COMMUNITY HEALTH WORKERS IN GAZANKULU

INTRODUCTION The aim of this paper is to give a brief description of the utilization of community health workers (CHWs) in Gazankulu. The general policy framework in which these health workers function is described briefly, and the way in which these workers function evaluated. This paper is intended as a modest contribution to evaluation research in the field of community health. DEPARTMENT OF HEALTH AND WELFARE


INTRODUCTION
The aim of this paper is to give a brief description of the utilization of com m unity health workers (CHW s) in G azankulu.The general policy fram ew ork in which these health workers function is described briefly, and the way in which these workers function evaluated.This paper is intended as a m odest contribution to evaluation research in the field of com m unity health.

THE GAZANKULU HEALTH SERVICE Geographical location of Gazankulu
G azankulu is one of the self-governing "national states" within the Republic of South Africa.Its territory is divided into four geographically separate blocks: the M hala area in the south-east between A cornhoek and Bushbuckridge; Ritavi 1 and 2 situated east of Tzaneen; and Giyani in the north-east.Giyani consists of three magisterial districts: Malamulele in the east, Giyani in the centre and Elim in the west.
At present the territory of G azankulu covers an estim ated 675000 hectares or 6 750 km2 -an area which will be increased slightly after consolidation.G azankulu shares borders with Venda to the north, Lebowa (both Ritavi 1 and 2 and M hala) and of course the RSA at various points.This division of its area into four blocks has a negative influence on the adm inistration of the territory and also on its health service delivery system.The de facto population of G azankulu was estimated in 1980 at 480000.The results of the 1985 census were not available at the time of writing.
•T h e research upon which this research is based, was undertaken w ith the aid o f a research grant from the University o f S o u th Africa. 1 also acknow ledge constructive criticism from my colleagues in the D epartm ent o f D evelopm ent A dm inistration and Politics.It should, however, be em phasized th a t the conclusions draw n and the opinions expressed are my ow n, and th a t none o f the abovem entioned persons or institutions are in any way responsible fo r any o f the shortcom ings from which this paper may suffer.
The annual rainfall is low (500-700 m m /year) and seasonal.An estim ated 80% of the rainfall occurs between the m onths of O ctober and April.The topography of the territory varies from mopani shrub lowveld in the eastern parts to hilly broken terrain in the Elim district.
From d ata compiled by the Rand A frikaans University (1983) it appears that the age profile of the population of G azankulu is relatively young.A pproxim ately 53% of the inhabitants are 15 years of age and younger -a fairly com m on trend in the Third World.

Health service structure in Gazankulu
The head office of the Gazankulu D epartm ent of H ealth and Welfare is in Giyani and is responsible for the organization and control of all health services in the national state.The territory is divided into a num ber of health wards.The hospital forms the nucleus of the health service in the health ward and is responsible for all levels of health care, including com m unity health.In each health ward a num ber of clinics are established to provide a more even distribution of health services to the local population.In the more remote areas visiting points by mobile units are also provided.The health services offered at such visiting points include m other and child care, im m unizations and other aspects of com m unity health, such as education on the im portance of proper sanitation, the necessity of pure water for domestic use, and nutrition. .
The organization of the health s e rv ic e s provided to the inhabitants of G a z a n k u lu can be shown graphically as follows:

OBJECTIVE OF THE STUDY OF COMMUNITY HEALTH WORKERS (CHWs)
The em ployment of CHW s may be regarded as a very im portant com ponent of prim ary health care.The question w hether this view is generally accepted borne out by the practice as observed Gazankulu.
Prim ary health care must be seen within the context of the definition of the c o n c e p t as defined in the declaration issued by the W orld H ealth O rganization (W H O , 1978): "Prim ary health care is essential health care based on practical, scientifically s o u n d and socially acceptable m ethods and technology made universally accessible to individuals and families in the com m unity through their full participation and at a cost th at the com m unity can afford to m aintain at every stage of their develop ment in the spirit of self-reliance and self-determination.It forms an integral part both of the country's health system, of which it is the central function and focus, and of the overall social and economic development of the com munity.It is the first level of contact of the individual, the family and the com m unity with the national health system bringing health care as close as possible to where people live and work, and constitutes the first element of a continuing health care process" (W HO 1978:3-4).From the above it is clear that primary health care forms an integral part of any program m e of com m unity develop ment which is aimed at improving the quality of life of the com m unity in general.It may also help significantly to eradicate indigence and to break the vicious circle of poverty in which Third World nations are caught up.
Health conditions in G azankulu bear close resemblance to that of other Third W orld countries.From interviews with people responsible for the introduction of CHW s to the G azankulu health service delivery system at that time (ca 1980) it is clear that they were heavily influenced by the thinking of the W H O as is reflected in the above definition.Introducing CHW s was an attem pt to make health care accessible to local com munities in ^B c.ankulu and at a cost that they could Tmord.
The em ployment of voluntary workers in health service delivery systems is generally regarded as one of the most im portant factors in the successful implem entation of a program m e of prim ary health care (W erner, 1977).In the "Blueprint of Objectives 1978 -83" issued by the D epartm ent of H ealth and Welfare in G azankulu provision was made for the creation of a new category of health worker, the com m unity health worker.This was done in an attem pt to bring health care within easy reach of every inhabitant of the territory.This intent was re-emphasized by the then M inister of Health (Rev.Kubayi), who said in her budget speech in 1984 "our policy still aims at im proving and expanding all health services with an increasing accent on com m unity health, and ensuring that the t ilation of G azankulu can afford and access to curative, preventive and lotive services at a cost th at everybody can afford, and within a reasonable distance of one's residence." In this paper it is m aintained that the com m unity health w orker has a positive influence on the health of a community.The aim with this research was to establish whether the above statem ent is valid or not.
After a study of the relevant literature, a num ber of visits were made to Gazankulu.Subsequently a questionnaire was drafted, in which use was made, inter alia, of the syllabus used in the training of com m unity health workers as well as of the duty sheet issued to each com m unity health worker when she commenced working in the village.1In the interviews with the respondents attention was given to degree to which particular health messages as well as other aspects of com m unity health have taken root or have been accepted by the community.

