Study site

This study was conducted in 40 communities in two rural health districts selected by random sampling from the eight health districts that make up Ona-Ara Local Government area (LGA), in south western Nigeria from July 2005 to January 2007. A well established collaboration exists between researchers in the College of Medicine, the State Ministry of Health and the health unit of the Local Government (LG) having carried out studies on malaria in the LGA in the past. At the time of commencing this study, ACT has not been deployed to the LGA health facilities by the Government and was not available in the private sector either. The study site has been described in detail by Ajayi et al. 2008 [10].

Health services in Ona-ara Local Government Area

The health care system (HCS) in Nigeria consists of three levels namely: primary, secondary and tertiary levels. The Local Governments are generally responsible for primary health care (PHC). In Ona-Ara LGA there is a Primary Health Care (PHC) unit at the LG headquarters and the coordinator of the unit is usually a medical doctor who oversees the activities of the health facilities in the LGA.

Each of the health districts has a health centre and maternity centre. The health centres are manned by a team of community health workers headed by a registered nurse/midwife. Few traditional healers and birth attendants (TBAs) provide indigenous health care in the districts. However, majority of the caregivers seek care from patent medicine sellers (PMS) and itinerant drug hawkers [10]. The populations of the villages in the two districts vary from 50 to 8,000 and are spread over a vast area with poor transportation and road network; making it difficult for the government health structures to be easily accessible or effective.

Study design

This qualitative study is a component of a quasi-experimental study to evaluate the feasibility, accessibility and safety of using ACT for home management of malaria. To conduct a situation analysis prior to distribution of AL for HMM, Focus Group Discussions (FGDs) and Key Informant Interviews (KIIs) were viewed appropriate [11]. These were used to explore caregivers' and community members' opinions on malaria as it affects children, the prevailing treatment seeking behaviour including practice of home management of malaria and the potential for change. The findings of the situation analysis were used in the planning of the intervention including mechanism for drug distribution. After one year of intervention during which AL was distributed by selected and trained community based medicine distributors (CMDs) in study communities, the communities' perception on the use of AL for HMM, strategies for its distribution and accessibility to the people as well as acceptability were explored. In addition, household survey was conducted (to be reported separately). CMDs in this study comprised patent medicine sellers, health workers and mothers selected by the communities to be trained as "mother trainers".

Sample population

Communities were selected for FGD and KII by random sampling technique from the 40 communities at baseline and after intervention respectively. The participants for the FGD which comprise mothers and fathers were selected using purposive sampling based on the recommendation by village head or opinion leader who perceived them to have sufficient knowledge and experience on the issues to be addressed and would be able to discuss freely [12]. In addition they must have a child less than five years or must have taken care of a child in the past. Focus group discussions (FGDs) were held in groups of six to nine women or men. The key informants were purposively selected based on their position in the community, their role in implementation, or because of the special information they possessed based on their involvement in the management of childhood fevers.

At baseline, a total of 14 FGDs comprising four groups of young mothers [aged <35 years], old mothers [≥ 35 yrs], and three groups each of young fathers [<35 years] and Old fathers [≥ 35 years] were conducted. Fourteen KIIs with six opinion/community leaders, four drug sellers and four health workers were also conducted. Post intervention, a total of 10 FGDs and nine KIIs were conducted. Three groups of young mothers, three of old mothers, two of young fathers and two groups of old fathers were involved in the FGD while three patent medicine sellers (PMS), four opinion/community leaders, and two health workers were interviewed during the KIIs. The total numbers of participants at the FGD sessions at baseline and post intervention were 103 and 89 respectively.

Intervention Activities

Treatment guideline

A treatment guideline developed for a past study on HMM using chloroquine [10] was adapted for ACT use in this study and pre-tested prior to distribution to every household in the study communities by the CMDs. The guideline was in cartoon format and had pictorial illustrations of some common clinical features of uncomplicated malaria, the correct dose and regimen of AL according to the age of the child [Fig 1]. Training was conducted for caregivers by the CMDs in their communities to improve their ability to recognise uncomplicated and severe malaria and on how to use the guideline.

