Primary health workers in North East Brazil

0277-9536/93 $6.00 + 0.00 Copyright Q 1993 Pergamon Press Ltd

Sm. Sci. Med. Vol. 36, No. 6, pp. 775-782, 1993 Printed in Great Britain. All rights reserved

PRIMARY

HEALTH

WORKERS

IN NORTH

CHRISTEL STOCK-IWAMOTO’ ‘Justus-Liebig-University,

Population

Nutrition

Institute

and

EAST

BRAZIL

ROLF KORTE*~*

of Hygiene,

Friedrichstr. 2, W-6300 Giessen, Germany and Health, Division, Deutsche Gesellschaft fuer Technische Zusammenarbeit (GTZ) GmbH, Postfach 5180, 6203 Eschborn, Germany

Abstract--One

approach to providing primary health care in developing countries is to implement programs which rely on minimally trained primary health workers (PHWs). Originally, such programs were primarily designed for rural areas with intact social structures-where a community-based PHW could easily establish a link between the community and the conventional health services (delivery programs). However, faced with increasing migration of the rural population to urban centers in develooina countries, the present need for elementary health care in urban slums raises the question of whether PHWs can also provide health assistance to the population there. This oaoer deals with the issue of whether PHWs in North East Brazil can improve health care delivery and the’health situation of the poor population in rural and urban areas. The investigations covered both PHWs and the population. Questionnaire-guided interviews and participant observation were used as research tools. The conclusion is drawn that the PHW-program there has succeeded, if the PHWs have increased the coverage of the health service delivery and also the level of health-related knowledge, attitudes and practices (KAP) in the population-although with significant differences between rural and urban areas (higher coverage and higher KAP in rural areas). The quality of training and supervision and the characteristics of the area and local framework were also found to influence the quality of the work (performance) of the PHWs. 1

I

Keg words-primary

health

workers,

urban,

rural

INTRODUCTION

Since the middle of this century [l] and especially since the 1960’s [2] more and more non-professionals have been recruited to provide health assistance, especially in developing countries [3,4]. This approach was developed after it was realized that even extending the conventional (formal) basic health services could not achieve full coverage of the population. Community-based, intersectoral, preventive and promotive activities were therefore developed to complement the curative and rehabilitative services. Until then, the formal services had in fact, been providing mainly curative services to a favored, elite urban population group. The term Primary Health Care (PHC), describing a new concept which explicitly emphasizes the utilization of lay workers (also called Primary Health Workers-PHWs [S] was acknowledged by members of the WHO at the conference of Alma Ata in 1978 [6]. It was to significantly contribute to achieving the goal: Health for all by the year 2000. PHWs were to be the gateway for the population to access the health system (Fig. 1). They were also to take the service to the people through regular home visits. Health professionals would supervise the PHWs, who refer patients to them and consequently become an extension of the formal health system. *To whom correspondence addressed.

and reprint

requests

should

be

health

services

Most early PHC-programs were developed for and initiated in the countryside, because the need for health assistance had seemed to be more urgent there. Moreover, the active involvement of the communities, a principal concept of PHC, can usually be obtained more easily in rural conditions. Only few attempts have been made to apply the concept to poor quarters in urban areas-probably for the simple reason that it was assumed that the population there had access to conventional health services. In recent times, many people have speculated that they would be able to earn their living more easily in the city than in the countryside. A massive migration to the big metropoles is therefore underway. Yet, people’s hopes of improving their situation are often not fulfilled, resulting in growing poverty-belts in the big cities, where it is difficult to organize and maintain essential health services because of the lack of an intact social structure required for effective PHC-activities. In these cases, for example, the urban communities offer little support to the Primary Health Workers (PHWs) and the results are that many give up their job after only having worked for a short time [7]. This was only one of the problems emerging in PHW-programs. The officers responsible for such programs also had to decide which tasks PHWs were to undertake, where their limitations were, and how the selection, training, supervision and remuneration of PHWs should be organized [4, 8-121. Variable

CHRISTEL STOCK-IWAMOTO and __

Selection

ROLF KORTE

+--l

Population

Access

10 health cilrc

Fig. I. The PHW

as a link (interface)

