correspondence Dr V.M. Marsh, KEMRI, Centre for Geographical Medicine Cost Research, PO Box 230, Kilifi, Kenya. E-mail: kemriklf@africaonline
Summarybackground Malaria control in Africa relies primarily on early effective treatment for clinical disease, but most early treatments for fever occur through self-medication with shop-bought drugs. Lack of information to community members on over-the-counter drug use has led to widespread ineffective treatment of fevers, increased risks of drug toxicity and accelerating drug resistance. We examined the feasibility and measured the likely impact of training shop keepers in rural Africa on community drug use.
methodsIn a rural area of coastal Kenya, we implemented a shop keeper training programme in 23 shops serving a population of approximately 3500, based on formative research within the community. We evaluated the training by measuring changes in the proportions of drug sales where an adequate amount of chloroquine was purchased and in the percentage of home-treated childhood fevers given an adequate amount of chloroquine. The programme was assessed qualitatively in the community following the shop keeper training.
results The percentage of drug sales for children with fever which included an antimalarial drug rose from 34.3% (95% CI 28.9%–40.1%) before the training to a minimum of 79.3% (95% CI 71.8%–85.3%) after the training. The percentage of antimalarial drug sales where an adequate amount of drug was purchased rose from 31.8% (95% CI 26.6%–37.6%) to a minimum of 82.9% (95% CI 76.3%–87.3%). The percentage of childhood fevers where an adequate dose of chloroquine was given to the child rose from 3.7% (95% CI 1.2%–9.7%) before the training to a minimum of 65.2% (95% CI 57.7%–72.0%) afterwards, which represents an increase in the appropriate use of over-the-counter chloroquine by at least 62% (95% CI 53.7%–69.3%). Shop keepers and community members were strongly supportive of the aims and outcome of the programme.
conclusions The large shifts in behaviour observed indicate that the approach of training shop keepers as a channel for information to the community is both feasible and likely to have a significant impact. Whilst some of the impact seen may be attributable to research effects in a relatively small scale pilot study, the magnitude of the changes support further investigation into this approach as a potentially important new strategy in malaria control.
Malaria remains a major cause of morbidity and mortality world wide. 90% of the estimated 1–2 million deaths per year occur in sub-Saharan Africa, where it is widely recognized that the situation is deteriorating rapidly. In these worst affected regions, malaria control relies primarily on the availability of early effective treatment, critical in a disease where many deaths occur within 48 h of the onset of symptoms ( Greenwood et al. 1987 ). Many studies have shown that the majority of early treatments for childhood fever are given at home, using shop-bought brand name drugs ( Igun 1979; Raynal 1985; Mwabu 1986; Van der Geest 1987; Foster 1991; Snow et al. 1992 ; Mwenesi et al. 1995 ). These treatments are usually incorrect or suboptimal ( Deming et al. 1989 ; Ramakrishna et al. 1989 ; Agyepong 1992; Slutsker et al. 1994 ; Adome et al. 1996 ). However, in many African countries, education on fever management is focused through health facilities; information on the use of over-the-counter drugs is not generally available to communities. This mismatch of supply and need for information leads to at least 70% of all children with fevers being treated inadequately, or even dangerously, in the first instance. It represents, at best, a missed opportunity to impact on malaria related morbidity and, at worst, an increasingly widespread and uncontrolled use of drugs in situations where health services cannot keep pace with peoples' needs. Drug retailers are, potentially, an efficient channel to communicate information on drug use to purchasers. However, the option of training drug retailers is not straightforward. It raises questions of motivation, sustainability and community acceptance. We report here on a pilot study designed to examine the feasibility of training shop keepers to offer information on the drugs they sell, and to determine the likely impact of such information on the use of over-the-counter drugs in the community.
Study area and methods
The study was conducted in Chonyi location, a rural area of Kilifi District in the coastal region of Kenya. The area has a scattered population of approximately 3500, mainly subsistence farmers, served by 23 small general shops selling a wide variety of household goods, including drugs. The formal health facilities in the area comprise a government health facility, a private clinic and a community pharmacy. Malaria transmission occurs throughout the year, with two main peaks in May to August and December to January, following the bi-annual seasons of heavy rain. Full descriptions of the area have been published previously (Snow et al. 1993 ; Nevill et al. 1996 ).
Baseline assessment and formative research
From December 1995 to January 1996 study field workers conducted structured observations outside the 14 busiest shops continuously from 0700 h to 1800 h over a period of 9 days in each shop. Following each observed sale of an antimalarial or antipyretic drug, the purchaser was interviewed briefly outside the shop to determine the age and identity of the person for whom the drugs had been bought. After 3 days, we visited the homes of children aged 6 months to 10 years to interview the usual caretaker of the child on the symptoms of the illness, subsequent health-seeking behaviours and drug use. We assessed overall health-seeking behaviour for childhood fevers in the study community by visiting all households in the study area over a period of 4 weeks and interviewing caretakers of children who had experienced a febrile illness in the two-week period preceding the survey.