Choice o f communities
It was decided to restrict the survey to the Giyani area (in other words to the magisterial districts of M alamulele, Giyani and Elim) due to the long distances between the four blocks comprising G azankulu.The inclusion of the Ritavi 1 and 2 blocks as well as the M hala area would have required a budget far greater than that available for this research.
M ore im portant, however, was that some work has been done on the CHW s in the M hala area (H am m ond & Buch, 1984) and also on the care groups2 in the Elim district (Karlsson, 1983) while no such work has been done on the CHW s in the Giyani area.
The present and form er Directors of the Elim Care G roup Project were interviewed to gain a better understanding of com m unity health in G azankulu.In addition interviews were held with senior officials of the D epartm ent of Health and Welfare in Giyani.
After reconnoitring the field in this m anner, four villages were selected for the survey: one in the Malamulele district; one in the Giyani district; and two in the Elim district.These four villages were selected on the following grounds: a. M atiyani in the Malamulele district was selected because it is fairly remote and no CH W is working there.To summarize -a CHW was employed in three of the four villages under the supervision of the com m unity health nurse from the nearest hospital.M atiyani could therefore be regarded as a control group in testing some of the knowledge of the local village com munities of certain aspects of com m unity health.
Choice o f the sample H lomela and W ayeni are essentially resettlement villages.After consolidation M atiyani and Hlomela were surveyed by officials of the D epartm ent of Agriculture, and a num ber of people were resettledmostly from Venda -in the two areas.A map of M atiyani was obtained from the local magistrate, showing the numbered stands.By using a random sample num bering table, 33 out of 315 households5,6 were selected, giving a sample of 10,5%.It was relatively easy to trace the stand num bers which could be cross-checked against the stand cards issued to each household.In cases where a stand was vacant its neighbour on the right-hand side was selected.
In H lom ela it would appear that no numbers have been allocated to the stands.A rough map of the village was therefore draw n on which each household was numbered.The random sample numbering table was also used in this case giving a sample of 30 out of 145 households, thus a sample of 21%.
In the case of Riverplaats and Wayeni a different m ethod was used to draw a sample.One reason was that Riverplaats was not surveyed and that although Wayeni is also a resettlement village, it was difficult to distinguish separate stands.
The method used to obtain a sample in Riverplaats was based on a method first employed by Kok & Kwamanga (1982) and subsequently used in a survey in the Elim district by Ijsselmuiden.Based on these assum ptions the following procedure was adopted.A cum ulative list was compiled of all the children attending Sub-Standard A at Wayeni and Riverplaats.
Subsequently the principals of the prim ary schools of Riverplaats and Wayeni were visited and permission obtained from them to draw a sample from the Sub-Standard A classes.Using these class lists the names of five children were selected, giving five clusters of five children each in the two villages.Each child was then asked to lead the researcher to his household.This house was excluded from the survey because it had "a Sub A school child" (Ijsselmuiden).Five households in close proxim ity to this household were then selected for the survey and the head woman in each household was interviewed.In this m anner 21 out of 361 households (5,8%) were selected in Riverplaats and 29 out of 225 (12,9%) households in Wayeni.
The following observations on the application of this sampling technique can be made: a.The exclusion of a household with a child in Sub-Standard A has some validity but does not apply in all cases.b.An even spread of households throughout the village was obtained.c.Selecting a house in close proximity to the first one did present problems.In Riverplaats the village is situated in a narrow valley on two ridges and obtaining a household close to the first one was quite difficult -sometimes stretching to a distance of 200 metres or more.In the case of Wayeni although there were one or two "straight" streets, households were unevenly spaced and it was quite difficult to obtain a cluster of five households in close proxim ity to one another.A young man who had recently completed his m atriculation exam ination was employed as a research assistant.All interviews were conducted with his assistance.Permission to conduct interviews was usually obtained from the senior wife; only in one instance was such permission refused.A week was spent in each village and interviews were also conducted with the local headmen, shopkeepers and the CHW s concerned.