Link

between

programs were designed to satisfy the different starting conditions. The main objective of our study into the PHW-programme in Sao Luis/North East Brazil was to look at two perspectives of the programme by thoroughly examining both the providers (the PHWs) and the clients (population) of the health care delivery system. One of the authors conducted most of the field work alone so as to get a direct and intimate view of the work of the PHWs. In this way interpersonal judgement errors were minimized. The principal question of the study was to find out whether or not the PHW program in Sao Luis/‘North East Brazil was able to improve the health care delivery. The study pursued the objectives of describing the actual life- and work-situation of the PHW in Sao Luis, setting up quality indicators of PHW work and identifying factors influencing the PHWs performance. Furthermore, the study intended to assess the potential impact of the PHW programme on the health situation in the poor quarters of Sao Luis and to identify problems impeding the progress of the PHW programme.

CONTEXT AND BACKGROUND

PHC in Brazil Brazil has been trying to resolve its health care problems for a long time. While efforts to reform the health system have been underway for more than 20 years, equal access to essential health services has not yet been achieved (according to WHO [13], in 1987, about 23% of the population had no access to conventional health services). Especially in underdeveloped regions like the North East, large sections of the population do not have access to adequate care in the case of disease, nor are the people informed about prevention. In this context, PHWs offer a potential for improving the delivery of health care to the population. Geographic context of the study Sao Luis, the island where the survey was conducted, belongs to Maranhao, one of Brazil’s 21 federal states. The city of Sao Luis, situated on the island of the same name, is the capital of the state. The Maranhao is in the region of the North East which is known for its poverty and underdevelopment. One million people (20% of the whole state’s

txtcn\1<,n

the population

and formal

health

system.

population) live on the island of Sao Luis--and about half of them in the city of Sao Luis. The infrastructure of the residential areas is far from ideal. Forty percent of the urban population dwell in the so called ‘belt of poverty’. Two of the quarters studied in this survey were in the city of Sao Luis. Two further study quarters were selected from the ‘interior’. They also illustrate the poverty and a poor health status of the population. In Maranhao, the infant mortality rate in 1983 was estimated to be approx. 15~-200 per 1000 live births [l6] and 3040% of the under-three’s were undernourished [14]. Diarrhea] diseases and upper respiratory tract infections show a higher incidence in the rainy season (January-June) than in the dry season [l5]. Diarrhea] diseases with dehydration contribute to 40-60% of all infant deaths in the Maranhao [ 161. Diarrhea is also closely related to malnutrition. the other main cause of infant death. Nutrition education of the population, the promotion of breast-feeding and Oral Rehydration Therapy (ORT) have positive influences in both cases. These tasks can be undertaken by non healthprofessionals. In 198384, it was therefore decided to implement a PHW-program in Sao Luis. UNICEF supported these efforts and together with local health authorities, set up the AISMIN program (Acoes Integradas de Saude Materno Infantil, i.e. integrated actions of mother-child-care). The German development organization GTZ also took part in the efforts to implement a PHC-program, through the PRODDIMA project (Programa de Doencas Diarreicas no Maranhao, i.e. Maranhao diarrhea] diseases program). The PHW-program was started in August 1984 in 4 sectors of Sao Luis. Fifty-three PHWs were trained. They were either selected by a community council or recruited after proposal by the local health centers. Although no definite criteria for their selection existed, most of them were between 20 and 60 years old, had children. could read and write. and had lived 8 or more years in the sector they were going to serve. The quality of basic training, refresher courses and supervision varied from sector to sector. The main tasks of the PHWs were almost identical to those of the GOBI-FFF-UNICEF-program [ 131:

I. health and nutrition

education/intervention. including promotion of breast feeding, monthly weighing of the under-threes, immunization status monitoring (preventive care);

Primary

health

workers

2. promotion and use of the ORT in cases of uncomplicated diarrhea (curative); 3. registration of census and health data (documentation). Up to June 1988, 400 mostly female PHWs had been trained; 337 of these were still registered in the AISMIN-program. They worked in 25 poor quarters of the island of Sao Luis. Their work was essentially volunteer, because plans for regular remuneration were never effectively implemented.

