Contents - Previous - Next

RAPing in the periphery of Rio de Janeiro: Application of anthropological procedures to the assessment of programmes of nutrition and primary health care

This paper also reports the experience of training and helping health agents to apply RAP in their programmes in Brazil. Such experiences have not only validated my own work as an anthropologist, but have also put me in the midst of what is really going on in terms of health and nutrition programmes.

The RAP guidelines [1] were partially applied in a semi-rural community known as Jardim Primavera (Spring Garden) in the outskirts of Rio de Janeiro city, in Brazil, during a period from August to November 1987 and April to July 1988, with the work of two anthropologists and an undergraduate student in social sciences.

The objectives of the investigation were to study the primary health care systems, to understand their rationale and functioning, perceiving the points of views of the health care providers and the users themselves, but focusing was on client/providers' relations and, consequently, the nature of health-seeking behaviour in the region.

Based on the RAP guidelines, a series of interview routines were planned in order to elicit information on income, number of residents in a household, children, sex, age, health needs, strategies for attaining health services, concepts about health and illness, belief systems and social networks, through open-ended questions. We were basically planning an ethnographic overview of community, focusing on the themes of health and healing.

Fieldwork was divided into three main tasks:

• Observation and interviews in the two health posts that serve lower to medium income families: two doctors, one laboratory technician, three nurse's aides, five users and two administrators were interviewed. Waiting room situations and medical consultations were observed. Similar procedures were applied regarding three folk healers - two women and a man. Pharmacists and the staff of the three local drugstores were interviewed as well.

• Observations and interviews in households with children under five years old: 15 families were interviewed, 10 of which were low income while the other five were lower-middle and middle class.

• Formal interviews with administrative personnel of the municipality, the Catholic Church, and community leaders.

Research results

THE SETTING Jardim Primavera (spring garden) lies within a network of small neighbourhoods belonging to the municipality of Duque de Caxias, in a district known as Campos Eliseos. Until recently these neighbourhoods were part of small farms such as the one from which Jardim Primavera inherited its name. Agricultural production declined immediately after World War 11 and the population increased at a rapid pace because of the heavy flux of migrants from Northeast Brazil, escaping the poor conditions of Brazilian rural life and a badly organized economic infrastructure.

Former rural workers in search of industrial and commercial jobs occupied, as squatters, most of the green areas of Campos Eliseos. The district now houses 27 slums, five of which are located in Spring Garden alone. Although these squatters' zones are not as crowded or sorry looking as the ones that burgeon on the hillsides of Rio de Janeiro city, they are characterized by very poor sanitary conditions and totally lacking sewage and drainage systems. Some houses are made of brick, or a combination of brick and wood, and are built at a certain distance from each other. With space becoming a scarce commodity, front and back yards are small, there are a few fruit trees and no vegetable gardens, with the exception of the larger households outside the poorer squatters' areas. However, most homes have a few medicinal herbs and trees.

In this peri-urban community, the lack of and/or the inefficiency of basic urban services is apparent. Transportation problems are dramatic. Because roads are unpaved, the heavy and continual traffic is in a constant state of calamity. When it rains, no matter how lightly, people have to wade through thick mud in order to cross most streets. Moreover, the city bus system is, for the most part, privately owned, and only a few buses service the whole area. Even to travel relatively short distances people have to take two or three buses and wait for long intervals between each. A direct trip to Rio de Janeiro is too expensive for most people, and they resort to crowded and poorly maintained trains that are always running off schedule.

These problems affect the health care delivery system, because people need mass transportation to reach the two low-income health clinics. Even for those who live within walking distance, it is not always easy to traverse the muddy, flooded streets.

During the study period in early 1988, heavy rains totally flooded Jardim Primavera and other parts of Duque de Caxias, destroying homes, and leaving the population isolated. People resorted to canoes in order to go through streets that had literally become rivers. Serious illnesses afflicted the population including diarrhoea, dehydration, high fevers, whooping cough and pneumonia. Medicines and oral rehydration solution packets were sent as gifts to the flood-stricken local population and the leader of one of the Neighbourhood Associations dispensed them from her house. Public health officers and governmental people had to resort to helicopters.1

THE QUEST FOR HEALTH The quest for health in the community was equated to the quest for food. A female informant put it very simply: "Our basic health problem here is lack of food." Food, however, is a category that is not necessarily associated with "nutrition." Any food whatsoever is viewed as nutritious, with very little or no regard for content, such as vitamins, proteins, etc. Except for middle-class families, there is very little scientific knowledge about nutrition. "Food is food," said another woman, adding that, "whatever 1 can get my children will be good for them." To keep someone, especially a sick person, from eating one type of food rather than another struck most interviewees as irrelevant.

