Health and Livelihoods in Rural Angola: a Participatory ... and Livelihoods in Rural... · Health and Livelihoods in Rural Angola: a Participatory Research Project Laura Habgood An

Health and Livelihoods in Rural Angola:a Participatory Research Project

Laura Habgood

An Oxfam Working Paper

©OxfamGB 1998

ISBN 0 85598 391 4

A catalogue record for this publication is available from the British Library.

All rights reserved. Reproduction, copy, transmission, or translation of any part of thispublication may be made only under the following conditions:

• with the prior written permission of the publisher; or• with a licence from the Copyright Licensing Agency Ltd., 90 Tottenham Court Road,

London W1P 9HE, UK, or from another national licensing agency; or• for quotation in a review of the work; or• under the terms set out below.

This publication is copyright, but may be reproduced by any method without fee for teaching purposes,but not for resale. Formal permission is required for all such uses, but normally will be granted immediately.For copying in any other circumstances, or for re-use in other publications, or for translation or adaptation,prior written permission must be obtained from the publisher, and a fee may be payable.

Available from the following agents:for the USA: Stylus Publishing LLC, PO Box 605, Herndon, VA 20172-0605;tel 800 232 0223; fax 703 661 1501; email [email protected] Canada: Fernwood Books Ltd., PO Box 9409, Stn. A, Halifax, Nova Scotia B3K 5S3;tel 902 422 3302; fax 902 422 3179; email [email protected] Southern Africa: David Philip Publishers, PO Box 23408, Claremont, Cape Town 7735, South Africa;tel. +27 (0)21 64 4136; fax +27 (0)21 64 3358; email [email protected] Australia: Bushbooks, PO Box 1958, Gosford, NSW 2250, Australia;tel. 02 4323 3274; fax 02 9212 2468; email [email protected]

For the rest of the world, contact Oxfam Publishing, 274 Banbury Road, Oxford OX2 7DZ, UK.tel +44 (0)1865 311311; fax +44 (0)1865 313925; email [email protected]

Published by Oxfam GB, 274 Banbury Road, Oxford OX2 7DZ, UK

Printed by Oxfam Print Unit

Oxfam GB is registered as a charity, no. 202918, and is a member of Oxfam International.

This book converted to digital file in 2010

Contents

Abbreviations / Glossary 4

Preface 5

Introduction 6

Map of Ganda Municipality 8

Part One: Methodology1.1 Background 9

1.2 Study design and organisation 9

1.3 Study area, sample, and sources 13

1.4 Methods and tools 15

1.5 Analysis and interpretation 18

1.6 External and internal factorsinfluencing the methodology 19

1.7 Influences on findings 20

Part Two: Findings and discussion2.1 History of Ganda 22

2.2 Social structures 23

2.3 Health-service providers 24

2.4 Sociocultural factors 27

2.5 Health beliefs 30

2.6 Health-related behaviours 32

2.7 Why do people makethe choices they do? 37

2.8 Health needs 40

Part Three:Conclusions and recommendations3.1 Health-related behaviour and the

use of health services within thesociocultural context of Ganda 42

3.2 Preventive health prioritiesof the communities and theirmost vulnerable members 44

3.3 The development of appropriatemethodological tools and health-status indicators 44

3.4 Recommendations to Oxfamconcerning programme directionand initiatives in Benguela Province 45

General recommendations for3.5working in communities 47

Appendices1. Population survey 49

2. Participatory methods and tools used 52

3. Example of a drawing used as a discussionstarter by research assistants 60

4. Causes of malnutrition —interview guide and diagram 61

5. Map of Ganda district 63

6. Health-service providers in Ganda 64

7. Research-project schedule basedon intermediate objectives 69

Notes 70

Abbreviations / Glossary

ACF Accion Contra Fome(Action Against Hunger)

CVA Cruz Vermelha de Angola(Angolan Red Cross)

DMPMF Delegagao Municipal paraPromogao e Desenvolvimentoda Mulher e Familia(a merger of UNITA's and theMPLA's women's organisations)

1CRC International Committee of theRed Cross

IESA Igreja Evangelica de Angola(Evangelical church)

MINSA Ministerio da Saiide de Republicade Angola (Ministry of Health ofthe Republic of Angola)

MPLA Movimento Popular de Libertacaode Angola (Popular Movement forthe Liberation of Angola

NGO Non-Government organisation

OMA Organizagao das Mulheres de Angola(the MPLA's Angolan Women'sOrganisation)

UNICEF United Nations Children's Fund

UNITA Uniao Nacional para aIndependencia Total de Angola(National Union for the TotalIndependence of Angola)

adobe mud

agimdente locally brewed alcohol

amigos chegados good family friendsbairro a neighbourhood comprising

several residential zonesbatuque. drum

cabewgrande traditional illness(potentially fatal bleeding occurs fromthe mouth and nose)

comunidade familiar family compounddesenrascar to scrape around for foodkandonga parallel marketkanjango extended family grouping

kimbanda common term for any healthpractitioner; the Umbundu equivalent isotchimbanda

lavra land distant from a rivermakulu used to describe illness caused by wormsmuhongo traditional pregnancynaca land bordering a river

olondele ancestorsolusongo scarificationondjango village or neighbourhood meeting placeotchitenlid 'lack of rains, then hunger' (Umbundu)pdssaro Portuguese for 'bird'; used to describe

an illness of childhood with symptoms ofconvulsions

planalto in-land plateau/central highlandsquimbo traditional villagequintal compound

santa/o female/male traditional practitioner(spiritualist)

seculo bairro elder or vice-soba

soba traditional chief, highest authorityin a bairro; sometimes appointed bythe Administration

tensdo de gota. the adult version of pdssaro

vanumuso tiny devil people which appearin dreams to attack the dreamer

walunguka a person with a particular capacityto understand and share an experience of life

Preface

Angola appears to be emerging from years ofsocial disruption, during which the people leastresponsible for prolonging the conflict havebeen most affected by the poverty that resultsfrom war. Planners developing basic social ser-vices face a situation marked by high levels ofpoverty-related morbidity and mortality amongpopulations returning to their homelands;scarce resources; and a lack of information toguide decision-making.

The Oxfam country programme in Angola,with several years' experience in working tomeet people's emergency and long-term needsfor water and sanitation, has been leading arecent shift in approach to community develop-ment in the country. This approach aims to helppeople identify and manage their problems, aswell as to cultivate a culture of information-sharing among programme staff. The 1997research project in Ganda which is the subject ofthis paper grew out of the need to gain a betterinsight into the lives of rural people with whomOxfam worked, particularly into their healthbehaviour,1 and use of existing health services.

Oxfam wanted to focus on the beliefs andperspectives of Ganda's communities, ratherthan to review health-service provision from theprovider side. In doing so, Oxfam also sought togain experience in information-gathering atcommunity level.

This work is the result of the combined effortsof several people: those who had the oppor-tunity and privilege of getting to know some ofthe many Angolans who will continue to look forways of coping with an uncertain future, andfriends and colleagues for whom such a life is'normal'. Thank you, Virgilio Joya, FilomenaRosalina, Hilaria Katumbo, Marion O'Reilly,Vincent Koch, Maria Catarina, Ana-Maria,Manuel Joca, Gustavo Manuel, Paulo Job,Avelino Rufino, Padre Adriano, Maria-AugustaPeixote, Gabriela Da Silva, Dona Maria Luisa,Manuel Gonga, Isabel Nimba, Aidan McQuade,Kate Home, Maria Emilia Barradas, OdeteAntonio, Fernanda Antonieta S de Carvalho,and the people of Ganda.

Laura Habgood, 1997

Introduction

Angola is a country devastated by many years ofwar. Those health and development indicatorsthat exist reveal human suffering on a scalenearly unparalleled world-wide,2 which cynicallymocks Angola's potential to become one ofAfrica's richest nations by virtue of its vast naturalresources. As a stable future seems possible, andsociety moves slowly from a state of conflict toone of rehabilitation and recovery, the attentionof international and local development organi-sations has begun to focus on plans for inte-grated rural programmes. Because of insecureconditions and difficult access, previous develop-ment initiatives have been confined to coastaland urban areas, whose population increasedrapidly as a result of large-scale war-induceddisplacement from the interior.3 The currentpopulation distribution within the country isunrelated to available natural resources or thebasic social-service infrastructure. Moreover,further population movements are anticipated aspeople return to their places of origin. Plannerslack experience and knowledge of ruralpeople's lives, whether settled or displaced. Inaddition, they lack the information necessary toguide programme development— although theyknow that the war caused high levels of ruralpoverty. People's livelihoods were lost, and thebasic social-service infrastructure was destroyed.

While a population's good health is funda-mental to development, isolated interventionshave little impact on improving health, which ispart of a complex and ever-changing interactionof social, cultural, economic and environmentalfactors. Oxfam U K/I4 has been working in Angolasince 1989 to restore access to basic facilities suchas water and health-care for people affected by thewar and living in poverty. Developments withinOxfam's understanding and management ofprogrammes since then reflect and respect thechanging needs and priorities of their targetpopulations. Previously, emergencies meantthat projects focused on providing handpumpsor temporary latrines for displaced people.Now the emphasis is increasingly on buildingand supporting viable rural developmentschemes. This goes along with a community-

based approach, for example by using parti-cipatory problem-definition techniques. If adevelopment initiative is to grow within thecommunity which it aims to benefit, it must alsobe located within the community's reality.

Problems cannot be defined without informa-tion. The challenge is to develop tools (andexperience in their use) to gather and understandrural people's vast traditional and local know-ledge, and to consider such knowledge at alllevels and stages of programme development.Health-related information too often consistsonly of formal service-providers' reports orofficial sources. The limitations of relying onsuch sources in the municipalities of Angola'sinterior, where the health-care infrastructurebarely functions, became evident in discussionswith health-service providers5 and in a review ofwritten records. They revealed the lack of anycoherent or meaningful system of informationmanagement at either community or facilitylevel. A reliable health-information system (HIS)is a basic requirement for the planning and deve-lopment of health programmes and services.6 Itis based on appropriate data-collection and data-processing methods, followed by interpretationand analysis. Information generated must thenbe passed on to relevant actors.

Therefore this research project was under-taken to contribute to improving people'shealth through the following means.

• Achieving a better understanding of the ruralhousehold, the target of Oxfam projects in BenguelaProvince. Merely targeting households will notnecessarily ensure that benefits reach those theyare intended for. Micro-level research is neededto understand inter- and intra-householdrelations, to identify inequities in distribution ofpower and resources at household level, and tolook for opportunities to redress the balance infavour of the most disadvantaged.

• Bringing about an awareness of health-relatedbehaviour which considers social, cultural, eco-nomic, and environmental aspects of people'slives as well as biomedical concepts of healthand disease. Health-related behaviour must beseen in its context, taking account of both

6

Introduction

internal and external influences, and of theirconstant interaction.

• Developing appropriate methodological toolsfor, and gaining experience in information-gathering and analysis at all levels. Informa-tion can stimulate communication, partici-pation, and development through a sharedunderstanding of the needs of target popula-tions — when it is able to represent their entirereality, not simply one set of problems.

• Improving local health workers' skills andcapacities to analyse the health situation ofthe populations they serve through jointplanning, training, and information exchange.It was envisaged that, with minimal supportfrom Oxfam's research team, the local publichealth team could expand its capacity todevelop both facility- and community-basedhealth information systems.

Map of Ganda Muncipality

N BALOMBO

CATUMBELA

HUAMBO

CUBAL

Part One: Methodology

1.1 BackgroundGanda Municipality, where Oxfam has sup-ported an environmental health project since1994, is located in the planalto region ofBenguela Province in western Angola. Dividedby the River Catumbela, the area consists ofsemi-arid, forested and cultivated lands (seemap). Ganda itself is a town surrounded by twohigh mountains and two small rivers, theIndongo and the Mbongo. The predominantethno-linguistic grouping is Ovimbundu, whoselanguage is Umbundu, although tribal dialectsand traditions within the Ovimbundu varyconsiderably. The municipality consists of fivedistricts, three of which have been controlled byJonas Savimbi's rebel force, UNITA, sinceOctober 1992 until the Government regainedcontrol in August 1997. Oxfam's project workhas been confined to Government-controlledareas. Since the Lusaka Protocol was signed inlate 1994, local security has been disrupted onlyby sporadic incidents of banditry. However, inthe months prior to the Government regainingcontrol, tension increased in the area, withfrequent rumours of impending UNITAattacksfrom the south, and reports of incursions intoGanda district across the nearby southernborders. The local population was nervous,talked of recruitment lists, and suspectedmovements of troops and heavy weapons. Atone point during this build-up of tension,Oxfam staff were withdrawn from the area untilsecurity had been further evaluated and foundsufficiently good for project work to continue.The area immediately surrounding Ganda hasnever been heavily mined, probably because ofthe kind of fighting, mainly by guerrillas, thatoccurred locally. The few landmine accidentswhich have occurred in the past year have beenattributed to newly laid mines; local people areaware of existing unsafe sites.

A grid of tarmac roads intersects the town,but the dilapidated state of most buildings bearswitness to deliberate destruction and years ofneglect. Moving out from the centre of townalong rough tracks, into the bairros, one encoun-

ters houses made of mud bricks and thatchedwith grass. Several small buildings are encircledby a wall to form a compound which enclosesliving quarters, cooking areas, a vegetable patch,and a variety of small animals — chickens,guinea pigs, pigs, and goats. Cattle are generallycorralled in the centre of residential areas forsecurity against bandits; crops arc sun-dried onroofs. Small informal markets line the principalroutes between bairros, and there is the unceas-ing movement of women carrying several tiersof firewood, tools, and food items on theirheads, their babies sleeping on their backs. In anormal year, the rainy season begins inSeptember, slackens in December, and peaks inFebruary before ending in April. From thenuntil the next rains, the weather is cold and dry.In 1996, the rains fell irregularly and heavilythroughout October to December and finishedearly. Most of the maize harvest, normallybegun during the rains, was lost as the plantsdried up before maturing. The sorghum, thesecond staple crop to be harvested, was infestedwith a pest which had proliferated becausethere was no rain to wash the growing plants.This, and the loss of the maize harvest, prompt-ed some people to bring forward the time oftheir sorghum harvest, and it was widely expect-ed that the communities would suffer hungerand hardship in 1997. The fieldwork wascarried out between May and August, whenusually least agricultural work is done.

1.2 Study designand organisation

Research objectives and approachThe research project intended:

• to increase knowledge and awareness ofhealth-related behaviour at household leveland of sociocultural factors which influencethe use of health services in Ganda;

• to assess the preventive health-care priori-ties7 of the community and its mostvulnerable members;

Health :incl Livelihoods in Rural Angola

• to gather information to guide the develop-ment of appropriate methodological toolsand indicators with which to monitor andevaluate preventive health-care;

• to make recommendations to Oxfam con-cerning the direction of future programmesand initiatives in Benguela Province.

The research aimed to explore and contex-tualise meanings, beliefs, and behaviours sur-rounding good and ill health. It involved thestudy of potentially sensitive issues at householdlevel such as unequal access to householdresources, and of issues which might implicateparticipants in criticism of the Government. Wechose to use qualitative research methods,which would enable us to explain, compare, andinterpret findings, rather than rely on direct,potentially conflictive questioning. Qualitativeresearch methods aim to gather information ina flexible and open-ended way, allowing forunanticipated discoveries and a wide range ofsociocultural factors.

The project had begun from a general idea,without a clear definition of how the processwould develop, and to what extent it would beinfluenced by the research team's resources andthe community's priorities. Consequently, it wascarried out in four phases; practical (inter-mediate) objectives developed in accordancewith the pace and direction of the research. Alsounanticipated at the beginning of the projectwas the development of a sub-project withM1NSA Public Health staff. We were asked toplan and carry out a population survey as thebasis for the development of a system forcollecting, interpreting, and reporting localhealth information (see Appendix 1).

The research team consisted of myself (anexpatriate with a background in public health indeveloping countries, who had lived in thestudy area for a year before the project started)as principal researcher, and two locally recruitedwomen as research assistants (RAs). Additionalsupport was given to the team by the OxfamBenguela programme health adviser, whovisited Ganda on several occasions during theresearch period, and by members of the OxfamGanda project team who work on environ-mental health, and food security.

RecruitmentThe research project was planned to include thetraining of two RAs in communication skills andresearch methods. It was anticipated that their

developing skills and abilities would influencethe process of information-gathering. Initialrequirements for recruitment were:

• being part of the local culture;• good communication skills;• willingness to learn and adopt new skills, and

to be part of a team;• written and spoken language skills in

Umbundu and Portuguese;• academic qualifications at high-school level.

The job description expressed a preference forwomen over 25 years of age, because the workwould focus on issues concerning householdproduction and reproduction, which aretraditionally women's responsibility. Followinglocal colleagues' advice, job notices inPortuguese and Umbundu were distributed tothe principal churches in Ganda, and theHealth, Education and Municipal Delegations,and advertised on a public notice board.Candidates were asked to fill in an applicationform, which would assist in selecting some of themfor interview with the principal researcher.

However, I soon realised that a requirementfor a specific educational level was discouragingapplications from otherwise suitable candidates,and decided instead to specify an ability to readand write well, with some understanding ofmathematics. I had hoped that applicationforms would help me select candidates forinterview, but in the end I had to interview all70 applicants personally in order to clarifyaspects of the job description. Many applicantswere keen to join the Oxfam Health Educationteam, which they thought would be teachingpeople in the bairros how to keep their housesclean; some disguised their lack of writing skillsby having friends fill in the forms. Most weresimply desperate for a job. Most of them wereunemployed young men, several were primaryschool teachers, but there was no one with anyrelevant technical abilities.

Eventually I offered the jobs to two youngwomen from Ganda, on the strength of theirwritten and spoken Portuguese and Umbundu,their apparent common sense, and their app-roachable manners. Although they clearly hadthe required personal qualities, this choicemeant a significant change from the originalconcept of working with RAs who were skilled inand familiar with academic work. This changeoffered other possibilities to the project, interms of developing appropriate trainingtechniques, but by necessity defined a differentstarting point and pace for the research.

10

Part one: methodology

TrainingA mini-induction course for the RAs includedan introduction to Oxfam as an organisation, itsinternational, country-wide, and local projects,and to the subject of public health. Because theRAs had had contact with NGOs in Ganda onlyas beneficiaries of emergency relief, it wasimportant for them to understand that workingwith communities is not simply about giving andreceiving handouts. Early discussions about theresearch project enabled me to assess the level ofthe RAs' knowledge and understanding, and togive them a sense of belonging to the projectprocess. There were new words and concepts totackle in discussing the project cycle, its aim andobjectives, and in planning project activities. Itwas helpful to visualise the project process as ajourney (see Figure I).8

Training continued throughout the researchproject, with an emphasis more on practicalthan theoretical work in an attempt to breakaway from the model of didactic teaching withwhich the RAs were familiar. Angolan class-rooms echo to the sound of voices reciting inunison as children learn their lessons by rote,with little opportunity to develop critical think-ing or questioning skills. An atmosphere of trustwithin the research team was essential to allowthe RAs to feel confident in their introduction toa new culture of information-sharing. Theirrelative youth and lack of exposure to alternativelearning methods meant that a high degree ofcontact time was required to achieve positivechanges in their ability and willingness toexpress spontaneous and independent thought.

Both RAs were active members of their churchcongregations and had participated in initia-tives to visit and assist needy people in theircommunities. It was useful to reflect on this

experience to draw on some of the communi-cation techniques which they had already dev-eloped and with which they were familiar.These included choosing comfortable and suit-able locations when talking to people (shade,stools, privacy), using a language that peopleunderstand, and speaking slowly, allowing timefor discussion and clarification.

The importance of being aware of theresearchers' own reactions to uncomfortable ordifficult situations was highlighted in the earlydays of fieldwork. During an interview with twoyoung women, their mumbled and monosyl-labic answers clearly irritated the interviewer.Glances and aside remarks between theinterviewer and note-taker did little to ease therelationship between the RAs and the infor-mants — nor did their tendency to read out thequestions like a shopping list, without checkingwhether they had been understood. Whensilences occurred or responses were delayed,the interviewer drummed her fingers on hernotebook, or talked 'at' the women about thetopic of the question, at one point giving them'amini-lecture on how to prepare a herbal tea forchildren with diarrhoea. And all this after theyhad carefully explained to the informants howvaluable their opinions and ideas were...

On other occasions, we acknowledged theRAs' use of positive techniques, such as helpingout each other by explaining questions,rewording questions from closed to open types,or changing the mood of a difficult discussion byintroducing a neutral topic. Both RAs hadfriendly manners and warm senses of humour,and, once they appreciated their own naturalabilities, were able to work on their owninitiative, rather than waiting to be told whatto do next.

Figure 1: The never-ending journey

assess progress:information analysis and

review of findings

identify destination:define problem and set

the objectives

plan how to reach destination:information-gathering activities

travel:carry out fieldwork

11

Heallli and Livelihoods in Rural Angola

Although 1 describe it as steps, the trainingwas a continuous process throughout theresearch project, involving constant review,reflection, and considerable overlap betweenareas of work. Much of it was practical training,so there was little emphasis on taking notes. Idecided to improve and build on existingcommunications skills as well as to develop newones as a foundation for the research methods.The topics covered and, where relevant, thetools used, were the following:

• role play which focused on the positive andnegative meanings of non-verbal and verballanguage;

• photographs and drawings of'good' and 'bad'situations to stimulate discussion about situa-tional influences on information-gathering;

• tape-recording each member of the researchteam while conducting a semi-structuredinterview, and critically reviewing the inter-view process and questioning techniquesafterwards (a very useful and levellingexperience for all team members).

Building on communication skills, the RAs wereintroduced to the main aspects of qualitativeresearch methods, including the following:

• interviewing techniques;• observation skills;• participatory methods and tools (see Appendix

2 for individual techniques and examples);• facilitation of group discussions.

Familiarity with these skills, and an under-standing of their purpose, were developed byvarious means:

• using different question types, listeningcritically to responses, probing, and followingup unexpected leads (in role play andfrequent reviews of fieldwork);

• practising observation-making, increasingsensitivity to surroundings using all senses;

• sharing information between all those teammembers who had been trained in Partici-patory Learning and Action (PLA) tech-niques9 at local and provincial level;

• exchanging experiences with colleagues whohad used focus-group methodology whengathering information during preliminarystages of another part of the Ganda project(on environmental health).

By working through their own experiences of,say, making treatment choices, to illustrate andpractise the use of participatory methods and

tools, the RAs became confident in the discus-sions that evolved and were able comfortablyto assume different roles as participant orfacilitator.

Fortunately, the office in Ganda has a largeporch with a concrete floor, which is ideal forchalk diagrams and conveniently close to thegarden for collection of materials. The frequentpassage of staff and visitors past their workprovided many opportunities for the RAs topractise new techniques and skills. As theirconfidence grew, they became more creative intheir designs and imaginative in their use ofsymbols.

