1. The study objectives

The main objectives of the study are: (1) to gain a better understanding of the roles of the VHVs and VHCs in community drug provision; more particular in the extent of the contribution they make to enhance the rational use of drugs by villagers, and in the extent their performance as community health workers is strengthened due their involvement in provision of drugs. (2) to investigate how the roles of the VHV/VHCs in drug provision are influenced by the context in which they work; more particularly by the availability of drugs and drug sources, by the villagers’ drug demands, and by the support they receive from the basic health services.

2. The study design

The study was designed to have an exploratory phase, followed by a phase focusing on in-depth investigation. The perspectives of the drug providers and of the drug consumers have both been included in the study. The first phase was more quantitative and provider-oriented in nature; it comprised of two methods: (1) a VHV/VHCs and Village Drug Funds (VDFs) Survey, and (2) a survey of the availability of drugs and drug sources. The second phase was, on the contrary, more qualitative and consumer-oriented; its main focus was on village case studies using a multi-method approach. Among the methods employed in this second phase were household drug use survey, focus group discussions, observation at groceries and VDFs, and non-structured/informal interviews. The duration of the whole study, excluding the reporting, was 21 months. Its actual operation started in August 1992 and the data analysis ended in April 1994. The field work of phase I and phase II took place, respectively, between November 1992 and January 1993; and between November 1993 and January 1994.

3. The methods

3.1 Survey with a mailed questionnaire (Phase I)

The mailed questionnaire was employed to appraise the current situation of the VDFs and the VHVs/VHCs. It was considered to be the most cost-effective and quick data collection technique to determine nationally representative figures on the VHV/VHC scheme. The data obtained by this method were to be used as a basis for further sampling, and as supporting information on the PHC programmatic context.

Samples and sampling

Eight provinces (approximately 10% of the total provinces in Thailand) covering all regions of the country, were selected for this survey. They were: (1) Chiangrai (CR) from the Upper North; (2) Uthaithani (UT) from the Lower North; (3) Mukdaham (MH) from the Upper Northeast; (4) Chaiyapoom (CP) from the Lower Northeast, (5) (AT) from the Central Central; (6) Prajinburi (PJ) from the East Central; (7) Chumporn (CPN) from the Upper South; and (8) Songkla (SK) from the Lower South. The process of selection began by dividing the country into eight sub-regions; the retention rate of the VDFs of each was, then, calculated on the basis of the MoPH’s official data. Finally, eight provinces were purposively selected, the most similar in retention rate and socio-economic and cultural characteristics of the sub-region to which they belonged.

Each village in each chosen province was sampled. The tambon (subdistrict) health workers and the district hospital health officers were requested to answer questions on each village’s VHV/VHC and VDF situation, as well as on their own personal and working situation. 1,019 tambons were surveyed. The distribution of the questionnaires began in August 1991; two reminders were sent in September and December. As of November 1991 when the field work of the Village Drug Provision Profile (VDPP) Survey was about to start, the return rate of the mailed questionnaires was 70.4%.

Respondents and instruments

The respondents of the survey were health workers at tambon health centers and district hospitals in the sample provinces. Most of them carried responsibility for PHC programme implementation. The instrument used in the survey was a structured, 4-page long questionnaire. It was designed to be used for one individual village and contained 21 questions divided into five main sections: (1) village socio-economic characteristics; (2) drug sources in the village; (3) number of trained and remaining VHCs; (4) the present VHV’s socio-economic background, and (5) the VHV’s role in drug distribution and his/her other activities.

However, it should be noted that this method has been yielding a rather poor quality of data. A high proportion of the returned questionnaires were incomplete. As to the completed ones, some of the data, specifically those concerning village drug sources and the VHV/VHC, are highly likely invalid when compared to data collected by rapid appraisal techniques in the village drug provision profile survey (VDPP survey) and by qualitative methods in the village case-studies’ phase (details of which are presented in the next section). Data about drug sources, in particular drug peddlers, injectionists and traditional practitioners, collected by mailed questionnaire survey, are under-reported. Only data on the presence of groceries and VDFs seem to be comparable to those collected by VDPP survey and village case-studies. In addition, data about the number of ever trained and presently working VHCs in each village are also very contradictory in the mailed survey data.

These methodological drawbacks are probably caused by two factors. First, our assumption that health center staff should know things that go on in all villages under their responsibility is unrealistic. As a matter of fact, this assumption cannot be applied to many circumstances. During our field-visits we found that, although many health center staff live in the villages, many stay outside. Some of them, seldom visit “their” villages. Second, it is probable that filling in five to ten 4-page-long questionnaires with information of five to ten villages (a questionnaire for each village) would be a very boring task, given the fact that health workers are requested by many organizations to fill in many kinds of questionnaires each year. This may partly explain why many of the returned questionnaires are incomplete and contain contradictory answers.

