Environmental sanitation provision among women with children in Kumasi, Ghana

Ghana is both one of the lowest coverage countries for urban sanitation, and one of the furthest ‘off track’ to achieving universal safely-managed sanitation in urban areas by 2030. It is also the country with the highest levels of urban shared sanitation, at 56.4%. Responding to the lack of focused research on the environmental sanitation priorities of vulnerable and marginalised people, the thesis identifies a population, women caring for young children in low-income and informal communities of Kumasi, Ghana, who are at increased risk of the negative health and social costs of poor environmental sanitation and are marginalised. This research takes the priorities of this vulnerable population as the starting point of inquiry, rather than imposing priorities based on disease prevalence or known rates of infrastructure use such as open defecation. 

The aim of this research is to explore and explain how carers of children under five years old perceive the relationships between environmental sanitation and child health. The findings identify key determinants of the behaviours women prioritise for child health. The study employs an exploratory sequential mixed methods design, and therefore begins with a qualitative objective followed by a quantitative objective. These are analysed separately, before being combined into a synthesis analysis bringing both qualitative and quantitative data together. 

Objective 1– Qualitative: To explore environmental sanitation priorities, beliefs and behaviours identified by carers as important for child health. 

Objective 2 – Quantitative: To validate the main qualitative findings, using a bespoke quantitative survey informed by the grounded theory, by measuring the prevalence of key practices and relationships identified in the qualitative data analysis. The survey data collection used a novel application of the random-walk sampling method to capture several hard-to-reach populations.

The grounded theory identifies the priority child-health concerns of women in Kumasi zongos as malaria and diarrhoea, and explains how these influence, and are influenced by, three categories of behaviour: hygiene, services and polluting. These behaviours are defined by five characteristics – their cost in time, in effort, or in money, the domain at which they influence environmental sanitation, and whether women are able to secure them (e.g. for hygiene, can they collect enough water; for services are they available in the community). 

Polluting behaviours exacerbate the neighbourhood-domain environmental sanitation problems that were identified by women as driving malaria and diarrhoea risks. Polluting behaviours of particular concern revolve around the combined impact of poor solid waste management, poor drainage, and lack of on-site sanitation, and in particular the disposal of faeces with solid waste.

This study quantified the impact of caring for children or infants on the likelihood of disposing of faeces with solid waste, and the impact of relying on sanitation services outside the home. Using a combined logistic regression model, it shows that caring for children increased the odds of faeces-disposal with solid waste by 23.95 times, and the lack of sanitation on plot increased the odds by 16.67 times.

Child faeces is expected to be disposed of with household waste, and not just infant (0-2 years old) stools. The storage of this waste in the home, even in an uncovered bucket or sack, is not seen as polluting behaviour. Rather it is both acceptable and convenient, as it allows women to postpone the expense of both sanitation and solid waste collection until finances allow for it.

Convenience is a crucial factor in determining whether women will use a particular waste disposal option, with 7 out of 10 respondents agreeing with this statement. Absence of sanitation on-plot, when combined with poor household waste management, poor drainage, and childcare, hugely increases the likelihood of faeces returning to both the home and the neighbourhood environment.

The relationships between the behaviours and their determinants fits within the matrix structure of the IBM-WASH model of WASH behaviour, so the findings from the grounded theory suggest that solid waste management can and should be researched using the IBM-WASH matrix as a guiding conceptual model. 

This study demonstrates how a mixed methods approach can be operationalised to reveal lived realities. The sophisticated combination of qualitative and quantitative data in sequential mixed methods allowed the deep and rich qualitative findings from in depth interviews and focus groups to inform quantitative data collection that further contextualises, explains and validates the qualitative analysis.