Following these shop and community-based assessments, the main sellers of drugs in the 14 most popular shops were interviewed informally to establish their current knowledge, attitudes and practices in relation to malaria. They were invited to discuss the planned training in meetings which covered attitudes to taking on an advisory role for over-the-counter drug use in the community, perceptions of the feasibility of this approach, perspectives on useful content and logistics for the training.
Shop keeper training workshops
In April 1996, 46 shop keepers from 23 shops were trained in a series of three workshops, each lasting 3 days. A full description of the training programme is available from the authors and will be published elsewhere. We placed emphasis on methods which would encourage active participation, provide practical training and allow us to evaluate the participants' skills. Training focused on the usual brand name drugs stocked by retailers. Shop keepers were trained to use two types of materials to give information to purchasers on how to use these drugs; dosage charts for chloroquine and aspirin/ paracetamol-based drugs and sets of rubber stamps depicting the correct way of using chloroquine in children of different ages. These stamps printed a visual reminder of the verbal advice given, and were retained by the drug purchaser. The shop keepers were also trained on symptoms which indicate the need for early assessment by a trained health worker. After the workshop, 1–2 h individual training sessions were held in the 14 study shops to allow direct observation of the shop keeper's skills in his/her normal workplace. A 2-day refresher training workshop was held after 6 months for study shop keepers.
Evaluation of impact
The impact of the training programme was evaluated in two rounds of observational studies at shops and home interviews during peak malaria seasons in June–July 1996 and December 1996–January 1997. The main outcomes were changes in the percentage of sales where an antimalarial or antipyretic drug was purchased, the total number of chloroquine tablets purchased and the reported use of shop-bought drugs. The shop keepers' dosage charts were based on locally derived weight-for-age charts and the number of chloroquine tablets which would give a total dose of 25 mg/kg chloroquine base. The use of five age bands meant that the definition of ‘adequate’ dose was quite wide, allowing the oldest/heaviest children in each group to count as adequate if they received the dose appropriate for the youngest/lightest and vice versa.
To corroborate the reported use of shop-bought antimalarial drugs following the training, we studied all children reportedly given shop-bought chloroquine for whom a home interview was achieved 72–96 h after a first dose had been given. 50 children who had reportedly been exposed to a full course of chloroquine at home, and where an interview was achieved in the specified time period, were recruited to this validation study. Following parental consent we removed 100 μl blood samples by fingerprick onto filter papers which were dried and stored for HPLC plasma chloroquine assay. These results were compared with plasma chloroquine levels in blood taken from children 72–96 h after they had received the first of a range of supervised doses of chloroquine, alongside standard antimalarial treatment of pyrimethamine sulphadoxine, as part of outpatient treatment for malaria at Kilifi District hospital.
Analysis of quantitative data was based on a comparison of percentages before and after the intervention, using EpiInfo to calculate percentages, 95% confidence intervals (Fleiss quadratic method to allow for some data sets with low numerical values) and compare percentages with χ2 testing for significance.
A broad based qualitative assessment was conducted over a 4-week period at the end of the programme to confirm the quantitative findings, describe factors underlying observed changes in behaviour and explore community responses to the programme. We conducted issue-focused conversations with all study shop keepers and focus group discussions with 8 community leaders, 8 traditional healers, 7 community health workers and 22 fathers of young children. Key informant interviews were held with 4 administrative leaders, the dispensary nurse, the public health technician and the private medical practitioner. Two groups of 12 school children participated in role plays and group discussions. A total of 60 mothers of young children were interviewed; 30 mothers involved in observed sales gave case narratives and a further 30 randomly selected mothers were interviewed informally.
Table 1, 2 and 3 give the results of the observational surveys for sales of antimalarial and antipyretic drugs and the reported home use of the drugs purchased before and after the shop keeper training. Frequencies, percentages and 95% confidence intervals for percentages are given for observations and reported behaviours before and during the 2 rounds of assessment conducted during consecutive peak malaria transmission seasons following the training. Percentages before and after the training have been compared and p-values are reported for changes in each of the post-training rounds of assessment.
