Characteristics and Performance of Village Animal Health Workers and Veterinary Assistants in Northern Malawi

INTRODUCTION Numerous primary animal health programmes have been initiated during the past 20 years in developing countries all over the world 1–5,10,14,16. The involvement of trained farmers (village animal health workers, laymen, keymen, barefoot vets) is one of the key-elements in almost every programme in order to develop practicable approaches for livestock health service delivery in remote areas. Despite the significance of this group, only a few studies have been published that describe the characteristics and experiences of these people in detail


INTRODUCTION
Numerous primary animal health programmes have been initiated during the past 20 years in developing countries all over the world [1][2][3][4][5]10,14,16 . The invovement of trained farmers (village animal health workers, laymen, keymen, barefoot-vets) is one of the key-elements in almost every programme in order to develop practicable approaches for livestock health service delivery in remote areas.Despite the significance of this group, only a few studies have been published that describe the characteristics and experiences of these people in detail 11,15,17 .
The Basic Animal Health Service (BAHS) Programme in northern Malawi was launched in the late 1980s by the Department of Veterinary Services, assisted by the German Agency for Technical Cooperation (GTZ), to provide a regional service to livestock owners.Self-help, demand-driven strategies and consequent cost-recovery are essential components of the programme's philosophy.All administrative matters are managed through a farmers' association (Foundation for the Improvement of Animal Health, which is registered as a legal entity under the laws of Malawi.More details about the set-up of the BAHS scheme can be obtained from Leidl et al. 13 .At present the BAHS programme operates within an area of 27 000 km 2 .The focal point of the programme is part-privatisation of Gover nment field veterinar y ser vices through establishing a Drug Revolving Fund (DRF), and deployment of trained farmers, appointed by their community, as village-keymen (KM).Figure 1 illustrates the DRF drug and money flow as part of extension pictorials produced by the BAHS programme.The trainees have to meet certain prerequisites before being accepted by both the respective village livestock group and the BAHS team.These criteria include literacy, experience with livestock keeping, age and reliability.After an initial 4-day training, KM are supplied with a kit (drug box), which allows them to treat village livestock and provide general advice on health and husbandry in their home area.Veterinary assistants (VA), who are Government employees, take part in the programme by handling a drug box that includes injectable drugs, which distinguishes it from the KM's box.In addition, VA provide backup and supervision to KM in their area.
The BAHS programme currently offers a range of 40 different drugs and additional items such as sprayers, ropes, calculators and tubes.Most of the drugs are tailored to meet the needs of the traditional farming sector, with emphasis on disease prevention.Many of them have been re-packaged into smaller and thus affordable quantities.The greatest demand is for dewormer-drenches (for prevention of helminthiasis caused by Toxocara spp.and Strongyloides spp. in young stock), pour-on formulae against ecto-parasites (mainly to prevent East Coast fever in cattle), and long-acting oxytetracycline for curative purposes 9 .
The aim of the current study was to describe characteristics, perceptions and performance of KM and VA involved in the BAHS scheme.The identification of factors limiting the effectiveness of the field service and motivation of field workers was considered to be important information that was required to develop strategies for continued enhancement and adaptation of the programme to meet the needs of the target population.The study is part of a comprehensive impact assessment of the BAHS programme 7 .

Selection of VA and KM and data collection
A total of 149 VA and KM were operating drug boxes from June 1998 in the study area, which covers about 60 % of the entire area of the BAHS operation.Of these, 42 KM and 84 VA with a minimum of 12 months' involvement in the BAHS programme were included in the study.The survey questionnaire used for structured interviews was pre-tested with 3 KM and 4 VA, and adjusted accordingly.Major sections included social and farm characteristics, BAHS workload, job satisfaction, difficulties experienced while performing BAHS duties, and selfperception.Three members of the BAHS team conducted the structured interviews between July and December 1998, each of which lasted about 30 minutes.
For detailed analysis of drug sales and treatment patterns, 6 KM and 12 VA were evaluated.The selection was based on completeness of annual accounting and prescription form data entries for the period of September 1996 to August 1997.While these results are not representative of the entire study population, they suffice to provide a general indication of the performance of KM and VA.