MATERIALS AND METHODS

The field study was carried out between January and June, 1988. Four study quarters, two rural and two urban ones, were selected. They all belonged to the PRODDIMA interventional areas because this assured easy access. The two urban study quarters were representative of poor quarters in cities of North East Brazil. Though health posts/centers often exist, people do not generally have access to well organized health institutions that are staffed around the clock. One of the selected urban quarters, Vila Palmeira, was the area where the first PHWs had been trained in 1984. Anjo da Guarda, the other quarter, was chosen because its program had just begun and, therefore, reflected the initial problems of PHW-implementation. The two rural areas, Sao Raimundo and Jose Camara, represented pilot study areas for the implementation of the PHW-program in the countryside. Both belong to the rural community of Sao Jose de Ribamar, situated about 30 km from Sao Luis. They differ from the urban areas in regards to their health-infrastructure and the living conditions of the people. In order to obtain representative results, all 83 PHWs working in the four study areas were selected to participate in the study. Fourteen PHWs, at least two from each quarter, selected by a process of random sampling, were accompanied during 4 days of participatory observation by one of the interviewers. Originally it was planned to include local interviewers in both observations and interviews. However, because of difficulties in recruiting suitable people, one of the authors performed the participatory observation and 80% of the interviews alone. This turned out to be an advantage, as standardized judgement was possible. A checklist had been prepared for the participatory observation after one of the authors had visited the study quarters several times. Observations focussed on the personality of the PHW, her social environment, her living conditions, behavior in health-related fields (e.g. hygiene, nutrition) and on problems related to her job, interaction with mothers and children during home visits and weighing, correct use of the Oral Rehydration

in North East Brazil

711

Solution (ORS) and correct documentation. At least 20 hr (distributed over 4 days) were spent with each of the 14 PHWs, including a day of weighing and a day of home visits. All 83 PHWs of the four quarters also took part in questionnaire guided interviews. The questionnaire had three parts. The first contained questions on age, marital status, children, level of education and profession. These aimed at obtaining data for possible variables relevant to the quality of work of the PHWs. The second part collected exact data on the number of families, children under 3, pregnant women, and for the preceding month, the number of home visits implemented, weighed children under 3 and referrals to the formal services. The third part was characterized by open-ended questions and gathered information about the work practices of the PHWs (weighing, health education etc.), their perception of training and supervision, and their ideas (suggestions) concerning possible program improvements. Actual knowledge of breast feeding and nutrition, growth monitoring and weighing, ORT and immunization was also tested. As a cross-check, 6.2% of the population of the study areas (255 households) were interviewed. These households belonged to two groups. The first group consisted of 98 families randomly selected from the census lists of the PHWs taking part in the participatory observation. The other 157 households were randomly selected from the population of the study areas. The household questionnaire was also in three parts. The first part contained questions about the number of family members, children, literacy, and about whether or not the household knew the PHW in the area and her responsibilities. Part two asked for details on the frequency of home visits by the PHW, the perception of her performance, wishes for additional health activities and also health related knowledge including breastfeeding, immunization, the meaning of regular weighing and ORT. In the third part, exact knowledge and skills in the preparation of an instant ORS and a home-made ORS were determined. The documentation records of the PHWs were used to validate the data obtained in the interviews with them. It consisted of: 1. The PHWs’ lists of households for the registration of all families in the designated area (Lista dos Chefes de Familia); 2. the registration sheets for each family, where each family member was listed, characterized and where the sanitary conditions of the family were described (Boletim de Cadastramento da Familia); 3. the record sheets for the PHWs activities with information on the number of home visits, referrals, weighed children etc. (Boletim de Atividade Diaria); and

778

CHRISTEL STOCK-IWAMOTO

4. growth charts (Cartao vaccination pass.

da Crianca),

including

a

The primary analysis of data obtained was carried out in Sao Luis at the end of the field study period. Detailed analysis and interpretation followed in Giessen at the University’s statistics center.

