach known as PHAST (participatory hygiene and sanitation transformation) has become the predominant model among nongovernment organisations. Although it is an imaginative attempt to involve communities in solving their own hygiene problems, PHAST is mostly an educational approach, is heavily reliant on the skills of trained facilitators, and is diffi cult to implement on a large scale. There are no rigorously collected data to support the eff ectiveness of PHAST programmes, and some evidence from Tanzania and Uganda indicates that the approach has limited eff ect on hygiene behaviour. Community health clubs were successful and cost eff ective in promoting sanitation and hygiene in two districts of Zimbabwe (panel 2), 76 largely because communal activities can change local norms.

Many programmes promote hygiene in schools. Although evidence of eff ect is scarce, data from a watertreatment and handwashing intervention in Kenya and an intensive handwashing educational programme in Chinese primary schools showed a reduction in absenteeism. 77, 78 The biggest obstacle to school hygiene might be the shortage of facilities; for example, studies in Kenya and Senegal showed that only 5-10% of schools had soap available for children to use. 79 Although all of these programmes might have helped to improve hygiene behaviour in their target communities, proven approaches to hygiene promotion that are eff ective on the large scale and that will help meet the Millennium Development Goals for child survival are needed. The most promising approach is that developed by the Global Public-Private Partnership for Handwashing with Soap (panel 3).

Far more is known about hygiene now than a decade ago. We understand the need to invest in hygiene and the key practices that require change, and we have appealing new ways of promoting hygiene. If hygiene promotion is truly the most cost-eff ective intervention for preventing disease in developing countries, 1 then it is extraordinary that hygiene features so seldom in international public health eff orts. What then holds back major investment in the improvement of hygiene?

The health sector needs to address four major challenges for hygiene to take its rightful place as a major issue within global public health. First, governments and ministries have to stop merely talking about the need for hygiene and instead act, investing in programmes that can actually change hygiene behaviour in villages and towns where children are dying from neglected diseases. Second, hygiene promotion has to fi gure in the job description for health agents, from the heads of health

Programme Saniya aimed to improve handwashing and stool disposal behaviour in the town of Bobo-Dioulasso in Burkina Faso. Based on principles of social marketing, including use of existing respected in-depth research, the programme was tailored to local customs and targeted specifi c types of behaviour, built on existing motivations for hygiene (social and aesthetic rather than health-based), and used locally appropriate channels of communication, including neighbourhood committees, street theatre, schools, and local radio. After the programme had run for 3 years, three-quarters of the mothers targeted had been involved with programme activities and half could cite the two main messages of the programme correctly. Although the safe disposal of children's stools changed little between 1995 and 1998 (80% before intervention and 84% after), handwashing with soap after cleaning a child's bottom rose from 13% to 31%. The proportion of mothers who washed their hands with soap after using the latrine increased from 1% to 17%. 2 The estimated household and societal cost savings associated with the programme far outweighed its costs. 4

The innovative method of community health clubs used in Zimbabwe signifi cantly changed hygiene behaviour and built rural demand for sanitation. 76 Villagers were invited to a series of weekly sessions where one health topic was debated and then action plans formulated. These proved highly popular with mothers. In 1 year in Makoni District, 1244 health sessions were held by 14 trainers, costing an average of US$0·21 per benefi ciary and involving 11 450 club members. In Tsholotsho District, 2105 members participated in 182 health promotion sessions held by three trainers which cost $0·55 for each benefi ciary. Club members' hygiene was signifi cantly diff erent (p<0·0001) from a control group regarding 17 key hygiene practices including toilet building and handwashing. The authors of the study concluded that if a strong community structure is developed and the norms of a community are altered, sanitation and hygiene behaviour are likely to improve. 76 services to the most remote rural community health worker. Third, massive eff orts need to be made to train health workers in the skills of hygiene promotion. This is important because otherwise they will continue to use outdated methods and health education approaches that are demotivating because they are ineff ective. Fourth, although we know enough to act now, gaps in our knowledge exist. Health research funders need to make up for some of the decades of underinvestment in hygiene. Support is needed for the research that will allow us to say with more certainty how to change hygiene behaviour on a large scale, what improved hygiene will cost, and what the fi nancial returns will be.