Because of these problems, data collected by this method are used with great caution in this report.

3.2 The village drug provision profile (VDPP) Survey (Phase I)

The operational definition of the VDPP adopted in this study was: the total range of sources which provide drugs in the village. The VDPP survey aimed to collect data on drug sources and drug items available at the village level.

Sample and sampling

The sample villages for this survey were selected from 48 subdistricts in 16 districts of the eight mailed-survey provinces. They were selected through a cluster sampling technique. First, two districts were chosen from the eight provinces; one was centrally located in the province, another peripherally. From each chosen district three subdistricts (tambons) were then selected by using the criteria of geographical distance from the district town, size, and the existence of a functional VDF. Finally, four villages from each sample tambon were chosen purposively. Village with less than 30 households, or where the district hospital or a health center was located were excluded. Priority was given to villages with functional VDFs. In conclusion, 24 villages from two districts and six sub-districts were selected in each province. Additionally, three more villages were included during the field work (two in CR and one in MH)¹. The total number of sample villages in Phase I was 195.

In each sample village, first, key informants were interviewed to gather information about drug supply lines, the total number of drug sources available, and village socio-economic characteristics. Next, for each drug source mentioned, the provider was interviewed and all drug items available there were recorded. In practice, however, they were only groceries and Village Drug Funds; the total number of which were 775 and 96 respectively (see Table 1, Annex 3) For each drug item, the information recorded was; (1) registration number; (2) trade name; (3) generic name or its formula; (4) legal classification; (5) expiry or manufacturing date; (6) preparation form, and (7) indications².

Instruments and data collection process

Rapid appraisal methodology was used in this survey. Four data collection instruments were employed in each sample village: a checklist, two short structured questionnaires, and a record form. The one page checklist only contained short questions. It guided interviews with key informants when the research team arrived in a village. Its main purpose was to help identify village drug sources and to rapidly appraise the village’s socio-economic characteristics. The respondents were mostly village headmen and other village leaders. One questionnaire was used to interview the grocery shop owner, the other one, the VDF caretaker. In the grocery shop owner’s questionnaire, questions were directed at kinds of commodities available and sources where they were obtained. Questions for the VDF caretaker, on the contrary, were more focused on the routine operation of the VDF. Questions relating to available drugs, how frequently each of them was sold, from which sources each was obtained, at whose suggestions each was chosen, etc., were included in both questionnaires. The drug record form was used in conjunction with the two questionnaires to record the drug items available in each grocery shop and VDF.

For the field operation, two teams of field workers were formed. Each comprised eight members: six interviewers, who were recruited from experienced graduate students with social science or health science background, and two field supervisors, one of whom was a social scientist from Mahidol University, the other a pharmacist from a local hospital in each sample province or from the Ministry of Health in Bangkok. Accordingly, in total eight pharmacists played the role of field supervisor, a different one in each province. A training workshop was organized to orientate the research teams and pretest the tools.

Table 1: Number of sample villages, VDFs and groceries

Source: Rapid appraisal survey and VDF field-visit.

3.3 The VDF field-visit (Phase I)

The field-visit to VDFs was an additional method to validate the mailed survey data and to gain a qualitative insight in the results of the mailed survey data. The visits were carried out during the same period as the VDPP survey operation. The VDFs visited were from among those reported functional and located in the 195 VDPP survey villages, to which VDFs reported functional in the remaining areas of the 16 sample districts were added³. Almost 400 VDFs were visited, but only 351 VDFs were found existing. Instruments used during the visit were a checklist and a drug record form. The checklist was used for a brief interview about the actual operation of the VDF with each VDFs caretaker. The form was used to record drug information.

3.4 The village case-studies (Phase II)

The study villages

The study villages were located in the surveyed areas of the VDPP study in Phase I, Yet, as the field work management was a concern, the case studies had to be limited to 16 villages in two provinces. The criteria for selection of the two provinces were: 1) number and types of the VDFs found in these provinces, 2) their socio-economic and cultural representativeness for the region concerned, and 3) logistic research considerations. On the basis of these criteria, the province Chiangrai in the upper North and Chaiyapoom in the lower Northeast were chosen.

In each province, one of the two former sample districts of VDPP survey was chosen and two of the three former sample subdistricts in that selected district were included. The criteria for selection of both were the same as those being used to select the provinces. Subsequently, all former VDPP surveyed villages in each selected subdistrict were chosen; the total of which were eight per province⁴.

Data collection methods

In each of the 16 studied villages five data collection techniques were used: (1) household drug use survey; (2) focus group discussion; (3) observation; (4) drug sale record, and (5) in-depth interview.

