Their first method involved direct,open-ended observations of ten house-holds, all of which included weaning-age children (12 to 23 months old).Each household was observed at leastfour times for a total, on average, of18 hours. Towards the end of thesecond visit, the researchers collectedseveral water and food samples, to-gether with swabs from the family'scooking utensils, and the children'shands (taken just before eating). Thesewere then analysed for faecal contami-nation.

The study team's second methodconsisted of open interviews with alarger group of mothers, to find out ifthe behaviour observed in the initialten households was prevalent in thewider community. Families with chil-dren aged between 12 and 23 monthswere identified through municipal re-cords, and 100 were selected at ran-dom. It was difficult to conduct whollyaccurate observations of the families'normal habits: partly because theychanged their behaviour when theyknew they were being observed, andpartly because the activities of particu-lar interest (such as handwashing and

used several methods: they observedfamilies at close range; they carriedout interviews with a few households;a survey was made; and specifichygiene recommendations were testedin the home environment.

The community of Ciudad Vieja hasaround 20000 inhabitants, all Span-ish-speaking Mestizos. Living condi-tions are precarious: most houses aremade of reeds and thatch, few streetshave been paved, and only a fewhouseholds are connected to the sew-age network. Most homes, however,do have indoor piped water, latrines,and electricity. The water supply isrelatively reliable, but its quality variesgreatly, from no coliforms to unaccept-ably large numbers of coliforms.

Research methodsThe initial task of the INCAP research-ers was to gather relevant information.

Weaning-food hygiene in aGuatemalan townby Sandra Saenz de Tejada and Floridalma CanoFood contaminated through poor handling,preparation, or storage can lead to diarrhoea.Can researchers target the everyday practicesripe for modification?

FOOD-BORNE DIARRHOEAS ac-count for a considerable proportion ofall diarrhoeas. One way or another,pathogens can contaminate food andmultiply in it and, if ingested inlarge-enough quantities, they can causeillness. Using a combination of anthro-pological and microbiological meth-ods, a team from the Institute ofNutrition of Central America andPanama (lNCAP) investigated thecommon domestic-hygiene practices- with special reference to weaningfoods - in the Guatemalan town ofCiudad Vieja, in Sacatepequez prov-ince, about 50kms west of GuatemalaCity.

The researchers' main objective wasto identify which specific activitiescontribute to the bacterial contamina-tion of weaning foods; with the ulti-mate aim of changing them throughan education intervention. Researchers

DEVELOPMENT MANAGEMENTThe Development and Project Planning Centre of the University of Bradford is offering two important new courses inthe area of development management in 1995:

Community Management of Development Projects: Water, Sanitation and Health7 May - 14 July 1995

This course is intended for practitioners in projects or programmes in water, sanitation and health, which includecommunity management. It will focus on the interaction of government, NGOs and private organisations with 'thecommunity' and bow problems arising at this interface can be overcome. Modules will include management skills,planning techniques, policy issues, appropriate technologies, computing and project planning.

Manaling Sustainable Projects and Programmes: Agricultural, Rural and Natural Resource Projects24 April- 14 July 1995

This twelve week course is intended for people involved in the management of sustainable development programmesacross the range of sectors which utilise both natural and human resources. It is divided into modules covering naturalresource processes; management skills; and managing sustainable development - financial sustainability, institutionaldevelopment, environmental management.

The opportunity to learn from experience and practice in different countries will be emphasised. For furtherinformation contact:

Professional Development and Training Programme SecretaJ)' (ATW)Development and Project Planning CentreUniversity of Bradford, Bradford, BD7 1 DP, EnglandFax: +44 1274 385280

24 WATERLINES VOL.13 NO.3 JANUARY 1995

A Ciudad Vieja mother washes nixtamal at the cement hasin.

food reheating) occurred only at cer-tain times, usually when visitors werenot expected!

One hundred mothers were inter-viewed by an anthropologist and threefieldworkers. The survey includedopen questions on water use, foodhandling and reheating, and maternaleducation. The team also closely ob-served the common methods of storingwater, mothers' handwashing prior tofood handling, children's handwash-ing, and the type of water used forwashing vegetables and for preparingcold drinks. This method allows peopleto respond with full explanations;instead of asking 'Did you eat beansthis morning?' the team asked 'Whatdid YOll eat today for breakfast?'

The last three activities listed areclassed as 'prompted behaviour', as theinterviewers asked the mothers to carryout specific tasks. The women wererequested to prepare a salad or a sweetdrink - with ingredients provided bythe research learn - while the re-searchers measured all the ingredientscarefully, pretending to be interestedin traditional seasoning. As most of theinterviews took place in the morning,the researchers told the women thatthey would like to contribute towardslunch - the family's main meal -and that they were interested to seehow much sugar (or salt and lime) thefamily usually used in their drinks orsalads; this diverted the mother's atten-tion away from hygiene.