Data analysis
Data were stored using Microsoft Access 97 (Microsoft Corporation, Redmond, USA).Global positioning system receivers were used to determine the precise locations of KM houses and VA stations within the study area.Maps were produced using the geographical information system software Arcview for Windows version 3.1 (ESRI Inc., Redlands, USA).Statistical analyses were performed using NCSS 2000 (NCSS Statistical Software, Kaysville, Utah, USA) and STATISTICA/W version 5.1 (StatSoft, Tulsa, Oklahoma, USA).Multiple correspondence analysis (MCA) was used to explore the data for relationships between categorical variables contained in simple 2-way or multi-way tables.The goal was to represent the entries in each table as distances between individual rows and/or columns in a low-dimensional graphical space.Another way of looking at MCA is to consider it as a method for decomposing the overall P 2 -statistic (or inertia = P 2 /total n).What is important are the distances of the points in the 2-dimensional display.Row points that are close to each other indicate a pattern of relative frequencies across the columns.Adding supplementary column points to the design matrix (in our case the KM/VA status), permits the equivalent of a multiple regression for categorical variables to be performed.The summary statistics (total inertia) provides an indication of how well one can explain the KM/VA status as a function of the other variables in the design.The column points in the final coordinate system should provide an indication of the nature (e.g.direction) of the relationships between the columns in the design matrix and the supplementary column points, indicating BAHS user-status 6 .

Locations of drug box holders
The study area covers approximately 18 000 km 2 of which about 7000 km 2 are forest reserves, national parks or game reserves.The KMs houses and stations of VAs involved in the BAHS programme are shown in Fig. 2. The map illustrates  the extensive coverage of the BAHS programme throughout the study area (Mzuzu Agricultural Development Division).

General description of VA and KM
All KM and VA were male.The average age of VA and KM was 37.4 years (SD 8.7) and 42.6 (SD 10.1), respectively.VA estimated their weekly BAHS workload to be 4.1 days (SD 1.7), which is higher than the estimate provided by the KM (3.7 days, SD 1.8).On average, VA had received their drug box 2.8 years ago (SD 1.37) compared with 3.4 years for KM (SD 1.83).Table 1 is a summary of additional keycharacteristics of both groups.More KM than VA own livestock themselves.Relatively more VA than KM were based in the plains area.The population in the study area includes people from 14 different tribes, with most belonging to the Tumbuka.There is greater variation in the tribal origin of KM than of VA.The Tumbuka tribe appeared to be underrepresented in the VA-group.
More than half of both groups were satisfied with the initial training they had received through the BAHS scheme.Fourty-four percent of VA and 43 % of KM suggested some changes, with most of these groups requesting regular refresher courses and practical demonstrations.Almost every VA and 88 % of KM had organised livestock farmer meetings in villages.Fewer KM than VA were satisfied with the outcome of the meeting(s) they conducted.Both groups were highly likely to include advice on livestock husbandry and management in addition to treatments as part of their field services.

Annual drug sales and treatment regime
Table 2 provides a summary of drug use, which reflects patterns across species summarised for the 12 VA and 6 KM drug box locations selected for a detailed survey.Most prescriptions were issued for treatment of calves, followed by adult cattle, chickens and small ruminants.A relatively small number of pig prescriptions were issued.The VA at Kachere appeared to be the most active person, followed by the KM in Mdolo, based on total number of visits.The average annual number of farm visits by the selected KM was 81.3 (SD 32.7) involving 42.2 (SD 23.6) farms, compared to 78.4 visits (SD 40.4) and 33.6 (SD 31.5)farms for VA.Comparing medians statistically did not show any significant differences between KM and VA.
Table 3 provides a summary of drug sales per selected drug box location.The calculation is restricted to the 4 drugs in greatest demand.There were no significant differences between drug sales by KM and VA.Since KM are not entitled to use injectables, a statistical comparison was not performed for long-acting oxytetracycline (LA-OTC).The average drug sales turnover for VA of *MK3271.5 (US$ 218) is almost twice as high as that for KM with *MK1859.3 (US$ 124), which is significantly different.

Job satisfaction and self-perception
Eighty-two percent of KM and 83 % of VA stated that they enjoyed their job.When asked for reasons, 59 % of KM and 55 % of VA mentioned drug availability after years of ineffective field services.Twenty five percent of KM and 31 % of VA felt they had gained knowledge about livestock production.Almost half of all KM (47 %) and a third of VA (34 %) reported an increase in self-esteem.All the KM and 96 % of VA indicated that they were satisfied with the job because farmers showed their appreciation.Figure 3 provides a graphical summary of these findings based on multiple correspondence analysis.The model explains 44 % of inertia using 2 dimensions.The 1st dimension mainly relates to BAHS programme aspects, whereas the 2nd separates the study population according to their relationship with farmers.The overall job satisfaction is clearly reflected by the closeness of a range of column points to both, KM and VA.These are statements such as 'I like my job as a  misunderstanding of the BAHS concept by farmers, namely the principle of prevention rather than cure, issues of self-help, lack of cooperation by the village livestock group and treatments on a cash rather than a loan basis.Thirty percent of all VA and KM stated that there are communication problems (getting farmers together, inability to explain certain technical issues, cultural/tribal barriers,) and 17 % said that farmers had problems in accepting the current drug and treatment prices.