Food restrictions or food taboos were considered only in relation to certain categories of illnesses and in relation to herbal remedies that place them into "cold or hot" categories. Such categories exist throughout Latin America especially in lower income, indigenous areas2 and for more than half of the interviewed families. "My children are perfectly healthy," was the assertion of mothers of children who had chronic earaches, puffed-up bellies, chronic running noses and coughing, as well as recurrent acute diarrhoea. Mothers tended to worry more - and therefore treat as "illnesses" - health problems like asthma, bronchitis, vomiting, and "nervousness3."

In Brazil, except for communities that define themselves as ethnically apart - such as Indian tribal groups and some European-descent communities - and live in relative isolation, ethnic category is not as relevant as social class. To be Black in a slum does not carry any more stigma, nor is it any more indicative of a social phenomenon, than being White in a slum. The distinctive mark is living in the slum, i.e., belonging to the so-called lower classes. Moreover, there were no language barriers to serve as a distinctive mark of ethnicity. Therefore, all questions and observations dealing with the theme of ethnicity or language were excluded. That was a very clear point.

Even though the need for educational programmes became evident, and people agreed on this need, the infrastructural problems they faced seemed larger than the needs. They questioned the validity of educational programmes for people who worked a double shift, in the fields, in the factory and at home, with no spare time. Even if they had some spare time, some noted that it would be difficult to reach the places where such educational sessions could be held.

Answers to questions dealing with medication reflected an ambivalence that forced us to return several times to the same point, rephrasing questions and adding new ones. "Medicines" (remedios, in Portuguese) that came from industrialized pharmacies were still to be distinguished as allopathic or homeopathic, as homeopathy is popular in that region as well as in the city of Rio de Janeiro at large. In fact, one of the doctors at the Catholic sponsored clinic is a homeopathic doctor. These "medicines" are supplied through the market economy, i.e. they have to be bought. There are also the "home medicines" from gardens not mediated by cash and that belong to the exchange system of neighbourhood networks. One question that definitely had to be reworded and split into sections dealt with children's illnesses because, as already explained, the "most common illnesses" are not readily perceived as such in Brazil.3

Certain questions at first did not seem to make sense to mothers, such as "What can a woman do to have a healthy baby? " Women seemed puzzled by it, since no steps or precautions ever seemed necessary to have a "healthy baby." The concept of a "healthy baby" was thus shrouded in mystery, for any child who is able to survive only on breast milk and manioc flour with water has to be healthy. The concept of "health" is viewed as the lack of obvious or more serious illness. A healthy mother is supposed to give birth to and raise a healthy child. A woman is considered healthy when she is fit to be on her feet in the early morning, doing her chores. An example of this was when an older lady appeared at the clinic with clear symptoms of TB. When asked why she had not come to the doctor before her condition became so serious, she answered that she had felt "fine" until then, although she had coughed a lot, because she was able to perform her household chores. Only when she felt so dizzy and weak that she could not stand on her feet was it suggested that she should see a doctor.


1. The investigator and colleagues had to refrain from going to the area until the waters had receded and buses had started running again.

2. This idea rests on a conceptualization of diseases that ascribes illness to an imbalance of heat and cold in the body and holds that the restoration of health can be accomplished only through the restoration of the proper or 'normal' balances of these as 'hot' illnesses are treated with remedies defined as 'cold,' and vice versa. In Spring Garden this imbalance also involves "hot" and "cold" foods for some illnesses, mainly the ones perceived as directly related to the blood which can be "thin" = "weak" or "thick" = "too strong" = "hot." Starchy foods as well as certain specific herbs are given to people who need "tonics" for "thin blood."

3. Nervousness is a disease category that involves several symptoms such as lack of appetite, disturbed patterns of sleep (either sleepiness or insomnia), irritability, fatigue and lack of interest in just about anything.


1. Scrimshaw SCM, Hurtado E. Rapid assessment procedures for nutrition and primary health care. Anthropological approaches to improving programme effectiveness. Los Angeles: UCLA Latin American Center, 1987.

10. Application of rapid assessment procedures in the context of women's morbidity: Experiences of a non-government organization in India

Sample selection
The ethnographic methods
Computer analysis of data
Preliminary analysis of free listing and pile sort data

By Shubhada J. Kanani

Shubhada Kanani is affiliated with the Department of Foods and Nutrition, University of Baroda and with the Baroda Citizens' Council, Baroda, India.