Handling waterTo see how children's hands werewashed, the team asked each motherat the end of the interview to wash oneof her children's hands in order toleave a 'souvenir' from the visit: animprint of the child's hand on somecard. While this 'display' for thevisitors might still reflect ideal behavi-our, the notes recorded during the visitdid provide useful information on thefamily's hygiene practices.

Each family in Ciudad Vieja usedthree distinct categories of water: waterkept in a cement basin outside; runningtap water; and water kept in containersinside the kitchen. Although the qual-ity of the piped water was relativelygood, the stored water was noticeablydirty, and microbiological analysisrevealed high concentrations of faecalcoliforms.

Water from the basin was used forwashing dishes, clothes, nixtamal,(maize cooked in lime from which thestaple, tortillas, are made), most vege-tables, and for personal hygiene pur-poses. The families drank water fromthe tap, and, because of the distance

between the kitchen and the cementbasin, food was prepared with theunsafe water stored in the kitchencontainers. When they were requestedto prepare the salad, almost a third ofthe mothers washed the radishes withwater from the basin, another thirdused tap-water, and the remainingwomen did not wash them.

Concepts of hygieneThe research team found that thewomen of Ciudad Vieja did believethat there was a relationship betweendiarrhoea and food hygiene. This ideaof hygiene, however, seemed to refermostly to dirt associated with domesticanimals, flies, and dust, and withinsufficiently cooked food, rather thanto their personal-hygiene practices.Their notion of cleanliness was tied tooutward appearance: mothers wereoften observed wiping children with adirty rag.

Although the mothers were alreadyaware of the benefits of handwashing,they rarely practised it. Even whenreadily available in the cement basin,soap was seldom used. Their hand-washing method consisted of rinsingand lightly scrubbing their hands in aplastic bowl filled with water from thebasin; drying them - usually in anapron - and only occasionally (butroutinely if an observer was present)with a clean cloth. The team's findingsshowed that, when asked to wash achild's hands, only 15 per cent ofcarers used soap, almost half scrubbedtheir hands, 13 per cent rinsed theirhands in clean water, and 53 per centdried them in a clean cloth. Whenasked to prepare the salad or drink. I Iper cent of mothers rinsed their hands.and none used soap. Mothers thoughttheir hands were kept clean as a resultof frequent contact with water, eitherby washing dishes or clothes. Theytended to wash their hands only if theywere visibly soiled.

WATERLINES VOL.13 NO.3 JANUARY 1995 25

A fieldworker demonstrates the easy-to-use hose for handwashing.

Preparing foodIn most Ciudad Vieja families, themother is responsible for preparing andserving food, and for feeding herchildren. Babies are introduced tosolids - food from the family pot -when they are about seven months old.Almost 60 per cent of all children agedbetween 12 and 23 months were foundto be breast-fed to some extent.

The child's diet is rather monoto-nous and, as in other poor ruralGuatemalan households, it was foundthat eight food items (sugaredcoffee, tortillas, rice, black beans,eggs, noodles, bread, and bananas)provided almost 80 per cent of thetotal calorie intake. For breakfast,young children usually eat breadand coffee; lunch is freshly madevegetable soup, or stew and tortillas;and, for dinner, they eat black beans, .tortillas, and the leftovers fromlunch. Only half of the childrenobserved ate more than three meals

a day; snacks, consisting of fruit,tortillas, bread, and coffee.

The family's meals were, almostexclusively, cooked; only a few rawitems were consumed. Most house-holds cooked tortillas once a day but,because of the high cost of fuel, somecould only afford to do it every 36 to48 hours. Several different sources ofcontamination were found in the tor-tilla preparation: on the grindingstones, in the water, and in the dough.Microbiological analysis revealed,however, that despite being preparedunhygienically, the freshly made tortil-las were not contaminated. Black beanswere cooked two to three times a week,each batch lasting about three days.When the beans were freshly cookedand still whole, young children weregiven broth as an alternative, as moth-ers believed that their immature stom-achs could not cope. The next day, thebeans were usually fried and pureedbefore being given to children.

Storing foodThe bulk of the family's food wasconsumed soon after being cooked, butblack beans and tortillas were oftenstored for over 24 hours. Mothers saidthat they preferred to eat only freshlyprepared foods, but it was observedthat, besides the staple foods, someother items (particularly rice, noodles,and soups) were deliberately cookedin quantities adequate for two meals.

The women were already aware thatinadequate storage led to food spoil-age. They defined spoiled food asanything that was clearly acidic, slimy,or mouldy. The families had no con-cept of germ contamination or prolif-eration, and the idea that food couldbe contaminated and still taste goodwas totally unfamiliar. Mothers didassociate rising temperature with de-cay, the onset of which they said theytried to allay, by storing food awayfrom the hearth. Observation showed,however, that in practice, they regu-larly failed to do this.

Tortillas are often kept in a basket,away from direct heat. Mothers under-stand that the moisture content oftortillas encourages their decay, sothey air-dry them before storage. Chil-dren are not always given reheatedtortillas: a third of the mothers inter-viewed considered them to be too hardfor children to digest.