Analysis

All interviews were tape recorded and transcribed in Yoruba the language which was used to conduct interviews and FGDs. Two of the research assistants and the social scientist (BO) and one of the authors (IOA) listened to the recordings, checked the accuracy of the transcripts and translated the transcript from Yoruba to English. Content analysis was performed independently by two of the authors IOA and BO the social scientist. They compared notes for congruency and where there was incongruence they re-read the transcript and made necessary corrections From the transcripts and field notes, the responses from different informants or groups (FGD) to questions asked to explore issues of interest were grouped together, coded and analyzed according to themes related to recognition of malaria, treatment practices, use and adherence to use of AL, adherence of CMDs, effectiveness and safety of AL, drug distribution mechanism and sustainability of the intervention. The findings were described, interpreted and reported in form of narratives.

Ethics

Ethical approval for the study was provided at national level by the Oyo State Ethical Review Board and the Ethics Review Committee of the WHO. Informed consent was obtained from community heads, household heads and the caregivers who participated in the study.

Knowledge and Recognition of Malaria in Children

An important exploratory study was conducted at the beginning of the FGDs to ascertain the study communities' definition and nomenclature for malaria. The locally recognized syndromes in children treated with antimalarial drugs by participants and the local terminology 'iba' were explored. Most of the FGD participants described "iba"- the local terminology for fever as high temperature associated with chills/rigors, vomiting, anorexia, headache and weakness. This definition is consistent with the symptomatology of malaria as seen in clinical practice. The FGD participants also mentioned that "iba" when literarily translated means fever, "iba" is as such also used for other febrile illnesses not presumed to be malaria. However, if a febrile illness is perceived to be from another cause the "iba" terminology is further qualified. Examples given by FGD participants are: a febrile illness perceived to be caused by measles, is called "iba ogbele". "Ogbele" is the Yoruba word for measles and "iba typhoid" if the fever is perceived to be due to typhoid fever. This preliminary finding was useful in ensuring that researchers and community members were in agreement in the use of the terminology for malaria.

Malaria in children was perceived to be a serious problem in the community and the general consensus among the participants was that it kills children fast if care is not sought promptly. A young mother said: "malaria is a major health problem because if it stays too long in a child's body, it drains the blood".

The participants demonstrated good knowledge of the symptoms and signs of malaria. The symptoms and signs participants associated with uncomplicated malaria included fever, chills, excessive weakness, anorexia, vomiting, excessive sleep, 'stretching of the body', concentrated urine, headache and dull looking eyes. For severe malaria, they mentioned very high fever, convulsion/fits, restlessness, prostration, excessive vomiting and weakness, yellow eyes and dark coloured urine. Some respondents mentioned they suspect severe malaria when a febrile child is not responding to the treatment first given.

Comprehension and Understanding of the Package of AL

In order to determine the FGD and KII participants' understanding of the illustrations on the package of AL and the appropriateness to the local setting, samples of the AL (Fig 2) to be distributed were shown to the participants at baseline. Most of the participants recognized the pack of AL as drug meant to be taken by children because of the picture of a child on the pack. However, their explanations on the use based on illustrations on the pack were mostly incorrect. Some said it is for malaria because the colour of the tablet is yellow and because a mosquito which was perceived to be a cause of malaria is drawn on the pack. Many participants could not comprehend how to administer the drug by just studying the drawing on the pack. The responses are that one tablet of AL is to be used three times a day for two days (1 × 6 pack); one daily for six days (1 × 6 pack) or twelve days (2 × 6 pack) and that the tablet is to be divided into two because each tablet has a line down its middle.

They requested to be given verbal instruction and training on how to use the drugs for better understanding and correct usage.

Assessment of intervention after one year of distributing AL

Acceptability

The caregivers, CMDs and other members of the communities in which this study was conducted were very receptive to the new drug (AL). The easy and convenient access of community members to AL at reasonable cost initially and latter at no cost and the observed rapid improvement in the state of health of children treated with AL must have contributed in no small measure to the acceptability by the community. The fact that the AL distributed was pre-packaged could also have contributed to the acceptability. The acceptability of the blister pack is consistent with results from a study in western Uganda [15] and rural communities in Nigeria [13]. These studies found that caregivers and drug distributors preferred pre-packed drugs over loosely packed tablets. The mentioned reasons for this were safety, cleanliness and possibly facilitate correct dosing. To further demonstrate the acceptability of AL is the demand for extension of the intervention to the treatment of malaria in adults and neighbouring communities. Community members were also willing to support any measure that could ensure continued availability of AL in their communities.