This paper describes the use of RAP by a non-governmental organization in India to obtain perceptions of slum women regarding health and illness and to obtain data for quantitative surveys and intervention strategies. In addition to ethnographic interviews with women it used cards representing illnesses that the mothers sorted into groups as a basis for further questioning. RAP was used to achieve a heightened awareness of the viewpoint of women: it showed personnel how to document their experiences, it contributed toward a realistic formulation of education messages, and it helped to improve intervention design. The principle difficulty was finding and retaining adequately trained personnel for ethnographic research. On the basis of this experience the researchers recommend RAP to other NGOs in developing countries. - Eds.

MANY HEALTH AND nutrition programmes are introduced in communities before any culture-specific information has been obtained. Rapid assessment procedures (RAP) make possible the quick assessment of health beliefs and health-seeking behaviour involved in maintaining health and overcoming disease using modern as well as traditional health services [1]. RAP may also provide the framework for planning intervention programmes - the latter being a primary concern of most voluntary organizations. There is a need to document and understand the strengths and limitations of RAP as commonly used by voluntary organizations, especially in the context of health and nutrition.

The present paper describes the use of RAP by a non-governmental organization (NGO), Baroda Council in Baroda, India during the year 1989. The Council used ethnographic techniques with the following objectives:

1. To obtain traditional perceptions of slum women regarding their health, morbidity and patterns of treatment, with a focus on specific commonly experienced illnesses.

2. To obtain data useful for planning both quantitative data collection phase and intervention strategies.

3. To evaluate experiences of the Council in the application of RAP.

This paper focuses on the latter two objectives.


The present project is being implemented in two disadvantaged communities, slum A and slum G. having a total population of about 6,500. Muslim families are predominant in slum A (97%) and Hindu families are the majority in slum G (90%). Besides having the common denominators of poverty (average per capita monthly income Rs. 150 or US$ 8), overcrowded and unhygienic living conditions, and low literacy levels of women (less than 50% female literacy), these two slums were purposively selected as to permit comparison of data between two different ethnic groups.

Sample selection

Since the primary aim of RAP was to obtain folk perceptions of women's morbidity, women subjects were selected, so that:

1. they were likely to have experienced, or heard about, women's illnesses arising from marriage and motherhood;

2. they proportionally represented Hindu and Muslim families at different sites in the two slums.

Thus, married women, in the age group 20-50 years with at least one child, participated in:

1. Focus group discussions (19 discussions; average group size 15 women).
2. Free listing and pile sorting (60 women).
3. Ethnographic interviews (50 women).
4. Narratives (50 women).
5. Key informant interviews (2 indigenous practitioners and 4 Traditional Birth Attendants [TBAs]).

The ethnographic methods

Focus group discussions


1. To build rapport with women
2. To obtain the general framework of women's morbidity - types. etiology, treatment


Initial focus groups were 'naturally forming' groups as curious women joined their neighbours who were discussing their health problems with the investigators. Later, the groups were systematized by selecting six to eight women per group: young mothers 20-30 years old or older women, 40+ years from a selected neighbourhood. The average size of the group, however, invariably increased to 15-18 women. Each meeting was conducted by a pair of investigators: a facilitator and a recorder. The tape recorder was deliberately not used so as to allay the suspicions of women who, at that time, were unfamiliar with the project team. The tape recorder was used later to record individual interviews.

Free listing and pile sorting


1. To elicit information from subjects on the range of women's illnesses, including local terms used to describe each illness and associated symptoms.
2. To understand how women categorize different illnesses.


Each woman was asked, in different ways, the names of illnesses common in their area and their associated symptoms. From the lists generated, frequently cited illnesses were pictorially represented on cards: one illness per card. The illness depicted on each card was explained and the women were asked to group the cards into piles (unconstrained pile sort). They were subsequently requested to explain the reasons for their grouping of illnesses. When women could not understand the concept of piles, the anchor-point clustering approach was used, i.e., women were asked to consider one card at a time and select other cards that went with the target card.

Key informant interviews with the health providers, i.e., traditional birth attendants (TBAs) and indigenous medical practitioners


To generate information about women's morbidity and health-seeking behaviour from health providers who have specialized knowledge about women's morbidity.


Once the 'popular' local TBAs and indigenous practitioners were identified through general observations of the project team and information had been provided by knowledgeable women, these key informants were then contacted and informally interviewed using an ethnographic guide. Interviews were tape recorded and later transcribed.

Ethnographic interviews with women


1. To obtain background data on the socioeconomic and political structure of community families.
2. To obtain women's perceptions about 'female physiology': adolescence, pregnancy, lactation.
3. To elicit detailed information about women's health-seeking behaviour.