COMMUNITY HEALTH WORKERS Background
The G azankulu " Blueprint of Objectives 1978 -83", to which reference has already been made, provided for the "creation of a new category of health worker, the Com m unity Health W orker: to be chosen by a com m unity, trained by the health services (but needed a minimum qualification of Std. 6) and paid by the health services.The Com m unity Health W orker would perform mainly preventive tasks in his/her village." Each com m unity health nurse was issued with instructions concerning CHW s in which it was stipulated that a CH W should be appointed at a public meeting to be held in the village.W hat happened in practice, however, was that the local headm an often appointed a relative (Karlsson 1983:728;H am m ond & Buch 1984:6).H am m ond & Buch (1984:6) even reported a case where the popularly-elected candidate was still waiting to be sent for training when a relative of the local headm an arrived back in the village as the trained CHW .
The instructions also required th at the CHW should be a m ature person, older than 27 and have at least Std VI.To this Karlsson (1983:729) added th at she should be a member of the local com m unity and have a house which she could use as a base for the execution of her duties.
CHW s were trained partly at the Nkhensani H ospital (6-10 weeks) and partly at the M pham bo Health Centre.Karlsson6 (1983:729), however, mentions a training period of 6-8 m onths of which 4-6 weeks was spent at the hospital.It is not clear w hat the actual length of training was -from inform ation obtained during the research it w'ould appear as if the firstmentioned period is the correct one.
Only three groups, each consisting of 12 trainees, were trained.According to Karlsson (1983:728) one of the CHW s never started to work, two left the service because their husbands moved out of the com m unity where they were placed, 8 elected to work in a clinic and not in a com munity and in four cases the local communities were not satisfied with the behaviour of the CHW s in their households and also dissatisfied with the general perform ance of their duties.
The training was stopped because a suitable person could not be found to conduct it and also because the G azankulu D epartm ent of H ealth and Welfare is currently rethinking the concept of CHW s and their role in health service delivery.It could not be established exactly why the D epartm ent stopped the program m e and answers to questions on this were rather vague.
In their evaluation of the role and functions of CHW s Hamm ond & Buch (1984:8) make the point that the person training these people should make use of an "appropriate adult educational approach".To this the authors added that these people should also have experience in com m unity work (the general principles of com m unity development) prim ary health care and adult education.It is however, unlikely that persons meeting these requirem ents will easily be found in an area where there is already a shortage of professionally trained people.However, my own survey should be seen as an attem pt to evaluate the system and propose suitable alternatives.H amm ond & Buch (1984:6) on the other hand found that in the M hala area the CHW s were often young women, frequently in their teens.At least in the Elim Health W ard this was not the case as well as was the case in one or two instances in the M alamulele and Nkhensani Health Wards.As a rule CHW s were placed in more remote rural areas with no other health facility nearby.

Training syllabus
The syllabus for the training of CHW s is prescribed by the SA Nursing Council.According to Karlsson (1983:729) there are two reasons for this, firstly that this formal recognition accords CHW s a legal status as part of the nursing profession which is im portant in the execution of their duties, and secondly because registration creates a professional career structure ensuring a measure of professional discipline.We shall return to this aspect later on in this paper.
The syllabus includes the following: ™ ethics, basic nursing (including anatomy), first aid; nutrition, hygiene (personal and environm ental); prenatal and antenatal care (which includes a com pulsory period of service in a labour ward); community development; health education and com m unication; and chronic diseases.
The practical work during training includes the following: gardening; the conserving of energy through the use of wonder-boxes and mud-stoves; drawing a plan of a village; com m unity diagnosis, home visits, care groups, drafting reports and home economics.
The guidelines for the presentation of the syllabus described the course's aim as follows: to train the CH W "to cope with ELEM EN TA RY prim ary health care in her own com m unity" (Syllabus, 1981).For this reason the training should be "com m unity based and orientated" and focussed on "prom otion of health and prevention of diseases".
To achieve this aim CHW s were encouraged to give priority to the following aspects (Syllabus, 1981) She should also keep her household and its surroundings clean, take care with her personal appearance and look after her uniform.

Evaluation of the syllabus
In a letter to the form er Secretary of Health and Welfare, Dr Erica Sutter (then D irector of the Elim Care G roup Project), com mented on the three training course of CHWs.
Her observations can be divided in two categories: the evaluation of theoretical training at the Nkhensani Hospital; and the practical work at the M pham bo Health Centre.
The syllabus used at Nkhensani for the first group was prepared by the head office staff of the D epartm ent of Health and Welfare with very little input from the hospital staff.At that stage an advanced course for the training of nursing assistants was being offered at the Nkhensani Hospital and the training of a category of ^K ^lth workers at such a relatively low of expertise was foreign to them.Consequently the hospital staff was uncertain as to what exactly was expected of them.
The whole issue was subsequently discussed with the hospital tutors and the syllabus for the training of the next group was amended accordingly.According to Sutter (1982) this new syllabus was better adapted to the situation prevailing in G azankulu at that time and gave the students a better insight into the functioning of the hospital as well as training them in basic nursing practice.
The third group was less fortunate as their syllabus got lost (Sutter, 1982).It is not clear why the previous syllabus was not used.The students were therefore only trained in sophisticated hospital care.
The practical training at the M pham bo Health Centre was handled by staff from head office of the D epartm ent of Health Welfare.Before they started with the ^^rc tic a l training the syllabus was discussed intensively with those members allocated to the training programme.The general feeling am ong the staff was that the lectures were quite adequate (Sutter, 1982).However, there was one problem, namely that the head office staff were not able to spend sufficient time with the group (it should be borne in mind that M pham bo is 30 km from Giyani).As a result the students were left to their own devices in the afternoons.
A nother problem was that the person made responsible for practical training had insufficient experience especially with regard to adult education and the organization of a training course.The staff attached to the Health Centre were either unwilling or incapable to participate in the training sessions, especially where this involved accom panying students on home visits.The fact that there was such a heavy reliance on local staff led to the em ployment of training personnel with insufficient experience and consequently the lowering of the general standard of training.T utors were also incapable of teaching students the skills and principles of com m unity development because they themselves had no experience of it.
One of the most im portant technical problems was that the transport facilities at the health centre were inadequate to meet the requirem ents of the students.Consequently no visits could be made to households at any distance from the Health Centre.Home visits were therefore restricted to com munities close to the Health Centre.These households complained that too many visits were paid to them while the more outlying com munites com plained of being neglected.
One very im portant problem was that the D irector of the Elim Care G roup Project lived at the Elim Hospital -80 km by a poor dirt road from the Health Centre.It became impossible for her to give adequate supervision and guidance to the programme.D uring this time the Secretary of Health and Welfare left Giyani to take up a position elsewhere and this was a further negative influence on the development of the programme.Frequent turnover of staff at head office also had an adverse effect on the program m e's development.

