Prevention and Management of Malaria During Pregnancy
Prevention and Management of Malaria During Pregnancy
The results presented in this article are drawn from a comparative study conducted at four sites in three countries. The study was undertaken by a team of researchers whose members were based across the sites and in Barcelona (Spain).
The study incorporated one country from each of the three main regions of Sub-Saharan Africa: Ghana in West Africa, Kenya in East Africa and Malawi in Southern Africa. Two sites with important regional specificities were selected in Ghana for several reasons: to collect data in at least one site of each of the MiP Consortium's main treatment and prevention activities; to include areas with different patterns of malaria transmission; and to examine intra as well as inter-country variation.
In central Ghana, fieldwork was conducted in two districts of the Ashanti Region: Ejisu Juaben and Ahafo Ano South. In both districts, agriculture is the main productive activity and there is a significant proportion of internal migrants, in addition to the majority ethnic group, the Asante. At this site, malaria transmission is moderately high and occurs throughout the year with peaks during the rains in May-October. In each district, data collection was conducted at the district hospitals, two to three health centres and several smaller clinics.
In northern Ghana, Upper East Region, the fieldwork sites were located in Kassena-Nankana District. This area is part of the Sahel and experiences only one annual rainy season during which people grow millet, maize and vegetables for subsistence. During the rest of the year, part of the population migrates to other regions. The Kassena and the Nankani, make up almost 90% of the population of the district. Here, malaria transmission is perennial but there is a seasonal pattern with a transmission peak that coincides with the major rains (May to October) and the low rates of infection during the dry season. Data were collected at a district hospital in Navrongo, (the capital), and outreach community-based services, which are common throughout the area.
Fieldwork also took place in Chikwawa and Blantyre Districts, in the southern region of Malawi. The main ethnic groups in Blantyre District are Chewa and Yao, whereas in Chikwawa they are Manganja and Sena. Most of the women in the area cultivate crops for subsistence and sale at the market. Both districts are in areas of high perennial malaria transmission. Fieldwork took place at three hospitals, and six healthcare centres that provide ANC services to the women in these areas.
Finally, in Kenya, fieldwork was carried out in Siaya District (Nyanza Province) where the principal ethnic group, the Luo, make up over 95% of the population. Livelihood activities include subsistence farming of maize, sorghum, millet and cassava. As a result of the relatively limited employment opportunities, migration to urban centres is common, particularly to Kisumu, the nearest city. Malaria transmission is high and perennial with the greatest disease burden borne by children and pregnant women. Data were collected at the district hospital and smaller health facilities where ANC is delivered.
At each of the sites, various clinical and non-clinical studies of MiP prevention and control interventions have been undertaken. Some of these studies overlapped with data collection for this research. Therefore, during data collection and analysis, efforts were made to exclude experiences of MiP prevention and control within clinical or non-clinical research. Furthermore, throughout this article, for reasons of brevity, the sites are referred to as "Kenya", "Malawi", "central Ghana" and "northern Ghana". This shorthand should not however be interpreted as any attempt at regional or national generalization.
An anthropological approach was taken to data collection. This entailed year-long (or longer) periods of fieldwork at each site, a range of data collection activities, including narrative and observational techniques (see Table 1 for a full list of data collection activities), and a flexible, reflexive and iterative process of tool design, data collection, and analysis. The use of multiple data collection tools with heterogeneous respondents ensured that findings could be triangulated and their reliability tested. To reduce the possible influence of individual bias on the study findings, at each site, several researchers collected data.
Fieldwork was carried out between April 2009 and August 2011, and lasted from one year in Malawi to more than two years in central Ghana. Assisted by two Barcelona-based researchers (AM and CP), fieldworkers spent extended periods of time in the settlements where data were collected and recorded their experiences of participant observation in field diaries. In the first phase, at each site, using free-listing and sorting exercises, the research team explored the main problems that pregnant women experience. Later, interviews and group discussions were conducted, several women (case studies) were followed and interviewed several times over the course of their pregnancies, and observations were carried out in the communities and at local health facilities. The language used to interact with informants depended on their preferences (English and different local languages). In-depth interviews and group discussions were recorded, transcribed and translated into English.
In-depth interviews tended to start with broad research questions related to pregnancy and ended with questions about malaria in pregnancy and experiences with malaria prevention and control. In contrast, group discussions often started with general questions about malaria, focusing later on groups particularly vulnerable to malaria and finalizing with malaria prevention and control. Other themes, related to MiP, such as miscarriage, stillbirths, pre-term deliveries, birth weight and anaemia – and their causes – were also explored during fieldwork. Data collection and analysis were carried out in parallel allowing the incorporation of emerging themes in the design of the tools, and questions' redefinition and attuning.
Members of the Barcelona-based research team made quarterly visits to the study sites. During these visits, a process of debriefing and reflection took place with fieldworkers. The Barcelona-based researchers were also able to participate in data collection, and provide ongoing training.