With the help of an ethnographic guide, each subject was contacted in her house and asked open-ended questions. Probing was minimal and non-judgmental; efforts were made to obtain unbiased responses. Responses were tape recorded and later developed into expanded field notes.



1. To obtain a step-by-step folk description of a recent illness episode.

2. To provide the range of health-seeking behaviors pertaining to specific, commonly encountered women's illnesses, i.e., malaria, anemia, leucorrhoea, menstrual disorders, infertility.


The above health disorders were selected on the basis of free lists, ethnographic interviews and the Council's health centre records. From these records, from among women who sought treatment within the past three months, 10 women for each disorder (total 50) were identified. These subjects were then asked to describe the illness episode: its symptoms, causes, treatment and home remedies.


The project personnel have received practical training in conducting ethnographic research through periodic workshops organised by the supporting agency, the Ford Foundation. Further, standard reference books and manuals [1-3] were followed to guide the data collection process. A protocol developed by the Johns Hopkins University for investigating women's reproductive health in India facilitated the application of the methods.


This section focuses on the strengths and limitations of each of the five ethnographic methods described above. The contribution of each method towards deciding intervention strategies is highlighted in Table 1.

Focus group discussions

Focus group discussions were of immense help in building rapport with women; in providing the framework for their health problems in general and reproductive health problems in particular. Through the process of group dynamics, i.e., women encouraging their neighbours to describe their problems freely, considerable data emerged on women's morbidity, on local terms used, and perceived etiologies and treatment patterns. An important outcome of focus group meetings was the setting up of a Health Centre for women in each slum, which was a felt need. "Open a health centre for us and you will know all about our health problems" was a frequent comment from the women. Limitation of this method stemmed from the reluctance of women to impart certain information in a group setting, e.g., health problems carrying a social stigma (infertility) or treatment patterns resorted to privately (visits to traditional faith healer, "Bhagat/Bhuvas").

Free listing and pile sorting

Free listing elicited from women locally used terminology about their illnesses, which subsequently facilitated better communication with the respondents. Literate women responded more readily than illiterate ones. The majority of the illiterate women initially expressed ignorance about women's illnesses in their community, saying, "I am healthy ... I do not ask others about their problems..." However, to the question, "What illness makes you visit the doctor?" a list of a few illnesses was obtained, to which the women could subsequently add more.

Another difficulty encountered was the overlapping of 'illnesses' and 'symptoms.' An illness was often recognized only through its symptoms. Further probing in this case sometimes did not bring forth much new information, perhaps because the team shared the same cognitive understanding of symptoms as the women.

Table 1. Application of RAP in Baroda Citizens Council - an Overview


Purpose Served

Possible application of Information gained



Focus groups (90 minutes; for group of about 15 women)

Helped in building rapport with women

Helped locate in an informal manner" women leaders" who could help with project

Women would not talk about sensitive issues like infertility

Method useful in initial stages of project

Helped decide interventions

Helped decide priorities for subsequent research

Provided data on terminologies used in morbidity by women

Free Sorting and Pile Sorting (20 minutes for each)

Local description and categorization of disease symptoms obtained

Grouping of symptoms may help local doctors diagnose uncommunicative women

Women reluctant to talk about illnesses especially if they are currently healthy

Selected groups should tar get mothers 30+ years (likely to suffer morbidity)

Pile sorting (activity based) easier than free listing

Knowledge of local terms helped workers communicate better with women

Women reluctant to explain piles (seemed to lack vocabulary to explain what they know)

Skill required to probe without biasing responses

Key Informant Interviews (30 minutes for doctors, 45 minutes for TBAs)

Gave supportive data on women's morbidity; helped verify women's responses

Data may help in training TBAs

Doctors reluctant to reveal information about treatments prescribed (trade secrets)

Should invite participation from initial stages of project(local health workers)

Provided insight into possible reasons why certain practitioners preferred by women

TBAs sometimes gave contradictory responses

Ethnographic (45 minutes)

Provided cultural context of women's morbidity (i.e. family support, self esteem, etc.)
Provided folk perceptions of female physiology, pregnancy, lactation, etc.

Data may help stress broader socio- economic issues as context for women's health care strategies

More time consuming than other methods hence practical problems of cooperation lower coverage, etc.

Combined with direct observations, insights gained on women's health attitudes and particular health seeking behaviours

Narratives (detailed descriptions of an illness episode; 2-3 one hour sessions)

Yielded illuminating and in depth data on specific illnesses (etiological factors, symptoms, types of treatment): The method's sharp focus allowed validation of responses

Data may be useful for patient education programmes; for modifying health care since women talk about expectations from health providers

Limited in scope as only women presently having illness gave meaningful responses; severity of illness affects responses.