Functioning of the community health workers
It is expected that CHW s will undertake regular surveys of the com m unity in which they live, by means of regular home visits.The duty sheet, however, failed to stipulate how many visits should be carried out in the course of a norm al working day.D uring a home visit the CHW is expected to be alert to any health problem needing urgent medical care, such as pregnancies; the health status of children under five years of age; whether there is a toilet and a refuse pit; the general condition of the household (environm ental health); the presence of chronically ill people; and whether TB and psychiatric patients are taking their medication regularly.
Where necessary the CHW should give advice on the use of the oral rehydration solution for the treatm ent of diarrhoea, elementary first-aid, hygiene, and m other and child care.G azankulu's clinics set aside one day per week for child care: this is generally known as the child health day.It is expected that on this day CHW s will assist the nurse in charge of the underfives clinic, usually with weighing and immunizations.
The CHW s should assist families to establish a vegetable garden, build mudstoves, and to use wonder-boxes.It is also expected that the CH W will work hand-inhand with the local care group, and if such a group does not exist in the village to assist the women of the village to establish one.Finally it is also expected that the CHW will work closely with local leaders, such as the headm an and church leaders.
On appointm ent every CH W is given a duty sheet containing general instructions on her duties.In this docum ent she is adm onished "to be friendly to every one and give the best care you are able to give, regardless of the patient's status or beliefs".She is cautioned against the use of any nursing practice for which she has not been trained, and told that anything beyond her capability should be referred to the nearest clinic or hospital.She is also reminded of the im portance of professional confidentiality.Most im portant of all, she is warned to listen to the people before advising or teaching any person or group.
The CHW have prescribed working hours.It is expected that the CH W will report every m orning to the tribal office, or, where there is no tribal office, as is the case in Hlomela and Wayeni, to report to the local headm an when she starts working.Likewise she has to report to the tribal office or headm an when going home in the evening.This is to ensure that the CHW may be easily traced when needed, as when the com m unity health nurse or one of the local doctors is visiting the village.
Each CH W is issued with a kit containing the following: first aid, a Salter scale, Road-to-health charts; tem perature strips; aspirin (both ju n io r and adult); malaria tablets and cotton wool and swabs.She is also issued with an ordering book for her kit and stationery and she is expected to keep files for m onthly reports, daily reports, family care as well as a personal file.

Evaluation of the training of community health workers
A num ber of provisional remarks on the training of com m unity health workers may now be made though this aspect will be dealt with more fully further on.
First of all, the syllabus for the training of CHW s is by any standards a dem anding one, bearing in mind that CHW s are at par with the lowest grade of nurses, viz. the nursing assistant.It is difficult to see how this syllabus can be covered effectively in the short space of time allowed to it, namely 6-10 weeks.The question could also be asked w hether the minimum level of schooling required, Std 6, equips them sufficiently to master this course.
A second issue relates to the time actually spent doing practical work.Is there a sufficient balance between the theoretical work taught at Nkhensani Hospital and the practical work done at the M pham bo Health Centre?Or was the theory given preference to practical work?
A third issue relates to the duties expected of the CH W when she returns to her village after com pletion of the training.One could also form ulate this as a question: does the training of the CH W equip her adequately to handle the responsibilities attached to this job?O f the CHW s met during the field work few com manded sufficient English to cope with elementary questions.This should be seen against the fact that the medium of instruction used during training was English.This language is also used in duty sheets and other written instructions.Therefore their relatively low level of education reflects seriously on their ability to understand instructions given to them and to handle this job.

THE COMMUNITY HEALTH NURSE
The com m unity health nurse is responsible for the general supervision of the work of all CHW s working in a specific health ward.She has to prepare the local com m unity before the CH W returns from training.The com m unity health nurse should then introduce the CH W to the local headm an, other members of the tribal authority as well as to the com m unity and explain the function and duties of the CHW to them and what they may expect of her.
Each com m unity health nurse is issued with a duty sheet detailing her responsibilities in respect of the CHW s.It is recommended that she should also spend at least a week full-time in the field with each CHW .She must also assist the CHW in the planning of her daily or weekly routine and in determining local health priorities.Discussing this aspect of her work, one of the com munity health nurses said that although she initially tried to visit all the CHW s in her ward at least once a week, pressure of work and the long distances involved forced her to reduce this to a single monthly visit (Karlsson, 1983:728).Such a visit normally lasted from half-an-hour to an hour.In another health ward the com munity health nurse spent from two to three hours on a visit to the CHW s under her supervision.
D uring these visits, the records of the CHW s are checked and problems solved, the com m unity health nurse accompanying the CH W to households with specific problems.In addition she has to visit the CHW s own household ensuring this is setting a good example to the women in the com munity.An im portant aspect of the com m unity health nurse's responsibilities toward the CHW s is to assist them in the planning of their work, on a weekly, monthly and annual basis.
It is difficult to see how the com munity health nurse can achieve anything substantial in a visit lasting a half-an-hour, leaving her little time for in-service training and following-up of cases.To name one example -a single trip from Nkhensani Hospital to Hlomela takes an hour.A return trip would mean that the com munity health nurse would be away from her office for at least a full morning which on the other hand shortens the time for her norm al office duties.