Different storage methods apply tothe refried beans and to the boiled,whole version. The whole beans areconsidered to be more vulnerable tospoilage, requiring thorough reheatingbefore being consumed. Traditionally,they were refried within 24 hours, butthis was often neglected because of thehigh price of cooking oil. Because oftheir fat content, the women thoughtthat refried beans were less likely tospoil, and would last up to three days.

The decision about where to storefood seemed to be more closely relatedto how soon it was going to be eaten,than to any cognitive concept aboutfood spoilage: foods to be eaten at thenext meal tended to be stored near thehearth, whereas those to be stored fora longer period were kept in a cupboardor on a table, away from the heat.

ReheatingFood was reheated for two mainreasons: to make it more appetizing,and to counteract its decay. Thewomen used four methods: warmingthrough, reheating thoroughly, boiling,and frying. They said that they usuallyreheated beans and tortillas and, to alesser extent, broth and rice, but, theyadmitted, they were not doing it for

26 WATERLINES VOL.13 NO.3 JANUARY 1995

long enough, or at a high-enoughtemperature (64 per cent of the foodcould be said to have been warmed).

The families always reheated tor-tillas (cold tortillas were consideredinedible), either by placing them di-rectly on cinders or, more frequently,on a tin or earthen slab (coma/). Theywere heated briefly, and rarely reached70°C, the temperature at which mostpathogens are destroyed. Forty percent of the reheated tortillas showedfaecal contamination.

Beans were heated in several ways:they were boiled, heated until warm,or placed cold between hot tortillas.Insufficient reheating was common,despite the well-known and acceptedfact that beans spoil easily if they arenot heated thoroughly. Cold pureedbeans were often given as a snack toappease a hungry child, even by thosemothers who would normally boil theirbeans for other meals. Over 60 percent of all stored beans showed faecalcontamination and, even after reheat-ing, 40 per cent were found to becontaminated.

TargetingAs a result of observing and interview-ing over one hundred mothers, and afterconducting behavioural and microbio-logical analyses, the research teamdecided that handwashing, and the

reheating of staple foods, were thehygiene practices that needed changing!improving 11]0stin Ciudad Vieja. Hand-washing was emphasized because chil-dren eat with their hands; there can beno reduction in the levels of contamina-tion while people's hands continue tobe significant vehicles of contamination.

While the mothers knew that thecombination of poor hygiene and inges-tion of decayed food caused diarrhoea,they did not perceive a clear cause andeffect relationship. This is becauseinfection is complex and it is not alwaysmanifested by illness. It depends on boththe amount of pathogens ingested, andon the immunological status of the host.Mothers also knew about the benefitsof handwashing but, overall, there wasa great gap between what they knew tobe correct behaviour, and everydaypractices. This can be explained partlyby their lack of specific skills, and alsobecause they had only limited knowl-edge about the consequences. For exam-ple, the advantages of running waterwere unknown: mothers were surprisedto find out that washing hands in a bowlof water was ineffective, and that dryingthem in a dirty rag would contaminatethem again.

DemonstrationAfter the 'key behaviours' were se-lected (in this case, washing children's

hands before they ate, and thoroughreheating of staple foods), the next stepwas to 'translate' them to mothers asmessages that would be understoodeasily. By spending time with mothersin their kitchens while they cookednormal meals, the researchers wereable to demonstrate how the staplesshould be reheated.

Twenty-five mothers were involvedin testing these recommendations intheir homes over one week. Handwash-ing was adopted enthusiastically: espe-cially after the project introduced asmall hose which was connected to thecement basin (pictured on previouspage). The families perceived this as asimple, easy-to-use device; soap usealso increased. When it came tochanging food-reheating practices, itproved easier for beans than for tortil-las. In the two-week trial, 14 out of 15households adopted the reheating rec-ommendations for beans, and II outof 15 adopted the suggestions fortortillas .•

Dr Sandra Saen: de Tejada is lllllllltirmpologi.l'£working at tire Instill/te of Nutrition of CentralAmerica and Panama. Cal:ada Roo.l·el·l'lI. lOlIllII. Aparllldo Postal 1188. t)1011 GuatemalaCity. Guatemala. Floridalma Cano is amicrohiologist who has worked ill foodmicrobiology for the last four years. She alsoworks atINCAP.

raising water from a depth of 100 metres.

SUPPORT SYSTEML IF E

which can be modified to act as either a pressure or suction pump

with PVC screens and casings. Pump equipment includes the SWN 80

WHATEVER YOUR NEEDS CONTACT US TO MEET THEM

complete lightweight hand drilling equipment and handpumps. together

operating to a depth of 40 metres. And, the SWN 81 pump capable of

With over 40 years of well drilling know-how, van Reekum Materials has

wells to tapping a water source at a depth of 100metres. VRM can supply

developed a range of simple-to-use pumping equipment suitable for the

THE

complete spectrum of environments. From the dewatering of hand-dug

WATERLINES VOL.l3 NO.3 JANUARY 1995 27