One of the skills the RAs acquired early onwas the ability to sift out unnecessary words andphrases from their note-taking, and to recordonly key points and useful quotes from thediscussions. Initially, they wrote in longsentences and included the complete questionsin their scripts. Their slow and deliberate way ofwriting, combined with the simultaneous effortof translating from Umbundu into Portuguese,meant that they could not record all the pointsbeing made. We tried to review the discussionsas soon as possible after each session, to accountfor 'missing' information before it wasforgotten. The following topics were covered intraining about how to record information:

• use of bullet points, short phrases, and keywords;

• use of tape-recorder simultaneously withnote-taking, and line-by-line translation toevaluate accuracy of written notes;

• organisation of field-notes;• design and presentation skills to make posters

and pictures representing common criticalsituations, which would act as discussionstarters during community feedback sessions(see Appendix 3 for an example: a man withthree wives, one of whom has a sick baby);

• basic introduction to computer use for wordprocessing and designing a simple instruc-tion leaflet in Portuguese. (The RAs areprobably the only computer-literate womenin Ganda.)

Analysis and interpretation of informationhappened in several ways. We reviewed thefindings by coding information and cross-matching it by source and method. Usingappropriate analytical models, we identified keythemes and a range of variables. The researchprocess was under constant review, and weevaluated our performance by asking regularly:What were the good things that happened? Why were

12


they good? What were the difficult things? Why were,they difficult? How might these difficulties beovercome next time? We also asked participants toevaluate our research by presenting ourfindings and interpretations to them.

A range of methods to help organise and inte-grate research work were used:

• planning and writing weekly activity plansfor wall display;

• presenting progress reports to weekly OxfamGanda project team meetings;

• planning, implementing, and evaluating com-munity activities such as community map-ping, in conjunction with the Oxfam Gandateam and MINSA colleagues;

• holding joint training sessions in facilitationtechniques and participatory information-gathering methods with the Oxfam Gandateam and MI NSA colleagues;

• sharing findings and emerging themes withthe Oxfam Ganda Environmental Healthand Food Security project.

The small size of the research team and theamount of time spent together working andduring breaks meant sharing gossip andpersonal concerns. Support given to one of theRAs during a worrying time of pregnancy-related complications, and familiarity with theirhome lives contributed to a high level of trustin the team. As the relationship strengthened,the RAs opened up to reveal their owntraditional beliefs, behaviours, and prejudices,which they had mostly denied or ignored earlieron. This new willingness to share deeply heldideas with an 'outsider' not only stimulatedrecognition of the importance of interviewersdeveloping rapport with informants, but madethem appreciate that the sense of rapport mustbe mutual.

1.3 Study area, sample,and sourcesThe study area comprised a geographicallydefined area including all 19 bairros of thedistrict of Ganda (see Appendix 6 for map), forwhich latest population figures give a total ofalmost 37,000 people.10 Of these, 19.3 per centare children under the age of five, and 12.2 percent are people aged 45 or more years. Thefigures are broadly consistent with the demo-graphics of developing countries." However,proportions of men to women (aged 15 years

and over) ol 42.8 per cent to 57.2 per cent mayreflect the excess of women in the area followingyears of war. Bairro populations range fromaround 600 to 3,000 people, with varyingproportions ofresidentand internally displacedpeople. A bairro community refers to all peopleliving in the bairro, although they may expressdifferences in their views and traditions. Most ofthe displaced people in Ganda had moved therebecause of threats to their own communitiesduring the war. Some have lived in Ganda formore than five years, but many express theirintention to return to their places of origin andsee themselves as temporary residents. Themost recent significant movement of peopleoccurred at the beginning of 1997, whenreturnees from Cubal halted in Ganda to await asecure opportunity to continue their journeyhome to the Chicuma area.

Sample and selection methodsThe study area was stratified to represent rural,semi-rural, and urban communities, reflectingdifferences in the origins and subsequent dev-elopment of the communities to facilitateselection of samples. Rural communities aregenerally small, located at the periphery ofGanda, retain traditional customs such as dancegroups, and have few displaced inhabitants.Semi-rural communities grew between the 1950sand 1970s, when the workforce of the expand-ing local food industry required housing. Thesemi-rural bairros have remnants of some basicservices, and accommodate larger proportionsof displaced people than rural bairros. Theurban community comprises resident anddisplaced inhabitants of the colonial-style towncentre, which has some intermittently func-tioning services (water, energy), tarmac roads,discotheques, and a video club.

Initial field work was undertaken in a ruralbaino community because of its small size, therelative homogeneity of its members, and thesupport of the soba (local chief), who had beeninvolved in previous Oxfam project work inGanda. All zones of the baino were included,and, unless meetings had been arranged inadvance, participants were encountered duringwalks that started from different points andcontinued in directions chosen at random.Complementary fieldwork was subsequentlycarried out in bairros representing the other twostrata. Informants in the city were 'selected' asthe RAs followed a map with a pre-drawnrandom walk. The samples had originally been

13

Health and Livelihoods in Rural Angola

selected by degree of urbanisation, but whenimportant themes began to emerge from theinitial fieldwork, we could then identify popu-lation subgroups to be targeted for further in-depth investigation. The themes that emergedincluded the following:

• nutrition and health;• women's health throughout their life cycle;• household coping strategies and vulnerability;• management and prevention of childhood

illness;• messages and sources of health education.

These themes influenced the subsequent choiceof a stratified purposeful sampling strategy,which would help to compare different socialrealities and health experiences within thesubgroups, and to explore key issues withparticular relevance to the research objectives.These were the criteria for selection to eachsubgroup:

• mothers with at least one child under fiveyears of age;

• women or men over 60 years old;• girls or boys between 14 and 25 years of age

without children.

Although we recognised that men, too, havespecific health experiences, the need to priori-tise research resources and time precludedtheir inclusion at this stage.

In order to ensure that interviews in theresearch area were carried out in a randommanner, we drew a simple map of the city'sstreets. The researcher started the walk at theoffice gate. He or she decided on whichdirection to take by reaching inside a containerand randomly picking one of the objects placedin it: a button (turn right), a seed (turn left), or aleaf (straight on). At each subsequent junction,the procedure was repeated, and the pathmarked on the map.

Stratified purposeful sampling was achievedby sorting folded pieces of paper with the namesof bairros into piles, according to rural, semi-rural and urban strata. For each of the threesubgroups, a paper was randomly selected fromeach pile.

Identification of key informants relied on theresearch team's prior knowledge of the studyarea and on information received during thecourse of fieldwork. (A key informant wasdefined as someone who, as a result of holdingan official or informal position in Ganda, wouldhave detailed knowledge on a particular subject.)

For example, a particular private practitionerwas mentioned during several interviews withdifferent community members. The selection ofa small number of'typical' malnutrition cases atthe Nutritional Rehabilitation Centre (NRC) forin-depth case studies, and of health-service usersfor semi-structured interviews, depended to someextent on self-selection by participants. Toreduce bias as much as possible, we approachedevery 'nth' (depending how many clients werepresent) person seen or met as they left the con-sulting room and invited them to participate.

Sources of informationInformation was obtained from both primaryand secondary sources. Primary informationsources included the following:

Community members• women with children (six groups)12

• young women and young men(three groups each)

• elderly people (three groups)• adults (12 groups of men, women or both)• families (three)

Health-service users• hospital inpatients and general outpatients• women attending antenatal clinics• clients attending other health facilities• mothers at the NRC

Key informants (health-service providers)• representatives of the Municipal

health delegation and health personnel• private practitioners• traditional practitioner (herbalist)• traditional practitioner (spiritualist)• traditional midwives• representatives of CARITAS

and IESA mission health posts• medicine sellers

Other key informants• Soba• representative of the Municipal delegation

for women's development• representatives of the Catholic Church

in Ganda• local project staff• colleagues from other NGOs working

in Ganda Municipality

Secondary information included observationsmade at health facilities, and our impressions ofenvironmental conditions, but was generallytaken from the following written sources:

14

Fan one: methodology

Municipal Health and Public HealthDepartment reports from 1995 to 1997;Municipal Health strategic plan for 1997;1CRC Nutrition survey reports for Ganda;Municipal Administration populationcensus 1996;MINSA/Oxfam population survey 1997;other NGO documents;maps of the study area.

1.4 Methods and tools

Preparing fieldworkA number of points had to be taken into accountbefore the start of the research and as part ofevery fieldwork session.

Previously established contacts with the localadministration and main institutions in Gandahad afforded us several opportunities to intro-duce the general idea of the research before theproject started officially. Subsequently, wediscussed the project's objectives, methods, andprogress with members of the administration,and submitted monthly progress reports. Whenseeking permission to work in the bairros, weheld meetings with the soba and his committeeof elders, at which our plans were presentedand discussed. In order to prevent participantsfrom expecting that our work would be relatedto a later distribution, it was crucial to explainclearly what the research objectives were, whatmethods would be used, what this would meanfor the people in terms of time and involvement,and to explain the proposed feedback mecha-nisms. As there is a fashion for documentationin Ganda — everything from a lorry to a chickenrequires a licence to be on the street — the RAscarried 'official' cards explaining who they wereand what they were doing when they were in thefield. We obtained permission from theMunicipal Health delegate before we reviewedMINSA documents, and only the mainresearcher reviewed internal health depart-ment documents.

The team's local knowledge of people'sactivities and commitments meant that mostfield visits were timed to coincide with thosetimes of the day when participants were leastlikely to be preoccupied or tired. Wheneverappropriate, visits were arranged beforehandso that participants would have time to take careof their daily tasks. It was important to remainflexible and sensitive to their needs and wishes:

a miserable baby or hungry children waiting tobe fed might prompt the participants to requesta return visit to continue the discussion at amore convenient time.

People favoured weekend meetings, as mostof their work duties had finished by Saturdayafternoons, but this had to be balanced againstthe RAs' own family commitments. At times, keycommunity members assisted with setting upinterviews and focus-group sessions, but thiswell-intentioned assistance from others couldhave its drawbacks.

One enthusiastic soba insisted on setting upan interview with a traditional midwife for us.When he was eventually persuaded to leave themeeting, she admitted to us how frightened shehad been to be summoned by the soba. Weapologised to her. Later on, when we triedarrange an interview with a traditional practi-tioner through another .vote, he told us that wewould have to pay both an entrance and an exitfee, so that neither our health, or that of theinformant, would be put at risk. We foundanother informant.

Facilitating the fieldworkWhen people meet in an everyday setting, theyexchange a formal greeting, and a generalenquiry about each other's well-being and thepurpose of the visit. Accordingly, at the start of ameeting, we would introduce each member ofthe research team, give a brief description ofOxfam Ganda, and explain the purpose of thework. We also took care to choose a suitable,comfortable site for the interview, causing aslittle disruption to people if they preferred tocontinue their work (such as sorting grains)while talking. The research team carried theirown small wooden stools to avoid either theneed for a 'chair-for-the-visitors search' or theupset caused to hosts by the visitors sitting onthe ground. In spite of wearing long skirts, theRAs preferred to cover their legs with lengths ofcloth in the traditional style during field visits.Refreshments were offered to participants afterfocus-group sessions and, when offered byparticipants, food was shared.

We obtained the verbal consent of theinformants and explained that although noteswere being taken, no names would be recor-ded, and that their anonymity was assured.Similarly, when we tape-recorded discussions,people spoke without revealing their identity.Although pictures were generally popular,some people refused to be photographed,

15


because they 'did not wish others to see themliving in such poor conditions'.

Our methods of investigation includedinformal and formal approaches (see Table 1).The approach chosen depended on the pur-pose of the encounter and the RAs' abilities.

Investigation toolsTo guide the early stages of information gather-ing, we devised a framework to help us to defineareas of interest according to the researchquestions. The framework was based on theconcept of a pyramid of information (see Figure2), which is a useful tool for building up a profileof a group of people who live in a geographicallydefined area.13 It describes aspects of the com-

munity's background at a specific point in time.The pyramid is a triangle divided into layers,each of which is subdivided into differentcomponents, defining the context in which thestudy sample live.14 In addition to usefullyorganising the topics to be covered, it facilitatesthe choice of sources and methods to beemployed for information-gathering.

Field guides, in the form of question or topicchecklists, were used for individual and groupinterviews. The first step in preparing a fieldguide with the RAs was a 'brainstorm' of ideasabout the topics to be explored and a discussion,based on their local knowledge, of the likelycharacteristics of the target sample. Practicalexercises were then carried out to better define

Table 1: Research methods used

Method Procedure Purpose

Interviews with individualcommunity members, groups,key informants, and colleagues

Focus groups

Observations

Matched case studies

Participatory methods(see Appendix 2)

Review of existing information

Formal survey (see Appendix 1)

A range of procedures, from semi-structured interview to casualconversation:• based on open-ended questioning

techniques• allows the informants to give

unstructured answers

Group discussion guided byfacilitator:• makes use of participants'

interaction

Unstructured recording duringbairro visits, interviews, and visits tohealth facilities

Interviews with mothers ofmalnourished children, matchedwith families living in similarconditions

Exercises:• introduced as planned or used

spontaneously in discussion

Archive reports and documents

MINSA/Oxfam collaboration

• to reveal key words which describe therange of cultural, environmental, andsocial factors that shape health-relatedbehaviour

• to explore sensitive topics such as howhouseholds are organised

• to provide detailed information on thebasis of particular knowledge

• to explore people's knowledgeand experiences of health, illness,and health services

• to gain insight into key problemsand issues put forward by participants,and to understand the meanings ofwords used

• to describe physical conditions• to verify information• to identify further subjects for enquiry• to record information on roles and

dynamics within groups

• to look for differences between familyresponses to hardship

• to explore the relationship betweenchild-feeding practices and poornutrition

• to describe patterns in time (e.g.chronologies) or space (e.g. maps),

• to establish perceptions andcomparisons

• to provide background informationon population, health status, healthservice provision, policies, and plans

• to collect demographic datafor population survey

16

Fan one: methodology

the key issues and to devise a logical sequence inwhich to approach them. For example, drawinga diagram which illustrated their ideas ofpossible causes of childhood malnutrition helpedthe RAs to appreciate the complex relationshipbetween nutrition and health, and provided themwith an understandable, real framework fordeveloping subsequent interviews (see Appendix4 for examples of checklist and diagram).

Field guides also facilitated comparison of thesame topics between different sources. Theywere initially written in Portuguese, and thentranslated into Umbundu and back intoPortuguese by the RAs to increase theirfamiliarity with the lines of questioning, and topractise interview techniques such as usingappropriate and understandable words. In theearly days, field guides consisted of an A4 sheetfolded into a notebook which contained a seriesof questions under topic headings. As fieldworkprogressed, the guides were reduced to a fewkey words or phrases to act as reminders to theinterviewer or facilitator. They were modifiedwhen a review of fieldwork indicated that par-ticular words or phrases caused difficulties forinterviewers or informants. Similarly, questionsor words that seemed to facilitate discussionswere noted for future use.

Aside from having available large pieces ofpaper, chalk, and coloured marker pens, theteam made up tools for participatory exercisesfrom whatever material was available at thetime. Circles of different sizes were cut out ofscrap paper for Venn diagrams, and materialswere collected at fieldwork sites for other dia-grams. A point to note about picking upmaterials is that many abandoned items, inparticular maize cobs, were once used as anal

cleaning materials — one soon learns to watchout for which items local people avoid!

Recording informationInformation generated in the research process andduring fieldwork was recorded in several differ-ent ways. The main researcher kept a daily diarynoting the aims for each day, whether they hadbeen achieved, what the team had learnt from theexperience, and how plans for the research wereaffected as a result. The discipline of setting asidea few moments to note down many thoughts,comments or actions which would otherwisehave been forgotten proved to be a valuableinvestment for later project review and analysis.

Rough field notes were written into the smallexercise books that every school pupil in Angolacarries; large papers and clipboards were avoid-ed except as tools in participatory exercises. Thenote-taker translated from Umbundu to Portugueseand simultaneously wrote down the main pointsof the discussion. Relevant quotes or commentswere recorded verbatim. We kept all rough notesand copies of participatory exercises, and in somecases displayed them on the office wall wherethey stimulated comment and discussion withvisitors. Some materials began to look a little wornafter several visits to the bairros for feedbacksessions and had to be handled carefully.

Formal field notes were written up inPortuguese and English after we had reviewedand discussed the points recorded in the roughnotes; they included any additional relevantinformation. They followed a standard format:

• a statement of the session's specific objectives;• a descriptive introduction (site of interview,

weather, number of participants, their sexand estimated age, and so on);

Figure 2: Pyramid of information

NEEDS

health-related problems,needs and hopes for the future

SOCIOCULTURAL FACTORS

health beliefs, behaviours and traditions,household and social support systems, education and learning*

RESOURCES

community composition, organisation and capacity,health services — availability, type, acceptability, health-service providers

17


• an account of activities carried out;• observations made, results of participatory

exercises and discussions (including quotes);• general remarks on the dynamics of session,

the performance of the research team, andcomments for improvement or changes.

The records were filed in chronological orderand also stored on computer disks.

Tape-recordings of several focus-group dis-cussions and semi-structured interviews weremade using a mini-cassette recorder and anunobtrusive 'tie-clip' microphone, which couldbe placed in the centre of a group while therecording was controlled from the periphery.We reviewed most tape-recordings in conjunc-tion with the written notes. Only once was anentire recording transcribed, which took almostfour hours for a 45-minute interview.

We took photographs of participants andtheir activities in the bairros and, wheneverpossible, gave copies to the people involved,who greatly appreciated them.

1.5 Analysis and interpretationGathering qualitative information can producerich results on which to base interpretations andhypotheses related to the research questions. Itcan also result in confusion as a mass of notesawaiting review accumulates. Therefore, it isimportant to begin analysing information as itis collected, in order to avoid losing sight ofthe direction of the research, and also in orderto identify emerging themes for further in-depth investigation.

During the Oxfam project, we reviewed field-work notes regularly to pick up on key words orphrases relating to the topics in the pyramid ofinformation. These were written on small piecesof paper which were stuck to a large triangle drawnon a poster. The papers were grouped into sub-categories within each layer, and new sub-cate-gories added. This method provided an accessi-ble visualisation of common themes in thefindings, of possible links between-them, and ofareas where further investigation was needed.For instance, one important theme emergedfrom preliminary fieldwork because we noticeda great variety of traditions surrounding weaningbabies. We later explored this topic in moredepth. Using the pyramid of information alsoallowed us to compare information obtained fromdifferent sources, or with different methods.

Information gathered through focus-groupdiscussions was reviewed and coded. Commonthemes were marked and later grouped. Wecompared discussions of similar topics bydifferent sample groups. Points on which therehad been consensus or disagreement were noted,as were findings which illustrated unusualopinions. In analysing the dynamics and inter-actions of group discussions, we considered theeffects of peer pressure, of who said what, and why.

Classifications and patternsConceptual approaches to understanding health-related behaviour must be broad, and commonlyused models do not assume a simple relation-ship, where health behaviours follow a patternof rational choices. In Angola's rural areas,many people may not identify health problemsas such according to the 'Western concept', normay they respond with health-related behaviourwhich would be considered appropriate in a'Western' context. In addition, there are problemsinherent in the researcher's inability to detachhimself or herself from a personal, outsider's per-spective when trying to understand and interpretmeanings of social phenomena. On one occasion, Iobserved that a child's red string waistbandsignified traditional beliefs about protectionagainst illness. This was met with much laughterfrom the participants. They were amused thatthe otchindele (white person) did not know thatthe cord held the child's nappy in place and that, ofcourse, vaccines protected against illness.

In order to draw from the findings the fullrange of factors that influence people's choiceregarding the use or non-use of health servicesin Ganda, I used an integrated framework,15

which seeks to capture the factors in thefollowing classification:

• characteristics of the person (for example,sex, social status and networks, assets);

• characteristics of the illness (for example,acute or chronic, beliefs about causation,expected outcome from intervention);

• characteristics of the health service (e.g.quality, costs, distance, staff attitudes).

The links between diese factors and the perceivedmorbidity are not static, nor does a single choiceresult from their interaction. To further examinethe process of decision-making, the findingswere explored in such a way as to analyse path-ways,16 illustrating the integration of central socio-cultural factors into the sequence of steps involved.

18


Checking validity of informationWe were excited by one traditional practi-tioner's description of what seemed to be atraditional vaccination, with the mother of ahealthy child 'buying' the measles illness of arelative's child. She would prick one of themeasles spots with a needle and rub the liquidobtained on to the skin of her own child in orderto prevent or to reduce the severity of a futuremeasles illness. However, try as we might, wefound no one else who was familiar with thismethod of illness prevention, other than onewoman who suggested that it had been used aswitchcraft in her family — a child with measleshad died shortly after its illness had beenbought by an uncle.

The strength of our research approach lay inthe opportunities it afforded for information tobe tested between a number of different sources,and for using a number of different methods.For example, we found out in interviews thatfood stocks were very low and that alternativefoodstuffs were being used. We later observedcooking pots containing unripe sorghum;women returning from the fields with smallbundles of wild plants for cooking; boys whohad caught field rats; and empty drying rackson the house roofs — all confirming thesereports. By using multiple sources andmethods, we were able to develop new lines ofquestioning, refine our questions, and identifyadditional sources or methods to test ourworking hypotheses.

Working as information-gatherers as well asanalysers made it easy to cross-check infor-mation and detect inconsistencies. In oneinstance, the men of one village had told us thatthere was no point visiting in the early morning,as nobody got up before 8 a.m. However, whenwe did come early, we found the village busywith people getting ready to go to their fields,women washing clothes, sweeping their yards,and milking cattle. In another instance, mothersmight during interviews deny any knowledge ofmethods used to prevent ill health ormisfortune, while their small children would bewearing the traditional red string braceletsbelieved to have many beneficial properties.

Statements appearing within sub-categorieswere tested against each other to validate orreject them. Feedback sessions with communityparticipants and other interested groups werearranged as the analysis identified clear com-mon themes and patterns. Where possible,visual aids focusing on a critical incident related

to the themes were developed by the researchteam to spark off discussion.

Analysing the validity of the research alsodepended on consideration of the potentialinfluences on both the process and the findings.

1.6 External and internal factorsinfluencing the methodology

Outside factors which had an impact on theresearch included people's concern for theirsafety and for their livelihoods. The researchitself was shaped by the level of the RAs' skillsand by the limited time and resources availablefor training.

Fears about the deteriorating security situ-ation in some areas inhibited people to par-ticipate in group work, so that we had to makegreater use of individual interviews. Access tothe bairros was limited in an atmosphere ofsuspicion. Especially in the city, it was noticeablethat people were reluctant to talk in groups,because they feared strangers and retributionattacks at night — as no one could be sure whohad relatives 'on the other side'. In the ruralbairros, people were more comfortable to talk ingroups, as they were familiar with the back-grounds and motives of other participants. Afew unpleasant incidents with aggressive anddrunk men suggested an antipathy towards'information-gatherers' in an atmosphere ofheightened security awareness; at times it wasnecessary to carry out fieldwork within sight ofour male driver. If possible, we made contactwith the bairro authorities at the start of eachfield visit in order to minimise potentialmisunderstanding. However, we then often hadto resist pressure to hold meetings in theircommittee building, which has powerful party-political significance.