The superior efficacy of AL in the management of malaria when compared to that of chloroquine which hitherto was the antimalarial drug of choice has been well established in clinical studies [8, 16, 17]. This study can only mention that caregivers perceived AL to be effective but cannot draw an objective conclusion on its effectiveness because it is a small scale study based on purposive sampling. However, the effectiveness of AL when distributed at the community level by community members was tested in an effectiveness study nested into this study (to be reported in another paper).

Drug adherence and safety

Two main concerns in introducing ACT for use in HMM context were adherence and safety of use. These were borne out of the fact that the effectiveness and safety of AL at the community level have not been evaluated in Nigeria. In addition, the practice of HMM using orthodox medicine has been shown to be largely incorrect in Nigeria [2, 13, 18]. Some of the challenges identified in past interventions carried out to improve HMM were also identified in this study. These include failure of caregivers to complete a full course of antimalarial drug including AL [2, 19, 20], provision of financial motivation to CMDs [21] and non-adherence of health workers to recommendations on the use of AL [22].

In this study, adherence to correct dose of AL by caregivers was attested to by participants at FGDs and KIIs. Although this was subjective and from the perspective of the consumers, there is a likelihood that this was the case as some caregivers who used the drug were able to correctly recall how the drug was used when asked by CMDs. This finding is reassuring in view of potential to development of resistance. However, the perception of some participants in this study that some caregivers sometimes genuinely forget to continue therapy once the ill child appears well or deliberately keep the remaining drug for latter use for another child who falls ill should not be overlooked. Unknown to caregivers, parasite clearance is not necessarily complete before the patient improves. This practice if uncorrected will encourage emergence of drug resistant parasites as was witnessed with chloroquine. There is thus the need to provide health education to the public and healthcare workers alike emphasizing the need to always complete the treatment regimen of not just AL but all ACTs.

Prompt treatment of malaria is an important factor in the prognosis of the illness. It is encouraging that caregivers in this study were said to have commenced treatment promptly. This finding is similar to reports in other past studies in the study environment [13, 18]. This attitude is favourable to implementation of home management of malaria strategy which emphasizes prompt and correct treatment. The malaria control unit of the Federal Ministry of Health, Nigeria should build upon this attribute of the caregivers in the introduction of AL into the community.

Throughout the study period, AL was perceived to be safe with no report of significant adverse event by the communities. Although this finding is similar to reports of earlier studies which reported mild adverse effect during clinical trials [16, 17] and among patients prescribed AL [22], it cannot be used to draw conclusion on the safety of AL because of the small scale nature of the study and the qualitative design both of which constitute a limitation to the generalizability of findings in this study. Despite this seemingly good safety, the Federal Government and all stake holders should put an effective pharmaco-vigilance system in place when AL is deployed on a large scale.

Many interventions to improve adherence to correct dose of antimalarial drug at home level have been demonstrated to be effective [14, 23–26] and this study is not an exception. However, sustainability of the correct practice is a major concern. This was addressed in this study by providing a pictorial guide which was comprehensible to the rural dwellers and served as a reference document even after the researchers might have moved out of the community. The availability of a trained resource person within the community who they can contact in case of doubt is another modality put in place to address the problem. Although these modalities appeared to have helped in ensuring adherence during intervention, there is a need to assess their effectiveness at regular intervals after the research team must have withdrawn from the communities.

Drug distribution

The drug distribution mechanism proposed by community members varied at baseline. While many reasonably favoured the use of the drug sellers (PMS and drug hawkers) who were the main supplier of drugs in these areas, there were misgivings by a handful of other participants who would not want them involved because of distrust and the perceived propensity of the drug sellers to maximize profit. However, this attitude was not demonstrated by drug sellers during intervention in this study. On the contrary, caregivers perceived drug sellers to be efficient and accessible. There was no report of drug sellers charging illegal fees at any time during the study. The reported good adherence by the drug sellers was probably enhanced by the incentives by way of commission they received on each pack of AL sold or distributed and commitment to their community's health and development.