RESULTS

The participatory observation and interviews of both the PHWs and the households provided comprehensive information about the PHWs in Sao Luis and the population in their districts. 1. Participatory

observation

Most of the 14 PHWs accompanied in participatory observation were young housewives with several children who shared the inadequate sanitary conditions and poor housing of the people they were caring for (4 of the 14 did not even have use of a simple latrine, half of them did not have any bathroom). On average. their families only had a monthly income of between U.S.$50 and U.S.$150. This is why many of them were forced to work as houscmaids or unskilled workers, in addition to their housewife duties and work as a PHW. Working conditions differed considerably between the urban and rural quarters. In the city, where people have easy access to the formal medical system. the services of the PHWs were less in demand and less utilized than in the countryside, where the distance to the next health post prevented many people from seeking help. Feedback from the community in the form of consultation, achievements like successful management of a disease (e.g. ORT for diarrhoea), and growing expertise, achieved through frequent performance of the tasks. stabilize the PHWs’ selfconfidence. This also proved to be stronger in the rural areas than in the urban ones. The same applies to supervision. In the rural Sao Jose de Ribamar a nurse was available daily for problems, and also visited each PHW at least once a month. In the city quarters, the frequency and quality of supervision varied depending on the person and the supervisor. Some supervisors visited their PHWs monthly but the majority did so less often. Some had not performed their supervisory role for more than half a year and reportedly could not answer questions properly, nor did they help the PHWs in their documentation. 2. Interviews

qf‘ PHWs

and households

and ROLF KORTE

In the ideal case, offer (supply) and need (demand) should be equal (total overlap of the two circles), which would indicate 100% coverage of health services. However, this can hardly be achieved and consequently the real coverage is represented by the overlapping area of the two circles in Fig. 2 as an indicator of the quality of a PHW program. The health-related knowledge of PHWs and population as a second indicator should also become identical in the long run. Coverage. In this study, coverage, as an indicator of quality, was measured by the number of monthly home visits (month previous to the interview) of the PHWs-according to data obtained from the PHW.r and the population. Cotlerage measured in the PH Ws. In the rural areas, each PHW made at least I I home visits per month but the majority made more than 50 home visits. In the urban quarters, on the other hand, the majority of PHWs had not made even one visit and no PHW more than 50 visits. The data on the average number of home visits in the month previous to the interview is shown in Table I. The ratio between the average number of visits per month and the average number of families or children under three, registered in the PHW list. showed 0.6 visits per family (even I .3 in the rural areas) and I .4 per child under three (Table 2). Coverage measured in the population. The answers given in the household interviews on PHW visits confirmed the generally positive findings on high coverage (63% for the total sample) and also the differences between rural (78%) and urban areas (56%) (Table 3). In the urban quarters, the majority of respondents could not recall any visit by a PHW, whereas in the rural areas the majority had even mentioned a visit during the last 4 weeks. Kno\taledge. Health-related knowledge was tested in the PHWs and the population. It indicates the results of the positive influence of the PHWs’ education. Knowledge ofrhe PHWs. The test of PHWs permitted a maximum score of 45 points for knowledge. KNOWLEDGE

KNOWLEDGE +

-

COVERAGE

(population)

In this study, both sides of the health care system, the providers (PHW) and the clients (the population) were investigated. Through interviews, data were obtained on health care coverage and health related knowledge, chosen as the main indicators reflecting the quality of the PHW program in Sao Luis.

Fig. 2. Supply (PHWs) and Need (population) .ren:ices.

of healrh

Primary

health

workers

in North

East

779

Brazil

Table I. Average number of visits/PHW in the different areas (refering to preceding month: answers of all 83 PHWs) Rural areas S. Raimundo 77.1

Urban A. da Guarda

V. Palm&a

82.9 I

8.1

II.1

I

Scores ranged from 19 (42%) to 42 points (93%) the average score was 30 (67%) with a standard deviation of 5.4 points. However, the results differed between the rural and the urban areas averages of 34.7 (77%) and 31.4 (69%) in the rural areas were higher than the 31.1 (69%) and 28.3 (63%) achieved in the urban ones. The differences between the study areas were statistically significant (P < 0.05). No correlation could be found between educational level and test knowledge. On the contrary, the PHWs with the highest scores (in the rural quarter of Sao Raimundo) usually had just a primary education and were housewives. There was a correlation between the number of visits implemented in the preceding month and the test score (P < 0.05). It confirmed the assumption made before that the number of home visits can be regarded as a measure of motivation, initiative and interest of the PHW, and that these factors influence retention of knowledge. With regard to the different topics in the test, it was found that the PHWs scored best in prenatal care; between 94% and 98% of the PHWs knew the answers to the questions of the prenatal prevention program, what to do in case of gestosis, and how to counsel a pregnant women on the precautions in pregnancy. The advantages of breastfeeding were also generally well known, especially the practical advantages [protects the baby from diseases (96%), is always ready and clean (78%) protects from malnutrition (74%) milk is always at the right temperature (66%), and is free of charge (52%)]. In growth monitoring, high scores were recorded concerning the necessity of regular weighing (78%) although the relation between nutrition, weight increase and health was understood by fewer PHWs (47%). The growth chart as a means to explain the importance of weight increase was only mentioned by 18%. of PHW home visits in the preceding the different study areas Visits/ family, month