Our awareness of people's preoccupationwith the failing maize harvest and the dailysearch for food influenced our expectation ofhow much time they (particularly the women)could give to participate in information-gath-ering. The word desenrascar (to scrape around tofind food) appeared frequently in discussions aspeople worried about how they might be able toprovide even a cup of maize meal for theirchildren. Living under such precarious condi-tions, and with recent memories of a devastatingfamine in Ganda in 1992-94, the priorities andconcerns foremost in people's minds inevitablyreflected their immediate needs.

19


It took much longer than planned for the RAsto achieve a level of skills which enabled them tomake effective use of the research methodsavailable to them, or to attain the confidenceand capacity to critically review the process andfindings. This meant that the research devel-oped at a pace largely reflecting the RAs'abilities, and limited the choice of methods tothose which they understood and could usemeaningfully. While they might have been ableto perform, for example, a circle diagramexercise, the value of doing so in terms of thetime available had to be weighed against thequality of the discussion generated. At times, itseemed that progress was very slow and that theresearch team could not respond to oppor-tunities generously given by participants.Although participatory methods, in their truestsense, may be seen as seeds sown for action andchange, in the context of the research resourcesand time available, they were used principally asa means of information-gathering. The rhetoricof action for change is uplifting, the sentimentsare true, but the reality has to be considered.When skills or resources do not match what isrequired, the honest approach is to recogniseand remain within one's capabilities, ratherthan to risk damaging the trust, confidence, andhopes of both participants and team members.

1.7 Influences on findings

Surroundings and audiencesThe arrival of the research team in a bairro couldcause a stir. There are few 'outsiders' in Ganda,and most people's contacts with NGOs havebeen related to distribution of food and goods.It was crucial to explain the team's presencefrequently, and to demonstrate that we were notdrawing up distribution lists. Honest explana-tion was also important in order to create anatmosphere in which people felt they' hadpermission to speak. The voice of the people ismost often heard through that of the appointedbairro authorities. Consequently, an approachthat deliberately sought the opinions and viewsof others had to be presented as above suspicionand respectful of social norms. The RAs felt thatin the early days of fieldwork, discussion wasinfluenced by the expectations of some parti-cipants, but that, with time, these diminished, astheir role and confidence became clearer tothem and to the communities with which they

worked. Because they came from Ganda andknew many people personally, interviewsoccasionally proved difficult for them. After aninterview with an acquaintance, they wouldadmit to feeling uncomfortable about probingdeeply, particularly into issues such asexpenditures or income sources.

It was difficult to avoid a crowd of curiousonlookers gathering around when carrying outa participatory exercise, as the activity arousedgreat interest and amusement. But in somecases, observing the reactions and interactionsof participants and audience compensated forwhat was lost in the quality of the discussion. Asoba, demonstrating the relative importance ofavailable health services with the help of a circlediagram, appeared to lose his nerve when agroup of elders arrived to watch. Rather thanrisk appearing uncertain, he ordered them tofinish the exercise. Similarly, during a mappingexercise, members of the bairro authorityroughly pushed children to the periphery of thegroup and told them to be quiet, while womengrouped in the background commented onwhat the men were doing.

Both women and men were plainly irritatedwhen heckling from drunks interrupted ses-sions. Young women would appear uncom-fortable and went silent during interviews orgroup discussions when men (of any age)stopped to listen or to contribute. We wouldmove to a 'neutral' topic until the men left, askthem directly to leave, or thank the participantsand end the session. When we anticipated that adiscussion might deal with sensitive topics, wemade particular efforts to find quiet and privateplaces to talk, or to insist that non-participantsstayed out of earshot. Experience showed thatwomen were uncomfortable to talk about sexualmatters in front of children and wouldthemselves ask the children to go away. At times,when large numbers of small children distract-ed the participants, I, as 'observer', would takethem away from the meeting site to makedrawings or paper shapes, leaving thediscussion to continue in relative peace. Thiswas preferable to the offer of young boys withlong sticks who wished to control the excitedchildren in a rather more aggressive way.

At times, group involvement could be anasset as comments and prompts from the wideraudience stimulated discussion among theoriginal participants. However, the negativeeffects included a few loud voices dominatingthe exercise, participants or facilitator being

20

Pan one: methodology

distracted, or being made to feel shy, whichprevented meaningful discussion. After themain discussion we made time for anyone elsewho wanted to speak in order to prevent badfeelings, and to avoid leaving the participants inan awkward position with other communitymembers. We learned to be aware of thepotential effect of who was listening on whatpeople said and on who participated.

Visiting and talking to people at healthfacilities required diplomatic skills, as we did notwant the staff to feel that we were a threat tothem, nor did we want the informants to feeluncomfortable in agreeing to participate. Onthese visits, we would go through formalities,introductions, and explanations; one of the staffwould implore the facility users to co-operatewith the research team - and then, inevitably,would choose the people to be interviewed. Atthis point, we would politely intervene andencourage the stall' member to continue withhis or her own work.

On some occasions, informants said that theywished to hold a party to thank the health stafffor all their good work - inevitably when severalnurses were within earshot. Even when we hadsecured a private setting for a discussionbetween health-service users, groups of peoplewho were strangers to each other were reluctantto make comments on the quality of the health

services. They feared that any negativecomments might be used against them, because'no one knows who has relatives working in thehealth service'.

LanguageRAs conducted most interviews in Umbundu,although all members of the research team wereable to conduct interviews in Portuguese if thiswas the informant's preferred language. Themeanings of key words and concepts werediscussed to ensure that the most appropriateand easily understood words were used duringfieldwork. Occasionally young people in thebairros would insist on speaking in Portugueserather than Umbundu, but little participationwould be achieved until they reverted tospeaking the language they were most familiarwith and in which they could communicate theirideas and experiences. A desire to use the officiallanguage could not overcome their nervousnessof making mistakes. Being familiar with commonusage of language was important when intro-ducing sensitive topics into discussions. It wasunusual for people to talk directly and freely oftheir own intimate experiences, preferring todescribe situations that related to 'a friend'. BothRAs use a vernacular form of Umbundu ratherthan a more formal kind spoken at church.

21

Part Two: Findings and Discussions

2.1 History of Ganda

The original tribal group in the area was theMunganda, whose name derives from the termused to describe cattle-raisers, Okukanda. Theywere reputedly a peace-loving people, butearlier this century, another Ovimbundu tribefrom the eastern region of Huambo passedthrough Ganda on their way to trade rubber onthe coast. They were attacked and robbed by theMunganda, who later also attacked the sur-vivors on their return journey with the productsof their trade. The area afterwards becameknown as 'die area of robbers'. The site of Gandatown was established with the construction ofthe Benguela railway line in 1912, and the townfounded three years later by the first soba of thearea, Tchilandala Kambia. His statue, bearingresemblance to that of a Greek god, remains onthe roof of the now dilapidated colonial-stylecultural centre.

With the development of the railway line andcommercial agriculture in the 1940s and 1950s,bairros for contract labourers were constructedaround the colonial town centre. By the early1970s, Ganda was a significant contributor toAngola's food industry, and sisal, eucalyptus,and sugar cane plantations covered huge expan-ses of the surrounding land. The Munganda,living in scattered mountain communities(quimbos), sold some of their land to thenewcomers who were contracted from outsidethe Ganda district and came from severaldifferent Ovimbundu tribes. People in thequimbos lived by traditional subsistence-farming,with some surplus production for trade. Theirproducts generally reached the town's popu-lation via the Portuguese 'bush traders', who setup village stores in remote areas. While theseprovided a convenient credit or exchangeservice for the rural population to obtainconsumer items and agricultural inputs such asseeds and tools, the terms of trade heavilyfavoured the bush traders.

The elderly of Ganda remember the pre-independence years as times of plenty when,with good rains, grain stores would be filled

with maize, ensuring that there was enoughfood until the start of the next harvest. Healthservices for the unemployed and the peasantswere provided by the state hospital, whileworkers and their families were entitled to thecurative services of privately run clinics. Inaddition, most commercial farms ran their ownhealth posts, at which local people's minorailments and injuries could be treated. Thoseliving in the quimbos relied mostly on their ownknowledge of herbal remedies and on theservices of traditional practitioners. A period offamine resulted from a drought in 1915; 20years later a devastating plague of locustsdestroyed crops and, according to one elderlyman, 'ate people'. But real hardship began withthe escalating conflict between the MPLA andUNITA in 1974-75. There was a large move-ment of people from the mountains into therelative safety of the bairros closer to town. At thetime of Angolan independence in 1975, Gandawas under the control of UNITA, and thePortuguese had fled, abandoning their factoriesand commercial farms. Following a year offighting, the MPLA gained control of Ganda,and there was a temporary improvement inliving conditions, although the Government'sYear of Agriculture in 1978 was also known asOtchitenhd — 'lack of rains, then hunger' — inUmbundu.

By the early 1980s, many displaced peoplefrom the southern area of Chicuma werearriving in Ganda. There were food shortages,and as conditions deteriorated, people used tojoke with black humour that even the ratshunted for food were pleading with theirhunters: 'Leave us, we are displaced as well.'People were forced to sell or exchange theirland for food, and there was an outbreak ofcholera in which some families lost four or fivemembers. The end of the decade saw Gandareturn to normality for a few years, with 'normalillnesses, no famine, and functioning institu-tions'. In 1992, UNITA sporadically attackedand then captured the town, and severalmonths of extreme isolation, starvation, andhigh mortality ensued. When Ganda returned

22

Part two: findings and discussions

to Government control, international relieforganisations arrived to provide food and basicservices for the thousands of displaced peoplewho lived in appalling conditions in the townand bairros. People talk of these years with tears.

2.2 Social structuresIt is important to understand how relations inAngolan society have changed over the past 100years or so. The top-down structures inheritedas a legacy of colonialism and perpetuatedduring years of Marxist rule continue todominate social structures. The Government'sauthority reaches into the heart of the bairrosthrough all the social actors. While in partreflecting tradition, the municipal soba systemimposes an artificial structure that can conflictwith tribal loyalties and customs. For the past 20years, the Government has claimed to act andthink on behalf of its people, and socialstructures such as the soba system haveperpetuated this attitude. As a result, peoplehave accepted a passive, subordinate role. Fewcommunity-based initiatives have developedthrough which the people might gain a say inthe decision-making processes that influencetheir lives.

The different experiences of people living inGanda's bairros illustrate that the loss oftraditional structures has had a profound effecton the cohesion of communities today. Those inwhich extended family groupings (kanjangos)remain largely intact retain a sense of trust andco-operation, whereas others experience suspi-cion and distrust between members. In anatmosphere where protective family bonds nolonger exist, people believe that malignantspirits have greater potential to affect their lives,and shared interests are neglected in favour ofindividual agendas.

Traditional institutionsA soba was traditionally the highest authority ina region consisting of several quimbos. He hadthe services of ministers, counsellors, and aprivate guard. Each quimbo had its own group ofelders, who had the position of vice-sobas. Thesoba was a man who the people considered to bewise and clever, whose judgements theyrespected, and whose decisions they supported.The elders chose his successor from among hissons; if they deemed none of them suitable, theywould chose one of the soba's sister's sons. It was

not customary for a woman to be considered forthe position. Quimbos were organised in familygroupings (das or kanjangos), or according togroupings of friends who 'had the sameunderstanding' (amigos chegados). All activities todo with managing the family grouping would befocused on a traditional building, the ondjavgo.This was a place of dialogue, of sharing know-ledge and goods, where traditions were record-ed, justice was sought, and conflict resolved.

Having been largely discredited during theearly years of Independence, the soba. systemhas been refashioned in both Government- andUNITA-controlled areas. Government author-ities appoint municipal sobas on the basis of theirability to undertake administrative duties in thebairro. Sometimes known as sobas de guerra('sobas of the war'), they are generally equatedby community members with the Government,and while the ideas of the people might beconsidered, the soba's authority in decision-making is publicly referred and deferred to.Where a bairro is composed of people fromdifferent tribal backgrounds, the appointed sobamay not enjoy the same esteem as his counter-part in a less mixed bairro. However, peoplegenerally feel that the soba, with the support ofelders, teachers, and church representativesalso 'put in charge by the authorities', is 'at thefront of the community'. From here he 'sees thesuffering of the majority and gives a solution toit.' A soba has the status of a rich man, but this,according to a group of rural men I inter-viewed, may be a disadvantage: 'A man's richesare only for his household and children, and therich never see the suffering of the community.They need sobas who do not have to work for theincrease of their wealth.'

People still see the ondjango as a bairro's centralmeeting point, although some now associate thebuilding with the Government party. Diffi-culties that arise within or between households,often concerning bewitching, or ownership offields and animals, are presented to the soba andhis circle of elders. Each elder will have his say,after which the soba gives the final judgement,which all agree with. The judges celebrateresolution of the problem with aguadenteprovided by the person who brought the case. Ifthe case is not resolved at this level, it may betaken to the Chief Municipal soba for judge-ment; failing that, it goes to the police, who work'according to the law of the land'. (The presentChief Municipal soba is a member of theMunganda tribe; this can at times cause tension

23


with those from other tribes, who suspect thattribal loyalties may interfere with justice.) A casethat cannot be resolved locally will eventually beheard before a tribunal in the provincial capital,at great cost to all involved.

Government institutionsMunicipal delegations of the ministries ofhealth, education, and agriculture are based inGanda town. The administrative structureextends to the bairros through the soba system;each authority has a secretary for each bairrozone and a representative of the AngolanWomen's Organisation (Organizagao das Mulheresde Angola/OMA). Since its inception in the mid-1980s, OMA has claimed to be the nationalpromoter of women's rights. It was originally amass organisation designed to transmit theMPLA party message to the people, but it hassince merged with its UNITA counterpart toform the Delegagao Municipal para Provwgao eDesen.volvime.nto da Mulher e Familia (DMPMF).The local DMPMF presents a lecture pro-gramme to communities in which it aims topromote equal rights for women and men, toinform women of their legal rights, to supportsex education in schools, and to highlight theconsequences of violence against women.

However, according to one person I spoke to,developing women's rights in Ganda is an uphillstruggle: women who live in the quimbos 'don'tknow anything, they are backward', and onlylisten, 'without speaking'. The DMPMFexpresses a strong intention to recruit ruralwomen into adult literacy programmes,although the women do not appear to be 'veryinterested in the classes, because they have toomany other worries'. Women are also repre-sented on the management committees of thefour agricultural associations in Ganda, whichlack equipment and resources, but which arerecognised by the Government and the peoplefor their existence if not for their functioning.The associations were developed in the mid-1980s in response to the failed nationalisation ofcommercial farms, in order to support farmersin increasing their maize production for supplyto urban areas. The farmers worked their ownland, but membership fees paid to theassociation enabled them to share resourcessuch as technical advisers, machinery, and mills.

Churches in GandaThe main Christian churches in Ganda areCatholic, Protestant, Seventh Day Adventist,

and Tochoista. The latter is a traditional Africanchurch founded early this century, whichcombines conventional methods of Christianworship with traditional song, dance, andmusic. The Catholic Church has the largestproportion of churchgoers; because of the localCaritas organisation, people view it as the bene-factor of the communities' poorest members.Operating at community level through anetwork of male and female catechists, Caritasdistributes food and material goods. Its work isclosely co-ordinated by the local leaders of theCatholic Church. The local branch of Promaicaorganises activities that aim to promotewomen's rights through the acquisition of skillssuch as sewing and knitting. A local Catholicwidows' support group offers Bible study andadvice sessions, and the chance to cultivatecommunal land, the produce of which is sharedwith the church. The pastor of the Pentecostalchurch holds regular meetings with groups ofwomen from his congregation to discuss'problems in the bedroom'. Young members ofall the churches help elderly, sick, and frailcongregation members with their housework.Youth groups are also enlisted by the town'sadministration to carry out environmentalclean-ups and have been involved inpreventive-health initiatives, such as anti-alcohol campaigns, through their churches.

Other organisationsInternational and local NGOs in Ganda supporta wide range of programmes: they distributematerial goods, agricultural inputs, and food;they work to rebuild the physical infrastructure;and they develop community-based initiativesto provide basic social services. Local com-munity groups include traditional dancegroups, which perform at the many officialcelebrations and sporting events in Ganda'ssocial calendar, and the municipal and bairrofootball teams.

2.3 Health-service providers

Health-service providers in Ganda can be dividedinto two levels — those which function withininstitutional structures and those which workoutside them, at community level. Community-level providers include kimbandas,11 home-birth attendants,18 sellers of both modern andtraditional medicines, and voluntary memberso(bairro health committees (BHC). Providers at

24


institutional level include state-run, church-related, independent, and other agencies (seeAppendix 6 for more details on all health-service providers in Ganda).

Some officials (for example, Governmentrepresentatives) consider the distinctionbetween health-care providers at communityand institutional level to reflect a difference inquality of service. Churches, NGOs, and otherinstitutions are regarded as legitimate actors inthe health field, whereas traditional providerssuch as medicines sellers and spiritualists arebranded as 'charlatans and liars'. The role ofchurches and NGOs as service providers duringthe emergency years means that officialsconsider them as igual do governo (equal to theGovernment). However, this official responsemay not necessarily be reflected in people'shealth-care behaviour and beliefs. For others,churches and NGOs form a relationshipbetween community and Government pro-viders, mainly by working with kimbandas whouse modern medicines, or home-birth atten-dants who have some links with the maternitydepartment of the state-run hospital.

Traditional health-care providersThe number of kimbandas currently practisingin Ganda is unknown, because the tools of theirtrade are usually kept secret — particularly inthe case of spiritualists, who are said to benumerous. However, people 'know who theyare'. In order to differentiate between them, it isuseful to consider their varying methods ofpractice (see Table 2).

The use of herbal remedies in health-care isnot confined to recognised kimbandas; manyelderly people and those who originate from themountain communities retain the requiredknowledge. Nor are methods of practicemutually exclusive; and kimbandas are not

averse to experimentation — they adapt treat-ment to each case's circumstances and theresources available. Several people claimed thata recent useful treatment for a stiff neck was amassage with the butt of a gun. (For some, whoconsider themselves more 'developed' thanpeople who believe in spiritualism, the power ofa sanla or a santo lies in their judicious use ofherbal remedies in conjunction with ritual,which they say only serves to distract thepatient.) In some cases a kimbanda will beknown for her or his success at managing aparticular health problem, rather than forusing a specific method of healing. A sanlaknown to treat women who have sufferedrepeated spontaneous miscarriages takes herpatient to a riverside, where she marks a sitewith ash and feathers. Traditional drums areplayed and there is singing. The patient lies inthe water and is washed with roots before acharmed cord is tied around her waist, which isworn continually to prevent miscarriage and toprotect future pregnancies. A local kimbandawho treats children thought to be late walkers isrenowned for his injections of antibiotics,dietary advice (he recommends beans, liver,and soup), and massage.

Institutions as health-service providersInstitutional health-service providers are linkedto some extent, because M1NSA regularlycollects data from health facilities in order torecord which activities were undertaken.Planning and managing vaccination campaignsalso involves collaboration at this level.However, while patients may occasionally bereferred from a state-run hospital to a privateclinic (for example, to obtain a prescription for amedicine the hospital lacks) or to a kimbanda. (ifthey find a condition unbeatable by Westernscience), there are few other links between

Table 2: Kimbandas and their methods of practice

Methods of PracticeHerbalism Spiritualism Prayer Modern medicines Dreams/visions

Private practitioner

Sanla/santo

Diviners

Home-birth attendants

Part-time practitioner

25


providers which would offer opportunities foran exchange of experience, joint planning, andthus for improvements in health services.

To everyone in Ganda, health servicesencompass those provided 'outside' the house-hold, and those within it, concerned withmatters such as cleaning and food preparation.Outside health services range from the purelycurative (for example, medicine sellers) to thepreventive (for example, hygiene education byBHCs or awareness campaigns promoted bychurch youth groups).

It is apparent from MINSA activity reports inGanda that state-run health-service providersare interested in the promotion of a primaryhealth-care programme. However, the fact thatmost elements of the programme are located atcentral facilities could reduce the impact andeffect of the promotion of health for all.19 Someinformants noted that health education, oncedisseminated by the churches, now reaches thepeople only through vaccination and cleaningcampaigns, or when they go to the hospital andare given medicines with explanations abouthow to use them. Table 3 identifies the specificelements of primary health-care and illustratesthe community's perception and experiencesconcerning its sources and provision.

Environmental health servicesMost of the population obtains its water fromtraditional wells. More than 600 of these wereprotected with cement headstands and woodentops in 1995. Although most of them provide ayear-round supply of water, some refill slowlyduring the dry seasons, yielding only smallvolumes of turbid water; some dry up com-pletely. Seasonal activities, such as the makingof mud bricks, put a strain on water sources.Alternative water sources are the two small

rivers which intersect rural bairros, and a smallnumber of working public tap-stands in thesemi-rural bairros. Rainwater is not collected bycustom, and large containers cannot easily beobtained. Water is supplied intermittently topublic taps which serve urban residents.

Several hundred temporary pit latrines wereconstructed during the emergency period of1994-95 to alleviate appalling sanitary condi-tions in the town, but there are many informal,open-air defecation areas currently in use. Despitethis, young urban residents state that people inthe bairros are 'underdeveloped because theydefecate in the open air'. The huge flat rocksthat scatter the landscape are popular sites fordefecation as well as for pounding grains; thereare separate (unmarked but well known) areasfor each activity. Alternative sanitation facilitiesfor bairro residents are household latrines (of amodified ventilated improved pit (VIP), soak-away, or traditional design). Sanitation in thetraditional villages used to be less of a concernfor people, who had 'no worries about latrines'because there was ample unoccupied land andanimals to eat the faeces (chickens in particularenjoy the worms contained in them).

Oxfam Ganda, as part of its environmentalhealth programme, has provided the means forpublic institutions to construct latrines. As yet,few of the schools and churches can claim tohave public sanitation facilities. But institutionalhealth facilities have responded by constructingpit latrines, some of which are maintained bydedicated employees. However, the care ofpublic facilities is generally problematic.

The Municipal Community Services Depart-ment is responsible for the maintenance of sixpit latrines located at the central market. As longas they had a 'guard', who collected a nominalusage fee and undertook daily cleaning of thefacilities, the latrines were popular with the

Table 3: Primary health-care services

Intervention

Education on common health problems

Promotion of food supply and nutrition

Adequate water supply and basic sanitation

Mother and child health

Vaccination programme

Provision of essential medicines andcurative services

Provider

Health-facility staff, bairro authorities, peers, family

Family, NGOs

NGOs, Government

Family, home-birth attendants, health-facility staff

MINSA

MINSA, church, kimbandas, family,medicine sellers, NGOs

26


market community: they were clean and hadwooden doors. After the guard abandoned hispost, the latrines rapidly deteriorated to a filthystate. Stallholders complain bitterly of the smelland (lies that now plague the small eating placesand meat stalls nearby, and of the health riskwhich the flies might pose. They suggest thatthe latrines should be torn down and new onesbuilt, for which people would be prepared toagain pay a reasonable usage fee. A protestaction will possibly develop to bring the prob-lem to the attention of the community services.

