Background: There have been few studies about the basis on which women in developing regions evaluate and choose traditional rather than western maternal care. This qualitative study explores the socio-cultural perceptions of complications associated with pregnancy and childbirth and how these perceptions influence maternal health and care-seeking behaviours in Kenya. Methods: Kalenjin women (n = 42) aged 18-45 years, who were pregnant or had given birth within the last 12 months, were interviewed. A semi-structured interview guide was used for data collection. A further nine key informant interviews with Traditional Birth Attendants (TBAs) who were also herbalists (n = 6), community health workers (CHWs) (n = 3) and a Maternal and Child Health (MCH) nursing officer (n = 1) were conducted. The data were analysed using MAXQDA12 software and categorised, thematised and analysed based on the symbolic dimensions of Helman’s (2000) ill-health causation aetiologies model. Results: Pregnancy complications are perceived as the consequence of pregnant women not observing culturally restricted and recommended behaviour during pregnancy, including diet; physical activities; evil social relations and spirits of the dead. These complications are considered to be preventable by following a restricted and recommended diet, and avoiding heavy duties, funerals, killing of animals and eating meat of animal carcasses, as well as restricting geographical mobility, and use of herbal remedies to counter evil and prevent complications. Conclusion: Delay in deciding to seek maternal care is a result of women’s failure to recognise symptoms and maternal health problems as potential hospital cases, and this failure stems from culturally informed perceptions of symptoms of maternal morbidity and pregnancy complications that differ significantly from biomedical interpretations. Some of the cultural maternal care and remedies adopted to prevent pregnancy complications, such as restriction of diet and social mobility, may pose risks to the pregnant woman’s health and access to health facilities whereas other remedies such as restricting consumption of meat from animal carcasses and heavy duties, as well as maintaining good social relations, are cultural adaptive mechanisms that indirectly control the transmission of disease and improve maternal health, and thus should not be considered to be exclusively folk or primitive.
The major focus of this qualitative study was on the cultural interpretation of pregnancy complications and the preventive and treatment remedies adopted. This study is part of broader research investigating the socio-cultural context of maternal nutrition and health in rural Uasin Gishu County in western Kenya. Data were collected between April and August 2015 from Kalenjin women, either pregnant or with a child of less than 1 year, seeking care at the government health facilities in the Maternal and Child Health (MCH) care section. The Kalenjin is the main ethnic population in Uasin Gishu County and comprises eight sub-ethnic groups (the Kipsigis, Nandi, Tugen, Keiyo, Marakwet, Pokot, Sabaot and the Terik) that share a common dialect and similar cultural traits. Among the Kalenjin speakers, each sub-ethnic group has its own distinctive dialect. The Nandi occupies the largest settlement in Uasin Gishu County, followed by the Keiyo. All 90 public health facilities in the county were included in the sampling frame. Quota and purposive sampling techniques were employed in the selection of a representative sample of health facilities for the study. The selection criteria included ensuring that the health facilities in all the six-quotas (sub-counties) are proportionately represented in the sample. All the health facilities must be in the rural area (outside the municipality territory) and have a catchment population mainly comprising at least 90% Kalenjin patients to enhance cultural homogeneity. This means that areas dominated by other non-Kalenjin ethnic groups and those within the municipal boundaries were eliminated. The last criterion is that the selected facilities should be spatially distributed from each other to diversify responses. In the end, a total of 23 health facilities were sampled for the study. All the Kalenjin women who come for routine antenatal and post-partum child welfare check-ups in the sampled health facilities were included in the sampling frame. They were recruited at the MCH clinics and in maternity wards. Eligibility criteria for the study participants depended on: being pregnant or having given birth within the last year, a Kalenjin by birth, willing and able to participate in the study, able to give informed consent [16] and willing to be audio recorded. This selection criterion eliminated non-Kalenjin women and those not willing to be audio recorded. Data were collected until the information reached saturation at a sample size of 42 women [17]. Nine key informants, including six TBAs who are also herbalists, one CHW, and one nursing officer in charge of MCH, were also selected for an interview. Quota sampling and purposive sampling techniques were used in the selection of key informants [17]. One TBA from each of the six sub-counties, who was highly mentioned by women respondents who had given birth at home or took herbal remedies during pregnancy, was selected and they could be reached at home or in the market centre. The CHW and nursing officer were selected from one of the largest rural facilities in the county because they are likely to encounter a wide range of pregnancy experiences and challenges given their large catchment area. In total, six TBAs who were also herbalists, one nursing officer offering MCH care, and one CHW were recruited. An open-ended interview guide (Additional file 1), divided into four sections, was used to elicit the information from the Kalenjin women. The first section presented demographic characteristics of the respondents including