Title: Data-driven decision making - Guiding organizations towards an “AI proof” status

Abstract:

Daniel Amobtiwon Amoatika is a nurse by profession and works with one of the private hospitals in Ghana. He holds a BSc in Nursing and currently pursuing a Master of Philosophy degree in Applied Epidemiology and Diseases Control at the School of Public Health, University of Ghana. Daniel has 2 years of public practice and 3 years of private practices as a nurse. He has also been involved in outbreak investigations of infectious diseases in Ghana. He has presented an abstract on money handling practices among food vendors in the University of Ghana at the 1th TEPHINET conference in Thailand. Daniel grew up in one of the remote villages of Ghana and is passionate about controlling infectious diseases.

Biography:

Cholera is a major health problem facing most developing countries. Globally, 132, 121 cholera cases were reported in 2016. About 54% of these cases were recorded in Africa.  Between June 2014 and January 2015, a total of 28,922 cholera cases including 243 deaths were reported in Ghana. WHO estimates that the true incidence of Cholera far exceeds the reported cases. We evaluated the cholera surveillance system to determine whether the system was meeting its objectives, and to assess its attributes. We evaluated the cholera surveillance system in the Ga East municipality. We interviewed staff of the GEMA on the operation of the system. We used semi structured questionnaire to assess the attributes of the system.  We reviewed data from the weekly and monthly IDSR and also from the district Health information management system from 2012-2016. We also reviewed annual reports and scientific papers. We applied the Centers for Disease Control and Prevention (CDC) updated Guidelines for Evaluating Public Health Surveillance Systems. Summarized descriptive analysis of qualitative data was done and presented in graphs and charts. The cholera surveillance is well situated in the IDSR. The case definition is clear, simple and easy to apply. The system is able to detect cases and notify the next level. The data matches with the case base forms. However, the entries in the case base forms were not complete. Positive predictive value could not be assessed as no single case was confirmed by laboratory test. CBSVs attrition was high in the municipality. However, Community health nurses were used as a replacement for the CBSVs. The system is meeting some of its objectives. The system is simple, flexible and acceptable. The system is fairly representative, stable but the data quality is low. Sentinel surveillance should be implemented as routine training of healthcare workers on reporting and proper documentation of suspected cases.