Applying a gender lens to understand pathways through care for acutely ill young children in Kenyan urban informal settlements

Background: In many African settings, gender strongly influences household treatment-seeking and decision-making for childhood illnesses. While mothers are often the primary engagers with health facilities, their independence in illness-related decisions is shaped by various factors. Drawing on a gender lens, we explored treatment-seeking pathways pre- and post-hospital admission for acutely ill young children living in low income settlements in Nairobi, Kenya; and the gendered impact of child illness both at the household and health system level. Methods: Household members of 22 children admitted to a public hospital were interviewed in their homes several times post hospital discharge. In-depth interviews covered the child’s household situation, health and illness; and the family’s treatment-seeking choices and experiences. Children were selected from an observational cohort established by the Childhood Acute Illness and Nutrition (CHAIN) Network. Results: Treatment-seeking pathways were often long and complex, with mothers playing the key role in caring for their children and in treatment decision-making. Facing many anxieties and dilemmas, mothers often consulted with significant influencers – primarily women – particularly where illnesses were prolonged or complex. In contrast to observations in rural African contexts, fathers were less prominent as influencers than (often female) neighbours, grandparents and other relatives. Mothers were sometimes blamed for their child’s condition at home and at health facilities. Children’s illness episode and associated treatment-seeking had significant gendered socio-economic consequences for households, including through mothers having to take substantial time off work, reduce their working hours and income, or even losing their jobs. Conclusion: Women in urban low-income settings are disproportionately impacted by acute child illness and the related treatment-seeking and recovery process. The range of interventions needed to support mothers as they navigate their way through children’s illnesses and recovery include: deliberate engagement of men in child health to counteract the dominant perception of child health and care as a ‘female-domain’; targeted economic strategies such as cash transfers to safeguard the most vulnerable women and households, combined with more robust labour policies to protect affected women; as well as implementing strategies at the health system level to improve interactions between health workers and community members.

This work was undertaken as part of a broader international, multi-disciplinary research network known as the Childhood Acute Illness and Nutrition (CHAIN) Network described elsewhere [23]. The CHAIN Network has nine sites in six countries within Africa and Asia, including Kenya. The overall aim of the CHAIN Network is to identify the biological mechanisms and the socio-economic factors that determine a child’s risk of mortality in the 6 months following presentation to medical care with an acute illness [23]. As part of this broader goal, a qualitative social science sub-study was undertaken in Nairobi in urban Kenya. This qualitative longitudinal study was conducted in two urban informal settlements of Nairobi County. Twenty-two families of acutely ill children aged 2–23 months with varying nutritional status who had been admitted to the study hospital and enrolled in the CHAIN observational cohort [23] were followed up over an 18-month period post-hospital discharge. Nutritional status is known to have a major influence on child survival, with undernutrition having a synergistic effect with acute illness [24, 25]. The initial intention was to purposively select equal numbers of families across three strata of nutritional status: severe wasting or kwashiorkor (SWK), moderate wasting (MW) and no wasting (NW); with varying socio-economic vulnerability (based on prospectively collected data on maternal education, household access to financial and social resources and household size). In practice, extended public health worker strikes and ensuing hospital closures significantly impacted our recruitment and led to us selecting families based on residence in two low-income settlements (Kibera and Mathare). We finally included eight, eleven and three households of children with SWK, MW and NW respectively, of varying socio-economic vulnerability. Kibera – the largest urban informal settlment in East Africa – was selected based on its proximity to the study hospital. The majority (18/22) of families in this study resided in Kibera. According to data collected by the Africa Population Health Research Centre (APHRC), as of 2012 Kibera had an infant mortality rate of 33.2 deaths per 1000 live births, post neonatal mortality rate of 45.3 deaths per 1000 live births and a total under-five mortality rate of 78.5 deaths per 1000 live births [17]. The remaining four families resided in Mathare, a populous informal settlement consisting of a constellation of slums, that is located a few kilometers north east of Nairobi’s central business district [26]. In addition to being the second largest informal settlement in Kenya, Mathare was selected based on previous existing relationships with key community stakeholders and gatekeepers, which helped to ease our entry into the community to undertake the household follow-ups. Each household was visited 2–3 times during the follow-up period, totaling 58 visits across the 22 households. For ethical reasons, household follow-ups (beyond a standard condolence visit) were discontinued in three of the households where the target child died following our initial visits. In-depth interviews conducted during the household visits were primarily with the children’s main carers and other family members, and covered a broad range of topics including: child health and nutrition; the child’s illness trajectory and related treatment-seeking and decision-making; experiences with the admitting hospital and the health system more broadly; as well as challenges faced during the child’s illness episode and coping strategies. All interviews were audio-recorded. Written informed consent was obtained from all participants in the initial household visit with verbal consent obtained in all subsequent visits to continue in the study. Non-participant observations were also conducted at the admitting hospital and at household level to give a sense of living conditions and family dynamics, community relations, as well as experiences and interactions at the hospital. Data were analysed using a modified framework approach. This entailed: extensive familiarization with the data (‘immersion’ in the data by reading and re-reading of transcripts, listening to audio-recordings and reading field notes); condensing the data into detailed summary sheets per household across visits and subsequently developing and refining a master summary for all households across all visits; consultatively developing a coding framework based on preliminary emergent themes and the study objectives; and coding the entire dataset into NVivo software to search for broader emergent themes. Comparison tables were also developed to identify patterns for example based on child nutritional status or household structure and headship; as well as ‘rich stories’ for each household to ensure that the broader narrative of each household was preserved. Concurrently with the framework approach, gender analysis was undertaken drawing on the gender framework by Morgan et al. [1] (Table 1). According to this framework, gender analysis can be incorporated into health systems research content, process and outcomes. For our analysis, we focused on the content and specifically examined gender relations at household and community level in the context of childhood acute illness, including exploring: who has what (access to resources); who does what (the division of labour and everyday practices); how values are defined (social norms) and who decides (rules and decision-making) [1]. We also reviewed our data for any illustrations of how power is negotiated and changed in households and communities. Gender as a power relation and driver of inequality

N/A

• Dima Health