2.4 Sociocultural factors

How do people earn a living?Most people in Ganda live by subsistence farm-ing, although recently they have also earnedmoney as contract workers, and although theland is better suited to grazing livestock andcommercial farming. Many rural people aspireto having the ability to raise cattle. Building up astock of small animals is the first step in thisprocess, but recent conditions for this have notproved favourable, so that farming is now analternative source of income. Land for cultiva-tion can be obtained from bairro authorities,which allocate portions from the few remainingagricultural associations or abandoned commer-cial farms,20 by private sale or rent, or throughthe family network. Displaced people in Gandausually rent or borrow land, because their ownland is too far away or continues to beinaccessible. It is not unusual for agriculturalland to be located half a day's walk from Ganda.Lavra land (at a distance from the river) yieldsmore than naca land (close to the river), becauseits harvest extends throughout the year fromJanuary until November, when the 'first hunger'111

begins and is fed by naca produce. Thetraditional staple crop is maize, although manypeople plant sorghum, a more resistant crop, incase the maize harvest fails. Other crops grownfor sale, exchange, and consumption includebeans, sweet potatoes, and cassava; vegetablegardens produce tomatoes, cabbage, and onions.

The local handonga (exchange market) systemprovides a livelihood for some, particularlyurban-based, people. At the same time, it offersrural people access to consumer goods whenthey have surplus produce for exchange. Anurban woman uses part of her husband's wageto buy items such as soap, clothes, and dry fish in

Ganda market. She takes the goods to anexchange market on the border with UNITA-controlled territory where she trades them formaize, sorghum, honey, beans, or peanuts. Onher return to Ganda, she sells these products(reserving some for her own household'sconsumption) and uses the profits to buy moreitems for exchange. In June 1997, lkg of maizecost 40,000kzr in Ganda. A skirt bought inGanda for 300,000kzr could be exchanged for20kg (800,000kzr worth) of maize.

Thus, the kandonga enables people toincrease the value of their products above theirmonetary value and to avoid the need or risk ofstoring their savings in currency. There are fewopenings for salaried employment or wageearnings locally, and most offer meagreincomes when compared with the prices of basicfood items.

Therefore, many people make a living bycombining activities that respond to changingcircumstances, although opportunities to do soare more readily available to those living close tothe urban trade centres. Displaced men buildwooden handcarts to use as taxis between thecentral market and the bairros, and small can-teens have opened along the main road. Thereare increasing numbers of small bairro distil-leries producing sugar cane liqueur, andwomen sell maize meal. Vitamin-rich husks aresold as animal food. Although grass mats,baskets, and clay pots are produced on a smallscale, there is no local handicraft industry. Someactivities are seasonal: November brings themango trees into fruit, June is the season forcollecting honey. An increase in the movementof people and goods between theplanalto regionand the coast has brought other opportunitiesto acquire cash and goods. Possibly as a result ofincreased trade, prostitution occurs — largelyin the town, with women working from known'houses' or concentrated around the marketplace. Urban residents observe that there areincreasing numbers of young girls from therural bairros and quimbos in the town.Remittances from outside of Ganda seem toplay a small part in household income, thoughgoods in the form of clothes and other non-fooditems may be sent from relatives. There is nostate-run social security system.

How do people manage their means?After several consecutive years of drought andpoor maize harvests, households have few

27


reserves left, and grain stores which wouldnormally have lasted the year sit empty. Womenespecially talk of the need to desenrascar,~ inorder to feed their children. In times of•hardship, the number of daily meals is reduced— first affecting women, then children — andthe variety of food is limited. Sorghum isharvested and eaten early, or staple dishes aremade from alternatives such as ground hardbananas and boiled green papaw. Wild plants,and whenever possible animals, are consumedas food. The focus of the household, and thuswomen's daily preoccupation, is to acquire food.Protecting one's assets becomes increasinglyimportant as the number of thefts of smallanimals and crops rises. Where close familymembers are in a position to help, they may givefood directly, allow relatives to work their fieldsin exchange for food, or take children into theirown households. (Apparently, younger chil-dren are preferred to teenagers because theyhelp with chores, whereas the latter 'cause onlyproblems'.) Elderly people who have no familysupport may beg for food and clothes in theroads. However, some rural bairros have com-munal fields, which community members farmonce a week in order to provide support to theirelderly and most disadvantaged neighbours.

Whenever people have surplus produce, theytry to build up their assets by investing in smallanimals or items such as clothes, which act as aninsurance policy for unexpected needs. Whenexpenses related to ill health cannot be met byexisting resources or by selling non-essentialhousehold items, a family initially seeks helpfrom their immediate relatives, followed by moredistant relatives, and finally neighbours. Whilethere is a sense of obligation between familymembers, assistance from neighbours usuallytakes the form of a loan. In order to repay theloan in cash or kind, the borrower may under-take paid work, sell essential items, or make asmall profit by trading in the market place.

The mother of a young child who had been illand sustained a burn in a fire borrowed moneyfrom a neighbour to pay for treatment from akimbanda. She planned to repay the loan bygathering firewood for sale but because she hadno other means to buy food, half of the profitwas spent on maize. Another woman who hadnothing left to exchange for food for her youngchildren regularly walked 12km to an exchangemarket, from where she carried sacks of maizeback to Ganda, making a profit of lkg in 10kg.

Although such ways of coping with povertyare in general regarded as survival strategies,

they are often simply part of resource man-agement, or a response to changing circum-stances. Selling firewood, carbon, or mangoes— which might initially be thought to representresponses to hardship — may be undertaken bychoice in order to buy food to protect maize inthelavras, which, though edible, would have lessnutritional and market value for the people ifharvested while still ripening. In otherinstances, the prostitution of young girls toobtain the means to survive, which occurredduring the recent years of war, would now beconsidered less socially acceptable, and mightbe undertaken only as a last resort to acquiremeans for a household. The impact of hardshipon a household will largely depend on itscapacity to employ strategies which could beconsidered reversible, such as borrowing fromrelatives or selling labour, and which do notincrease the likelihood of being unable torecover its resources, such as selling tools,animals or land.

Who manages the means?When describing a household, most peoplerefer to a unit that includes all those who eattogether around the same fireplace (lareira).'23

Members of a household are also said to sharea plan for economising resources and thesame worries in difficult times. Several house-holds can exist within a compound, becauseby living in a comunidade familiar (family com-pound, see Figure 3) with family members orclose friends, people feel they will not be'exposed to the vices of strangers, nor will theybe responsible for breakages.'

It is widely accepted in rural areas that theman who owns the house (or is married to orpartner of) the woman living in it is responsiblefor making decisions about economisingresources, although most questions will first bediscussed with household members. Theprocess of discussing problems is important, asis an 'understanding' between a couple; but asone woman said, when times are hard 'no oneneeds to discuss, because everything goesstraight into the pan.'

Ultimately, however, the man will decidewhether an item of clothing may be exchanged,an animal sold or killed, or resources pooled tobuy food. He generally decides which seedsshould be planted in the fields, and whether andhow payment may be made for medicaltreatment. Although his wife may have physicalcontrol of the money to avoid it disappearing on

28


Figure 3: Example of a family compound made up of three households

= Household(casa familiar)

/Molher-in-law\ /^Bachelor soii\\ ^ (Mgra) J P I (soliteiro) J

Father, mother, children** hearth

Widowed cousin,with children

** hearthHousehold

drink, if she disagrees with her husband, shemust sacrifice her own belongings, for exampleher clothes, to pay for what is needed. In theopinion of a group of rural men, 'The womanonly has the care of the children, the house, theclothes, and the food for all the family in thehouse. The man is able, say, to have five women;the decisions of these households depend onhim — food, clothes, health.'

A man with more than one wife, although heis absent from the house, is consulted about alldecisions apart from those which concern thehousehold routine (such as when to fetch water,go to the fields or prepare meals). Among men,polygamy confers stature, as a man demon-strates that he has sufficient means to supportmore than one household. As in most ruralsubsistence societies, in Ganda men are seen asthe primary providers, while women are resp-onsible for carrying out the most labour-inten-sive activities which maintain their households.Elderly household members, women in partic-ular, are often expected to work in the fields, orto care for children and the sick during the day.Younger members are also expected to con-tribute to the income of rural households, eitherthrough working in the fields when not atschool, collecting water, or sharing profits madefrom trade, although they are not obliged to doso. They recognise that their input entitles themto their parents' good care and concern, who inturn expect to be cared for in their later years bytheir children in return for the investmentmade in their upbringing. A group of youngrural children planning their professionalfutures outside Ganda agreed that they wouldone day return to help their parents farm.

Obligations to one's immediate family arestrong and, in some rural households, stronger

than those defined by conjugal ties. When timesare hard, in-laws are sometimes the last onesprovided for in a household, and even if awoman's husband's family lives nearby, sheconsiders herself alone if her own family isdistant. In past times, a man might haveconsidered his sister's children closer to himthan his own children, because they were of hisfamily bloodline. Maternal uncles were oftenresponsible for the counselling and support oftheir nephews and nieces; it was also said thatthey had the right to sell them into slavery. Lawsof inheritance still ensure that a large portion ofa man's property is returned to his blood familyafter his death, although his children — femaleand male equally — occasionally benefit. Hiswife can be forced to return to her own familywith her children, having been stripped of herhome. If a woman dies, however, her husbandand children inherit her property. The followinganecdotes, told by young women, illustrate theapparent lack of confidence and commitmentsurrounding conjugal relationships:

At times a man arranges traditional medicinefor his wives to remain friends and not to maketrouble or fight between themselves, but there isalways disagreement between them.

Sometimes, a woman arranges traditionalmedicine to make her husband stay with herand forget other women. At the beginning, thereare no problems, but later the husband becomeslike a donkey in his infatuation for her, as themedicine kills his heart and he does everythingthe woman says, washing the children's clothesand dishes. The medicine fills up the abdomenand contaminates the body; the man's head isbroken, and he forgets and neglects his ownfamily, and dresses in dirty clothes. With thecorrect medicine [given by his family, who

29


realise what has happened to him] he is able tovomit everything out [and be saved].

A change in a woman's menstrual pattern canindicate that she has a muhongo (traditionalpregnancy), which lasts longer than a normalpregnancy. As monthly bleeding continues, the'baby' cannot develop properly, and the so-called pregnancy sometimes lasts for up to fiveyears, until the woman takes traditionalmedicine. Her abdomen swells and shrinks, butthe baby never actually appears.

Because the concept of muhongo allows for anunpredictably long pregnancy, and the actualdate of conception could therefore vary quiteconsiderably, it also provides a convenient wayof explaining a real pregnancy that results froma non-conjugal relationship.

2.5 Health beliefsand behaviours

What is health?Good health for people in Ganda almostuniversally means having food and being able toeat well. Good health is having water and soapto wash oneself and one's clothes at least everyday. Good health is found in the hospitals andhealth posts, which should have a plentifulsupply of medicines to treat all illnesses inpeople of all ages. Good health is not to have toworry about any of these things, to estar avontade, and to have everything 'in conditions.'These are sentiments expressed by men andwomen, both young and older, displaced andresident in Ganda, although women apply theconditions to their children rather than only tothemselves. For a group of rural women, to 'visitrelatives and have communication betweenthem all' was necessary for the health of a family.After having survived so many years of short-ages and social disruption, it is hardly surpris-ing that, for the moment, Gandans equatehealth with the security of having enough food,water, and health-care to meet .their needs.Urban-based young people, whose basic needsare more likely to be satisfied, might add thatgood health for them also means having thepossibility to play sport, to study, love, dance,and gossip.

What are the causes of ill health?When people talk of ill health, they differentiatebetween 'normal illness' and 'traditional illness',

although the categories are not apparentlydistinct and what begins as a normal illness canbecome a traditional illness, depending on thenature of its development or chronicity. Normalillness is described in terms which stress thephysical nature of the sensations experienced —headache, fevers, and chills — and often each ofthese is considered as an illness in itself. It isunusual for people to refer to a group ofsymptoms (such as that indicating malaria) or asyndrome as a single illness, or to attempt toexplain the cause of a normal illness. Table 4lists the causes commonly given for some'normal illnesses'.

Causes largely reflect most people's livingconditions; overcrowding has resulted in theappearance of'new' illnesses (particularly amongdisplaced people) such as yellow jaundice,measles, and cholera. The presence of faeces inthe road causes smells and attracts flies, which inturn provoke vomiting in children, as 'the smellaffects their hearts'.

Threadworm is the literal translation of theUmbundu term makulu, but the life-cycle andeffects of makulu are quite different from aWestern understanding of threadworm infec-tion, in which transmission is effected viaunwashed hands, and the infestation causesonly symptoms of peri-anal irritation. K makuluinfestation results from eating dry bread orleaves cooked without oil, of from sitting on thedamp ground, or from the union of a male andfemale roundworm present in a person'sabdomen, which breeds thousands of worms.Once makulu has eaten everything inside thesufferer's stomach, it moves through the body tothe spine, which it climbs. When it reaches theperson's neck, the bones of the neck becomeweak and break. Death follows.

The origins of some episodes of ill health are,at their outset, unknown, and only becomeapparent as the patient either recovers (with orwithout treatment) or gets worse. For example,a healthy child who fails to thrive after anepisode of diarrhoea may have been bewitchedby another mother, who was secretly envious ofthe child's previous good health and cast badthoughts. The misfortune of a household whosechildren suffer repeated episodes of illness suchas fevers and diarrhoea might have its roots inan unresolved interpersonal conflict, betweenpeople and their ancestors or between inhabi-tants of the living world. Ancestors used to forma constant and crucial part of people's lives.Olondele (the forebears) accompany their livingdescendants, advising, chiding, and consoling

30

Pan Uvo: findings and discussions

Table 4: Perceived causes of some 'normal illnesses'

Illness

Headache

Backache

Cough

Fever

Nasal congestion

Abdominal pain

Simple diarrhoeain children

Cause

Threadworms, nasal congestion, fever, carrying heavy weights on the head,many thoughts and worries

Sitting on the ground all day, working in the fields, fever, threadworms,carrying heavy weights on the head

Salty food, dry fish, carrying heavy weights, cold weather, hunger(causing tuberculosis)

Threadworms, nasal congestion, wounds, mosquitoes, climate change

Smoke, new grass, dust, climate change

Indigestion from sweel potatoes, beans, raw or badly cooked food, wild plants,maize from mill, threadworms

Weak breast milk, eating wild plants every day, sorghum, change in dietduring rainy season

them. If traditions were not properly respected,and the olondele felt offended or disregarded,the individuals or households in question wouldbe punished. The punishment was bad luck, illhealth or even death, unless reconciliation andappeasement was sought. Although people saythat the influence of the church encourages themto disregard such beliefs, the deep need remainsto understand and to explain why some peoplesuffer more than their fair share of misfortune.

Conflict between a husband and wife caninduce a 'nervous illness' in the woman, which ismanifest in apathy, withdrawal, a twisted face,and even, at times, a uterine prolapse. Similarly,lack of harmony between the parents of anunborn child, or the father's infidelity, canharm the pregnancy, the delivery, or the child'ssubsequent health and development. Envy ofsomeone else's good fortune can also lead tohealth problems, because the aggrieved personeither uses intermediary agents such as poison-ed flies sent to settle on the other person's foodor body, or administers traditional substances:

An urban man who was so fortunate as tohave a paid job began to experience a change inhis lifestyle. He took to drinking in local barsand to generously entertaining his friends. Hishousehold suffered as his earnings disap-peared, but he refused to listen to their warn-ings and pleas. At times, he beat his wife.Fearing that he had been bewitched, his familytricked him into taking a remedy against thespell by explaining that he needed treatment forjaundice, as his eyes were yellow. Only when the

medicine caused him to vomit out all the 'illness'did he realise what had happened to him. Heand his family blamed his behaviour change ona poison placed in his food or drink by someonewho wanted to benefit from his loss of controland decline into alcohol dependency.

Some illnesses are more likely than others tobe attributed to a traditional cause, and thecircumstances surrounding their onset will beconsidered as part of the explanation. Convul-sions in children are attributed to the harmfulinfluence of a bird that has passed overhead andcaused the illness pdssaro.-4 If the child survives,any subsequent convulsions will be attributed tothe continued effect of the illness.

Harmful influence may also be exertedthrough dreams, in which a vision appears ofthe person casting a spell, or in which vanumuso(tiny devil people) appear to fight with thesleeper, using a sharp stick or a knife. Onwaking, the dreamer experiences a sharp,localised pain in his body or coughs up blood —indications that an injury was sustained. Noteveryone, however, is susceptible to the effectsof traditional illness, and those who escape it aresaid to have a body and blood with certaincharacteristics. Entire families appear to beimmune to traditional illness, because inheritedcharacteristics offer them protection.

People use complex explanations of illness,which are sufficiently flexible to take changingcircumstances into account, to make sense oftheir health and that of those around them (seeFigure 4).

31


Figure 4: Perceived causes of ill health

Evil eyeIndirect: birds, poison, wind,insects, other agents, dreamsDirect — a malevolent gaze

Behaviour of othersFather of unborn child

Family conflict

Spiritual inheritance

Natural causesFood/breast milkWork burdenClimate change

Agent-induced causesDirty flies, makulu

Constitutional susceptibility

The disruption and conflict which resultedfrom the war have led to an increase in theincidence of illnesses related to evil eye in somecommunities. Family groupings have beenseparated; there is hate and suspicion betweenpeople; and a few thrive while others continueto suffer. The familiarity with which people talkof well-publicised illnesses such as cholera,measles, and TB suggests that even if people donot know the microbial cause of these illnesses,they are aware of the environmental conditionsin which they spread. While some causes of illhealth are perceived to be avoidable, forexample, by having sufficient and appropriatefood, others are probably not avoidable, such asillness induced by evil eye. Explanations of someforms of ill health, such as that caused by alcoholdependency, are, perhaps conveniently, used toabdicate responsibility.

2.6 Health-related behavioursWhen describing health-related behaviour,people differentiate between three categories ofbehaviour: behaviour which helps maintain goodhealth; actions taken specifically to preventillness; and actions taken to restore health.

Actions taken to maintain healthThese broadly cover what people describe as'health services within the household', andinclude care of living space, care of food andwater, and personal hygiene. People from allbackgrounds and age groups recognise that forcertain individuals and at certain times of life,special care must be taken to ensure that goodhealth is maintained. Special care is considered

important for young children, the chronicallyill, and pregnant women. How well thesegroups are cared for depends on the availabilityof health-care, which varies according to thedegree of urbanisation.

Care of living spaceThe first task of the morning for the women andgirls of a household is to clean the compoundand living quarters. Girls are introduced earlyto the traditional female role as carers withintheir households, undertaking tasks appropri-ate to their ages. Blankets (if people have any)are hung out in the sun and dishes washed 'sothat the flies are not able to provoke illnesses.'Sweeping out rubbish is said to reduce thenuisance of flies and dust; cutting back grass andvegetation removes hiding places for snakesand scorpions; and clearing away children'sfaeces reduces bad smells. When the ground isnot too hard, faeces is buried. Householdrubbish is buried in pits, although most plasticbags, bottles, and tins are reused.Cleaning is generally seen as each household'sresponsibility, and some people claim not toknow what others do, but a woman who toler-ates a dirty or untidy environment will be talkedabout by her neighbours. Displaced men feelparticularly aggrieved that the disruption tonormal routine and the overcrowding of theirliving conditions has resulted in a neglect ofcleanliness and in negative effects on people'shealth. An accumulation of rubbish outside thecompounds, in communal spaces and roads,suggests that shared spaces are no-one'sresponsibility, unless the work is undertaken ina clean-up campaign organised by the authori-ties. Notably absent from discussions with rural

32

Pail two: findings and discussions

people was mention of latrines, perhapsbecause they are still familiar with the custom intraditional villages of allowing natural degrada-tion or removal of excreta by animals.

Care, of food and waterThe desirability of providing a diet that includesa staple, plus vegetables or meat cooked in oil, iswell accepted by women, men, and children,regardless of background. When families liveon stomach-filling rather than nutritionallyvaluable food, it is because they lack the moneyor opportunity to provide good food, notbecause they are ignorant about nutrition. Attimes, 'good food is anything that appears'.Some mothers believe that a child's hungerpains indicate an organism living inside itsstomach which is beginning to eat into thechild's body.

Maize porridge, which provides vitamins andenergy, is served to all age groups, as a thingruel for infants and as porridge for the sick.Depending on the stomach and age of theperson, it is eaten with or without husks; forexample, maize milled with its husks is thoughtunsuitable for children. Women are responsiblefor deciding who eats what according to customand for ensuring that leftovers, if there are any,are boiled before eating. The ability to feedothers at social occasions such as weddings,funerals, and during times of mourning isperceived as an important indicator of aperson's (particularly a man's) position insociety and of his or her commitment to familyobligations. In anticipation of such occasions,and because chickens are valuable as offerings,exchange, and food, eggs are kept for breedingrather than eating.

Water is carried from source to home invarious containers, usually plastic or metal basinsand pans. Boys and girls help their mothersand, in rural areas, spread large leaves on thewater's surface to minimise spilling it. Once insidethe house, water is occasionally transferred tosmall clay pots which are kept raised off theground for cool storage, although it is rarelyallowed to settle before drinking. Despitewidespread knowledge that drinking watershould be boiled, particularly for children andthe sick, this is not a common practice because ofthe amount of pans, fuel, and time involved.

Personal careAccording to young urban boys, 'girls havemore hygiene' as they bathe three or four timesa day, compared with their two. Young urban

women consider the urban areas to be moredeveloped than others because they have awater system and washrooms. Women in ruralbairros prefer to collect water once, late in theafternoon, to wash their children 'who spendtheir whole life playing with the soil', becausewithout bathing, 'they will not sleep'. Soap is ahighly valued commodity for everybody, withoutwhich washing is considered ineffective. Wash-ing in public is disliked by men and women; ifthe compound has no washroom structure, theyprefer to wash at a secluded riverside, whereclothes can be washed and laid out to dry, orafter dark in some privacy. Displaced menworry that women who are not able to wash posea health hazard to their children.

Personal care also refers to how one controlsone's body. Young people, especially those wholive in towns, say that they ought to resist thetemptation to take drugs, to smoke, 'whichburns the lungs', and to drink strong alcohol,'which leaves the head mad and the bodywithout strength'. Young people also say thatthey ought to control their sexual behaviour bynot having many partners, which can lead toserious problems such as gonorrhoea and AIDS— which come from men and sex, or womenand sex, depending on who is talking.According to a group of young women living ina semi-rural bairro, 'a boy's first girlfriend (theone he is likely to many) may drop him if he seesother girls, because she doesn't want to getillnesses... she may speak with him first to advisehim not to get an illness. She says that other girlshave sexual illnesses and she is going to leavehim'. They mention HIV/AIDS informationcampaigns on the radio and at hospitals beforethe war in 1992-94 and are open about theirconcerns that they might be at risk of infection.However, they say about condoms that only 'themore developed, open, and unmarried youngmen would be able to use them'.