Five main categories of respondents were interviewed (Table 1): pregnant women, their relatives, community members, opinion leaders and healthcare providers. Purposive sampling was used to ensure the interaction with a wide range of experiences. Married and unmarried pregnant women of a range of ages, parities and gestational ages from across the different settlements (within the field sites) were interviewed. Relatives included mainly mothers, mothers-in-law and husbands of the pregnant women. The sample of opinion leaders was made up of a variety of religious leaders, traditional and political authorities, and relevant women in the local communities. Finally, ANC staff, pharmacists and drug sellers, traditional birth attendants (TBAs), and other healers (who attended to pregnant women or dealt with malaria) were interviewed at each site. Respondents were identified in ANC clinics and via contacts in the local communities, which developed as fieldwork went on. The final number of participants was a result of the directed sampling and the point of saturation, whereby no further novel insights were identified from interviews.
At each site, a first phase of data analysis ran in parallel to data collection. Using Atlas.ti 6, flexible codebooks were developed and revised using a combination of established categories based on the original research questions and themes that emerged from the data. Particular attention was paid when analyzing the interviews with case studies to identify changes in a woman's responses over the course of her pregnancy, for example, with regard to ITN use. The preliminary results obtained from this site-specific analysis were compared and discussed amongst the members of the team in periodic meetings throughout data collection. In a second phase, data associated to the codes relevant to malaria in pregnancy perceptions, were extracted, collated and discussed between authors one and two, looking at the similarities, differences and variations between and within the different sites.
Overall ethics clearance was obtained from the Clinical Research Ethics Committee, Hospital Clinic-University of Barcelona. Separate local ethics clearance was obtained at each site: in Ghana, clearance was obtained from the Institutional Review Board of the Navrongo Health Research Centre, Navrongo and the Committee on Human Research Ethics, Kwame Nkrumah University of Science & Technology, Kumasi; in Kenya, clearance was obtained from the Institution Review Board of Centers for Disease Control and Prevention, Atlanta and from the National Ethics Review Committee, Kenya Medical Research Institute, Nairobi; and in Malawi, clearance was obtained from the College of Medicine Research and Ethics Committee. As approved by all ethics review committees and institutional review boards, informed consent was obtained orally from study participants. Oral rather than written informed consent was obtained because the study procedures posed minimal risk to study participants and to avoid the possible negative influence of a written consent on rapport between researchers and respondents. With the agreement of participants, verbal consent was voice recorded prior to each interview or focus group discussion.
Methods
The results presented in this article are drawn from a comparative study conducted at four sites in three countries. The study was undertaken by a team of researchers whose members were based across the sites and in Barcelona (Spain).
Settings
The study incorporated one country from each of the three main regions of Sub-Saharan Africa: Ghana in West Africa, Kenya in East Africa and Malawi in Southern Africa. Two sites with important regional specificities were selected in Ghana for several reasons: to collect data in at least one site of each of the MiP Consortium's main treatment and prevention activities; to include areas with different patterns of malaria transmission; and to examine intra as well as inter-country variation.
In central Ghana, fieldwork was conducted in two districts of the Ashanti Region: Ejisu Juaben and Ahafo Ano South. In both districts, agriculture is the main productive activity and there is a significant proportion of internal migrants, in addition to the majority ethnic group, the Asante. At this site, malaria transmission is moderately high and occurs throughout the year with peaks during the rains in May-October. In each district, data collection was conducted at the district hospitals, two to three health centres and several smaller clinics.
In northern Ghana, Upper East Region, the fieldwork sites were located in Kassena-Nankana District. This area is part of the Sahel and experiences only one annual rainy season during which people grow millet, maize and vegetables for subsistence. During the rest of the year, part of the population migrates to other regions. The Kassena and the Nankani, make up almost 90% of the population of the district. Here, malaria transmission is perennial but there is a seasonal pattern with a transmission peak that coincides with the major rains (May to October) and the low rates of infection during the dry season. Data were collected at a district hospital in Navrongo, (the capital), and outreach community-based services, which are common throughout the area.
Fieldwork also took place in Chikwawa and Blantyre Districts, in the southern region of Malawi. The main ethnic groups in Blantyre District are Chewa and Yao, whereas in Chikwawa they are Manganja and Sena. Most of the women in the area cultivate crops for subsistence and sale at the market. Both districts are in areas of high perennial malaria transmission. Fieldwork took place at three hospitals, and six healthcare centres that provide ANC services to the women in these areas.
Finally, in Kenya, fieldwork was carried out in Siaya District (Nyanza Province) where the principal ethnic group, the Luo, make up over 95% of the population. Livelihood activities include subsistence farming of maize, sorghum, millet and cassava. As a result of the relatively limited employment opportunities, migration to urban centres is common, particularly to Kisumu, the nearest city. Malaria transmission is high and perennial with the greatest disease burden borne by children and pregnant women. Data were collected at the district hospital and smaller health facilities where ANC is delivered.
