Examining the spatio-temporal patterns of diarrhoeal disease incidence from routinely collected clinical data

Diarrhoeal disease amongst children under five years of age is a significant public health issue in Kenya,
and has previously been linked to geographic risk factors. Health facilities collect and generate data to
support the delivery of routine patient care, and could therefore be a viable source of population-level
health intelligence to optimise and target interventions to those people and localities with greatest need.
A better understanding of demand and utilisation at individual facilities could also reveal gaps in
coverage and improve the overall efficiency of the health system.
However, routine health data has historically been under-utilised in sub-Saharan Africa owing to concerns
around its completeness and accuracy. This study will evaluate the quality of the routine data generated
by health facilities in Kenya, and will develop methods to facilitate its use for analytical purposes. It will
then quantify the incidence of childhood diarrhoeal disease by health facility, describe the variation
between facilities, and create statistical models describing the association with patient-level and
geographic risk factors.
As real-world data will be used, this study could have immediate impact in Kenya if its recommendations
are adopted by health service managers, and prove useful in the development of interventions to reduce
incidence and improve health outcomes of childhood diarrhoeal disease. Developing working
relationships with study collaborators in Kenya could also help to build analytical capacity within the
health service, and could contribute to a broader case for greater investment in systems facilitating the
recording, storage and analysis of routine health data.