DISCUSSION
The fact that the communities tended to select comparatively older farmers as KM is part of a culture of respecting elders in their communities.Meemark 15 reported the average age of keymen in northeast Thailand to be 39 years, almost identical to the situation in Malawi.The BAHS programme team advises against selection of KM of an advanced age (e.g., 4 KM are older than 60 years) because of the physical demands of service delivery.
An estimated BAHS weekly workload of 3.7 days for KM as compared to 4.2 for VA is remarkably high, considering the additional commitments of KM as farmers themselves.Although KM as well as VA are provided with a bicycle as part of the initial drug-box delivery, it is known that distances and absence of roads can result in every farm visit becoming a tiresome and time-consuming job.Woods 18 found proximity to the clientele to be a crucial determinant of the quality and frequency of livestock services for the performance of veterinary technicians in Zimbabwe.Apart from farm visits, activi-ties of KM and VA include bookkeeping, group-meetings and cash accounting and transfer, which involves hand-over of cash to the treasurer of the village livestock group, but also stock-taking whenever requested by either the BAHS team or the village group.For a basic animal health project in northeast Thailand, KM bookkeeping was limited to filling in forms for drugs received/drugs sold 15 , whereas it is much more demanding in the case of the BAHS programme in Malawi.The differences in KM/VA locations according to ecological zones (Table 1), including distances to their BAHS clientele have been evaluated recently 8 .
The majority of both groups enjoy their role in livestock service delivery.Almost all of the VA and 88 % of KM organized meetings to promote the programme, and about 2 3 of both groups thought that they had performed this task successfully.The BAHS training, however, was not perceived to be adequate by the KM and VA.Their first suggestion was regular refresher courses, followed by more applied demonstrations and sessions on development of extension skills as shown in Table 1.It should be noted that the BAHS programme conducted regular refresher courses until 1997, and then committed VA and their supervisors to play an active part in KM training and supervision.This has apparently not worked well.One of the reasons is certainly due to changes (reduction) in the number of staff involved with the BAHS team.
Jones 11 referred to an initial 2-week training for village animal health workers in southern Sudan, followed by a refresher course after another 6-12 months.Apart from the initial training of 4 days mentioned earlier, the BAHS team places much emphasis on ongoing supervision of KM, in particular by veterinary field staff.The regular BAHS drug delivery tours provide an additional opportunity for supervision and discussion, as does the annual stock-taking event attended by each livestock group.
In terms of service delivery, cattle as the economically most valued animals receive most of the attention.Pigs receive the least input in terms of drugs.This is not surprising, because efforts by BAHS to promote enhanced pig production in villages so far have been rather modest.It is noticeable that KM as lay practitioners on average visited more farms more often than VA.There are, however, no significant differences between KM and VA with regard to average visit numbers, drug sales or treatments performed.KM apparently spent more visits on fewer animals when compared with VA, while VA at the same time sold more drugs.The latter figure is increased by sales of antibiotics and other relatively expensive injectables through VA.Although VA on average had almost twice the turnover of KM, the large data variation and small sample size makes it difficult to obtain a clear picture of the situation.The fact that the number of VA and KM included in this calculation is not representative of all drug-box holders may be relevant in this context.
Seventy-three percent of VA and 43 % of KM said that they sometimes had problems when dealing with farmers.VA and KM attributed this to inability to attract sufficient farmers on one hand and poor understanding of the BAHS concept by farmers on the other.Government field staff in general have a reputation of passivity and inefficiency, which is inherited from a history of pure state services 12 .It is also possible that tribal issues influence farmers' interest in the scheme.This is supported by the fact that almost 2 3 of VA do not belong to the Tumbuka, who predominate in the study area (Table 1).This factor should be taken into consideration during future phases of the BAHS programme, but remains to be further investigated.
When interpreting the responses that VA and KM provided in this survey, a tendency to give positive rather than negative information in relation to the programme must be taken into account, as VA and KM themselves receive a financial benefit from BAHS.Subjectivity and potential response bias do not invalidate the conclusions of the analysis, which provides an important summary of VA/KM perceptions in relation to BAHS.
As illustrated in Fig. 3, the overall job satisfaction of VA and KM involved in the BAHS programme is the most striking pattern that emerged from this study.It resulted to a significant extent from the encouraging response they received from their clientele.The feeling of support by the BAHS team and enjoyment from talking to farmers was common in both groups.This was closely related to personal perceptions such as increased selfesteem and positive learning experiences.It is clear that for community-based livestock programmes to be successful in the short-term and sustainable in the long term, village animal workers must receive a perceived benefit from the programme, which need not necessarily to be measurable on purely economic grounds.

Fig. 1 :
Fig. 1: The drug and money flow of the Drug Revolving Fund, the core piece of the BAHS programme (pictorial as part of BAHS extension material).

Table 3 : Sales of most wanted BAHS drugs per selected location between September 1996 and August 1997
(using Kruskal-Wallis 1-way ANOVA to indicate statistically significant differences between KM and VA, P < 0.05).