overall average 30

i

10

80

Table 2. Frequency

areas

J. Camara

In the area of dehydration/diarrhea treatmentORT, 96% of the PHWs mentioned hollow eyes as a symptom of dehydration; 90% mentioned dry, wrinkled skin and 80% a sunken fontanelle. Over 95% of the PHWs could tell how to mix and use an oral rehydration solution (ORS), but only 54% had instructed mothers on how to use it and less still, 16%, thought that it was important to educate the mothers about diarrhea prevention e.g. by hygienic behaviour. Available vaccinations for immunization were also well known by the PHWs: poliomyelitis (94%), measles (93%) tuberculosis (88%) tetanus (82%), pertussis (81%) and diphtheria (70%). Knowledge of the population. The study found that the PHWs had conveyed their health-related knowledge well to the population. There were differences between the study areas in the sense that in the rural areas more correct answers were given. However, the variation was not as sharp as with the PHWs. A possible reason could be that the urban population has better access to the media and additional sources of information. After having divided the interviewed households into groups according to the time of the last home visit (‘never’, ‘only one visit for registration’, ‘visit more than one month ago’, ‘visit during the last month’), the group of households which were visited by a PHW in the previous month showed a higher level of knowledge than the average of all households. This seems to be due to the activities of the PHWs. The differences in household knowledge were statistically significant in the areas of infant nutrition/ breastfeeding (P < 0.002) and weighing (P < 0.001). Family knowledge scores. Breastfeeding as the ideal nutrition for the infant was mentioned by 53% of the whole sample (Fig. 3) 59% of the families with children under 3 (not shown in the figures below) and 68% of those visited by a PHW in the previous month (Fig. 4). In rural Sao Raimundo (very active PHWs) 71% were convinced of the superiority of breastmilk to formula nutrition and in urban Anjo da Guarda 66% (both results not shown in the figures below)

month in

Visits/ family, month

S. Raimundo J. Camara Average for rural areas

1.4 I.2 1.3

2.6 2.0 2.2

A. da Guarda V. Palm&a Average for urban areas

0.2 0.2 0.2

0.6 0.7 0.6

Total

0.6

1.4

Table 3. Percent coverage with PHW services according lation (number of respondents visited)

to popu-

Rural areas (n = 86)

Urban areas (n = 169)

Total (n = 255)

Never visited by PHW

22 (19)

44 (74)

36 (93)

Visited for registration Visited > I month ago Visited in preceding month = Coverage

5 15 58 18

9 18 29 56

8 17 39 64

(4) (13) (SO) (67)

(16) (30) (49) (95)

(20) (43) (99) (162)

CHKlSTtL STOCK-~VAM~TO

70

r

-

r 60

I

40 30 20 10

-

0

Total

Rural

q

q

Formula

Fig. 3. Infant

Breast

mtlkitorm.

0

Only

nutrition---opinion of the rural population (% of respondents).

hrea\tmilk

and

urban

also gave similar responses. The latter result may be due to the initiative of a breastfeeding support group whose members visit and intensively support pregnant women and young mothers. With regard to monthly growth monitoring, 86% of the whole sample asked to give their opinion on weighing showed that they had understood the importance (they answered that regular weighing provides information on whether or not the child increases weight and is consequently healthy). In the city quarters, the percentage was even higher (87 and 95%), probably because weighing had been practiced there for a longer time. By taking the percentage of children under 3 who had actually been weighed in the previous month, the answers of the population could be validated and a review made of how far the theoretical knowledge was put into practice. It was interesting to find that, on the whole. 76% of the children under 3 were actually weighed-but even though the theoretical knowledge had been higher in the urban areas, more children were in fact weighed in the rural areas (83 and 73%) than in the urban Vila Palmeira (74%). Considering the group of families with children under 3 that had been visited by a PHW in the preceding month, as many as 90% had had theil child weighed, and in the rural quarters even 9 1% and 94% compared with 83% and 90% in the urban ones. The respondents of the interviews also scored very well as to immunization-related knowledge. Ninety68