Table 1. Observational survey results for antimalarial (AM) and antipyretic (AP) drugs sold through study shops
Table 2. Reported home use of over-the-counter chloroquine (CQ) for childhood fevers before and after the shop keeper training
Table 3. Reported use of aspirin-containing drugs (ASA) over a 24-h period
Figure 1 demonstrates the results of the chloroquine HPLC assays performed for validation of the caretakers' reported use of chloroquine at home. This shows the plasma chloroquine concentrations in randomised groups of outpatient attenders given supervised doses of chloroquine approximating to 15 mg/kg (2 doses on day 0), 20 mg/kg (3 doses on days 0 and 1) and 25 mg/kg (4 doses over days 0, 1 and 2) as well as 50 children from the study area, reportedly given an adequate dose of shop-bought chloroquine over 3 days, in whom plasma chloroquine results were available (labelled CQ4). The means and 95% CI for the 4 groups of children are shown. The mean and distribution of chloroquine plasma level for the group of children reportedly given an adequate dose of chloroquine at home is not significantly different from that of the group of children given 25 mg/kg chloroquine under supervised conditions, but it is significantly different from that of the group given 20 mg/kg chloroquine under observation.
We recorded major changes in the way that study shop keepers sold drugs before and after the training. Before the training, shop keepers very rarely either gave any information about the drugs sold through their shops or asked about the user. Following training, shop keepers gave advice on the type and quantity of a drug to buy and how to use the drug at home and asked questions about the age of the drug user and the symptoms of their illness in the majority of drug sales. These results have not been presented in full here, but will be published separately as part of a detailed report on the training components of this study. Care is needed in interpreting these observational data which reflected the shop keepers' abilities after training but not necessarily their ‘normal’ behaviour.
Full results of the qualitative assessments of community beliefs, practices and responses to the training programme will be published elsewhere, including an analysis of factors underlying noncompliance with a shop keeper's advice. Important findings are summarized here. We confirmed that uncomplicated fevers in this community were normally first treated at home using shop-bought drugs; the main factors underlying this practice were the proximity of local shops, which had important advantages in terms of cost, time and convenience as well as saving the user fee required to attend the local dispensary. Prior to the training, there was a strongly perceived need for information on over-the-counter drugs in the community. In the absence of other sources of information, shop keepers were usually considered reliable informants. Against this background of unmet need, the community strongly supported the aims of the training programme and reported increased confidence in the information offered by shop keepers following the training. The shop keepers in the study area were keen to participate, both before and after the training. The perceived benefits of the training were increased status in the community, increased knowledge in treating family members, increased confidence in selling drugs and increased profitability of their businesses. The mean cost of treating a child with fever rose from KS 8.12 (US $0.15) to KS 13.18 (US $0.25) 3 months after the training and KS 11.60 (US $0.21) after a further 6 months.
Throughout Africa, a large majority of malaria episodes in children are first treated at home, using shop-bought drugs ( Igun 1979; Raynal 1985; Mwabu 1986; Van der Geest 1987; Deming et al. 1989 ; Ramakrishna et al. 1989 ; Foster 1991; Agyepong 1992; Snow et al. 1992 ; Mwenesi et al. 1995 ; Adome et al. 1996 ). The reasons underlying this practice include both ease of access and perceived deficiencies in the performance of formal health services ( Hassouna 1983; Kroeger 1983; Igun 1987; UNICEF 1991; Foster 1995). Our baseline results are typical in showing that the majority of such home treatments are inappropriate in terms of the drugs used, the dosages used or both ( Deming et al. 1989 ; Ramakirshna et al. 1989; Agyepong 1992; Adome et al. 1996 ; Slotsker 1994). In examining drug usage, we used wide limits in designating dosages as ‘adequate’; nonetheless, only 4% of children given shop-bought chloroquine could be said to have received an adequate total dose of the drug for a fever in our baseline surveys. Even fewer (2%) received chloroquine over the recommended 3-day period. Usage patterns of other drugs also gave cause for concern. Aspirin was used almost universally (236 of 250 fever episodes) for the treatment of childhood fevers; among these children, 22% received potentially toxic dosages over 24 h. This finding was reinforced by the presentation of cases of severe salicylate toxicity at the local District Hospital (English et al. 1996) . The pattern of widespread inappropriate early treatment of febrile illness has serious implications. It represents a major missed opportunity to reduce morbidity and mortality from malaria in Africa. In addition, it contributes to morbidity through the toxicity of the drugs themselves and increases the pressure for drug resistance ( Swartz et al. 1983 ).
The most obvious channel for changing community use of shop-bought drugs is the shop keeper. However, there are a number of reasons why such an approach may not work. Shop keepers are primarily concerned with increasing profits, not standards of health in the community. It is unlikely that any change in behaviour which would result in a loss of profitability, including increased use of time, would be acceptable. Similar feasibility questions surround the acceptability of shop keepers as health advisors to the community. However, we found that rural shop keepers can be motivated to take on an advisory role in the community where this does not diminish their income. The combination of increased knowledge, increased social status and increased profits proved a powerful incentive. Support from the community for the approach was also strong since the need for information was already well recognized. However, an important requirement from the community was that shop keepers should be trained by a trusted agency.