While young urban women consider itundesirable to become pregnant 'before theright time' or to have many pregnancies'rapidly', how many children a woman has, andwhen they are born, is in 'God's hands'... anddepends on the orders of the man. There is abelief that the preordained number of childrenin a woman's abdomen must be finished, the'true number sometimes being 14'. But ruralwomen feel that in reality their circumstancescannot support having many children. If thereis 'an understanding' in the household, awoman may be able to tell her husband that sheis 'near her time' and thus postpone sexual

33

l-lenltli and Livelihoods in Rural Angola

intercourse. In the absence of calendars, womencalculate their cycles by the phases of the moon,which may or may not coincide with their timesofovulation. Some young rural women believedthat being 'near their time' meant that they wereabout to menstruate and sex was therefore unsafe.They were eager to gain a better understandingof their bodies and to learn how to work outtheir own calendars. They also anticipated thattheir knowledge would enable them to helptheir women friends, and to speak with moreauthority to their boyfriends. Contraceptivepills, which are not available in Ganda, arebelieved to ensure that, if there has been'contact', the fertilised egg leaves a woman'sbody with menstruation. Young urban womenknow of locally available 'barrier' contraceptives(for example, pastes made from a plant mixedwith soap and lemon, or tablets of aspirin,chloroquine, or penicillin which are usedintravaginally before having sex).

Early years of childhoodBreast-feeding usually begins immediately afterbirth. The act of sucking ensures that the yellowmilk is removed quickly to allow the 'good' whitemilk to flow. Some older rural women prefer thatthe yellow milk be expressed, although youngerwomen try to speed up the change in the milk bywashing their breasts with warm water. Thebaby's first green faeces are a sign that theyellow milk has passed through the child's bodytogether with the remains of the food which thewoman ate during her pregnancy. Baby-feed-ing bottles are uncommon in non-urban areas,although from the age of one or two monthsalmost all babies are fed a thin maize gruel inaddition to breast-feeding, which women fromall backgrounds say should continue for twoyears. If a mother believes that her breast milkhas become weak through lack of food and weakblood, she will immediately eat raw cassava androasted dry fish, if she can afford it. Feeding ababy with weak breast milk is believed to cause itdiarrhoea and ill health, and this combinationof factors often precipitates early weaning. Aswomen who describe their milk as weak usuallyappear to be reasonably healthy and with a goodsupply of milk, they might adopt such a strategy'unconsciously to balance the many demands ontheir lives; it gives them a legitimate excuse tostop breast-feeding.

It is also widely believed that breast-feedingshould stop when the mother becomes preg-nant again, because each pregnancy has its ownsupport system within the mother's body —placenta, blood, and breast milk. An elderly

rural woman explains that 'some children diebecause they have fed on the milk of anotherchild. Having babies rapidly means that theytake the milk of the child before, and if thesituation continues, all may die because theylack their own milk. There used to be atraditional medicine for the woman to take tocontinue breast-feeding — it was normal for awoman to have children close together, becausethe elders knew which traditional medicinecould help. But nowadays, without traditionalmedicine a pregnant woman is not allowed tocontinue breast-feeding when she is pregnantagain, but the baby depends on the mother'smilk. Some women are clever and wean theirchild as soon as they see their menstruation hasstopped; but there are some who are ashamedto say that they are pregnant again and do notwean; others know, but they don't accept it andcontinue to breast feed until the new baby dies.'

There are no traditional weaning foods; butcessation of breast-feeding is said to be accomp-anied by the passage of all the milk from thechild's body, who then needs four or five mealsa day to recover his or her strength. A child whohas been displaced by another pregnancy mustalso be protected from the heat of the mother'sbody and is placed to sit apart from her.25 Insome rural families, the child is rubbed with acloth that has been used by its mother to cleanherself after having sex with the child's father.The massage is believed to strengthen the child'sbones and ensure good health.26 Likewise,bathing the new-born while it is strapped on itssibling's back minimises jealousy and strengthenthe relationship between them. Heat, semen,and blood all have a powerful influence onhealth. Young urban woman talk of a commonpractice to rub a child's joints with its mother'sfirst menstrual blood, while thin new-bornbabies are given sips of their own bath water todrink, as it contains the heat from their body.

Chronic illnessSufferers of common and chronic illnesses suchas pdssaro and its equivalent in adults, tensao degota, must observe certain rules in order tomaintain their health. They are not supposed tostay near open water, in crowded spaces, ornear the fire. Neither are they allowed to eatmeat from male animals, fresh fish, mealscooked in blood, and certain red or sweet foods,such as tomatoes and honey. They and theircarers should not visit a house where there is anunburied corpse, as the heat from the body iscapable of exacerbating the illness, nor should

34


they request embers from the fire in the houseof a deceased person. This demonstrates thatpeople feel their actions can in some wayinfluence the course of an illness, even if theybelieve it is almost incurable.

PregnancyWomen are very sensitive to their pregnanciesand to influences on the developing baby.Interviewees from all subgroups talked of theimportance of care during pregnancy. For agroup of semi-rural women, antenatal careshould be undertaken because 'every baby hasits own bed in the mother's abdomen, and thequality of the beds changes as the first ones, thebest, leave with each delivery; after the third orfourth the mother's body changes, she suffersmany illnesses and births are difficult'.Younger women tend to link their care withservices provided by the midwives at the hospi-tal, where they are given vitamins and medi-cines, while older women are more concernedwith the kinds of food they eat and the heavywork they do. It is considered normal to carry40-50kg, but they believe this should bereduced to 20kg in pregnancy in order not toprovoke an abortion, a premature or a difficultdelivery, or an unhealthy baby. Some womenbelieve that hospital midwives are able to seefrom the shape of their abdomens that theyhave been doing heavy work, and will scoldthem, saying they will not have the strength todeliver. During pregnancy, the baby is said tohave its own tastes, and although greenvegetables and fruits are widely recognised assuitable foods for pregnant women, food thatdisagrees with the baby will make the motherfeel nauseous and vomit. Eating rabbit is pro-hibited, as the baby will be born with a face 'tornlike a rabbit'; eating tortoise meat will prolongthe length of time the child crawls on its belly.

Some rural traditions, especially in relation toa traditional pregnancy {muhongo) involve thepregnant woman placing a stone or stick, orbanging a rattle, at any crossing in the road thatshe passes, to ensure that the baby does not 'staythere'. The parents may also talk directly to theunborn child, or the father carries out a ritualbefore leaving the house to protect the unbornbaby in his absence and not to 'take it with him',i.e. cause an abortion. A commonly availableplant may be taken by the mother as a medicinethroughout a normal pregnancy, to protect theunborn child from the harmful effects of itsfather's relations with other women. As with theissues surrounding weak breast milk, while

women feel responsible for the well-being oftheir child, they are also liable to be blamed ifsomething goes wrong with its health anddevelopment. But rural people believe that ulti-mately the outcome of a pregnancy is outsidehuman control, because 'there is an organismwithin the woman that bites the baby belore itdelivers and leaves its water in the body of thebaby without a mark; the baby dies aged a fewmonths or years'.

Actions undertaken to prevent illnessHealth-care providers and community mem-bers explain that in order to prevent an illness,one must know the nature of the threat in orderto institute appropriate measures against it.There seems to be a trend for people in Gandato associate preventive actions with health-careproviders, so that the actions are the result ofusers' and providers' combined efforts.

VaccinationWhile people do not generally understand themicrobiological working of modern preventiveinterventions, they have credibility because theyare linked with diseases well-known to be majorcauses of child illness and death. A vaccination issaid to protect the child from a certain illness, orto ensure that the child suffers only a mild formof the illness. The local vaccination service andperiodic campaigns have raised levels of aware-ness among both the displaced and the residentpopulation. However, rural women say of thosestill living in remote traditional villages: 'Thewomen are not used to vaccines as they havenever had them and now don't give any value tothem'. All subgroups of the population talk ofthe need to vaccinate children several timesfrom the first day of life until the fifth year, untilthey have had all the necessary jabs. A newlydelivered baby will often be taken by relativesand friends from the bainos to hospital for a firstvaccine, but some confusion was reported amongmothers, because 'the nurses now say vaccinesshould begin at three months'. Intervieweesknow that each vaccine treats a specific illnessand has its own name, and name vaccines formeasles, tetanus, fevers, diarrhoea, and whoopingcough. After receiving a vaccination, a child isfeverish, cries at night, and its arm is swollen —signs that the medicine is working in the body.

Young urban women relate their attendanceat the hospital antenatal clinic to the availabilityof vaccines 'to prevent problems' and 'to avoidcertain [unspecified] illnesses', but others who

35


have had uncomplicated deliveries in the pastfeel there is no need to consider vaccines. Thebelief in the power of the needle is greatlyexploited in the practice carried out by someyoung urban men, who privately receive coursesof antibiotic injections to protect against variousillnesses, including fevers and sexual illnesses.

Traditional, methodsModern and traditional methods of illnessprevention are not mutually exclusive, andmany children who receive vaccines also weartraditional protective tokens. Traditional methodsof protection are also called upon whenchildren suffer from chronic ill health, and thiscustom is not necessarily confined to rural areas.A red cord bracelet is said to protect a baby fromrashes, and often carries a tiny piece of wildanimal hide to guard against illness caused byenvy, evil eye, and the wicked intentions ofothers. It can also help to thwart the effect of aharmful wind that makes a healthy baby loseweight. A small bag attached inside the baby'sclothes, containing the baby's dried umbilicusmixed with herbs and a piece of a feather, issupposed to guard against the illness pdssaro,which is caused by a bird's harmful influence.Usually, the tokens and medicines are providedthrough the services of the elders and spiritu-alists. As part of a consultation, asanta prescribestraditional protective medicines, which thepatient and his or her family place at strategicplaces in the home — for example, around doorframes, close to sleeping mats, or near the fire.

Recovering healthWhen people talk about treatment, their firstresponse is to refer to the medicine that isrequired, whether it is modern (tablets, oint-ments, and injections) or traditional (herbal).However, treating and managing illness meansmore than a simply a cure with medicines:patients and their households take actions whichaim to prevent the condition from worseningand to promote rapid recovery.

A child who is suspected to have measles isplaced in the sun, so that the heat will quicklydraw the illness in the spots out of the body.During the cold season, the child should bedressed in red clothes for the same effect. Usingspecial instruments 'to look inside the child'sbody' (these appear to be auroscopes), someprivate practitioners can diagnose measlesbefore the spots appear, giving an opportunityfor appropriate measures to be taken. Theinside of the body — including the bones and

intestines — is affected to such an extent thatwounds form in the throat, which closes, leavingthe child unable to speak or eat. Once thespots have appeared, the child is not allowed toleave the house, nor should its mother attendthe funeral of a child who has died of measles,for fear that the heat from the death willexacerbate her child's illness. When the childbegins to recover, a paste of maize husks isrubbed on the skin to encourage exfoliation. Toopen the wounds in the throat and recover loststrength, the child is given a thin maizeporridge to eat, as well as young chicken meatif the family can afford it.

Early symptoms of illness are interpreted inthe context in which they occur. According tothe perception of the sufferer and those aroundher/him, a decision is reached on whether thereis a health problem, and second, whether itneeds treatment. Diarrhoea in adults duringthe rainy season — when diets change and foodis scarce — is so common that it barely meritsattention. So, too, is eating sand, which ruralwomen and men consider a normal reaction, byadults and children alike, to the smell of theearth after the first rains. The desire to drinkwater early in the morning may be a sign thatthe afternoon will bring a fever, and lying inthe sun may help to draw the fever out of thebody. If someone wakes to a dream in whichvanumuso have appeared, the sleeper must notwash his or her face, for fear of forgetting whosent the dream. Sometimes the onset of anillness is so dramatic — as with cabeia. grande, atraditional illness in which potentially fatalbleeding occurs from the mouth and nose —that there is no doubt about the need for rapidtraditional treatment.

There appear to be preferred treatment-seeking patterns for the most commonly experi-enced 'normal' illnesses across all categories ofthe population. Children with simple feversshould be treated at the hospital with theappropriate medicine. Likewise, adults withfever and headache will in the first instance seektreatment from a hospital, health post or clinic.But if a fever is accompanied by a convulsion,the initial treatment sought in most cases —particularly in rural areas — is traditional. Inthese cases, modern medicines (includinginjections) are believed to prejudice recovery, ifnot actually cause harm to the patient. Simplediarrhoea can often be treated at home, withpotions made from the leaves of common trees,roots, special stones, or mango bark, mixed withwarm water and given in sips several times a

36


day. Few people mention giving oral rehydra-tion solution specifically to increase fluid intakeduring an episode of diarrhoea, although someyoung urban mothers cited water and salt as apossibility. Sugar is difficult to obtain locally,and available alternatives, such as honey or fruitjuices, are not mentioned. A maize gruel is fed topatients, and treatment mixtures change withthe symptoms. Every illness is believed to haveits own treatment, but the initial treatment inalmost all cases of illness involves some form ofingestion, purging, application, or inhalation ofa remedy — whether it be modern, traditional,or a combination of both.

Optimising conditions for recovery involvesboth patient and family in an understanding ofhow and when the treatment is to beadministered, and what rules they shouldobserve regarding the patient's diet. Severalhealth-service users of all ages and backgroundsmentioned that they had not been given dietaryadvice during their consultations with a health-care provider. Such advice is consideredindispensable, as 'a remedy taken without foodmay become like a poison', and an incorrectdiet might prejudice recovery. Patients'expectations that recovery will begin within aday or two of treatment seem to be higher withmodern treatments than with traditional ones,but an alternative treatment may be soughtsoon after the first one appears to fail.

A young urban man who was diagnosed withmalaria was still feverish after the first day oftaking the prescribed medicine. He visited anelderly relative in a rural bairro, who gave him aherbal remedy for the makulu which the patientbelieved was causing the fever. A commonalternative treatment for makulu, is peri-anal orvertebral scarification, after which the badblood containing the makulu is expressed andmedication is rubbed regularly into the woundsto heal them.

Spiritualism and divination are not often thefirst choice of treatment for most illnesses whichpresent with physical complaints, althoughsome ritual may be involved (for example, inmanaging convulsion-related illness). Yet whena person or family suffers repeated episodes ofill health, a prolonged or debilitating illness, orwhen there is a history of interpersonal conflict,help from the spirit-world may be sought tounderstand the cause of the problem and toguide treatment.

2.7 Why do people make thechoices they do?

Health-related behaviour — the process ofdecision-making — can be seen as an interactionbetween three predominant elements: infor-mation, access to resources, and beliefs. Each ofthese elements is in turn shaped by a range offactors. 1 he dissemination of information isinfluenced by the sources and means of trans-mission. Costs and seasonal effects determinepeople's access to resources. Beliefs abouthealth-behaviour are influenced by concor-dance of views, personal benefit, and ability tomake changes.

InformationThe belief in the importance of family ties isstrong in Ganda, particularly in the more ruralcommunities. Tradition held that young peoplelearned from their elders, and that theextended family was the source of one's supportand counselling. Girls still learn throughaccompanying their female relatives in the careof the household, while boys cluster around theondjangos to hear the rural elders discuss theirlives. Within the family, one person — young orold — may be thought of as a wahinguka, aperson with a particular capacity to understandand share an experience of life, gained throughcareful observation of his or her surroundings(rather than through inherited powers). Awalunguka has the courage to speak out on issuesrelating to the solution of household problems,courage which others lack through shame orembarrassment, and his or her advice will oftenbe sought.

Young people, particularly in rural commu-nities, view the elderly as valuable sources ofknowledge with regard to traditional practicesand herbal remedies, and elderly woman arevalued for their experience gained in years ofassisting at childbirth. However, people whohave been separated from their own families inthe disruption and displacement caused by thewar say that they feel lonely and isolated, andare at times unable to seek help from their in-laws and neighbours. The suspicion and mis-trust that now pervades some communities haserected barriers between people who considereach other strangers, whose motives cannot beunderstood. In some circumstances, there is atendency to see representatives of officialstructures as punitive rather than supportive;catechists who deal with naughty children, sobas,

37


and teachers who punish the young. Thedifficulty sometimes experienced in therelationship between health-service providersand users is expressed by young rural mothers,who had 'felt shamed and annoyed by nurseswho spoke rudely to them and blamed them' fortheir children's ill health.

Much of what people learned traditionallyfrom the elders was transmitted through story-telling which took place in the ondjangos.Without books and pictures to refer to, theyspoke of the real pictures in their heads. A lowlevel of adult literacy, a paucity of writtenmaterials for those who can read, few oppor-tunities to reach secondary-level education, andthe concentration of the few existing radios intown, mean that information is perhaps stillmost reliably conveyed by word of mouth.Women today sing songs telling stories of thewar years while they pound their maize. Dramasare acted out with dance. The small, indistincthealth-education posters stuck high on the wallsof some health facilities are less attractive andeffective. People say they are more likely to trustinformation if they hear it repeated from severaldifferent sources which they respect, and if theinformation is consistent. This is illustrated bycommunity members' explanations of the suc-cess of some childhood vaccination campaignsin Ganda. Staff at the health facilities give priorwarning of an 'oncoming illness or healthproblem' to the bairro authorities, who thendeliver the message to people in their bairros.Other bairro members who visit the healthfacilities receive the same information from staffand relay it to friends and family. Anotherannouncement may then be made via amicrophone or, occasionally, on the radio.

However, some health-care providers' ten-dency to talk of health issues in terms whichimply a superior biological and technical know-ledge puts others at a great disadvantage. This isone of the potential barriers to developingrelationships in which health knowledge isrespected and shared. While brief lectures areheld for waiting patients at health facilities,there is little encouragement of discussion andquestioning, which could be an important partof the learning process. Likewise, health-careproviders who reduce people's complex under-standing of what causes illness to a singleexplanation — such as blaming microbes fordiarrhoea — will be treated with some scepti-cism, because people's views have been formedby years of experience and rationalisation inorder to arrive at satisfactory, meaningful

understandings of their changing world. This isnot to say that local people are unwilling tolearn, or that it is impossible, with anappropriate approach, to build on their existingknowledge of factors which influence theirhealth, and explain their relevance to their lives.

Access to resources and costsState-run health services are provided free,although transport costs (by wheelbarrow or,rarely, by vehicle) are incurred by the families ofseriously ill patients, the frail elderly, andwomen in labour. The time people spendtravelling to obtain free modern health caremust also be considered, as should the timespent waiting for a consultation — an average ofthree or four hours. Although the state-runchildren's health services are a preferred firsttreatment option for most 'normal' childhoodillnesses, their use as a preventive or positivehealth service could be restricted by the manyother demands on women's lives. Mothersvisiting the clinic rarely have someone at hometo help with their work while they are away. Thishas implications for the effectiveness andcompleteness of vaccination schedules, andlimits the opportunities for learning more abouthealthy children's development — they onlycome in when they are ill. The non-monetarycost of treatment also increases the pressures onsome household providers, particularly women.When faced with long-term hospital treatmentsfor themselves or their children, they have tobalance the knowledge that the illness is seriouswith the needs of their households. In the wordsof one young displaced woman, who abandonedtuberculosis treatment in Cubal as soon as shefelt better: 'I needed to cultivate my fields.'

Women, more than men, talk of the adviceand care provided by close family members,elders, and friends which often concernreproduction-related health problems, such asfailure to conceive and sexual illnesses. Theprice for a 'sexual illness' consultation is twicethat of any other at one private clinic in Ganda.However, even where elders would once freelyhave given advice on traditional remedies, a lotof them now demand payment, perhaps areflection of the difficult economic conditions inwhich most people find themselves. As fewpeople in Ganda, except perhaps traders, haveimmediate access to cash, the ability to pay inkind or to pay on credit is an alternativeappreciated by most rural and semi-rural usersof fee-paying health services. For them, most

38

l'arl two: findings and discussions

traditional consultations are undertaken as asecond stage of treatment when home orhospital management of a normal illnessappears to have failed. Providers of fee-payingservices record a decrease in patient attendanceat times between payments of Governmentworkers, and when trade with coastal townsslackens. They have noted that, but cannotexplain why, the number of consultations perweek throughout the year has declined by morethan a half, although no corresponding increasein absolute numbers is recorded at the state-runfacilities. By contrast, services offered byherbalists and spiritualists often do not requirepayment until a cure has been effected. Thisoffers advantages to both parties: it allowspatients to spread the costs of treatment overtime, and reduces expectations that a cureshould be immediate because they pay up front.However, a goat or cow is then required aspayment for apparently successful treatment ofcabeza grande or terisdo de gota.

A 15-year-old girl from a rural baiiro presen-ted to a private clinic with her mother, withwhom she lived alone. She had a six-monthhistory of chest pain, 'as if there was a woundinside', and a painful body. She had beentreated with home-made herbal remedies manytimes, had attended the central hospital onseveral occasions, and had made threeconsultations with kimbandas in the bairros,paying 1 million kzr, 700,000kzr, and 50kg ofmaize (maize at 50,000kzr/kg, July 1997 prices).She had also visited the Catholic sisters' healthclinic on four occasions, with costs totallingalmost 3 million kzr. She was now attending theIESA clinic, where she received four packets oftablets and six injections. In order to obtainmoney, her mother had sold sugar cane andmaize from their fields, all their chickens, andsome of their fields. They were about to selltheir mud-brick house to live in a smaller grasshut in the compound.

The father or husband of a household hasthe last word on critical decisions about themanagement of household resources and theirconsumption. While a woman may, for example,appreciate that a latrine in her compound couldreduce the nuisance of smell and flies caused byfaeces, she would be unable to have one withouther husband's co-operation. In some circum-stances, a woman managing her householdresources without the influence of a man is lessdisadvantaged than one who is married and hasno such control over the means which areprimarily provided by her.

In addition, patterns of perceived healthproblems change throughout the year, togetherwith the changing demands on people's liveswhich follow the agricultural calendar andopportunities for trade. This influences notonly how people respond to ill health, but alsohow much time and resources they have forpreventive and positive health-care. The drymonths are normally a time for making repairsto the compound, for making mud bricks andfor accumulating goods through the kandongasystem. The rains bring the time of mostintensive agricultural labour, with some womenspending up to ten hours a day away fromhome. They worry that their children are morelikely to be sick at this time, requiring more carewhen there is less time and energy to give it. Thetype of foods available change, and their varietyis reduced. Women also say it is the worst time tobe in the late months of pregnancy.

BeliefsPeople's general health knowledge and theirunderstanding of how their healthy bodiesfunction influence the actions they take tomaintain health. This is illustrated by theexplanation of many women in Ganda that eachof their pregnancies exists with its own supportsystem. Some believe that the sharing of breastmilk between infants is undesirable, whileothers see an increased need to give extra carefor higher-parity pregnancies. The belief thatmuch traditional illness is caused by some formof contamination, by an organism or otheragent, means that traditional remedies arebased on expulsion of the cause or the 'rubbish',through purging, bleeding, and exorcising.Thus a combination of modern and traditionalremedies, to treat both the 'illnesses' and theircauses, can be very successful.