70

and ROLF KOKTE

Sl) 40 30 20

10 0 ., Rural

q

Cannot

Fig. 5. Knowledge

Urban

q

Only

ORS

and use of the ORS

[7

Both ORS

in the rural and

urban population (O/Uof respondents). three percent of the entire sample and 97% of those with children under 3 were convinced of the importance of vaccination. It was interesting to find the highest percentage in the urban quarter of Vila Palm&a, which reached 98%. By comparison, only 85% of the respondents in the rural Sao Raimundo said that vaccinations were essential. By checking the records of the immunization cards of the under 3 it was found that despite the theoretical knowledge on hand it was often not put into practice (same trend in all study areas). More than half of the children had been fully immunized, but about 22% did not even have an immunization chart and 26% had only been vaccmated sporadically. Theoretical and practical knowledge concerning the preparation and use of ORS was not so impressive (Fig. 5). Although two-thirds of the respondents could mix the instant ORS. only 13% could also prepare the homemade solution. The percentage was higher in the rural areas where 2 1% were familiar with the home-made ORS compared to 9% in the urban areas. Though not statistically significant, a higher percentage (approx. 10%) of respondents who were visited by a PHW during the precedent month could prepare both types of ORS (Fig. 6). DISCLSSION

In this study, different aspects of quality of the PHW-program in Sao Luis were examined. The data

bX

67

instant

Total

63

60 50 40 30 20 10 n

Rural

q

Formula

q

Urban Breast

Fig. 4. Infant nutrition PHW

milk/form.

0

Only breastmilk

opinion of the families visited by a in the preceding month (% of respondents).

Urban

Rural

‘otal

q

Cannot

q

Only

instant

Total

ORS

0

Both

ORS

Fig. 6. Knowledge and use of ORS by the families visited by a PHW in the preceding month (36 of respondents).

Primary health workers in North obtained from the participant observation and the PHW and population survey demonstrate high PHW coverage rates as well as good health-related knowledge among PHWs and their clients, the population. However, considerable differences between the rural and urban study areas also became apparent. The rural/urban difference was perhaps the most important finding. Although health problems do not differ considerably between rural and urban poor quarters. Investigating the reasons for the impressive performance of the rural PHW’s work offered some clues on how to design successful PHW-programs in both urban and rural areas. In the rural areas of the Sao Luis program, not only several favorable conditions for the implementation of PHC were fulfilled, but the excellent training and support system offered by the small scale PRODDIMA project also enhanced success. As in other successful programs [3,4,8, 17, 181 it was found that there is not one single factor determining whether the activities of non-professionals achieve their objectives, but that the combination of several factors is important. Intensive training and supervision along with the support of a social homogeneous community and a limited target area (40-60 families)* were ideal conditions for high quality work by the PHWs (Sao Raimundo), whereas lack of support and communication, and social distance between PHWs and population (Anjo da Guarda) proved to be inhibiting. The assertion [3, p. 4431 that small-scale projects have a greater impact on the health situation of the population was also confirmed by the results of this survey. Higher coverage and knowledge were found in the PRODDIMA-supported rural areas. This probably resulted from manpower motivation) of the small project, whose staff trained the PHWs intensively, arranged frequent meetings and regular refresher courses. Nevertheless, from the experiences obtained in interviews and especially the participatory observation it was concluded that the local life situation (country or city) was the decisive factor determining the work conditions and quality of the PHWs. Between residents in the urban slums communication is usually limited to superficial contacts whereas in a rural community there is more unity and also exchange of advice when a problem arises. In Sao Jose de Ribamar, for example, where the next basic health post is about 5 km away, rural PHWs are often the only people who provide health assistance

*Responsibility for an excessive number of households 200 or even 1000, like the PHWs in India in the ’70s [20, p. 14781 are, of course beyond the capacity of a single voluntary worker and tend to paralyze initiative. Even about 100 families as in Jose Camara frustrated the PHWs as they could not fulfil their tasks carefully. tin comparison with attrition rates of 30% and more [6] this rate is low and indicates that, overall, the Sao Luis PHW-program is successful.