THE SURVEY
As indicated earlier on the aim with this paper was to establish whether or not the CHW s has a beneficial influence on the com m unity in which she works.
As previously explained the survey involved the adm inistration of a question naire in the four villages, Hlomela, M atiyani, Riverplaats and Wayeni.The questionnaire was structured as follows: d ata on the family structure -question 1-7, 9-12.history of illnesses in family -question 8. d ata on nutrition -question 13-20.1.d ata on personal and environm ental health -question 22-27.2.
An observation schedule was also developed to measure certain practical measures which CHW s are supposed to teach to women in the community.These include the construction of toilets and refuse pits and the conservation of energy through the building of mud-stoves.

Family structure
The question was asked whether the husband lived at home or not.The response given was as follows: No These figures and the percentages in each case correspond with those obtained by Ijsselmuiden (1982:3), who obtained an absenteeism rate of 68% during a survey of the village of M ahatlane.Males living in these villages at the time of the survey were either locally employed (tribal or govern ment service), on leave, on retirement, at school and in one case the husband was mentally disturbed and unfit for employment.

Size o f families in the fo u r villages HLOMELA:
Average num ber of children per family:

2,8.
N um ber of children under 5 in the sample: 64 (92% of all children counted).

MATIYANI:
Average num ber of children per family: 3.6.
N umber of children under 5 in the sample: 43 (43% of all children counted).

RIVERPLAATS:
Average num ber of children per family: 3.6.
N umber of children under 5 in the sample: 18 (24% of all children counted).

WA YEN I:
Average num ber of children per family: 3,5.
N um ber of children under 5 in the sample: 26 (34% of all children counted).
These figures are much lower than those reported in the Rand A frikaans University Report (1983).It is impossible to explain this discrepancy as it is not clear what method the RAU Report used in their calculations.
In Hlomela 4 out of 30 families were polygamous; in Matiyani 6 out of 33; in Riverplaats 5 out of 21; and in Wayeni 5 out of 29.

A ction taken when members o f fam ily are ill
The reason for this question was to establish from respondents the num ber of children who became ill during the immediate period of two m onths preceding the survey, what action was taken in case of illness, what kinds of illnesses were mostly prevalent and whether the assis tance of the CH W was sought before remedial action was taken and what kind of action was taken.a. HLOMELA Ten children of respondents were currently ill in the village, 17 not and in three cases the question was not answered.In nine out of ten cases the m other sought assistance, and one treated the child herself.There was one case of pneum onia, five of influenza; two of diarrhoea; and one with an eye problem.Of these nine children, five were taken to the hospital, and four to the clinic -presumably the mobile clinic.In three cases people asked the C H W 's advice before taking action, five made their own decision and one was a member of the local care group (it is possible that she discussed the problems either with the CHW or with other members of the care group).

b. MATIYANI
In nine cases children were ill and in 24 not.None of the respondents treated the children themselves -two were taken to hospital, five to the clinic, one woman took her child to her church and in one case it is unknow n what form of action was taken.There were five cases of influenza, one of diarrhoea, one sore throat and possibly also kw ashiorkor and in two cases the cause of the illnesses were unknown.This village does not have a CHW , so there was no-one to give advice before action was taken.It should also be borne in mind that this village is 15 kilometres from the nearest clinic.

c. RIVERPLAATS
Nine children of the respondents were ill in this village and twelve not.Three of the children had a cough, three diarrhoea and three had influenza.None of the respondents treated her own child -four were taken to hospital, three to the clinic (possibly the mobile unit) one to a private doctor and in one case it was not clear what steps were taken to treat the child.

d. WAYENI
Only six children of the respondents w e rjA ill during the period under discussion, a n 23 not.Two suffered from diarrhoea, one from gastro-enteritis, one of influenza, and one had an eye problem and one a back problem.None of the respondents treated their own children, in three cases the child was taken to hospital, in two to the clinic and in one to a private doctor, in four cases the children were treated by a relative, in one by a member of the local care group and in one by the nursing sister at the nearest hospital.
To summarize -it is clear that most of the com plaints could be considered to be relatively m inor and therefore not in need of specialized medical care.The possible exceptions to this were the one case of pneum onia, the eye problems and the child who had a back com plaint.
O f the com plaints listed above, CHW s were only trained to treat diarrhoea.Their lack of adequate curative skills has led to all the problems of loss of credibility and 'prim itive' services . . .' (H am m ond & Buch 1984:12).
With the exception therefore of the cases of diarrhoea listed which could normally be treated in the early stages with an oral rehydration solution the CH W is not equipped to handle any m inor illnesses or emergencies.The first-aid kit issued to her is woefully inadequate.In the M hala region people went to the extent of offering the CH W the norm al clinic fee (R2,00) for treatm ent to save them the expense of travelling to the nearest hospital or clinic (H am m ond & Buch 1984:12).No incidents of the same nature were found during my research.
It should also be kept in mind that most of the CHW s live in remote villages, with little or no com m unication with the nearest hospital or clinic.In cases of real emergency patients have to be transported at risk to the local hospital and it is doubtful whether the CH W would be able to give first aid-treatm ent in the case of traum a.With the meagre stocks issued to her she would hardly be able to apply firstaid when necessary.
The frequent reference by respondents to the "clinic" in the case of Riverplaats and Ji'ayeni may be misleading.D uring the ^■ trview s it was observed that some respondents when using the word "clinic" were actually referring in some cases to the office of the CH W or to the mobile clinic.It was difficult to establish to which of these respondents were actually referring.Reference to "clinic" in the answers tabled should be viewed against this fact.