Types of behaviour vary greatly according to age of duration of interview, etc.

2-3 visits may be required for each illness episode

Pile sorting helped determine the degree to which different signs and symptoms were perceived by women to cluster together those symptoms that they themselves (or family members) had experienced, either simultaneously or one after another. Cause-effect relationships of illness - symptoms also emerged from the piles.


Fever - headache - body ache;
Leucorrhoea - backache - weakness

Visual aids helped the illiterate women to identify the health disorders. Though pile sorting, because it was an activity, was easier to administer than other interview-based methods, considerable explaining was required to enable women to understand the concept of pile grouping. Difficulty was also experienced in eliciting from women the reasons for the pile groupings made, perhaps because the women 'knew' but lacked the vocabulary to express themselves.

It is envisaged that the grouping of symptoms may help local health practitioners and the Council doctors to diagnose health disorders in the female patients who often are shy and uncommunicative.

Key informant interviews with health providers

The traditional birth attendants (TBAs) or "dies" contributed information mainly related to dietary dos and don'ts during pregnancy and lactation and their method of conducting deliveries. With regard to women's morbidity, their responses were confused and uncertain; further, they usually referred sick women to doctors and limited their advice to simple home remedies. The local doctors, on the other hand, responded with clarity regarding etiology and symptoms of disease but were reluctant to elaborate on specific treatment prescribed, saying, "It depends on the individual cases."

Both groups of key informants corroborated certain information obtained from women; e.g., perceived etiology, treatment patterns, home remedies and attitudes towards health.

Limitations of these key informants arose chiefly from:

• variations and contradictions in response, especially regarding home remedies, amongst TBAs;
• the reluctance of doctors to spare time for the investigators and to share their specialized knowledge regarding treatment they prescribed.

Ethnographic interviews

These interviews provided the contextual family information that helped us to partly understand women's health-seeking behaviour, e.g., the busy household routine and presence of young children being real constraints to seeking treatment; inherent apathy of several women towards their own health; lack of any aspiration for themselves; indifferent family (husband) support, and so on.

The interviews also provided an insight into folk perceptions of female physiology: adolescent development, various organs in the body; pregnancy, lactation; preferences of women for home deliveries and hospital deliveries, etc.

The drawbacks of this method stemmed mainly from some of the abstract questions asked, which most women could not answer meaningfully. Examples: quality of life (past and present), concept of positive health. Further, women could not answer some questions simply because they had never given a thought to themselves - their bodies, their health.

Data from these interviews can help emphasize the need to integrate other developmental (e.g education) and supportive (e.g., child care) programmes with health care services, which the present project is already attempting to do. Reasons for preferences of women for home deliveries and hospital deliveries will help the project team to design more convincing messages to promote hospital deliveries, which is feasible in an urban setting.


Narratives yielded illuminating data on specific illness episodes (leucorrhoea, menstrual disorders, infertility, malaria, anaemia): perceived etiology, progression of symptoms, sequence of health-seeking behaviour, preferred health practitioners, home remedies, constraints to seeking treatment, and other aspects. Narratives, by and large, confirmed the usual sequence of events known to health providers and, in addition, revealed interesting information on folk etiologies of disease (e.g., "excessive heat in body" causes disease), home remedies and nature of family support. Women's expectations from health providers regarding treatment and reasons for premature discontinuation of treatment were also highlighted.

Narratives, however, were very time-consuming (Table 1) and could be asked only of women currently undergoing treatment, as most women cured of an illness (despite its being recent) could not recall well the sequence of events or they denied ever having any illness.

This methodology can help design relevant patient education messages to overcome some perceived misconceptions about etiology and treatment of disease; it can enable medical practitioners to appreciate folk beliefs and the constraints of women seeking care.

Use of tape recorders in ethnographic interviews


1. Several significant pieces of information were available to us later that otherwise might have been missed.

2. The women felt quite important speaking into a recorder and their interest in answering questions was sustained.

3. The recorder facilitated a free flow of information between informant and investigator.


1. Busy staff were overwhelmed by the large amount of time consumed in transcribing the taped information into expanded field notes. For an NGO, with limited manpower and several time and resource constraints, this might prove to be a real handicap.

2. In this project it is intended to use the recorder on a subsample as a check on the quality of data.

Computer analysis of data

Generating reports through the computer became time-saving and meaningful after:

1. the field team learned the use of computers and personally fed the data into the computer;
2. the computer programmer spent a few weeks in the field meeting women and collecting data.

Contents - Previous - Next