East Brazil

781

and are therefore consulted and utilized far more frequently than PHWs in urban districts. These contacts in rural areas increase the experience and successes of the PHWs and act as a stimulus and there were practically no drop-outs of PHWs in Sao Jose de Ribamar, whereas about 10% of the urban PHWs from Sao Luis stopped their activities within 4 years, although this also has something to do with a non-existing remuneration.? However, this issue was not further investigated in this study. Many of the PHWs who abandoned their jobs after a few months did so for financial reasons, indicating that this factor should be examined in more detail. A favorable factor was that ideal local conditions and a well-structured intensive training and supervision-system coincided in Ribamar, where the most successful PHWs worked. Other studies had reached the same conclusion that the quality of training and support influence in the PHWs’ performance [3, pp. 450, 51, 54, 57-58, 7, p. 525; 18, 19, p. 228; 21, p. 71; 22, p. 222; 23, 241. In the other areas, deficits similar to those demonstrated in the survey of Owuor-Omondi [24] were found. They were mostly related to the supervisors not frequently visiting their PHWs, assisting and correcting their activities on the job, and being available in case of problems. Insufficient knowledge and poor teaching experience of the tutors were other shortcomings which could be eliminated by better training of the supervisors. Although the results of the survey demonstrated that the PHWs generally perform well, their difficulties with the documentation system became apparent during the field study. These could also be resolved through improved training. Another important finding in this study was that the coverage (measured by the frequency of home visits) and knowledge of the PHWs and population were directly proportional and consequently demonstrate the effect of the PHWs’ work. Though they do not describe the impact or outcome of the program, they can provide valuable information as process indicators as has been shown in other studies [3, 17-191. High contact rates between PHWs and the population (compare Tables 2 and 3) especially in the rural areas (1.3 visits/family, month and 2.2 visits/child under 3 and month) can be considered a success of the program, especially when compared with the results of other programs, like that of the Narangwal project in India [ 191where one visit/month and family was observed. They reflect the PHWs contribution to the extension of health services and the growing health service coverage of the population. This is even more important as the PHWs take the service to the people rather than waiting for the people to come to them. Seeing the family in their daily environment, they also have many opportunities to influence prevention-oriented measures such as improving hygienic and sanitary conditions. Another advantage of

782

CHRISTEL STOCK-IWAMOTO and

home visits is that regular contact is maintained and health problems can be tackled at once. Serious diseases might be identified early enough to be referred to specialized services of the formal system*, provided the cooperation between the two is good. An interesting finding that differed from the results of Marchione’s survey [18] was that the number of home visits correlated with the score in the PHWs’ test. He had found that the better scoring PHWs had a tendency to neglect home visits and rather concentrated on documentation on theoretical jobs. In this study however, the highly-motivated PHWs who often had only had elementary level education showed the best results in the test and also visited homes more often than lower-scoring colleagues. This was probably due to the fact that the knowledge required in the test was identical to what had been taught and what they were actually performing on a daily basis. The test scores of the PHWs and the population indicate that through the PHW-program a fairly high level of health consciousness had been achieved in Sao Luis. Assuming that increased knowledge also leads to changes of attitude and behaviour, the health situation of the people should improve in the long run. With about two-thirds of the answers in the test correct, the PHWs of Sao Luis are in line with their peers in Jamaica [IS]. and even leave health workers in Indonesia behind [17]. Though the objective for the implementation of the PHW-program in Sao Luis was to reduce the high infant mortality rate, information about changes in this rate was not gathered in the present study. The same applies to more extensive investigations on whether the health-related knowledge of the PHW and the population does indeed influence attitudes and practices in this field. Further research on the reduction of morbidity in the PHW served areas would provide interesting information on the effectiveness of PHWs in both urban and rural areas and should be considered. For health planners the results of this study are highly relevant. With adequate training, supervision and motivation PHWs can perform well and influence knowledge, attitudes and practice in the community especially in rural areas. Similar services for urban populations are more difficult to achieve good results. In many circumstances, the deployment of PHWs may therefore be an effective tool to bring health services to the underprivileged at least in the short and medium term. Urban and peri-urban societies are, however, subject to a rapid change in service conditions and social values. This dynamic situation should be monitored closely to avoid upholding outdated services that are no longer fully accepted by the community. There may be a rapid trend to seek _..~ ___ _~~ *In the Narangwal study [19] only about 10% of all cases had to be referred.