People's focus on chronic problems such asrecurrent: pregnancy loss or failure to conceivesuggests that these are important psychologicalas well physical concerns for both women andmen. From a medical point of view, this maypartially reflect the consequences of inappro-priately managed sexually transmitted diseasesand of chronic undernutrition, particularly inwomen. However, people's preference fortraditional spiritual and herbal care is alsoexplained by their need to apportion blame forwhat appears inexplicable, and to embark onlong-term care in which they, through therituals they undertake, are part of the process ofhealing. The process of healing chronic illness

39


often involves important members of thepatient's social network, whereas the manage-ment of a simple, symptomatic condition whichis perceived to respond to modern treatmentswill not require the mobilisation of friends andfamilies to the same extent.

External and uncontrollable factors, such asthose imposed by years of drought and inse-curity in Ganda, have contributed to peopleshowing increased signs of resignation andpassive acceptance of fate. But the existence ofbeliefs that certain actions can prevent or main-tain health indicates that most people (particu-larly women) still feel that they have controlover some aspects of their lives and those oftheir household members. Some urban youthclearly express the desire to avoid 'vices', such asalcohol, smoking, and early pregnancies, inorder to maintain health; but for othersubgroups of the population such issues remainlargely unaffected by individual actions.People's expectations of individual benefitinfluence whether certain health-relatedactivities are undertaken. Some women decideto attend antenatal clinics at the hospital inorder to receive an attendance record card,without which they feel they will be punished bymaternity staff, if they eventually require ahospital delivery.

2.8 Health needs

Major felt needsThere is an almost universal desire expressed bypeople in Ganda that there should be moremedicines and more health posts available to allmembers of the population. Gandans are alsogenerally dissatisfied with the current prescrip-tion services. Tablets that are given in fractionsare believed to be less effective than whole ones,and treatments that consist of only one type ofmedicine are not thought capable of curing theseveral illnesses that the patient may havepresented at the consultation. This does notnecessarily mean that people wish to replacetheir own systems of coping with traditionalproblems, but that they consider the institu-tional health systems inadequate for conditionswhich require specific modern treatment. Suchan expression also implies that people wish for akind of health-insurance system, for the securityof knowing that adequate and appropriatetreatment will be available should they need it inthe future. Frequent references to the loss of the

knowledge of traditional practices and herbalremedies which the elders have held forcenturies are rooted in similar concerns abouttraditional health-care systems, threatened byfamily separation, economic hardship, and alack of opportunities to pass on knowledge.Some people's concern about health servicesincludes not only medicines, but also the qualityof care on offer. Before the war, 'patients werenever left in pain', an elderly woman explained;and 'there were not many maternal deaths,perhaps one a year died in childbirth. Thenurses and midwives treated the patients andwomen well; there was more love between them[but] now there are negative cases where thebabies or mothers are dying in the hospital. Awoman arriving at hospital with pains has herwhole body massaged and the baby is borndead. Because of the war, there is only hate andenvy'.

'Illnesses' are a significant problem in thelives of most respondents, and much of it isperceived to be related to insufficient andinappropriate food. Having the ability toproduce food in the present context of Gandatakes priority over primary needs such as water,sanitation, and education. While people areaware that the food situation does not comparewith the years of starvation during the early1990s, there is a real sense — among displacedpeople in particular — of the precarious natureof their lives and of the daily preoccupation withproviding or obtaining food. As one elderlyrespondent commented: 'If the stomach isempty, how can the head learn new things?'Learning new things, such as how to managetheir own business, is a way in which youngwomen traders seek to improve the security oftheir livelihoods and open up opportunities. Acommon theme of comments made by youngmen from all backgrounds is the wish for apeaceful life, without war or gossip.

Institutional health-care providers' concernsfocus on the population's high demand forcurative services and the providers' inability tomeet this demand in a way which maximisestheir scarce resources. If increased materialsupport were available from other agencies, oldnetworks of MINSA traditional birth attendantsand health promoters in the bairros could bereactivated and additional public health activ-ities, such as communicable-disease education,could be carried out by centrally based mobilehealth teams.

Evidence from other countries shows that thedegree of community participation in the

40


development and implementation of com-munity-level health initiatives is essential totheir success. Generally, national initiativeswhich lack clear policy, a real commitment tocommunity involvement, and the long-termfinancial support to ensure training, super-vision, and monitoring, have been lesssuccessful than small-scale programmes,-'7

which have been largely supported by NGOs.However, while such programmes do contri-bute to the development of community-levelhealth services in some areas, an absence ofclear Government policy and commitment toservices will result in unequal provision ofhealth-care, lack of accountability, andfragmentation of the health services.

Some health workers, notably private prac-titioners, appreciate that the continuedjudicious use of traditional, herbal remedies iscomplementary to other forms of health-care,and should be supported by focusing on otheraspects of care such as dietary advice. Othersthink that using traditional medicines should bediscouraged by means of community health-education programmes, because there are toomany unknowns in terms of safe dosages andunderstanding of pharmaceutical effects. At thecommunity level, curative health services areclearly more lucrative than preventive orpositive health services, but informants typicallyfeel that health-care providers at all levels needto co-operate more, by means of shared trainingopportunities and feedback.

Real health needsIt is clear that the main health needs in Gandareflect a high level of poverty, as in othersocieties which have been socially, economically,environmentally, and politically disrupted formany years. And as in other societies, povertymanifests itself for the most part in healthconditions that can be prevented if the primaryneeds of all members of the population aremet. The most commonly diagnosed medicalcondition — throughout the year, in all agegroups, and at all institutional facilities — ismalaria. It is the diagnosis made in 40-60 percent of paediatric consultations and in a slightlylower proportion of adult cases. By comparison,pdssaro is also cited as a common childhoodillness at community level; clinically, the

convulsions described in many cases arcprobably a manifestation of the fevers andcerebral complications caused by a malarialillness. A combination of factors relating to theproper diagnosis and appropriate treatment ofmalaria may be responsible for the recentemergence of local cases, which appear to beresistant to all commonly available antimalarialmedication.-'8 This has important implicationsfor all health services. Other frequent diagnosesin adults and children are respiratoryinfections, gastrointestinal infections, andconjunctivitis, which accounts for approximately20 per cent of 'emergency' consultations.However, with a wide range of diagnosticcategories — up to 50 items in one month ofhealth-facility consultations — and varyingdegrees of diagnostic skill in health-facility staff,interpretations must be made cautiously.

Although a large volume of health data isregularly recorded and reported in Ganda, it isdifficult to interpret, which limits its usefulnessfor improving our understanding of localhealth trends. Without the possibility ofcomparing reliably gathered figures from alllevels with an accurate demographic base,opportunities are lost for the development ofmonitoring systems and for better-informedhealth-service planning. Age and sex-specificindices of mortality and morbidity, which mightreveal crucial differences between levels ofhealth experienced by subgroups of thepopulation, can only be guessed at. Certainlythe frequency with which local friends andcolleagues request assistance with wood andnails for coffins, the regularity with whichfunerals are held, and the personal histories ofmultiple child and infant deaths, are indicationsthat deaths are too common. Likewise, birthassistants speak of a 'high number' of spon-taneous abortions and premature deliverieswhich occur as pregnant women carry heavyloads or trade between markets. But com-munities' actual levels of morbidity andmortality are not clearly known, as there are nofunctioning verbal or other information-gathering systems. The involvement of com-munity members themselves in the collection ofhealth information aims to increase theiropportunities for participation in the planning,provision, and monitoring of health services,'-'9

according to their own needs.

41

Part Three: Conclusions and recommendations

3.1 Health-related behaviourand the use of health serviceswithin the sociocultural contextof Ganda

Sociocultural contextGanda is considered rural in the wider Angolancontext, but within Ganda the experiences,beliefs, and behaviours of those who live intowns and those who live in traditional villagesdiffer markedly. Returnees from other areashave added to the variety of perceptions andpractices. Urban/ rural differences are reflectedin the way households manage and prioritisetheir affairs and in the degree to which theyfocus on individual or community 'survival'.

In community organisation, power andauthority are wielded in a vertical, top-downway, and most of the structures which touchpeople's lives, including international NGOsand churches, are perceived to be backed by theauthority of the Government. There are fewexamples of social mobilisation through issue-based community groups or other community-level initiatives in Ganda, but some communi-ties retain mechanisms for communication andproblem-solving through traditional structuressuch as the ondjangos.

The years of conflict and social disruptionhave generated an atmosphere of mistrust,suspicion, and envy. This is felt more deeply bypeople who live in communities comprisingmany unrelated households, and it has thepotential to provoke accusations of witchcraft,blaming others for misfortune, and to hinderdevelopment.

Smaller, more traditional bairros tend to haveoverlapping household networks, which canfunction as an informal social support systemand increase household security. Bairros with ahigh proportion of isolated households havelinks and responsibilities outside of thecommunity in which they are based.

Men exercise power at all levels in society.Men are seen as controlling the household'sresources (whether they are present or not);

women cannot make critical decisions. Yet thesurvival of a woman's household depends on aher skill as resource manager and primaryprovider throughout her life.

Most people in Ganda are poor in terms ofinadequate reserves, capital assets, and nutri-tional opportunities, but they are extremelyexperienced in and capable of managing withwhat is available. Women in particular have avast knowledge of how to earn a living bycombining activities, planning, taking risks, andusing the kandonga system.

At times of stress — when social obligationsthrough family blood-ties override those createdthrough sharing the same cooking hearth — thehousehold proves to be a valued but fragilestructure. Such obligations can hold back someindividuals' or households' capacity to develop,but recipients of such support can avoid othercoping strategies which carry long-term costs,such as selling essential assets.

Young people tend to view their relation-ships with people in positions of authority(teachers, catechists, health staff and so on) aspunitive rather than facilitative; much criticallearning takes place through informal contactswith peers or through close family relation-ships. Many young Gandans are involved intrading from an early age, but the dearth of localopportunities for higher educational orvocational training in Ganda means that theycannot easily acquire useful professional skills.

Health services and providersHealth-service providers exist at institutionaland community level in Ganda. There is a widerange of community-level, non-institutionalproviders, using a variety of methods ofpractice, and functioning largely independentlyfrom each other and from institutional-levelhealth-care providers.

Health-service provision in Ganda focuses onthe plans and resources of institutional healthfacilities rather than on the community'sperceived health needs or on observed changesin their health status.

42

Pail three: conclusions and recommendations

Preventive health-care at primary level isprovided mostly by women and mothers withinhouseholds, but also by health-care services (forspecific interventions, such as vaccinations) andby traditional practitioners. State-organisedprimary health-care (PHC) initiatives such ashealth-promoter or traditional midwife systemsreceive little recognition within communities,and nor does their role in illness prevention andthe promotion of good health.

A tendency among health-service providersat institutional level to manage PHC as thedelivery of services from provider to communityputs at risk the essence of the PHC approach:that there should be a visible and effectivepartnership between both. Such a partnershipwould make it possible to recognise andappreciate existing health resources in Ganda,such as local people's knowledge and beliefsabout health matters.

People are already paying for some curativeand preventive health services, in terms of feespaid to individual providers and at privatefacilities, and in terms of non-monetary costs,such as time. They appear to be willing, thoughnot necessarily able, to contribute towards whatis perceived to be an effective service.

While most people in Ganda have reasonableaccess to water all year round, it is not consis-tently available at the sources closest to people'shomes in sufficient quantities to meet theirperceived needs.

Family and public latrines which offer privacyare acceptable sanitary alternatives to defe-cation in the open air for many people. Some,however, have kept the customs of life in thetraditional village, although the relative over-crowding and lack of space in the bairros do notallow for safe and natural degradation of waste.

Health-related behaviourHealth-related activities form a continuum fromthe seeking of health-care to the maintenance ofgood health, with a common perception thatpeople are responsible for and capable ofinfluencing some aspects of their lives, bothwhen healthy and when ill. This perception isreflected in taking special care of vulnerablepeople, especially children and pregnantwomen; in rules around the care of people withchronic illness to prevent a worsening of thecondition; in knowledge of nutrition and eatinghabits (such as food for different age groups,food in pregnancy, food in illness andconvalescence, and food taboos).

Women of all age groups, particularly thoseliving in rural communities, currently dependon support from their family, friends, andneighbours for health-care. They lack directaccess to resources within their households andrisk losing personal assets to pay for treatment.Women's entitlement, as individuals, to effec-tive health-care is undermined by the manifolddemands on their time, by social pressures, anda lack of services which specifically address theirhealth needs.

Many health beliefs and some child-carepractices which are potentially harmful (such asearly weaning as a result of another pregnancy)include the concept of blame. Because womenboth assign blame and are blamed for healthproblems, they are under considerable socialpressure to conform with these beliefs andpractices.

Every illness is believed to have its own treat-ment: some are clearly recognised to requiretraditional management (for example, pdssaro),others are initially treated with modern,allopathic medicines (for example, fevers andheadaches) although herbal remedies for suchillnesses are available. The cause of illness isoften defined by people in terms of the patient'sresponse to certain types of treatments.

Some health problems that are believed torequire specific, non-allopathic interventions,such as 'nervous' illness, are attributed to a lackof harmony in interpersonal relationships; helpis therefore sought from appropriate sources. Aperceived increase in the prevalence of 'tradi-tional' illness in some communities is related tothe negative influences on people which haveresulted from years of war and social disruption.

Chronic illness and ill health is often attri-buted to traditional causes, and although acondition such as tuberculosis is commonlyunderstood to be a serious illness requiringspecific treatment, its symptoms and long-termeffects are not widely recognised. The burden ofcare for the chronically ill falls on women intheir role of household carers and providers,and on non-institutional, community-levelhealth-care providers.

Patients' expectations of recovery areinfluenced by whether, how, and when they payfor treatment. When people are charged feesfor health services, they expect to receivesufficient medication to treat all the illnessesthey present, while deferred payment mightallow a longer time-scale for recovery — as is thecase in treating traditional illness.

43


3.2 Preventive health prioritiesof the communities and theirmost vulnerable members

Vulnerable households are headed by a singleadult, separated from their close family, lackland, and have few options for earning a living.Within a vulnerable household, sustenance ofchildren will often take precedence over that ofadults; when resources are scarce, the elderlyare respected but not generally protected.

Displaced households are not homogeneous:some have more opportunities to earn anincome, for example as paid labourers andtraders, than others who remain isolated. In thecurrent political climate, the threat of losingassets, such as cattle and crops, through theftrenders many rural households vulnerable.

People want their needs for food security,water supply, and curative health-care to be metfirst. The impact of any additional interven-tions, such as health-education and sanitation, islikely to be increased if planners consider howbest to meet primary needs alongside these.

The concept of prevention is well establishedin Ganda, as is illustrated by the importanceplaced on the care of people at vulnerable stagesof life, on diet and nutrition, childhood vac-cination, and on maintaining a clean livingenvironment. The study highlighted theemergence of specific concerns — about youngpeople's reproductive health, about nutritionaladvice at times of ill health, and about women'shealth and reproductive health — whichsuggest that Gandans have an interest in andfeel a need for opportunities to improve theirknowledge of preventive health-care.

It appears that people's health-care prioritiesare curative services, whether from traditional,spiritual, or allopathic sources; patients under-stand that each of these curative services makesan individual contribution to health, and thatthey are not interchangeable. In general,allopathic services in Ganda are unable to meetpeople's expectations due to lack of resources.

Allopathic services seem to be predominantlyconcerned with managing preventable illnesses,which has led to both providers and usersdemanding more curative services. Obviously,repeated ill health due to preventable ill-nesses such as malaria or intestinal infectionsreduces individuals' and households' capacityfor productive work, depletes scarce resources,and increases people's vulnerability to furtherhardship.

The potential positive effect of preventivehealth-care activities such as vaccinations couldbe diminished because external factors stoppeople from using them effectively: children areoften taken to central health facilities only whenthey are sick, because there are other demandson women's time and resources, and incon-venient clinic hours or long waits furtherdiscourage routine visits for vaccination andgrowth-monitoring.

3.3 The development ofappropriate methodologicaltools and health-status indicators

A qualitative approach was appropriate for theexploratory phase of this study, which wasintended to collect information covering arange of beliefs, knowledge, and behaviour. Itallowed people to express their opinions andknowledge in a non-threatening, informalatmosphere and identified issues whichpreviously 'outsiders' might not have suspectedto be community health priorities.

We counteracted the problem of interpretationbias within the research team by working withlocal research assistants who shared the studypopulation's beliefs, cultural background, andlanguage. With time and growing confidence,the RAs were able to discuss findings and feltcomfortable to be probed deeply on somesensitive issues, which would have been moredifficult with participants.

However, the study was affected by limitedtime and resources both in the research teamand the local community. Ideally, we wouldhave liked to complement the qualitativefindings with quantitative research, in order toincrease their explanatory power and genera-lisability, and with further in-depth work toexplore the associations between qualitative andquantitative approaches.

Thus far, the research methodology hashighlighted areas for further investigation andprovided a model that can be adapted to guidefuture research initiatives. It accommodates theneed to create an atmosphere of trust inworking relationships, the realities of projectresources, and the legacy of a top-downapproach to information-sharing in Angola.

This project was planned specifically togather information rather than to initiateinteractive processes and community action.However, experience gained from training

44

Part three: conclusions and recommendations

assistants in and applying participatoryresearch techniques shows that such projectsopen up opportunities for real dialogue. Goodfacilitation skills, a clear understanding of theirobjectives, and frequent evaluation of personalperformance are essential in order not to wastethese opportunities.

It was difficult to monitor and evaluate theimpact of local health-related interventions.Health-care providers and planners seldomattempt to define and agree on expected out-comes with beneficiaries in a way which reflectstheir understanding and experience of healthmatters. They have also failed to take into consi-deration the long-term and multifactorial causesfor the most frequent illnesses in Ganda.

Although routine health data is collected inabundance at health facilities, its usefulness inmonitoring the health status of the populationand of specific subgroups is diminished becauseno recognised corresponding system of routinedata collection exists at community level.

3.4. Recommendations toOxfam concerning programmedirection and initiatives inBenguela Province

Approaches to communitiesThe following points should guide howcommunities are approached both at the startand during the course of a project:

• Start off with small initiatives which developconfidence on the part of the communities andproject staff, and which consolidate existingfield skills and information-gathering techniques.Think about what people are saying and why,and take the information back to them in a formthat invites further discussion and shows respectfor their own ideas and initiatives.

• Use research findings to define starting pointsfor further discussions with communities.These may validate or reject your findings, orhighlight issues and differences which requirefurther exploration in order to contribute to abetter understanding of the communities.

• Demonstrate the ability and willingness toinvolve a range of community members,including the least vocal and visible, in identi-fying community needs and in developingsolutions acceptable to all subgroups. This willincrease levels of trust and open up possibilities

for information-gathering between staff andpartner communities.

• Respect people's priorities and take intoaccount other demands on their lives. Make sureyou recognise when particular interventions areappropriate; for example, address diarrhoeatreatment in the rainy season and promotelocally available products to prepare oralrehydration solutions with.

• Identify and work at project level with thoseindividuals (for example, from Governmentservices) who show initiative and potential withregard to approaching communities, to identi-fying their needs, to solving problems at com-munity level, and to encouraging the develop-ment of communication and analytical skills.

• Maintain good relations with the localauthorities to keep open communicationchannels and to maximise opportunities forlobbying and advocacy.

Programme content and developmentOxfam should work with and develop localgroups in order to meet the basic needsidentified by the community, such as buildingand managing water points, or organisingagricultural associations. Projects should aim toenable communities to support themselves bymeeting the basic needs of displaced peoplereturning to their homes, by increasing oppor-tunities to make a living, and by providingtechnical and organisational training. It isimportant to look at feasible cost-recoverymechanisms, so that community initiatives willbe sustainable. For instance, a group of womenborrowed money to start a bakery in a tradi-tional village on credit and will use the bakery'sprofits to repay the loan.

Oxfam must remain prepared for emer-gencies. The political situation in Angola is stillprecarious, and most people's reserves areseverely depleted or threatened by continuinglocal insecurity and theft. In order to reactquickly when people's livelihoods are threat-ened, we need to develop systems to monitorpeople's income which are context-specific. Oneeffective system in Benguela Province wouldmeasure the proportion of time that womenspend during a week collecting firewood as analternative or as the only source of income.

Health-care projects should make positiveuse of the community's existing practices andbeliefs, for example, working with women'sexperience and pride in maintaining a healthy

45


environment for their children and households.Project workers should discuss with women andmen, separately and together, how women'sworkload can be reduced, look for where ideasmeet, and create sufficient trust in theirrelationship with the community to challengeharmful practices.

Food-security projects which addressnutrition ought to make use of the knowledgethat people, especially women, already have offood production and resource management.Rather than repeating standard nutrition andhealth messages, extension workers ought toknow and talk about what people are growing intheir fields. They also ought to use familiarlanguage and concepts. For instance, peopleconsider smell and flies as tangible evidence ofsomething unhealthy; telling them theoreticallythat latrines reduce diarrhoea contradicts theirexperience of its causes, and they cannot see anyevidence for it. At the same time, workers oughtto avoid reductionism and simplification ofmessages, such as 'flies cause diarrhoea', bymaintaining diverse approaches in discussion.

In order to promote consistent, positivemessages about health matters and reach thewidest possible audience, Oxfam must work withcentral sources of information such as churches,youth groups, schools, teachers, and otheragencies involved in environmental health. Itmust also lobby staff at health-care facilities touse illness episodes to increase patients'knowledge and awareness of preventive health-care measures.

Rather than use standard indicators of healthstatus and the success of health-educationactivities which may be confounded by manyfactors and different interpretations, Oxfam'sprojects ought to develop and use indicatorsthat reflect people's perception of health statusand can chart a process of change. Suchindicators should examine an individual's orhousehold's perception of their well-being;people's sources of information and ways oflearning about health-care; changes in theproportion of people who believe that com-munity health is the responsibility of others;changes in the proportion of women who wouldcontinue to breast-feed if pregnant; andchanges in young men's and women's aware-ness and knowledge of locally available methodsof contraception.

Project workers ought to consider workingwith youth groups as peer educators on envi-ronmental health matters. This could broaden

health-education activities around specificimportant issues such as reproductive health bytraining and developing young people's skills,with minimal input of extra resources.

A priority in Angola is to increase theavailability of curative care across the country;Oxfam should promote and/or lobby for a systemthat utilises existing health-care resources (suchas knowledge of herbal remedies) andcollaborates with traditional practitioners, whilealso meeting people's expectations forimproved allopathic curative care.

Further research to gather household-levelinformation about livelihoods should examinethe following topics:

• resource flows within and between households;• opportunities for women to both manage

and control resources;• changes which would reduce women's work-

load but which men would considerbeneficial to the household;

• how people balance options and risks whenthey do have choices and when they don'thave choices (for example, selling maizeseeds or hoe from distribution);

• at what point social obligations overridefamily or household obligations (for example,when payment for a relative's funeral takesprecedence over children's school fees);

• how appropriate credit schemes are in thecontext of exchange-based systems and in thecurrent security climate.