There are encouraging signs that, although investment still remains low, the topic of hygiene is moving up the political agenda. As pointed out by the former director of the World Bank Jim Wolfenson, hygiene is no longer seen as a joke. Inspired advocacy events, such as the Global Handwashing Day organised by the Global PPP-HW, have enhanced the global profi le of hygiene. Celebrated every year on October 15, the day involves imaginative high-profi le activities organised by public and private players from around the world. To become the focus of real investment, rather than good intentions, hygiene needs champions at all levels: from global, right through to village, and especially national ministries of health. Hygiene needs to fi nd a place in national health plans and in poverty reduction strategies. Donors need to actively solicit hygiene promotion programmes and bring companies interested in promoting hygiene into the public health fold, rather than treating them with suspicion, as is sometimes the case.

Coordination is a key issue in hygiene improvement; each country needs to designate a focal point to provide eff ective management of diverse eff orts. Greater impact could be achieved if the many agencies, donors, nongovernmental organisations, companies, and government and citizen institutions with hygiene in their mandates could agree upon a few simple principles and harmonise their approaches. Every mother who has contact with a health worker during pregnancy or in the neonatal period needs to learn about the importance of hygiene, and handwashing in particular. Similarly, every family member who prepares food needs to know a few basic rules of food hygiene. If the coordination of outreach to the community is seen as crucial to eff orts to combat HIV and malaria, the same should also be true for hygiene promotion.

In increasingly decentralised countries, policy building work needs to take place at national, regional, and provincial level. Creating this framework is challenging given competition for attention in the relevant ministries, their limited human and logistic resources, and shortage of skills. These problems are often worst in the remote and poverty-stricken areas that would benefi t most from improved hygiene. Global leaders need to get involved to help show that hygiene is not a dirty contaminated topic, but one that can be attractive and popular, increasing votes, attention, and resources.

For action on hygiene to become part of the remit of health workers, greatly increased investment in the development of capacity is needed. Training in up-to-date methods of communication is lacking at all levels in health ministries. Tertiary institutions that can provide this training need training themselves, and this is an area that could be addressed by external funders; although for some reason funding such skills development has, unfortunately, not been a priority for donors in recent years. Marketing expertise from the private sector has been helping to fi ll the skills gap, by designing state-of-the-art hygiene communication programmes and helping to train health offi cials in the techniques of marketing. Links between government health bodies and private organisations could be developed on a wider and more formal basis. The

Conceived as a way of combining the expertise and resources of the private and public sectors, the Global Public-Private Partnership for Handwashing with Soap (PPP-HW) has been building coalitions and national programmes in over 17 developing countries. Usually based out of ministries of health, the programmes aim to work at national scale. The use of formative research to investigate the determinants of handwashing behaviour, 1 use of professional creative agencies to design coherent, attractive, and outstanding national communications strategies with advice from industrial marketers, and attempts to evaluate programmes rigorously are also key features. Good preliminary results were achieved after 1 year of programme activities in Ghana and the model is convincing and innovative enough to have attracted substantial funds, but questions about eff ectiveness remain.

In Ghana, Peru, and Senegal public health authorities and soap companies were not always easy bedfellows, and national partnerships can be hard to sustain for longer than just one campaign. Partnerships require constant attention from a full-time coordinator skilled in reconciling public and private stakeholders. Coordinators must be able to secure commitment for the approach at the highest levels and keep the vision alive through constant change of personnel in both sectors (for example, from 2003 to 2009: Peru had six ministers of health).

It is a key objective of these partnerships to gather rigorous public domain evidence about the eff ectiveness of large-scale handwashing promotion eff orts. However, gathering such data is proving challenging, especially when interventions cannot be centrally controlled because they depend on the goodwill of various partners.

For the future, some of the most promising initiatives arising from the PPP-HW's work are coming from the soap companies themselves. Unilever, for example, has made handwashing part of the social mission of the Lifebuoy brand and has pledged to bring handwashing to one billion people by 2012. Procter & Gamble and its Safeguard brand has reached 35 million children with one-on-one, school-based handwashing education and is planning to reach 100 million more by 2012. 8 Colgate runs a "Clean Hands-Good Health" campaign in many countries to promote proper handwashing practices among school children with developmentally appropriate materials and a classroom curriculum. Local soap companies such as GeTrade in Ghana are also contributing their eff orts. If promoting hygiene in emerging markets makes good business sense to private companies, because they can improve sales and enhance their reputations, this could lead to sustained long-term improvements in public health, as it has in developed countries. 9

Review interface between programmes and health research is also problematic, as it is for health development in general. Local universities are the obvious institutions for designing and evaluating hygiene promotion programmes, but very few have that capacity at present.