CMDs' performance

"Mother trainers" and PMS were perceived by caregivers and other community members to be effective in this study. Not much was said about the efficiency of the health workers probably because they did not have to go to the health facility and consequently had minimal interaction with them. However, drug sellers and "mother trainers" who had to go to the health facility to replenish their drug stock had a lot of reservations about their effectiveness. They reported that health workers were often absent from their duty post and were perceived to hoard the study drug (AL). This attitude of the health workers made the CMDs to prefer drug distributed to them directly through other channels such as directly from the LGA drug store and work under the supervision of community leaders rather than health workers. The use of community members to assess efficiency of health workers and vice versa could be fraught with biases and prejudice existing prior to the study. This not withstanding the negative attitude of health workers if unchecked poses a major threat to the deployment of AL at the community level in Nigeria. Home management should not be seen as an alternative to strengthening health systems. The Primary Health care system needs to be strengthened and effective monitoring put in place in order to ensure a sustainable home management program. Following these a functioning linkage between the health facilities and the CMDs as well as community leaders should be fostered. This linkage is important to support HMM strategy because the health care facilities are needed for supervisory role and management of referred cases.

Sustainability of program and CMDs' performance

This study has demonstrated that community volunteers and patent medicine sellers in the communities constitute an effective channel for intervention which is similar to findings by Marsh et al. [25]. The CMDs effective performance in this study was facilitated by the intervention in form of training provided to them prior to assuming their role as drug distributors and provision of incentives to sustain behaviour change [27]. However, the sustainability of their involvement and the regular supply of drug is a major concern. One suggestion to sustaining the involvement of the CMDs is maintenance of training and supervision of their activities. However, where sellers lack financial incentives to behave as they have been trained, they tend to revert back to inappropriate practices [27].

Incentives which may not be sustainable contributed significantly to the good performance of the CMDs during this study. Four strategies have been advanced for improving the practices of retail pharmacies in developing countries: information, persuasion, incentives and coercion [28]. The use of incentives to promote interventions and change in behaviour remains a controversial issue [21, 29]. The outcome of this study suggests that if used appropriately incentives can positively affect performance. In the study by Wasunna et al. [22], health care workers at facility level demonstrated concern on the sustainability of a program that provides expensive drugs free to the people and the fear of drug being out-of-stock resulted in non-adherence by the health workers to distribute drug as recommended. In an effort to ensure stocks last, the health workers prescribed AL only to those they perceived to have serious illness. Could the perception that health workers hoarded drug in this study be related to the perceive fear exhibited by health workers in Wassuna et al. study but unspoken?

One way to address the issue of sustainability of community based programme is community partnership and participation strategy [30]. Efforts were made to address this during the study. The engagement of the CMDs including selected mothers was a demonstration of community participation. However, more could not be done due to the prevailing free drug policy by the government. A caveat to the use of this strategy is that organizers of most community based interventions seem to assume that because malaria is a major cause of mortality community members will volunteer their time free of charge to be trained, or train others, to act as health educators and drug distributors [19]. However, this has been shown to be fraught with problems. This was suggested in this study in which the CMDs suggested the need to provide commission to make up for the profit they would have made selling other antimalarial drug and time used in carrying out the assignment despite the fact that it was made known to them that it was a community volunteering work.

To scale up HMM using ACTs the government needs to consider these challenges seriously and addressed them in order to achieve successful implementation. In Nigeria, the private sector, which consists of a great variety of providers including CMDs, is used by a wide cross-section of the population [31]. Using the private sector to distribute drugs is a good option to address the challenges faced by consumers in the public sector. However, the practices of medicine sellers in low-income countries have been demonstrated to be fraught with many incorrect practices and these constitute a major challenge to working with them. Infringement of health-related regulation is extremely common [32]. Most stores lack permit, stock unregistered and/or expired drugs, stock prescription-only medicines, sell drugs at inflated prices and the staff do not have the minimum qualification required. The evidence is limited as to which approaches work best for public-private collaboration to improve medicine sellers-based treatment. Eliminating regulatory infringement is unlikely to be feasible and could be undesirable if access to essential medicines is reduced [32]. Alternatives include bringing official drug regulation closer into line with locally legitimate practices; greater use of positive incentives for providers; and consumer involvement [32]. In this study as mentioned earlier, training and provision of incentives were used and suggestions were made to the involvement of community heads or opinion leaders in supervising and monitoring drug distribution at the community level. However, there have been many references to other approaches such as social marketing, accreditation, franchising and contracting [33]. Further research is necessary to determine which of these strategies would be well suited for use in the context of HMM in Nigeria.