These household-level data can inform futuredebates on user-fees for health services, becauseit sheds light on seasonal variations in access toresources, and on the low relative value of moneycompared with the value of assets surrenderedto obtain health-care. It will also help to identifysubgroups who should be exempted frompaying fees.

Project managementIn order to ensure that the research carried outas part of a project will be useful in the longterm, researchers should keep in mind thefollowing points.

• Make field-notes, keep diaries, record whatpeople are actually talking about (rather thansimply that they say), so that this valuableinformation can be shared within Oxfam andwith other local organisations, and used forlearning about the process of development inpost-conflict situations.

46

Pan three: conclusions and recommendations

• Monitor the level of skills within the team,and make frequent use of opportunities todiscuss constructively with others theirexperiences and what they have learnt duringand between specific projects.

• Within programmes and together withagencies or organisations, identify points ofaction through which links can be developed.For instance, if a food-security project alsoworks to meet the community's identified waterneeds, this increases the population's level ofinterest in the intervention, which can thenpromote a more comprehensive health-carepackage including food security, environmentalhealth, and health education.

• Formalise an exchange of information withother agencies and health-service providersthrough regular meetings. A greater awarenessof topics such as local eating habits and women'sdaily activities could be useful in improving servicessuch as supplementary feeding programmes.

• Develop Oxfam's role as an intermediary,enabling people to demand basic rights byincreasing the community's awareness ofopportunities to address and resolve problems.Develop specific resources at project level tosupport local lobbying and advocacy activities,and promote information exchange on a widerlevel by delegating an advisory team to otherorganisations or Government agencies.

• Assess what decisions have been made as aresult of the research project. One way of doingthis is to draw up an action plan, for example atworkshops, and follow it up at local, organisa-tional, and inter-agency level.

3.5 General recommendationsfor working in communities

External project staff must try to understandand respect the differences between and withincommunities. Rather than using project 'blue-prints', they ought to develop strategies fortheir work which demonstrate consideration forand an understanding of the community'scomposition and structures.

Before starting the project, it is useful togather information from different sources suchas local authorities, other agencies working inthe same area, and important non-institutionalcommunity figures. Project workers shouldmake efforts to maintain facilitative and honestrelationships with local and Government

authorities, which will also open up avenues forlobbying and advocacy work at a later stage.

Project staff should be open to learn abouttraditional structures such as the ondjango and,where appropriate, use them as a base forinitiatives, instead of creating new ones whichmight be less long-lasting. Being sensitive to acommunity also means identifying how, and bywhom, information can be disseminated at com-munity level in a way that respects traditionalmechanisms. Project workers should 'listen tothe batuque'30 (respect and consider other people'sknowledge) and emphasise existing positivehealth-care behaviours and practices, ratherthan reinforce blame by focusing on what is notdone. The community might develop their ownsystems of sanctioning perceived unhealthybehaviours once their awareness of and respon-sibility for their own well-being is increased andtheir control of the future enhanced. This canbe achieved, for instance, by strengthening acommunity's capacity to make decisions aboutits own health priorities. It is vital to encourageindividuals and communities to take respon-sibility for positive health-care withoutforgetting that there are things which areunavoidable or beyond their capacity to change.

Methods of information-gathering, analysis,and feedback should be adapted to matchcommunities: while a more 'authoritarian'approach may be appropriate for initialcontacts with communities where individualagendas predominate, participatory approach-es would be more acceptable and transparent tomore cohesive communities. In every case, theproject's aims and objectives must be clear to all,because the involvement of community membersfrom the start of planning is important for latermonitoring and evaluation of interventions.

The needs of young people ought to beaddressed specifically. Researchers should lookfor opportunities to determine young people'sneeds in terms of information and skills, anddevelop material which is suitable for them.Peer-education networks and participatorytechniques — unlike some health-care settingsand other relationships that are viewed asauthoritarian — offer young people the chanceto grow more confident, to deal with a range ofproblems within their communities, to findsupport from others in similar difficult socialsituations, and to learn through constantdiscussion and questioning.

Men should be encouraged to identify andunderstand the problems that affect women's

47


lives, because new initiatives are more likely togrow if men also benefit in some way. Bringingpeople together at community level will enablethem to recognise shared problems and todiscuss opportunities for their solution.Community action may also have a role in therecovering confidence and trust, and inrebuilding disrupted social support networks.

Lobbying and advocacyNGOs and Church organisations can play asignificant part in lobbying for an effective,appropriate, and officially recognised structure(for instance, a network of community healthagents or workers) which facilitates communi-cation, health-service development, and infor-mation-management between community andinstitutions. A lobbying agenda should includethe following points:

• the public health sector's focus should bewidened to include existing non-institutionalcommunity level health resources;

• Government policy should be clear andinclude a firm commitment in terms of financeand support to make such a structure part of thepublic health sector;

• within such a structure, health workers' status,and roles as well as their relationship with otherlevels of the health sector must be clearly defined;

• Government and NGOs must collaborate toensure effective use of minimal resources (forexample, the integration of parallel healthprogrammes, such as those supported byCaritas and CVA).

Oxfani and others have to realise the limitationsof Government institutions in the currentpolitical climate. They must continue to lobbyfor the provision of basic services by Govern-ment in the knowledge that local and outsideorganisations will function more or lessautonomously for the near future. Once theoptions and extent of possibilities of collabora-tion with state-run institutions are clear, NGOscan support state-run services by:

• developing alternative strategies for service1

provision;

• supporting the development of clear Govern-ment policies, as an essential component forwork with all Government institutions;

• providing in-service training for publicservice staff in order to improve practice anduse scarce resources effectively;

• encouraging collaboration in defining statusindicators which will be useful to determine theimpact of services and interventions at all levels;

• promoting community organisations whichoffer opportunities to voice needs, and byimproving the negotiating power of communityorganisations with institutions (for instance,through bairro health committees);

• improving the health-service base byemploying existing resources in rural areas. Forexample, traditional practitioners could betrained to recognise syndromes for commonillnesses such as TB or sexually transmitteddiseases; market medicine sellers could receivebasic pharmaceutical training.

48

Appendix 1: Population survey

When Oxfam met staff at the Angolan Ministryof Health (MINSA) during the initialpreparation of our qualitative research project,they asked us to plan and carry out a populationsurvey as the basis for the development of asystem for collecting, interpreting, andreporting local health information.

PlanningThe first stage of the sub-project that developedwith the MINSA Public Health team had severalspecific objectives:

• to carry out a population survey by agegroup and sex in order to establish usefulpopulation denominators for the Government-controlled area served by Ganda Municipalityhealth services;• to train a core team within the MINSA PublicHealth Department in data-collection methodsand analysis of survey data, and to carry outverification procedures using participatorytechniques such as community mapping;

• to feed back information generated on method-ology and results to all levels and relevant actors;• to identify health-information needs in orderto plan and develop a facility-based health-information system;

• to identify with MINSA staff and withcommunity authorities opportunities for thedevelopment of a community-based health-information system.

The core survey team consisted of the MINSAPublic Health delegate and three members ofthe Public Health staff representing thevaccination, sanitation, and health-education

programmes. The Oxfam Public Healthresearch team provided support in the form oftechnical advice, visits to bairros, and somelogistics (transport and stationary items).

Using a map of Ganda Municipality, theGovernment-controlled area was divided intothree sections: the first covered all 19 bairros ofGanda, the second the outlying villages, and thethird the Babaera District including AltoCatumbela. A simple A4-size data collectionform was drawn up to gather the informationneeded for MINSA health-monitoring activities,and to collect baseline information for futurequantitative research. Each form could accom-modate information on 35 houses. A decisionto define a 'house' by physical structure wasbased on the assumption that the communitysecretaries and those carrying out the countwould have knowledge of who lived regularly intheir areas of coverage. They would thus beaware of men responsible for more than onehousehold, and would thus include them onlyas members of the house of the first wife.Members of the MINSA team felt strongly thatpeople associate lists of names with distributionlists, so collecting names was avoided. Thisapproach, combined with clear explanations,would diminish people's expectations of animminent distribution of food or goods andtheir accompanying tendency to exaggeratepopulation numbers.

Implementation of population censusWorking systematically from south to northacross the first section of the survey area, amember of the Public Health team visited eachof the bairros to arrange a meeting with the soba,

Table 5: Example of a data-collection form

# women # women # children # children # children< 1 vears

1

2

3

I

1

1

I

I

II

I

1

I

11

11

II

1

I

1

49


to which he and his committee of zone repre-sentatives and secretaries were invited. It waspossible to arrange two or three meetings eachday. The survey team then visited to address themeeting, explaining the reasons for under-taking the census and the methods to be used.The bairro secretariat discussed how and whento carry out the data collection, after which aninitial sample of houses was surveyed by theteam, together with secretariat members, todemonstrate how to fill in the forms correctlyand to identify any difficulties. For example, torecord the number of children in each agegroup, the informant would be asked to listall the children by age because asking 'Howmany children of age... are there in this house?'would result in more errors. This form ofposing the question evolved after the surveyteam attempted to document the members oftheir own houses — they concluded thatcounting numbers of children is difficult. Eachhouse, when accounted for, was marked withcoloured chalk. The team felt that paint wouldnot be acceptable to house-owners, and duringthe dry season, chalk would be fine. A sufficientnumber of forms, chalk, and pencils were leftin each bairro and proposed data collection dayswere noted by the survey team. Where thesurvey team members anticipated that theremight be difficulties with the accuracy of thedata collection, they tried to accompany thezone secretaries during the count to monitortheir progress.

Data analysis and verificationOnce the forms were returned, the patterns ofdata were briefly examined to check for anygross discrepancies. If any were discovered, thesurvey team returned to the zone or bairro inquestion to discuss the forms with the respectiveauthorities, and the count was repeated. To fur-ther process the data, the forms were countedby two independent counters. The results werecompared and recounts were carried out ifthere were discrepant results. The results wereaggregated to give totals for each category byzone and by bairro.

One set of data from a small rural bairro wasidentified as clearly erroneous, having two ormore members of each category living in everyhouse. A member of the survey team returnedto the bairro to talk with the soba. Together withother members of the bairro authority the countwas repeated, with very different results. The'discovery' of the false results was resolved

amicably, but it took a degree of diplomacy andtact to deal with an embarrassing andpotentially confrontational situation.

Verification of data was planned by means ofcommunity-mapping exercises, during whichinformation on the houses and the number ofadults living in each would be collected andcompared with data from survey forms.Initially, five bairros within Ganda were targetedin collaboration with the Oxfam environmentalhealth and sanitation project, which wasdeveloping links with the communities of thebairros through participatory work. It was feltthat the project and survey teams could use-fully work together on the process ofcommunity mapping, in order to minimise theamount of disruption to the communities and tomaximise the learning experience for all theteam members. Before beginning fieldwork, ajoint session was held to evaluate previousexperiences of participatory methods and toincorporate lessons learnt into the planning ofthe community mapping.

OutcomesThe majority of bairros within the first sectionhad completed and returned their data-collection forms to the survey team within twoweeks of the initial visit. At the time of writing,final counts have been completed for these andfor the bairros of the third section. As a result ofthe improved local security situation, it becamepossible to survey the villages close to UNITAareas, and plans for similar surveys in formerUNITA-held areas of Chicuma, Casseque andEbanga are now developing. There is currentlya lack of demographic information for theseareas. Evidently, the better security climate alsoencourages displaced people to return to otherareas, giving more weight to the need for asystem which monitors population changes aswell as providing isolated 'snapshot' totals.Work is now in progress to feed back the surveyresults to bairro authorities via the MINSAPublic Health team, and to explore with themthe possible mechanisms for ongoing datacollection at community level. This involvessome training sessions with the survey team oninformation presentation, use, and interpreta-tion, to identify the most effective means ofcommunication and the information needs ofboth the health services and the communities.

Verification of the data by mapping has notbeen completed. Following an unsuccessfulattempt to map a large zone in one of the semi-

50

Appendix 1: A population survey

urban bairros with members of the community,the team learned several lessons. Participatorymapping is time-consuming, particularly whenthe participants have other demands on theirtime. In planning such a project, you mustconsider that what people expect to get out ofthe exercise will inform what they put into it.For the benefit of both facilitators and partici-pants, the purpose of the exercise — be it togather data for operations, for research, or a

combination of both — and its expected out-comes must be made clear. If there are severalexpected outcomes (for example, to collectdemographic information, to discuss quality ofservices, and to develop activity plans) it issensible to select a manageable area or unit formapping and to work with those residents only.Later on, you can select another, building up apicture of the community rather thanattempting to do everything at once.

51

Appendix 2: Participatory methods and tools used

The participatory techniques used were adapt-ed from those used in the Participatory Learn-ing and Action method. They were combinedwith other research methods such as semi-structured interviewing to achieve a morecomplete appreciation of local characteristics,processes, and perceptions. Techniques wererepeated with a range of individuals or groupsof people, and results presented in visual formto encourage discussion and facilitate compari-son. Most activities made use of locally availablematerials such as small stones or mango cores todevelop diagrams on the ground, which werelater copied onto sheets of paper.

Broadly speaking, the techniques learnt andused in the research can be divided into fourgroups, according to the purpose for which theywere used (see Table 6).

Table 6: Use of participatory techniques

Purpose

Descriptivediagramming

Chronology

Comparison

Perceptionand explanation

Technique

Community mappingHousehold-linkage mappingBirth historiesDaily activity charts

TimelinesSeasonal calendars

Matrix of options and criteriaWealth-ranking

Circle diagramsCause-and-effect diagrams

MappingA community map showed the physical layout ofa bairro with zones and roads, resources, andservices (water points, schools, and so on),thelocation of key community figures (soba, health-care practitioners). It also included areas whereindividual houses were identified by name ofthe woman who lived in it for a later wealth-ranking exercise. A household-linkage mapseeks to illustrate relationships within andbetween households, and to indicate the typeand quality of the relationships.

Family historiesTo facilitate discussion of birth histories or offamily composition, stones of varying sizes wereused to indicate each pregnancy or familymember. Verification of facts is made easier forboth informant and facilitator by providing avisual representation.

Daily activity chartsThese aim to illustrate the relative use of time,facilitating comparisons between differentseasons or community subgroups and aidinganalysis of where the main constraints oropportunities for change may lie.

52

Appendix 2: participatory methods and tools used

Figure 5: A household map drawn up by a young woman from an urban bairro

a a

\

V ia Aa

* *

TCHiMBOA

CODE•& FR.lEN.DS

• HUSBAND

O MALE RELATIONS

£> FEMALE RELATIONS

FINANCIAL HELP

CL.OTH£S/SHO£S/

MATERIAL <JoofcS

SALARY

FOOb P«06L/C.TS

FAMILY Tl£S

Figure 6: Daily activities during the current dry season for women and menwho live in a rural bairro

CLEAlO,5WE6P,MAKE

OP TIME

A CCOITo

WASH CLOTHES..lls/ EATHE IN R.\VEie,

C O L L E C T IA/ALIC. ToTo SELL OR.

FoofcSPREPAREMAIZE

„ MEAL

C00< ,PREPARE

MEALSLEEP

VISIT FtfiENDS

DAWN MIDDAY SUNSET MiqHT

lA/ASH,EAT

IN

BATHE.,R.EST

WATER

RJ

, HouSe

RESTEATE E N

MEALVISIT

FRIENDS

SLEEP

DAWN/ MIDDA1/ SUNSET

53


TimelinesThese were developed from the discussions disasters, population movements, and local polit-held with older community members. ical and military events. The old people describedInformation was cross-checked between sources, changes in living conditions, health-serviceand a profile built up which detailed significant provision and people's general health status, andevents in Ganda's history including natural gave reasons or explanations for the changes.

Table 7: A timeline of Ganda from 1912-72 according to a very old manin a semi-rural bairro

1912 Construction of the Benguela railway line

1915 Foundation of Ganda, in the name of the traditional soba, Tchilandala Kambia,who came from Kuma

1915 Otchitenha (drought): there was much hunger and no water

1927 Foundation of the Saletina's Mission

1936 Plague of locusts which ate maize and people

1938 Construction of a new bairro in Ganda to house workers for the railway

1952 Establishment of food industry in Ganda with large farms and processing plants

1960 Arrival ol Sr. Loinunba (a legendary Angolan hero) in the Ganda area

1970 Ganda had one state hospital and two private hospitals

1972 Movement of people employed from outside Ganda; Government builthouses in new bairros; many farms and industries producing for export

54


Seasonal CalendarsSeasonal calendars visualise an 18-month cyclecharting month-by-month changes in climate,local agricultural activities, food availability, andillness patterns, and quantifying their effects. Theyhelp to understand potential links between

different components of life in Ganda for variouscommunity subgroups. They are also intended tobalance informants' tendency to focus on thoseproblems or aspects of life which are most currentand pressing.

Figure 7: A seasonal calendar of a normal year (no drought, no war) drawn up by a groupof women in a rural bairro

DATESNEW CAHNIVAUYEAR. MPL.A EASTER.

HOT - - —

MAIZE /SoR-̂ HUV

" - ~ — ^

CHILDRENSI>AV

COLD

\ HAftVEST\

/Sow BEANS SOW MAIZE/HAfcVEST HARVESTJ SW6ET PcrrA-To BEAMS

FIWALEXAMS

M

/PREPARE <Sow MAIZE

\

\

5^> mio.li . » |J£W CARNIVAU

" HOT

,SO«C,HUM. 8e*>wS BEAMSW E E D I N G

/ SO\A/ ^ \/KAievEST SWEET \/ MAIZE POTATO \

y

q « £ A T E S T ~ ~ ~ ~ ^ — ~ ^

EASTER

WEATHER

LAVRAS

W 0 < K ,M F,£LbS

WACAS

/ AVAlLABlMTy

WOR.K

55


Options and criteria matricesThe first step when developing a matrix was toidentify a range of options available to partici-pants. This was achieved by discussing the topic

of interest, and drawing up a list of criteriabased on participants' reasons for choosing oneof the various options over another.

Table 8: Options and criteria relating to health facilities in Ganda, described by youngwomen in the city

\ . options

criteria^.

confidence

instruments

medicines

injections

diagnosticequipment

no paymentneeded

works dayand night

able to treatminor illness

hospital

••••

:*.••

v..•••

%•v/0 «

• •«

>rivateiractitioner

0 0*

• * :

••••

herbalist

4

0

• : ••••

piritualist

#

marketeller

• •

«

Catholichealthpost

*••

Evangelicalhealthpost

J:

«

Which health provider is preferred overall and why ? Hospital, because it is able to treat many illnesses.

56


All adaptation of the matrix was made tofacilitate the discussion of sources of advice orsupport for various groups of people, givendifferent situations, and to elaborate on house-

hold decision-making responsibilities. In theselatter diagrams, there was no quantification ofpreferences, simply an illustration of choicesthat prompted further explanation.

Table 9: Main sources of advice and support for young boys in a semi-rural bairro

^\. Adviser

Problem^^.

love <^<^>affairs

building ahouse 40Rfe

avoiding a •pregnancy 4

gifts for 4!gkgirlfriend ^ 3

money fej-matters *-*&

health ^

priest teacher cathechist soba father

•

•

•

grand-parent

uncle

•

•

mother

•

cousin brother

•

•

•

god-parent

•

•

aunt

•

•

friend

•

•

// is interesting to note that initial discussion brought out the 'official replies' (that the priest or teacher are the mainsource of support); only subsequently, as real situations and preoccupations were explored, did boys disclose theiractual sources of advice.

Of all the participatory techniques intro-duced, options and criteria matrices proved themost conceptually difficult for the RAs, in par-ticular the recognition of differences betweenpositive and negative criteria. The use of

representative piles of pebbles was easier thannumerical scales when comparing preferences,so it was necessary to repeat the exercise severaltimes before they felt secure in its applicationand interpretation.

57


Wealth-rankingDuring a community-mapping exercise, housesin a defined area were marked individually, andthe name of the woman of the house was writtenon a small piece of paper. The papers were thendivided into piles according to the participants'criteria for having 'possibilities for a good life'.They also identified reasons why some houses

were considered die least well-off. This was a time-consuming and potentially conflictful exercise,and required a great deal of explanation. When wefollowed up one ranking exercise with otherresearch (observations made in the same commu-nity) it revealed interesting relationshipsbetween people's perceptions of poverty andtheir actual circumstances.

Table 10: Wealth-ranking results in a zone of a rural bairro

Ranking:possibility of

a good life

Reasons

Most

possibilities

Cattle bleeders,

people whoseextendedfamily ownscattle

Some

possibilities

Those who own

a few cattle

No possibilities

Couples(includingpolygynousrelationships),young ormiddle-agedwith children

but no cattle

Difficulties

Separatedwoman or man

left with smallchildren, widowor widower

Many

difficulties

Old people and

the widowed

Poorest

Orphans

Circle diagramsThis exercise makes it easier to discuss therelative importance of different health-serviceproviders to participants during a definedperiod of time, because pre-cut paper circlesare a tool to represent their ideas. Once the

comparison was made and different-sizedcircles assigned to each provider mentioned,the informants placed the circles in a way whichindicated degrees of linkage between theproviders.

Figure 8: Circle diagram made by a group of women in a semi-rural bairro

PRIVATEPRACTITIONER

POLICLINIC

HOMEBltfTHASSISTANTS

PART-TIME

ASSISTANT(MALE")

HERBALISTS

MARKET

58


Cause-and-effed diagramsThese were used to connect ideas about theorigins of certain problems with their effects,and to explore the sequences of events that mightfollow from a certain situation. They usuallystarted from a focal point and provided opportu-

nities to review and check information that wasgiven by participants. Wherever possible, symbolswere used in place of words, but researchers aswell as participants need to be imaginative andhave a good memory for symbols.

Figure 9: A cause-and-effect diagram showing a child with diarrhoea as starting point

CAUSES EFFECTS

SELL 0*WEAKBREAST MILK

HERBALREMEblES

CLIMATECHANGE

WEIGHTLOSS

59

Appendix 3: Example of a drawing used as adiscussion starter by research assistants

60

Appendix 4: Causes of Malnutrition

Guide for interviews with mothers ofmalnourished children

General information

Name of village

Age of child

Household compositionRelatives living in same village, in Ganda,or outside of Ganda, and type of linksbetween them

Child's historyHow was the pregnancy?How long was the child breast-fed,and when was maize gruel started?Why was breast-feeding stopped?

Previous illnesses and how did thefamily treat themFeeding during illnesses and whycertain foods were chosenHow the current illness started andwhat was done at home to treat itWhat was the provocation of thecurrent illness?

When did the child receive the firstvaccination?

Health of other household membersHealth of other members, in particularthe motherIf there have been recent illnesses,how were they treated?If there have been other children with thesame illness, what happened to them?

Information about household resourcesDo they have fields, what products are grown,how is this year's harvest, how was last year'sharvest, what do they sell, what do they store?

Ownership of animals, household memberswith professions

Who works in the fields?