Breast-feeding
A large section of the questionnaire was devoted to the gathering of inform ation on nutrition.This inform ation will be used elsewhere but one aspect will be high lighted in this paper, namely breast feeding.This aspect was singled out because of the im portance of breast feeding in com bating m alnutrition in infants in the Third World.
The following inform ation was obtained: From the above results it is clear that breast-feeding does not constitute any problem in those villages.
Subsequently respondents were asked who gave them the advice to breast-feed and from their answers the following picture emerged: It is interesting to note [hat so few of the respondents actually said that they were advised on breast-feeding by the CHW (d); or a member of the local care group (e).The usual reply was "it is our custom ", or "we know that we must breast-feed our babies", hence the large num ber under (0-A few respondents gave no reply to this question (g).
Action taken when a child had a "loose stomach" The use of the oral rehydration solution (ORS) has been strongly advocated in rural areas to com bat diarrhoea and the large num ber of positive answers is indicative of the success of this campaign.
The logical question to ask then was where did they learn to prepare the ORS?The following replies were given in answer to this question.In Hlomela and Riverplaats the majority of respondents gave their children O RS in the case of diarrhoea.The fairly large num ber (18) in the case of Wayeni who said that they took their children to the clinic could be ascribed to a m isunderstanding of the word "clinic".It is most likely that they actually refer to the office of the CH W -as explained earlier on -and not to the nearest clinic at Gavaza.Although there is no CHW at M atiyani, the large num ber of people who knew how to use ORS ( 16) could be ascribed to the influence of the sisters at the clinic at Mhinga.Those who responded that they knew how to use ORS in the case of diarrhoea knew also in the m ajority of cases how actually to prepare the solution -a fact ham mered home by the clinic's sisters throughout G azankulu.W hat is interesting however, is that so few people (H lom ela 4 and Wayeni 1) actually said that they received this inform ation from the CHW .

Responsibility fo r administering medicine in case o f illness
The aim of this question was to establish whether the CH W was consulted in case of illness, bearing in mind th at one of her duties was surveillance of possible illnesses in the community.In both Riverplaats and Wayeni all the respondents said that they alone were responsible for adm inistering medicine to their children.In M atiyani a large num ber of respondents said it was the sister at the clinic at M hinga who administered medicine to their children and in Hlomela the majority did not respond to this question.However, it is doubtful whether the respondents actually understood this question even though it was explained over and over again and hence no valid conclusions could be derived from their answers.
Asked what action they themselves took in the case of illness the m ajority replied that they either went to the nearest clinic or hospital.

Environmental hygiene
In the following group of questions the existence and use of pit latrines were tested together with questions on the presence of flies as im portant vectors of com municable diseases.
Asked whether the household possessed a pit latrine or not, the following results were obtained: In the observation schedule (which results will be discussed further on) the actual existence and condition of pit latrines were checked by the author or his assistant.The following results were obtained: It is difficult to explain the difference between the answers obtained from the respondents and what was actually found.It can perhaps be attributed to the fact that the tribal authority is responsible for the enforcement of the digging of pit latrines in the villages.This fact was borne out by the replies given to the question: W ho told you to build a pit latrine?
The replies are reflected in the following table: As could be seen in the m ajority of cases respondents were advised (instructed?) by the local tribal authority (chief/ headman) to build a toilet, namely Hlomela (36,7%); Matiyani (54,5%); Riverplaats (33,3%); and Wayeni (31%).In the case of Hlomela (36,7%) and Wayeni (34,5%) the women interviewed said that they were advised to build a pit latrine by a member of the local care group.One could therefore argue that in the case of Hlomela and Wayeni the com m unity health w orker did have some influence in the erection of pit latrines through the local care groups.
Flies play an im portant role in the transm ission of diseases such as trachom a, diarrhoea, etc. and for this reason respondents were asked whether flies could be considered bad for one's health?From these answers it can be seen that the m ajority of respondents did have some knowledge about the effect of flies on health.In the following question they were asked why they were bad and again in the majority of cases the response was that they cause diseases.Asked where did they learn this, the following inform ation was obtained: It is clear that in the case of Hlomela and Wayeni this knowledge was gained through the local care group.Again it can be argued that the CH W did have some influence through the local care group in their respective villages.The large num ber of respondents at M atiyani who did not give a conclusive answer (18/33) but said they learned it "somewhere" refers to know ledge/inform ation passed in the village by word of mouth.O f the 18 three indicated that they learned it at school and one that she was advised by the local headman.
In the last question in this group respondents were asked what measures they took against contam ination of their food by flies, particularly when they were handling food.The replies can be reflected as follows: The fact that people indicated that they kept their food covered, was confirmed by observation during the interviews.
A final aspect under this heading is the existence of refuse pits.Inform ation on this was obtained through an observation schedule.The following results were obtained: In most of the cases these refuse pits were uncovered shallow holes in the ground.Only in a few instances were attem pts made to cover the holes with grass or something similar.Usually the refuse pits were ju st outside the homestead -sometimes at the back and sometimes on the side of the hut.The high incidence of uncovered refuse pits tended to give the villages an untidy unkem pt appearance.In Hlomela the CH W did have a refuse pitlarge and deep -but the refuse was left uncovered.