Household drug use survey

The survey was intended to assess health-seeking behaviour and drug use of households. It focused particularly on the relative importance of CHW/VDF and other sources of drugs in relation to the villagers’ demand. Five tracer illnesses (diarrhoea, cold/cough, fever/headache, stomach and bowel up-set, and muscle pain) were taken as points of reference. Moreover, to maximize the information relating to use of prescription drugs, five tracer drugs were used as well. They were: antibiotics, anti-inflammatory drugs, analgesics, injectables, and “YaChud”.

About 42 households with at least one school-age child (< 14 years) per sample village were randomly selected for four consecutive interviews on illness episodes, each using a one-week recall period. Before starting the household survey, a whole week was used to establish rapport with villagers, collect basic community information, and administer a tracer drugs survey in the sample households. Altogether the research team spent five weeks in the same village. With only few exceptions, mothers were the respondents during the survey.

Focus group discussion (FGD)

FGDs were used to validate the data collected by other techniques, particularly those obtained from the household drug use survey. Important themes for discussion were the household attitude towards various drug sources, the drug use preferences related to certain complaints, and the culture of drug use prevailing in the community.

For each FGD seven to nine housewives were selected from sample households of the drug use survey. Their age was between 30 and 50. Another criterion for selection was being a good informant. In each village, at least one FGD was held, about two weeks after finishing the household survey.

In every village of Chaiyapoom, another FGD for male heads of families was set up as well. This was because the data showed that male heads of families had important problems of health and drug use (e.g., the use of pain killers and anti-inflammatory drugs). Participants in the male FGD had the same characteristics as those in the female group.

An FGD was facilitated by the field supervisor, two to three interviewers took note; tape recordings were also made. Usually the FGD was held in the evening at the house of the village headman.

Observation

Two grocery shops (one small, one big) were selected for observation in the sample villages. Each shop was visited twice in a day, for about one hour in the morning and in the evening. Covert observation was used. What the observer recorded were the number of buyers, what kind of drugs they bought, and the process of transaction, including what was asked, and what the seller told his/her customers. Not only the grocery shops but also the village drug funds were observed in this way. Drug sale observation was conducted during the third and fourth weeks of data collection in each village.

Drug sale record

In each village, the interviewers selected two cooperative grocers, asking them to record all types of drugs sold in the whole week. Day-by-day the grocers received a blank sheet and gave the record of the previous day to the interviewers. In addition, drug sale receipts given by the drug stores where the grocers procured drugs were also collected.

In-depth interview

The informants for these in-depth interviews can be divided into two groups; the VHVs/VHCs or others who operated the VDFs, owners of grocery shops, heads of villages, injection doctors and informative housewives form the first group; the responsible tambon health workers (either working at tambon or district public health office), the second. The interviews of the VHVs/VHCs focused on the operating status of VDFs, their own performance and problems of drug use in the village; the interviews of the grocery shop owners, on their drug sale activities, drug sources, and the buying characteristics of villagers. For injectionists and village heads, the interviews were on villagers’ common health problems and health-seeking behaviour.

Issues of public health policy, and the supervision and performance of VHVs/VHCs were put to the health workers to comment on. They were also asked about their opinions on improper drug use and distribution.

Research teams and field management

The case studies were carried out by two research teams, each of which was responsible for the eight villages in one province. Each team comprised nine members: eight interviewers were recruited from among those who held a bachelor degree and did fieldwork before, one field supervisor who was a researcher from the Center of Health Policy Studies, Mahidol University. To prepare the research teams, techniques of qualitative data collection were practiced and research instruments were pretested during a five day training course.

During the first two weeks, in each team, the interviewers were divided into four sets of two. Each of those was assigned one village. At the beginning of the third week, one member of each set was moved to another village, the remaining interviewers continued to collect data in the initial four villages. The reason for this procedure was to standardize the performance of all interviewers, particularly during the beginning of the field work. The field supervisors gave advice and checked field notes. In addition, the supervisors accompanied interviewers regularly on data collection visits. Weekly workshops of the team members of each provincial team were held to discuss data, provide technical supervision, and plan for the next round of data collection. The standardization of the data collection of both teams was reinforced through the close contact of the field supervisors and through three workshops held by the Principal Investigator: one in Chaiyapoom and two in Chiangrai.

Table 2: Summary of methodologies used in the study

NOTES TO ANNEX 3:

1. The three villages were added because they are the same communities as those which were sampled. They are situated close to each other only there is a small road in-between; they used to be the same village but were later divided into separate villages by local authority for administrative reasons.

2. In practice, this information was recorded only for newly found drugs. For the frequently found ones, the information collected was only registration number, trade name, and expiry or manufacturing date.

3. The sampling frame can be summarized as follows:

4. One village in Chaiyapoom was dropped after data collection had been carried out during four weeks. A field worker was found not to be qualified enough to continue her work. The field supervisor decided to exclude all relevant data from the analysis as it could affect the quality of data.