At each of the sites, various clinical and non-clinical studies of MiP prevention and control interventions have been undertaken. Some of these studies overlapped with data collection for this research. Therefore, during data collection and analysis, efforts were made to exclude experiences of MiP prevention and control within clinical or non-clinical research. Furthermore, throughout this article, for reasons of brevity, the sites are referred to as "Kenya", "Malawi", "central Ghana" and "northern Ghana". This shorthand should not however be interpreted as any attempt at regional or national generalization.
Data Collection
An anthropological approach was taken to data collection. This entailed year-long (or longer) periods of fieldwork at each site, a range of data collection activities, including narrative and observational techniques (see Table 1 for a full list of data collection activities), and a flexible, reflexive and iterative process of tool design, data collection, and analysis. The use of multiple data collection tools with heterogeneous respondents ensured that findings could be triangulated and their reliability tested. To reduce the possible influence of individual bias on the study findings, at each site, several researchers collected data.
Fieldwork was carried out between April 2009 and August 2011, and lasted from one year in Malawi to more than two years in central Ghana. Assisted by two Barcelona-based researchers (AM and CP), fieldworkers spent extended periods of time in the settlements where data were collected and recorded their experiences of participant observation in field diaries. In the first phase, at each site, using free-listing and sorting exercises, the research team explored the main problems that pregnant women experience. Later, interviews and group discussions were conducted, several women (case studies) were followed and interviewed several times over the course of their pregnancies, and observations were carried out in the communities and at local health facilities. The language used to interact with informants depended on their preferences (English and different local languages). In-depth interviews and group discussions were recorded, transcribed and translated into English.
In-depth interviews tended to start with broad research questions related to pregnancy and ended with questions about malaria in pregnancy and experiences with malaria prevention and control. In contrast, group discussions often started with general questions about malaria, focusing later on groups particularly vulnerable to malaria and finalizing with malaria prevention and control. Other themes, related to MiP, such as miscarriage, stillbirths, pre-term deliveries, birth weight and anaemia – and their causes – were also explored during fieldwork. Data collection and analysis were carried out in parallel allowing the incorporation of emerging themes in the design of the tools, and questions' redefinition and attuning.
Members of the Barcelona-based research team made quarterly visits to the study sites. During these visits, a process of debriefing and reflection took place with fieldworkers. The Barcelona-based researchers were also able to participate in data collection, and provide ongoing training.
Respondents
Five main categories of respondents were interviewed (Table 1): pregnant women, their relatives, community members, opinion leaders and healthcare providers. Purposive sampling was used to ensure the interaction with a wide range of experiences. Married and unmarried pregnant women of a range of ages, parities and gestational ages from across the different settlements (within the field sites) were interviewed. Relatives included mainly mothers, mothers-in-law and husbands of the pregnant women. The sample of opinion leaders was made up of a variety of religious leaders, traditional and political authorities, and relevant women in the local communities. Finally, ANC staff, pharmacists and drug sellers, traditional birth attendants (TBAs), and other healers (who attended to pregnant women or dealt with malaria) were interviewed at each site. Respondents were identified in ANC clinics and via contacts in the local communities, which developed as fieldwork went on. The final number of participants was a result of the directed sampling and the point of saturation, whereby no further novel insights were identified from interviews.
Data Analysis
At each site, a first phase of data analysis ran in parallel to data collection. Using Atlas.ti 6, flexible codebooks were developed and revised using a combination of established categories based on the original research questions and themes that emerged from the data. Particular attention was paid when analyzing the interviews with case studies to identify changes in a woman's responses over the course of her pregnancy, for example, with regard to ITN use. The preliminary results obtained from this site-specific analysis were compared and discussed amongst the members of the team in periodic meetings throughout data collection. In a second phase, data associated to the codes relevant to malaria in pregnancy perceptions, were extracted, collated and discussed between authors one and two, looking at the similarities, differences and variations between and within the different sites.
Ethics Statement
Overall ethics clearance was obtained from the Clinical Research Ethics Committee, Hospital Clinic-University of Barcelona. Separate local ethics clearance was obtained at each site: in Ghana, clearance was obtained from the Institutional Review Board of the Navrongo Health Research Centre, Navrongo and the Committee on Human Research Ethics, Kwame Nkrumah University of Science & Technology, Kumasi; in Kenya, clearance was obtained from the Institution Review Board of Centers for Disease Control and Prevention, Atlanta and from the National Ethics Review Committee, Kenya Medical Research Institute, Nairobi; and in Malawi, clearance was obtained from the College of Medicine Research and Ethics Committee. As approved by all ethics review committees and institutional review boards, informed consent was obtained orally from study participants. Oral rather than written informed consent was obtained because the study procedures posed minimal risk to study participants and to avoid the possible negative influence of a written consent on rapport between researchers and respondents. With the agreement of participants, verbal consent was voice recorded prior to each interview or focus group discussion.