PersonaI hygiene
In this group of questions respondents were asked when they have to wash their hands and where they learned this.
When asked when they have to wash their hands the following replies were given: In the case of Hlomela 43,3% said that they keep their food covered or did nothing -the usual response being "we ju st sweep the flies away with our hands".In the case of M atiyani 54,5% said that they keep their food covered, while 45,5% did nothing.In the case of Riverplaats 66,7% kept their food covered and 33,3% did nothing.In the case of Wayeni 37,9% kept their food covered, while 61,2% said that they did nothing.
The results of Wayeni -where there is a CH W -is interesting, especially when it is com pared with that of M atiyani, where a larger percentage said that they keep their food covered than those who did nothing.As could be read from the replies the majority of people knew they had to wash their hands when handling food, during its preparation and also when eating.However, the num ber of children who ate or handled food with unwashed hands observed during interviews was quite high.
In the next question respondents were asked if they have to wash after relieving themselves.Their replies are reflected in the following table: Here again it could be argued th at the large num ber of persons who said that they received this inform ation through the local care groups (Hlom ela -16, Wayeni -9) was due to the work done by the CHW .The large num ber of people at M atiyani (22) who could not give any reply to this question could be indicative of the remoteness of this village and possibly also because there is neither a care group nor a CHW working in the village.

Knowledge o f one particular disease
In this group of questions the knowledge o f respondents about one particular disease was tested.M alaria was chosen because Gazankulu falls within the "m alaria belt" of the Transvaal and there is a high incidence of the disease particularly during the rainy season.The responses to these questions were as follows: The The few who knew the answer to this question had learned the inform ation from the hospital (2); the clinic (2); school (2); member of the care group and in the one case the CHW was the respondent and knew the correct answer.
The last question was that if there were mosquitoes in and around the homestead what did they do about it?The replies given can be reflected as follows: The large num ber of respondents who did nothing about mosquitoes are quite significant Hlomela (33,3%); Matiyani (48,5%); Riverplaats (67,7%); Wayeni (72,4%) bearing in mind the high incidence of the disease in that areas.The usual method of com bating the pest is by spraying an insecticide, but although many respondents said that they use Doom, when asked they did not have any in the house.It is regular practice that all the huts are sprayed by m alaria teams which made lack of knowledge about the disease even more rem arkable considering also the replies given to d above.O f these the results of Wayeni, where there is a CH W , is quite remarkable.It should be borne in mind that building a mud-stove is no easy task but still the relative absence of mud-stoves in the three villages where a CH W is working is rather difficult to explain.