R~LF KOR~F.

the highest possible level of health care resulting in high cost for the provider and the user unless precautions are taken. Cost effectiveness studies should accompany this development.

REFEHENCES

I. Enge K. Ecaluurion.

Health promutcr programs, Ministry of Health. Peru. Management Sciences for Hlth, Boston, 1984. 2. King M. Medical care in developing countries. A SyncD,o.yium fiw Makererr. Oxford Universitv Press. London. ‘1966. 3 Berman P. 6’1 ui. Community-based health workers: head start or false start toward health for all’? Sot,. Sci. Med. 25, 443 59, 1987. Newell K. W. (Ed.) Health hj, I/IC People. WHO, Geneva, 1975. World Health Organization. Fomutkr~fing .S/rure~qies./or Health for All hl, the Year 2000. WHO. Geneva. 1975. World *Health Organization. dlmu ,4r(r 197X: PrimcrrJ Health Core. WHO. Geneva, 1978. Harpham T. PI ul. Urban community health workers. In rhe Shador, of the City,. Chap. 13. Oxford University Press. Oxford, 1988. 8. Bender D. and Pitkin K. Bridging - - the pap: the village health worker as the cornerstone of the-primary health care model. Sot. .Si. Med. 24, 5 I?- 528. 19x7. 9. Fendall N. R. E. Au.vilic1rie.v irr Ncrrlrll Care. Progrums in Dewk@~~ C’ounrrie,s. The Johns Hopkins University

Press, Baltimore, lY73. IO. Morley D. Pediu/ric Prrorilrc.\

I I.

in rhc Dewlopmy

H&/h

Twm,

1973. K. Hcu/I/? Axrilirrrirc pp. 137. 17.5 (‘room Helm.

World.

Butterworths, London. Skeet M. and Elliott

und

/he

I.ondon.

1984.

I?. Werner D. 0 Agentc de Saude. S~~ude t/r 13. 14.

15. 16. 17. 18. 19.

p. 200. EP. Sao Paulo. 1984. Grant J. P. The SIU/C> of the

IVorid’.v

Comumdadc.

Children

198X.

Unicef. New York. 1988. Shrimpton R. Ecologia da desnutricao na mfancia, analise da evidencia de relacoes entre variaveis socioeconomicos e o estado nutritional. Serie Instrumentos para acao 3. Centro National de Recursos Humanos de IPEA;Unicef, 1986. Ypiranga L. Beziehungen zwischen dem Ernaehrungszustand von Kindern und dem Auftreten von Diarrhoe. Dissertation, Giessen. 1989. GTZ: Unofficial report from PRODDIMA 1983. Plano de Operacoes, PRODDIMA. 1983. Berman P. A. Village health workers in Java, Indonesia: coverage and equity. Sot. Sri. Med. 19, 411 422. 1984. Marchione T. Evaluating primary health care and nutrition programs in the context of national development. Sot. Sci. Med. 19, 2255235. 1984. Kielmann A. er ul. Child and Marernal Health Sert~ices in Rural India. The Narangwal Espcriment, Vol. 1. In/rgrated Nutrition and Health Cure. The Johns Hopkins

Universitv Press. Baltimore, 1983. R.-The community health volunteer scheme in India on evaluation. Sot. Sci. Med. 17. 1477.-1483. 1983. 21. Habicht J. P. Assurance of quality in the provision of primary medical care by non-professionals. Sot Sci. Med. 13B, 67-75, 1979. K. Will primary health care efforts be 22. Heggenhougen allowed to succeed? Sot. Sci. Med. 19, 217.-224. 1984. 23. Mueller F. Rural health. Ideas und Action 145, 612,

20. Maru

1982.

24. Owuor-Omondi

L. et ~1. The Changing Role of Family Welfare Educators. National health status evaluation. monograph Ser. I. Ministry of Health, Gaborone. 1986.