Who is responsible for care of food, includingconservation, preparation, and distribution?How many times a day did the householdeat each day before admission of the sickchild, how many times did the child eat, whatwas eaten?

How does the household overcome problemsrelated to feeding?

How do they deal with the childrenwhen there is little food?

What can the mother do when thereis a poor harvest?

Where does food for sale or exchangecome from?

Who works the fields when thereis illness in the household?What opportunities do the householdmembers have for obtaining food? (Forexample, what is exchanged, do they collectand sell firewood, do they work for others, dothey ask for credit with family/neighbours, andon what terms is credit granted?)

Ideas about prevention of this illnessWhere does the information come from?What can a household do if they have theproblem often?

How can a household promote good healthfor its members?

61

Health and Livelilioods in Rural Angola

Figure 10: Diagram illustrating the causes of malnutrition

ILLMESS/N

YEA* OLh

HOUSEHOLD

LACK OF KNOWLEDGE

OA BAD AWICELACK OF

THE

IPESTS

S

FotfTREATMENT

LACK: OFpcssisiEg. PLAN?

LACK 0F:

COAT fc/T/oMS- F/ELDS- M O N E Y

62

Appendix 5: Map of Ganda district with bairros andinstitutional health facilities (June 1997)

CODECHPH PPPM

independent private clinicMI NSA hospitalMINSApolyclinicMI NSA health postprivate health postsmarket

scale: i 1 approx. 1 km

63

Appendix 6: Health-service providers in Ganda

Community-level, non-institutionalhealth-care providersThese may be regarded as a first point of contactwith a health-care system. ] exclude home- orself-treatment such as relatives performingscarification or prescribing herbal remedies.

Traditional private practitionerA kimbanda (private practitioner) generally worksfrom home, providing a range of curativeservices for all age groups. Some practitionersspecialise, and are known to have particularskills for treating certain traditional illnesses,while others tackle anything from childhoodailments to childbirth. He or she may use allo-pathic medicines (injections and tablets) acquir-ed from the market place, herbal remediescollected from the fields, or a combination ofboth. Herbal remedies consist of fresh or driedplant parts, barks, stones, and animal skins orexcreta. Although private practitioners some-times take notes on patients, describing symp-toms and treatment, they keep no register ofattendance figures or of diagnoses made. Thereis little work-related contact between the practi-tioners and other health-care providers, nor dothey work with official recognition from the state.

Most herbalists acquire their understandingof remedies and procedures through experi-ence and ad hoc training with older relatives orfriends. While many ordinary people are able toperform procedures such as olusongo (scarifi-cation) and operations to remove threadwormsfrom the anal margin, the in-depth knowledgeand understanding of herbal remedies is widelyrecognised as a skill nowadays mostly held bythe elderly. Previous personal experience of ahealth problem that has been successfully cured,such as infertility caused by a 'shut womb', addsto the credentials of a traditional practitioner.Some practitioners have received basic techni-cal training in the state health system or inreligious missions, while a few have previouslyworked as health monitors in the outlyingvillages, and provided first-aid services to thecommercial farm workers during colonialtimes. Some have more dubious qualifications;

one learnt his maternity practice from a cousinwho had been a health assistant in the army.Payment for services is made in cash or in kind,and prices often reflect the severity of thecondition treated. Sometimes charges are notcollected in full until the patient recovers astreatments can be prolonged.

Traditional spiritualistHealing with the assistance of the spirit world islargely the domain of the santas and santos (maleand female spiritualists). Their powers may beinherited from relatives, derived from dreams,or from the ancestors, most often on the mater-nal side and sometimes jumping several gener-ations. A person may not know that she or hehas such power until falling ill with a maladythat fails to respond to any form of traditional ormodern treatment. An experienced spiritualhealer can diagnose the cause of the illness andreveal its nature. After elaborate night-longrituals beside a river with singing, drumming,and feasting on sweet foods, the santa takes hernovice to the bush to instruct her or him on howto find and use herbal remedies. Others, whoalso have these powers, may be drawn to the siteby the sound of the drums. During the war, suchrituals had to be held at dawn because of thenight-time curfews.

Santas and santos work primarily as mouth-pieces for the spirits, entering their world andcalling to them for help through rituals involv-ing water, perfumed soap, bowls, white cloth,plates, and cups. Patients visit the home of thesanta or santo with a group of close friends andrelatives. A typical patient might be a womanwho has developed a severe pain in her chestand believes that a distant, jealous relative hascaused the illness. The santa prepares thepatient and her instruments, places water in abasin and dissolves soap in it. In the froth, thesanta can see the spirit on whom she calls forassistance. As the santa falls into a trance, thewords of the spirit flow from her mouth, givinginstructions to all present on how the illnessmust be treated. Each of the patient's compan-ions concentrates hard to remember the details,clapping and chanting with her, because they

64

Appendix 6: health service providers in Ganda

know that the sankj will not recall her words oncethe spirit has left her. Suddenly the santa, in afrenzy, splashes the soapy water on all present,and sinks into silence. The spirit has gone.

Prescriptions may involve the use of herbalremedies, some of which are ingested, and someof which are used to massage the affected part ofthe body; others are placed in strategic places inand around the patient's home. A less experi-enced santa or santo may refer difficult cases to heror his mentor, or request help when an unusualremedy is required. Someianto may also invokethe help of God, blessing their instruments andplacing a crucifix beneath them. In addition tocuring ill health, santas can discover who mightbe the cause of an illness by calling upon thespirits to create an image of the perpetrator in abowl of water. Having done so, the santa, mayalso offer to provide a medicine or counter-spellagainst that person for a higher fee.

When a spirit enters a sanlo's head, he is drawninto the bush to hunt the wild animals that appearto him in a vision or dream. After he kills them,his assistants help to carry the animals back tothe ondjango in the compound, where a feast isprepared. The animals' fur is displayed on thesanto's waistband, and he hangs red or whitecloth alongside his bow and arrow on the wallsof his sleeping area. Solutions to his patients'problems appear to him in dreams. He is able toadvise them about the appropriate treatmentsand actions for cure, although one formerpatient said: 'Whether he is able to treatdepends on his head and the luck of the patient;when he starts to tremble perhaps the explana-tion will come or perhaps through his dreams.'

Diviners also use visions to solve problemsand to receive instructions for remedies whichthey communicate to their patients. Theysometimes use a piece of glass or a mirror inwhich images appear to them.

Traditional home-birth attendantsFemale relatives and close friends often per-form as home-birth attendants, as they have donefor generations. A young woman 'with a goodhead' who witnesses deliveries-and their man-agement might be interested to learn more, andgradually builds up her knowledge and exper-tise by accompanying older women, who arehighly respected in their communities. Ratherthan undergo formal midwifery training in theWestern sense, these women learn by seeingand doing, through stories of how previousproblems were resolved, and through their ownexperiences as mothers. Some receive under-

standing and guidance through dreams. Anexperienced birth attendant can predict thecourse of labour by examining the shape of apregnant belly, and prepare for anticipatedproblems with traditional herbal medicines.Payment may be offered for the services provided,depending on the relationship between thewomen. A neighbour who has been called tohelp, but who is not 'close', is likely to be paidmoney. There are certain routine practices thatshould be respected during childbirth, andpossible complications that should be dealt withaccording to traditional practices. For instance,the umbilical cord should rarely be cut until theplacenta has been delivered, unless it is tied to astick in order to prevent the cord from re-entering the womb and harming the mother. Aplacenta which is slow to deliver may be gentlypulled from the uterus with one finger,although the first action is to force the woman toretch by placing a stick in her throat. If a baby isborn still inside its sack, this must be opened inthe correct place behind the baby's head toprevent the baby from drinking the water insidethe sack, which will kill it. At times, an extra 'cord'may appear after the placenta is delivered. Thismay only be removed from the womb bysingeing it with hot ashes, because it wouldotherwise re-enter the womb and cause harm. Itcan be a sign to others that a woman has notlooked after her pregnancy well.

Part-time private practitionersHealth staff working at state facilities may workpart-time as private practitioners, offering out-of-hours services, usually from their homes. Privateconsultation and prescription of a range of oraland injectable medicines offers a convenientservice for some patients as well as a supple-mentary income for staff, who say that they haveeconomic difficulties. They may also attendcomplicated home deliveries, for example bygiving powerful injections to stimulate delivery.

Bairro health commissionsSince early 1996, Oxfam has worked to developlinks with MINSA health-education and sani-tation programmes. Health-education trainingcourses focusing on environmental health andcommunication techniques were undertakenwith central community figures, such as sobas,members of the national women's organisation(OMA), technical supervisors, church and edu-cation representatives, and MINSA healthpromoters. Participants committed themselvesto work as voluntary members of bairro health

65


committees, which Oxlam continued to supportby providing education materials and trainingat community level. With the development ofthe environmental health programme, it became

•apparent that the commitment with whichmembers represented their communities variedconsiderably. Communities and staff met to dis-cuss how much time volunteers could realisti-cally give and what the specific local environ-mental conditions were, and decided to revisethe composition, structure, and function of thecommittees. They now work in groups of volun-teers who plan their activities carefully, takinginto account everyday demands on their lives,especially women's lives, in Ganda. Most volun-teers are women, although a few interested and'acceptable' (to the women) men represent smaller'familiar' areas rather than whole bairns.

Medicine sellersTraditional and herbal remedies are availablefor sale in the central market place in Ganda.Their uses are many and varied, ranging from thetreatment of constipation in children to streng-thening of a man's 'power', and protection fromdeath. The sellers who collect them from thebush often travel far to find them, avoidingunsafe areas. They say that when the fields areburnt prior to the beginning of a new plantingseason, leaves are destroyed, but roots areundamaged. Old people in Ganda say thatbefore the war, modern medicines could only bebought from the pharmacies in the town. Thetraders today come from coastal cities and areknown as 'official traders', but in general theyhave no formal health-related qualifications.The cost of two tablets of a simple analgesic isequivalent to a kilogiam of maize; ampoules ofinjectable medicines cost five times as much,excluding disposable needles and syringeswhich are more difficult to acquire. Althoughthe medicines are generally sold within theirexpiry dates, the lack of storage facilities maynegatively affect their quality.

Institutional structures

State-runThe 80-bed hospital offers free in-patient adultand paediatric services, a maternity department,a dental clinic, and a nutritional rehabilitationcentre for malnourished children. Routine minorsurgical procedures, such as treating abscesses,are performed, and the local MINSA staff havethe experience and skills to deal with traumacases where immediate intervention is needed

to save lives. When transport is available,complicated or serious cases arc evacuated tothe privately run Hospital Chabungo in Cubal50km away, or to Benguela, which is a journeyof 200km. Suspected cases of tuberculosis arealso transferred to Cubal, where they mustremain as patients for the required months oftreatment. Although the hospital maternity staffare female, most other staff members at alllevels are male. The hospital's laboratory wasdestroyed in 1992 and has not functionedsince.31 There is a day and night 'casualty'service for urgent cases of all ages, andoutpatient clinics for adults are held eachmorning — patients arrive at dawn in order toregister. An 'ambulatory treatment' serviceprovides daily follow-up doses of medicines/injections after initial consultations, but patientsdo not hold their own treatment cards.Approximately 3,500 outpatients are seen eachmonth, about half of whom attend as out-of-hours or emergency patients. The hospital israrely full with inpatients. Walking to thehospital from the peripheral bairros takes abouttwo hours and involves crossing a gully andnegotiating potholed paths.

PolyclinicThe polyclinic Centro de 26 de Julho is situ-ated lkm across the town from the state-run hospital. It houses the Public Healthdepartment, the child health clinic, the centralpharmacy and the Municipal Healthdepartment offices. Children under 15 years ofage are seen at the polyclinic which opens onlyon weekday mornings. The one 'fixed'vaccination post for the Ganda Municipality isbased at the polyclinic, where all vaccines areadministered, including anti-tetanus jabs forpregnant women, although ante-natal clinicsare held at the hospital.

The MINSA public health department isresponsible for co-ordinating health-educationactivities at central and community levels; theirstaff give daily health-education talks to theassembled patients before consultations at allfacilities. Themes include the care of water andthe importance of boiling water for drinking,correct care of food and personal hygiene,cholera, the use of oral rehydration solution,and vaccination. After this, mothers go throughweighing, vaccination, consultation, andtreatment. First doses of treatment are usuallygiven on site, and pharmacy staff areresponsible for demonstrating how to preparesubsequent doses.

66

Appendix 6: health service providers in Ganda

MINSA health posts and health promotersThere are MINSA health posts in the rural bairroof Atuque and in the communes of Babaera andAlto Catumbela, and small health posts in theoutlying villages of Chacuma and Tchimboa.Health posts offer curative services, and AltoCatumbela has a small maternity department atwhich ante-natal clinics are held. The supply ofessential medicines from the central pharmacyin Ganda to peripheral health posts depends onwhether enough has been received to cover thehospital and polyclinic services.

There are, on record, 11 MINSA health prom-oters, most of them men who were involved inlocal training courses undertaken by 1CRC in1994/95. (When the original concept of primaryhealth-care promotion at community level hadtaken its form in the national health-promoternetwork, the number was double. Several of theremaining promoters had been recruited andtrained by MINSA seven years before; otherswere recruited through the bairro authorities.)After the ICRC training, the promoters wereprovided with printed health-education materialsand a small kit of essential drugs. They workedvoluntarily, and their activities were monitoredthrough monthly reports and follow-uptraining sessions. Although support from ICRCended when the organisation withdrew fromGanda in 1996, the health promoters are official-ly reported to continue providing some servicesat community level, and to play 'an importantrole in raising awareness and giving hygiene-education lectures in schools, churches, and atParty committee buildings.' In addition to theirroles as educators, they were also trained to treatcommon conditions such as malaria, worms,anaemia, and minor injuries in the community.More serious cases are referred to the healthfacilities. However, the few health promoterswho remain in the health sector at presentappear to be seconded to work at the hospitaland polyclinic facilities, and all levels of healthstaff say that the MINSA health-promoterprogramme has failed because of 1 ack ofmaterial support and incentives.

Traditional midwife (TM)A state-recognised TM programme was startedin Ganda in 1989 following an initiative by thenational health department to expand primary-level health services. Province-wide trainingseminars paved the way for a local midwifetrainer to recruit women from most of the hairrocommunities. Most of them were chosen throughthe OMA system with the support of the bairro

authorities. The TMs underwent three monthsof formal training to learn how to supervise safehome deliveries, following standard, Western-style hygienic procedures, to recognise compli-cations arising during childbirth, and to refer suchproblems to the hospital maternity department.The training was based at the hospital andcomprised lectures in Umbundu and practicalsessions. The women were provided with basicbirth kits containing scissors, aprons, and so onfrom UNICEFand OMA, supported with othermaterial incentives. The small number (six) ofTMs who continue to work on a 'voluntary' basisreport their monthly statistics to the maternitydepartment; in return, they may receive soapand occasionally disposable gloves. They do notexpect financial reward from patients, but mayaccept payment in kind. In general, TMs onlyattend complicated deliveries and do not offerante-natal care services, but some haveknowledge of herbal remedies for treatment offertility-related problems.

Community services:environmental health and sanitationTeams of women, paid a small monthly salary bythe Municipality, work with the CommunityServices to sweep the streets of the town. Menfollow to collect the piles of rubbish, or work inthe municipal gardens and lido. The large pilesof rubbish that accumulate in the market placeare collected less regularly, which causes discon-tent among the market traders. The centralwater supply is regulated by Community Services;the main pipelines and system have sufferedfrom deliberate destruction and neglect, somost residents of the town and parts of the semi-rural bairros rely on intermittent daily suppliesto public tap-stands. The town's sewage systemwas also destroyed, although some buildingsretain septic tanks which at present lack regularmaintenance. The Community Services Depart-ment is mandated to carry out sanitaryinspections at bairro level.

Church-relatedThe Catholic and Evangelical (IESA) mission runtwo private, fee-charging health posts staffed bytrained personnel in one semi-rural bairro. Theywork under the auspices of the MunicipalHealth Department and submit regular reportsof service activity. The health posts charge fixedprices, equivalent to several kilograms of maize,for consultation and treatments. The moneyearned goes towards the purchase of medicinesand materials. Payment for consultation may be

67


received in kind, or, in the case of theEvangelical mission, may be deferred until sucha time when the patient is able to pay. Curativeservices are provided with consultations, dress-ings and injections, although neither post hasin-patient facilities. Diagnostic equipment isbasic, but the IESA staff have access to a simpleinstrument to analyse blood, which is promi-nently displayed in the consulting room.Complete oral treatments are dispensed at thetime of first consultation, and patients returndaily for repeat injections as required. Clinicsare held during the mornings only, and neitherfacility offers community outreach or specificpreventive health-care activities. Individualcounselling takes place during consultations;for instance, patients presenting with sexuallytransmitted diseases are advised to trace theirsexual partners. Pregnant women and mothersof young children are told to attend MINSAhealth facilities for vaccinations, which are notprovided at the mission posts.

Independent piivate clinicSituated in a semi-rural bairro is a private clinicrun by a mission-trained practitioner of morethan 20 years' experience. He works with twomembers of his family. The fees paid forconsultation and treatment services cover thecosts of buying medicines and materials.Because the clinic has official recognition fromMINSA, it has access to subsidised supplies.Prices charged vary according to treatmentsgiven, and are often calculated once treatment

is complete, but they are generally cheaper thanat church-run health posts. Those unable topay, such as orphans and the elderly poor, aretreated free, while others pay in kind or deferpayment. Regular activity reports are submittedto MINSA, and health-education talks areoccasionally given to assembled patients by staffbefore the day's work, but most advice is givenon an individual, ad hoc basis during consulta-tions. The practitioner focuses on diet as part oftreatment and recovery of strength after illness,and advises the use of herbal remedies whereappropriate to complement prescribed allo-pathic therapies. He gives detailed instructionsto the patient's carers.

Other agency support to health facilitiesThe hospital and polyclinic have recentlyreceived assistance from Accion Contra Fome torepair their buildings' damaged and deterior-ated physical infrastructure. They providedfurnishings and equipment, and currentlysupport the central MINSA facilities with twoexpatriate health personnel. ACF also supplythe essential drug stock for all the state-runhealth facilities in Ganda and have organised in-service training for facility-based local healthpersonnel, to cover elements of primary health-care such as vaccination, maternal and childhealth care, and management of common healthproblems. UNICEF and ORA International havesupplied drugs and supplemented equipmentin the past, and the former continues to providevaccines and medicine kits.

68

Appendix 7: Research-project schedule basedon intermediate objectives

Phase one (six weeks) Phase two (eight weeks) Phase three (14 weeks) Phase four (20 weeks)

• Preplanning and contactwitli local authorities

• Formulation ofjobdescriptions and localrecruitment of tworesearch assistants (RAs)

• Induction and initialtraining of RAs incommunication skills

• Discussion of researchactivities and data-collection framework

• Preparation of interimreport

• Training of RAs inresearch methods

• Fieldwork with initialinformation-gatheringand review

• Identification of centralissues and sources ofinformation

• Selection of focus groups

• Review of secondaryinformation

• Discussion and planningpopulation census withM1NSA public-healthstaff

• Preparation of interimreport

• Participation in strategicplanning

• Continued training ofRAs

• Carrying out matchedcase studies

• In-depth focus groupwork

• Interviews with keyinformants

• Review and analysisof information

• Population census

• Feedback to participantsand verification offindings

• Preparation of draftreport of researchmethodology

• Health-educationworkshop with RAs'participation

• Validation of findingsand interpretations '

• Documentation offindings and conclusions

• Continued trainingof RAs

• Presentation of reportto all interested parties

• Discussion andformulation ofrecommendations

• Writing up andtranslation of report

• Dissemination of finalreport

69

Notes

1 'Health-related behaviour' (HRB) includeshealth behaviour and health-seeking beha-viour — those actions taken to maintaingood health and to restore health when ill.

2 UNICEF(1997): 'Angola: Socio-EconomicIndicators Data Sheet'.

3 IUCN The World Conservation Union (1992)'Angola: Environment Status Quo Assess-ment Report'.

4 In 1998, Oxfam UK and Ireland becameOxfam GB and Oxfam Ireland.

5 'Health-care providers' includes those at alllevels who provide health services to thepopulation of Ganda.

6 Curative and preventive health-care services.7 'Preventive health care' encompasses the

prevention of ill health and death by con-trolling communicable diseases, maintain-ing mothers' and children's health (withimproved care before, during, and afterpregnancy), monitoring nutrition and breast-feeding, carrying out immunisation, andmaintaining young people's health.

8 Hubley, J Communicating Health: An actionguide to health education and health promotion,The MacMillan Press Ltd, 1993.

9 Pretty, Guijt, Thompson, and ScoonesParticipatory Learning and Action Handbook,International Institute for Environmentand Development, London, 1995.

10 Oxfam UK/I/MINSA population survey,Ganda 1997.

11 Vaughan JP and Morrow RH Manual ofEpidemiology for District Health Management,World Health Organisation, 1989.

12 The number of group participants ranged fromtwo to 17, most commonly from five to eight.

13 World Health Organisation: 'Guidelines forrapid appraisal to assess community healthneeds' (WHO/SHS/NHP/88.4).

14 Werner D and Bower B Aprendiendo a Prom-over la Salud, Fundacion Hesperian, 1984.

15 Kroeger A 'Anthropological and Socio-Medical Health Care Research in Develop-ing Countries', Soc. Sci Med., Vol.17. No. 3,pp. 147-161,1983.

16 ibid.17 A term commonly used to denote any health

practitioner; the Umbundu equivalent isotchimbanda.

18 This term refers to female relatives, friends,and neighbours who assist at home deliveries.

19 Macdonald, J Primary Health Care: Medicinein its Place, Earthscan Publications Ltd,London, 1992.

20 Under the present Angolan Land Law, sobasare permitted to do this with the consent ofthe municipal administration.

21 'First hunger' describes the period duringthe rainy season when food stocks aregenerally at their lowest while agriculturalactivity is at its most intense.

22 Literally, to scrape around for anything to eat.23 Translated as oxiuri in Portuguese.24 Pdssaro is the Portuguese word for bird: a

child having a convulsion is said to look likea flapping bird.

25 A pregnant woman may bathe her child twicein water that has remained overnight in a panthat had been used to prepare pirdo; shethen transfers a string that has been worn bythe child from her waist to the father's waistto take away the child's attention from her(Balombo, personal communication).

26 A similar practice in Balombo aims to bond asmall child with the new boyfriend of itsmother.

27 Health Systems Trust (1997): 'Communityhealth workers in South Africa: informationfor policy makers'.

28 ACF Ganda, personal communication.29 Oranga, HM and Nordberg, E 'Partici-

patory community based health informationsystems for rural communities' fromParticipatory Research in Health: Issues andExperiences, Zed Books, 1996.

30 The batuque is a drum; the saying remindspeople to listen to others.

31 ACF plans to open the laboratory to performessential diagnostic tests.

70

Health and Livelihoods in Rural Angola: a Participatory ... and Livelihoods in Rural... · Health and Livelihoods in Rural Angola: a Participatory Research Project Laura Habgood An

Documents