CO NCLUSION
Earlier in this paper it was stated that the aim of this study was to establish whether the CHW has a positive influence on the t <h status of the local com m unity where iias been posted.It is expected of the CHW s that they encourage the construction of pit latrines.But it is also one of the duties of the tribal authorities to ensure a basic standard of environm ental hygiene in the villages.It is therefore difficult to establish how many pit latrines were built in a particular village since the CH W started to work there.In Hlomela (76,7%) of respondents did have a pit latrine, Riverplaats (71,4%) and Wayeni (79,3%), but in M atiyani where no CHW is working (72,7%) of the respon dents had pit latrines.It is therefore very difficult to state unequivocally that the CHW s were responsible for the high incidence of pit latrines in Hlomela, Riverplaats and Wayeni.
The same argum ent could be used in the case of imm unizations.Observation showed th at in Hlomela the total num ber oO m m unizations increased after the ^^p e was included in the itinerary of the mobile unit, but it is very difficult to give the exact percentage of this increase.
There is also another factor to be taken into account when evaluating the work of CHW s.The use of CHW s is a relatively recent phenom enon in South Africa.Com m unity or village health workers are employed in some of the so-called national states or TBVC countries, but in most cases purely on an ad hoc basis.It was often a case of the local medical superintendent taking the initiative in the training and employment of CHW s, as in KwaZulu at the Benedictine, Bethesda, Charles Johnson and Manguzi Hospitals.
G azankulu, on the other hand, is the only "national state" where the CHW actually forms an integral part of the health service delivery system.This practice is also fairly recent -the first group was only trained in 1980, and, as has already been indicated, the training program m e has been reconsidered.Karlsson (1978:728) is of the opinion that the CHW s deliver a measure of prim ary health care to remote rural areas.N ot all of them were successful, and some of them did in fact have a negative influence on the communities where they were placed (Karlsson, 1983:728;Ham m ond & Buch 1984).A few observations can, however, be made.
On the level of personal hygiene it could be argued that the CHW s had a positive influence.People knew when to wash their hands, that flies carry germs and that one should keep food covered.No significant differences could be observed between Hlomela, Wayeni and Riverplaats on the one hand, and M atiyani on the other.
W ithout exception streets in the four villages were unkem pt and untidy, with large puddles of water around the stand pipes -ideal breeding places of mosquitoes.This accords with the conclusion of Ham m ond & Buch (1984:16) with regard to Mhala.
In Hlomela, Wayeni, and Riverplaats few people knew what the com m unity health worker is.She was more often known as the "nurse".This was reinforced by the blue uniform she was wearing -the same as that of the hospital staff.In one of the villages it was rem arkable that even people living in close proxim ity of the CHW could not point out the homestead of the CH W or did not know what her name is.H am m ond & Buch (1984:7) went so far as to say that the fact that she wears a uniform brought distance between her and the community.
The question "w hat causes m alaria?" was deliberately chosen to test the knowledge of the local population about at least one disease.M alaria is endemic in G azankulu and it was indeed rem arkable how few people in the four villages knew that the disease is caused by mosquitoes.Nevertheless homesteads are regularly sprayed by the m alaria teams.
It is expected of the CHW s to visit each household at least once a year.It is extremely doubtful whether this aim is ever fulfilled.In many of the households in one village where interviews were conducted the people present did not know the CH W although she had been working in that particular village for more than three years.
The health support system of the CHW could be improved.The CHW is only visited once a m onth by the com m unity health nurse and often less than this.The fact that the monthly mileage travelled by these nurses has been curtailed also has an influence.A possible solution would be to appoint a separate supervisor for the CHWs.
Com m unication of the CHW with the nearest clinic, health centre or hospital is very poor.In three of the villages no radio or telephone com m unication existed between the village and the nearest health facility.In addition the kit issued to her is insufficient to handle routine com plaints let alone an emergency.Shops in the villages had a better supply of patent medicines than the CHW (H am m ond & Buch 1984:12).This left the people with the impression that the local shopkeeper is better equipped than the CH W to handle routine complaints.
One could but agree with the conclusion of Ham m ond & Buch (1984:10) that ". . . of all rural health workers, a lone village based w orker is potentially the most geographically, socially and intellectually isolated."It has been shown th at this was often done deliberately, but som ething will have to be done to improve this state of affairs.W ithout sufficient support from the com m unity health nurse and other medical staff it is difficult to see how CHW s can deliver a proper prim ary health care service.
In the Nkhensani health ward the com m unity health nurse brought the CHW s together for a refresher course.This also happened in other health wards (Karlsson 1983:728), but this practice should be extended to all the health wards and organised on a regular basis.
A nother m atter which deserves urgent attention is the issuing of medication by the CHW .This m atter is under investigation by the SA Nursing Council but this body is not in the position to relax the conditions under which CHW s should issue medicine.However, the question could be asked whether it is sensible to apply First W orld standards to Third W orld conditions w ithout the suitability of these standards being examined?
All in all I would like to agree with Karlsson (1983:728) that there is indeed a place for the CH W in the health service delivery system.M any of the remote rural villages cannot easily be reached by the health service delivery system.In these com munities the CH W can contribute to the raising of the health status of the local communities especially in personal and environm ental hygiene.If the CH W is appropriately trained she could make a valuable contribution in the early detection of diseases, and pregnancies, support and care of the aged and the chronically ill.
The sense of isolation in which the CHW works should be addressed in some or other way because this is a major contributing cause of frustration and a general lowering of the already poor standards of the health care service she is supposed to give.Karlsson, 1983and Sutter, 1979, 1983).

The main conclusion of the research by
H am m ond & Buch (1984), is that the CHW s do not meet the expectations which their introduction as an im portant auxiliary in com m unity health in Gazankulu raised.Karlsson (1983) merely gave an overview of the establishm ent and functioning of the care groups with little attem pt at evaluating their work.It was hoped that this piece of research will supplement the scant inform ation on the CH W s as an im portant facet in the health service delivery system of Gazankulu.
4. D uring the research, the author established that Hlomela was included in the itinerary of the mobile unit after a request of the CHW .
5. " H ouse" or "household" means the num ber of huts occupied by one family irrespective of whether it is m onogam ous or polygamous family.This implies that a household -judged in terms of its size -would consist of four huts (the norm al pattern is four).
In the case of poor families or newlyweds, there is only one hut."H u t" instead of "house" is used because the latter has a specific Western connotation.
6.A random sample num bering table was used to draw the sample.

Karlsson was Secretary of H ealth and
Welfare during the time the first group of CHW s was trained.
at clinic c = Sister at hospital d = com m unity health worker e -member of local care group f = other g = no reply Tsonga d = Friend/relative e = Sister at clinic f = M em ber of the local care group g = Com m unity health worker h -O ther

*
In this case m ore than one answ er was given to the question.a = Prepare food b = Eat c = Feed a child th an one answ er was given to this question.a = Clinic b = Hospital c = School d = Radio Tsonga e = Friend/R elative f = Sister at clinic g = M em ber of the local care group h = Com m unity health worker i = O ther them out d -huts are sprayed by the m alaria team CHW s are taught during training to encourage the use of fuel-saving devices such as the construction of mud-stoves or the use of wonderboxes, as a means of conserving energy.Use of mud-stoves can also be an aid in preventing serious burns, particularly during the winter, when young toddlers often fall into open fires in the cooking huts.The existence of mud-stoves was therefore included in the observation schedule.The following results were obtained: EN DN O TES 1.The author acknowledges the contribution of officials of the Division N utrition of the D epartm ent of N ational H ealth and Population Development in the form ulation of questions on nutrition.Copies of the questionnaire are available from the author on request.2. The care groups give a direct input to com munity health in G azankulu (cf.
first question was what disease is caused by mosquitoes.The following replies were received: