The role of water in cholera dffusion: improvements of a cholera di�usion model for Kumasi, Ghana

Doldersum, T. (2013) The role of water in cholera dffusion: improvements of a cholera di�usion model for Kumasi, Ghana.

Abstract:Cholera is still one of the most feared infection diseases, especially in those countries where clean drinking water is not available o the local people. In order to develop an e�ective strategy against the spread of cholera it is important to understand the behaviour of cholera which can be studied by simulating it. Therefore, Useya (2011) developed an agent based model that simulates the spread of cholera for the outbreak of 2005 in Kumasi, Ghana. Although the the number of cholera cases. Crucial for the model performance to simulate the geographical distribution well is a process where the probability to get infected with cholera depends on the living location within the study area. Furthermore, the model shows that removing dump sites that are situated close to the river resulted in a decrease in the number of cholera cases. Finally recommendations for follow up studies are: � Pay attention to processes that: contaminate rain water that drops on the dump site and describe the behaviour of V. cholerae in the river. � Extend the activity model with activities that can explain the upstream movement of the cholera epidemic. For example the mobility of individuals to work, markets or family. � After the model improvements have been done, perform a validation for either another catchment of Kumasi for which data are available or another cholera outbreak in an area with characteristics similar to this study area. developed cholera model was able to reproduce the epidemic curve of cholera, it contained some parts, particularly the hydrological part, that should be improved before it can be used for more practical purposes. This study focused on the improvement of these parts. Therefore the research objective is: Improve the model of Useya (2011) by implementing the hydrological processes that play a major role in the spread of V. cholerae, to gain more insight in the spread of cholera via water and use the model to evaluate different scenarios to make the strategy against cholera more effective. An analysis of the original hydrological procedure shows that: (i) the velocity of the water was unrealistically low and (ii) the study area was rather small. To improve this procedure the velocity of water is now based on the formula of Manning and the study area is enlarged to the catchment area. The next step was to calibrate the procedure. Since no discharge data were available, the simulated discharges were compared with the outcomes of a generally accepted model: the Curve Number method. The calibration resulted in a Nash-Sutcli�e of 0.95 and a Relative Volume Error of �0:2% for the epidemic period. The model was validated for the same period of the years 2006, 2009 and 2010 resulting respectively in a NS of 0.92, 0.94 and 0.94 and for all years the RVE was lower than �0:3%. The main improvements made to the original cholera model are: (i) removing inconsistencies between the model and the report of Useya (2011) and (ii) adapt the model's procedures to the enlarged study area. This is done to compare the simulated and available data to study the cholera di�usion within the study area. The model was calibrated on the relative contribution of each community (eleven are taken into account) to the total number of cholera cases. Since it is widely accepted that there is a substantial under-reporting of cases, the use of absolute values would not be appropriate. The calibration resulted in a r2 of 0.87. This means that the model was able to reproduce the geographical distribution well. The resulting epidemic curve and the contribution of the transmission mechanisms are both comparable to what is reported in the literature. The main contribution to the number of cases is caused by transmission of cholera via water. The model analysis shows that within the cholera model the hydrological procedure is very important, because 75% of the cholera cases get infected via river water that is contaminated by the runo� from the dump sites. Due to the Environment-Human (EH) transmission procedure the model clearly shows a random spatial pattern of the dffusion process while this is not expected from literature. The model is quite sensitive to the scheduling of the daily activities and changes to the bacteria procedure. The scenario analyses show that there is a strong relation between the epidemic curve and the rainfall. When it rains river water gets infected and the model shows a strong increase in
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
Subject:56 civil engineering
Programme:Civil Engineering and Management MSc (60026)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page