Another question for policy makers is whether hygiene should be promoted alone or in concert with eff orts to improve water and sanitation infrastructure. The introduction of a new water supply to a community is a perfect opportunity to raise the issue of hygiene. However, large-scale engineering programmes are rarely equipped to handle what they call the software (ie, the behavioural) side of development. Equally, the most eff ective use of a hygiene budget might be to cover larger areas by use of mass media, rather than to restrict eff orts to villages in the process of acquiring new water facilities. Hygiene messages should always be integral to eff orts promoting improved sanitation. Ministries of health can play a part by insisting that it is not acceptable to build toilets in schools, health facilities, workplaces, and homes without appropriate hand washing facilities.

Finally, good professional practice requires continual advancement in a feedback loop of learning and knowledge development whilst doing. Much can be done now. Far more could be done with serious investment to fi ll some of the important knowledge gaps about hygiene Trials of interventions to change key hygiene practices Randomised controlled trials are needed to test interventions to improve hygiene practices, including handwashing, safe stool disposal and food hygiene. Such studies should use objective outcomes such as clinical infection or mortality.

Small-scale testing of approaches in a laboratory or community setting, as well as large-scale screening, can provide answers about what works best to change hygiene behaviour and assurance of eff ectiveness before interventions are rolled out at a large scale.

The eff ectiveness, cost-eff ectiveness, and diff erential impact of diff erent channels? An analysis of the diff erent routes of communication used in the Ghana PPP-HW campaign suggested that TV and radio had greater reach and impact than community events. Further analytical studies into the eff ectiveness and cost-eff ectiveness of diff erent channels of communication are needed. We need to know more about the diff erential impact of diff erent approaches on the low-income sections of society, which are at greatest risk of death from diarrhoeal disease and have fewer resources to commit to hygiene. 80 We also need data to calculate dose-response curves: how much intervention produces how much behaviour change, and hence what level of investment is most cost eff ective?

The process of turning insight about behaviour into effective behaviour-changing communication is still more of an art than a science. More needs to be understood about what makes communications attention-grabbing and memorable, as well as motivating. Habit clearly has an important role in hygiene and many other health behaviours, but the topic of how to create and change habits has been little studied. 6, 7 Methods and models for hygiene promotion at diff erent scales Proven model approaches to hygiene promotion are badly needed by decentralised authorities and non-government organisations. Such agencies are often willing to implement hygiene promotion, but rarely have the specifi c expertise or capacity to develop the approaches themselves. Several examples of simple, eff ective, attractive, and costed activities and materials that have been tested and have been shown to work, are needed so that organisations can adapt these to local circumstances.

Even when we are successful in changing hygiene behaviours we still do not know how persistent such changes are, 11, 12 or the sort of investment that is needed to maintain the gains in a given population. Perhaps the most important tasks facing hygiene promoters and soap manufacturers are to work out how to make hygiene a matter of habit and a social norm. Once hygiene is established, improvements in behaviours will be truly sustainable.

If we cannot accurately measure changes in hygiene behaviour we cannot measure the eff ectiveness of interventions in trials or evaluate the delivery of behaviour change in programmes. Because hygiene behaviour is private and morally loaded, simple questionnaire surveys give overestimates of behaviours such as handwashing, whereas direct observation is cumbersome and intrusive, and technological fi xes, such as Smart Soap (containing accelerometers that record usage) have drawbacks too. 14 Simple, cheap, and widely applicable methods of measuring hygiene behaviour change are still needed.

Although simple technologies, such as water-saving taps, nappies, potties, and child-friendly toilets, can help families to live more hygienically, little eff ort has been made to develop and market hygiene-helping products that are appropriate for the consumers with low income. Three things are needed: exploration of the design space for the products that the poorest consumers need and want, 16 the adaptation or creation of technologies, products, and services that meet those needs, and the development of business models that can operate profi tably and be sustained on a large scale.

that still remain. Panel 4 sets out crucial questions that need answering urgently if we are to be able to deliver better hygiene programming in the future.

Though the evidence base is far from complete, the information we do have strongly suggests a need to improve handwashing behaviour, stool disposal practices, and food hygiene in particular when weaning. We know that hygiene can be promoted successfully through conventional health channels, water and sanitation initiatives, schools, and by commercial companies.

The fi rst priority for any new resources allocated to hygiene is the design, management, and rigorous evaluation of large-scale hygiene promotion programmes (using randomised trials, where possible).

Second, we need more medium-scale programmes, operating at rural or urban district level. Such programmes provide the opportunity to learn more of the basics of hygiene promotion, how to turn insight about hygiene into eff ective promotional campaigns, how to invest to get the most behavioural change, which channels to use, how best to reach the most vulnerable, how often and how much to intervene and how to sustain behavioural changes. The capacity to implement medium-scale programmes needs to be built through learning by doing. Programmes of research led by local universities, with international support where needed, can begin to tackle these multiple issues. Because there are many diff erent ways to promote hygiene, having more diverse and properly evaluated programmes will build a body of knowledge as to what works best in changing these persistent habitual behaviours.

Third, we need more dedicated epidemiological research funded by international donors and research councils. Many RCTs investigating effi cacy and eff ectiveness are needed to provide rigorous evidence of the importance of improving individual hygiene practices. Food hygiene has the best claim to be tackled fi rst, but all of the hygiene practices we have discussed have been neglected relative to the eff orts for malaria or HIV research. For example, there is no evidence to indicate whether the practice of using animal dung to smooth fl oors and walls (common in Asia) is injurious to family health.

Fourth, we need to set new and more ambitious targets for the coming decades. It is unacceptable that, in the 21st century, most schools in developing countries still do not have sanitation and hygiene facilities, or that health centres, hospitals, maternity facilities, workplaces, and public institutions still cannot off er water and soap to their users. It is unacceptable that birth attendants and outreach workers are not always trained to wash hands with soap and do not systematically promote handwashing and hygiene to mothers. Handwashing and hygiene should be promoted at least as aggressively as vaccination. For the future every child should have the right to live in a household that is protected from disease by good hygiene.

In the next 5-10 years we have a window of opportunity to develop the high impact programmes which will bring about mass scale changes. If these programmes are successful in leading the members of all societies to adopt hygienic habits as a matter of course, then hygiene will be able to take its rightful place as one of the foundation stones of global health.

VC wrote the fi rst draft and subsequent drafts, revised and fi nalised the paper. SL made substantial contributions to the content and conclusions of the paper, reviewing and fi nalising. WS made contributions to the content, reviewing and fi nalising. RF made contributions to the content, especially concerning policy issues. OT made contributions to the content in the food hygiene section. AB made contributions to the structure of the paper.

VC has received a research grant from Unilever, has been a consultant in a think tank on hygiene for Kimberly Clark, and has been a consultant for a hygiene resource for health workers at Colgate Palmolive. AB has receiveda grant from Unilever to evaluate a hygiene intervention. RF, SL, WS, and OT declare that they have no confl icts of interest.

We searched Medline, from 1970 to 2009, regardless of language, using the search terms "[diarrh(o)ea AND hygiene]", "[respiratory AND hygiene]", "[food AND hygiene]", "[pneumonia AND hygiene]", "[stool AND disposal]", "[waste AND hygiene]", and "[animal faeces AND diarrh(o)ea]". We searched the reference lists of relevant articles and contacted authors and experts for further identifi cation of articles. Table 1 draws together our assessment of the available evidence from available reviews (and other key papers) concerning the four sources of evidence, biological plausibility, risk modelling, observational studies, and randomised controlled trials.

Socioeconomic status and health in blacks and whites: the problem of residual confounding and the resiliency of race

Empirical evidence of bias in treatment eff ect estimates in controlled trials with diff erent interventions and outcomes: meta-epidemiological study

Household water treatment in poor populations: is there enough evidence for scaling up now?

Eff ect of washing hands with soap on diarrhoea risk in the community: a systematic review

Hand washing for preventing diarrhoea

Handwashing and risk of respiratory infections: a quantitative systematic review

Eff ect of handwashing on child health: a randomised controlled trial

Maternal and birth attendant hand washing and neonatal mortality in southern Nepal

A review of the evidence for the 'F' and 'E' components of the SAFE strategy for trachoma control

Ascariasis and handwashing

Evaluation of three years of the SAFE strategy (Surgery, Antibiotics, Facial cleanliness and Environmental improvement) for trachoma control in fi ve districts of Ethiopia hyperendemic for trachoma

Risk factors for trachoma in Mali

Transmission ecology of the fl y Musca sorbens, a putative vector of trachoma

Impact of face-washing on trachoma in Kongwa

Eff ectiveness of handwashing in preventing SARS: a review

Physical interventions to interrupt or reduce the spread of respiratory viruses: systematic review

Studies of food hygiene and diarrhoeal disease

Growth and survival of Shigella fl exneri in common Bangladeshi foods under various conditions of time and temperature

Contaminated weaning food: a major risk factor for diarrhoea and associated malnutrition

World Health Organization. Prevention of Foodborne Disease: Five Keys to Safer Food

The eff ects of postnatal health education for mothers on infant care and family planning practices in Nepal: a randomised controlled trial

Promotion of exclusive breastfeeding is not likely to be cost eff ective in West Africa. A randomized intervention study from Guinea-Bissau

An educational intervention to promote appropriate complementary feeding practices and physical growth in infants and young children in rural Haryana

Implementation of the Hazard Analysis Critical Control Points (HACCP) method to improve microbiological food safety in peri-urban Mali

Feces, fl ies, and fetor: fi ndings from a Peruvian shantytown

Defecation practices of young children in a Peruvian shanty town

Incidence and etiology of infantile diarrhea and major routes of transmission in Huascar

Children's feces disposal practices in developing countries and interventions to prevent diarrheal diseases. A literature review

Compound hygiene, presence of standpipe and the risk of childhood diarrhoea in an urban settlement of Papua New Guinea

Paediatric campylobacter diarrhoea from household exposure to live chickens in Lima

The epidemiology of acute diarrhoea in a rural community in Imo State

Survival of Salmonella in bathrooms and toilets following salmonellosis

Disinfection and the prevention of infectious disease

Achieving hygiene in the domestic kitchen: the eff ectiveness of commonly used cleaning procedures

Eff ect of antibacterial home cleaning and handwashing products on infectious disease symptoms: a randomized, double-blind trial

Reducing absenteeism from gastrointestinal and respiratory illness in elementary school students: a randomized, controlled trial of an infection-control intervention

Rubbish index and diarrhoea in Salvador, Brazil

Poor housing quality increases risk of rodent infestation and Lassa fever in refugee camps of Sierra Leone

Impact of fl y control on childhood diarrhoea in Pakistan: community-randomised trial

Trachoma and fl y control

Eff ect of fl y control on trachoma and diarrhoea

Community-based hygiene education to reduce diarrhoeal disease in rural Zaire: impact of the intervention on diarrhoeal morbidity

The weight of evidence: a method for assessing the strength of evidence on the eff ectiveness of HIV prevention interventions among young people

Marketing hygiene behaviours: the impact of diff erent communications channels on reported handwashing behaviour of women in Ghana

Structured observations of hygiene behaviours in Burkina Faso, validity, variability and utility

Comparing the performance of indicators of hand-washing practices in rural Indian households

Associations among handwashing indicators, wealth, and symptoms of childhood respiratory illness in urban Bangladesh

A simple index to measure hygiene behaviours

Planned, motivated and habitual hygiene behaviour: an eleven country review

Household characteristics associated with handwashing with soap in rural Bangladesh

Marketing hygiene behaviours: the impact of diff erent communication channels on reported handwashing behaviour of women in Ghana

Hard to handle: understanding mothers' handwashing behaviour in Ghana

Formative research for hygiene promotion in Kyrgystan

Eff ect of washing hands with soap on diarrhoea risk in the community: a systematic review

Experimental pretesting of hand-washing interventions in a natural setting

Hygiene in the home: relating bugs to behaviour

Dirty hands: bacteria of faecal origin on commuters' hands

Formative research for hygiene promotion in Kyrgyzstan

Dirt and diarrhoea: formative research for hygiene promotion programmes

Data to action: using formative research to develop intervention programmes to increase physical activity in adolescent girls

Three kinds of psychological determinants for hand-washing behaviour in Kenya

Why healthcare workers don't wash their hands: a behavioral explanation

Lifebuoy fi ve occasions study

Can the emotion of disgust be harnessed to promote hand hygiene? Experimental and fi eld-based tests

Evidence for behaviour change following a hygiene promotion programme in West Africa

Health education interventions in developing countries: a methodological review of published articles

The eff ects of postnatal health education for mothers on infant care and family planning practices in Nepal: a randomised controlled trial

The eff ectiveness of interventions aimed at increasing handwashing in healthcare workers-a systematic review

Creating demand for sanitation and hygiene through Community Health Clubs: a cost-eff ective intervention in two districts in Zimbabwe

A cluster-randomized controlled trial evaluating the eff ect of a handwashing promotion program in Chinese primary schools

The impact of a school-based safe water and hygiene programme on knowledge and practices of students and their parents

Can hygiene be cool and fun? Understanding school children's motivations to use their school toilets and wash their hands with soap in Dakar

Determinants of handwashing practices in Kenya: the role of media exposure, poverty and infrastructure