age, educational level, parity, ethnic affiliation and gestational age at the first ANC visit, marital status, and tribal affiliation among others. The other sections contained questions about food restrictions, recommended food, activities restricted and activities encouraged during pregnancy. Every practice mentioned was probed to obtain an insight into the underlying reasons. The respondents were further questioned about their opinions regarding these cultural practices and whether they indeed practised them. Face-to-face individual interviews were conducted in a private room. Each woman was interviewed once and the interview lasted between 30 and 60 min, depending on her responses. Key informant interviews (KIIs) followed later to provide clarity on the issues raised during the interviews. The KIIs lasted between an hour and 2.5 h. They were also questioned about the kind of advice they give pregnant women and health challenges they face when providing care to pregnant women. Important notes were taken and at the same time responses were audio recorded. The study was approved by the National Commission for Science, Technology and Innovation (NACOSTI) in Kenya and a research clearance permit number: NACOSTI/P/15/2335/5353 dated 2 April 2015 was issued to facilitate the research process. As approved by NACOSTI, the permit was then presented to the Uasin-Gishu County Commissioner, County Director of Education and County Director of Health, for their approval to conduct the study in the County. Further, appointments were booked with the respective officers in charge of the various facilities visited. Participation was voluntary. The respondents were informed of the aim of the research, confidentiality and anonymity of their responses, and then gave their signed consent to participate. Permission to audio record the interview sessions was sought from each respondent. Only voices for those who consented were recorded. Recorded responses were transcribed and, together with field notes, were studied by way of content analysis using MAXQDA 12.0.3 software. Helman’s [15] classification of lay-illness aetiologies model was adopted as the initial coding guide. Meanings attributed to various adverse pregnancy outcomes were established in the data and were classified into four major categories based on Helman’s [15] symbolic classification of lay-illness causation aetiologies model: individual, natural, social and supernatural, as illustrated in Fig. 1. The categories were further classified into sub-categories and themes as interpreted below. Sites of illness aetiology (Helman, 2000:120) These include lay theories that locate the meaning of pregnancy complications in the individual woman for “not taking care” of herself in terms of diet, dress, hygiene, lifestyle, relationships, sexual behaviour, smoking and drinking habits, physical exercise, emotions or doing something abnormal or incorrect. An adverse pregnancy outcome is, therefore, evidence of “carelessness” and the woman should feel guilty and responsible for causing it. However, in some rare circumstances, individual causes can result from external forces over which the victim had no control such as bad luck, economic power or hereditary factors. In this category, an adverse pregnancy outcome is thought to be caused by the natural environment, both living and inanimate. Common in this group are climatic conditions, such as excess cold, heat, wind, rain, snow, damp, cyclones, tornadoes, eclipse or severe storms. Others include accidental injuries which originate from the “natural environment”, or are caused by animals, birds, insects, or infections caused by micro-organisms, such as germs, bugs or viruses. This category involves blaming other people for causing adverse pregnancy outcomes and is a common feature of non-industrialised and smaller-scale societies, where interpersonal conflicts are frequent. The common forms of these are witchcraft, sorcery and “evil eyes”. In witchcraft, certain people are believed to possess a mystical power to harm others and this power is inherited, either genetically or by membership of a particular kinship group. Sorcery, as defined by Helman [15], is the power to manipulate and alter natural and supernatural events with the proper magical knowledge and performance of rituals, and this is different from witchcraft. Sorcery is often practised among one’s social world of friends, family or neighbours, and is often based on envy. Evil eyes, or a “wounding eye”, relates to the fear of envy in the eyes of the beholder. The possessors of evil eye are usually believed to harm unintentionally and are often unaware of their powers and are unable to control them. The influence of evil eye, as explained by Helman [15], is avoided or counteracted by means of devices calculated to distract its attention, and by practices of sympathetic magic. The social aetiology of illness also includes physical injuries, such as poisoning or battle wounds, inflicted by other people. Furthermore, it can be stress or actions caused by spouse, children, friends, employer or colleagues and neighbours. It can also be contagious diseases transmitted by other people. Here a pregnancy complication is ascribed to the direct actions of supernatural entities, such as gods, spirits or ancestral shades. In the gods aetiology, illness is described as a reminder from God for a behavioural lapse or sinful behaviour. The cure in this case involves acknowledging the sins and vowing to improve one’s behaviour. In the case of spiritual causes, disease-bearing spirits strike unexpectedly causing a variety of symptoms in their victims. Their invasion is unrelated to the individual’s behaviour, who is therefore considered blameless and worthy of sympathetic help from others. In the case of ancestral shades causes, a pregnancy complication is ascribed to spirits of the ancestors whom they have offended and diagnosis takes place in a divinatory séance.