Flooding and Diarrhoeal Disease Incidence in Urban Poor Communities in Accra, Ghana. Investigating the Effects of Risk Perception Measures

Abstract

Flooding is an environmental challenge faced by many nations in the 21 st Century. Studies have shown that recent floods in cities are responsible for the outbreak of cholera and non-cholera diarrhoea, but there is little work on the risk perceptions of diarrhoea resulting from flooding. Also, it is not known how risk perceptions of diarrhoea mediate the relationship between flooding and diarrhoeal disease. The risk perceptions people develop resulting from their exposure to floods is the key driver of the measures they will employ to avert the health effects of it. Measurement of risk perceptions has, therefore become an important topic to policy makers concerned with risk management, communication and safety issues. Accurate measurement of risk perceptions is considered a crucial aspect of managing hazards as it directs the development of hazard mitigation strategies. This study examines how three risk perception measures of diarrhoeal disease explain the flooding and diarrhoeal disease relationship in urban poor communities in Accra. Two main sources of data were used for the study. Firstly, time series data on rainfall from the Ghana Meteorological Agency and data on reported diarrhoea cases from the Center for Health Information Management, Ghana Health Services, for Ashiedu- Keteke sub-Metropolitan Area of Accra were used to examine the relationship between flooding and diarrhoeal disease in the study communities. Second, cross-sectional data from two urban poor neighbourhoods (Agbogbloshie and James Town) in Accra, Ghana, consisting of five enumeration areas (EAs) from Agbogbloshsie and nine enumeration areas from James Town, were used to examine the effects of the three risk perception measures on the relationship between flooding and diarrhoeal disease. A total of 401 households involved in the study were asked to rate the chance of a member of their household being diagnosed with diarrhoeal disease within the first four weeks after the October 26 th 2011 flooding of Accra on a 0-100% numerical risk perception scale and a verbal and comparative risk perceptions scale with five descriptive categories each. The study employed the Granger causality test to examine whether flooding predicts incidence of diarrhoea in the Ashiedu-Keteke sub-Metropolitan area of Accra. To examine the effect of risk perceptions on the relationship between flooding and diarrhoea, binary logistic regression analysis was used. The study revealed that flood Granger cause incidence of diarrhoea in the Ashiedu-Keteke sub-Metropolitan Area of Accra. Experience of the October 26 th , 2011 floods predict incidence of diarrhoea in the study communities. The numeric, verbal and comparative risk perception measures were all significant predictors of incidence of diarrhoea in households. Overall, the numeric risk perception measure did best in identifying households at high or low risk of diarrhoea than the verbal and the comparative risk perception measures. Other significant predictors of diarrhoeal disease were household source of drinking water, distance to the nearest refuse collection point and presence of cockroaches in households. The study recommends research into risk perceptions of people in addressing critical global issues and also calls for education on water and sanitation, which are critical in diarrhoeal disease transmission in the event of flooding.