The shop keeper training programme had a dramatic effect on the practices of both shop keepers and purchasers of over-the-counter drugs. The percentage of drug sales to treat a childhood febrile episode for which an antimalarial drug was purchased rose from 34.3% (95% CI 28.9%–40.1%) to 84.0% (95% CI 78.5%–88.3%) after 3 months and to 79.3%(95% CI 71.8%–85.3%) after a further 6 months. The percentage of antimalarial drug sales where an adequate amount of chloroquine was bought rose from 31.8% (95% CI 22.6%–37.6%) to 82.7% (95% CI 76.3%–87.3%) after 3 months and to 89.9% (95% CI 82.7%–94.4%) after a further 6 months. We believe that this latter change is important because it relies on observational data, which are less susceptible to bias than reported data, and, whilst purchase of an adequate amount of antimalarial drug does not guarantee later appropriate use, it is a necessary prerequisite. Reported use of an adequate dose to treat a febrile illness rose from 3.7% to 65.2% 3 months after training and to 75% after a further 6 months. This represents an increase by 62% (95% CI 53.1%–69.3%) after the first 3 months and by 71.3% (95% CI 62.6%–80.0%) after a further 6 months. Although reports on behaviour are subject to many biases, our confidence in the reliability of mothers' accounts is strengthened by our finding that the distribution of plasma chloroquine levels in the group of children reportedly given full courses of chloroquine at home was not significantly different from the group given 25 mg/kg under direct supervision in the outpatient department in Kilifi, but was significantly different from the group given 20 mg/kg. Gains were not restricted to antimalarial drug usage; the proportion of cases in which a potentially toxic amount of salicylates were used fell from 22% to 2.9% by the final evaluation. The aim of the training programme was to optimize treatment in those who had already taken the decision to use shop-bought drugs, not to encourage their use as an alternative to other sources of health care. It is important to note that no shift in the proportion of cases treated by any one modality (traditional treatment, dispensary, private clinic, hospital) followed the training programme.
There are a number of caveats to be borne in mind. As with all studies using behavioural changes as an outcome measure, the potential for a research effect is high. However, the key question in this study was whether community members would change their use of shop-bought drugs as a result of advice offered by trained shop keepers. This assessment could only be made where we were certain that appropriate information had been given. In this respect it is encouraging that the level of appropriate action was maintained, or even improved, at the second round of assessment, conducted 9 months after the original training, and following a 4-month period when no research activities had been conducted in the shops or the community.
It may be asked whether improving the pattern of chloroquine use in the community can be expected to result in real health gains, given the levels of drug resistance in East Africa ( Bloland et al. 1993 ; Nevill et al. 1994 ). The training programme focused on chloroquine because this is currently the only antimalarial drug legally available through shops in Kenya and is, in fact, the only antimalarial drug available in the majority of shops in the study area. The main patterns of resistance seen currently in coastal Kenya are R1 and R2, so that chloroquine is likely to remain of value in preventing subsequent severe morbidity in communities where children have some degree of immunity ( Anabwani et al. 1996 ). However, from a strategic perspective, the choice of drug in this study was less relevant than the aim to demonstrate the effectiveness of shop keepers as channels for community education on drug use. For this purpose, chloroquine, a drug with a regime for which it is difficult to maintain compliance, is an excellent model for any multidose successor.
This study has demonstrated that health education through shop keepers can lead to major improvements in the way that shop-bought drugs are used at home in treating fevers in rural African children. Since this is the commonest first-line treatment for children with fever, the potential contribution to malaria control strategy is clear. At a time of widespread concern over the deteriorating malaria situation, this approach is notable for harnessing human inclinations and resources that are already in play rather than requiring new technologies. We have described this research as a pilot study examining the feasibility of training shop keepers and the likely impact of such a programme on community drug use. The need now is to find out whether similarly impressive results can be achieved and sustained in a large-scale operational programme incorporating a controlled trial design.
This paper is published with the permission of the Director, Kenya Medical Research Institute (KEMRI). The investigation received financial support from KEMRI, the Department for International Development of the United Kingdom and the UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR). However, the Department for International Development can accept no responsibility for any information provided or views expressed. The authors are indebted to many colleagues for advice and support. We are particularly grateful for the support of Dr John Ouma, Director, Division of Vector Borne Diseases, Dr Anderson Kahindi, MOH Kilifi District, Dr Norbert Peshu, Director Centre for Geographical Medicine, Coast, Dr Robert Snow, Dr David Nyamwaya and Dr Carol Jenkins. The study could not have been conducted without the commitment of our field team or the enthusiastic participation of the community and the community leaders in Chonyi. KM is a Wellcome Senior Research Fellow in Clinical Science.