Design and deployment of relational geodatabase on mobile GIS platform for real-time COVID-19 contact tracing in Ghana

Abstract

This study reviewed the design and deployment of relational geodatabase on mobile GIS application, using collector for ArcGIS and survey 123 for ArcGIS platforms for COVID-19 contact tracing in Ghana during the lockdown. The study assessed whether cases spread by physical neighborhood contacts, defined by a 2km buffer of initial known 60 cases location. The application was deployed on the android tablet, which was used by field workers. Application Post-deployment review shows that from 30^th March to 4^th April 2020, 828 samples were collected with 34 confirmed cases, of which 61% occurred outside the 2km buffer. From 1-30 April 2020, 8,748 individuals with 16,087 contacts were tested within the physical neighbourhoods, 2.4% turned positive. Similarly, 7,501 individuals with 17,071 contacts were tested outside the physical neighbourhoods with 4.3% positives.  Results suggest that more infections occurred outside the case’s physical neighbourhoods possibly due to; (1) existence of unknown cases prior to lockdown; (2) cases were moving outside their physical neighborhood and infecting others; (3) panic movements of cases within the 3 days window between announcement and enforcement of lockdown; (4) movement of cases into the country through unapproved routes.  New cases were identified outside the lockdown areas, which could not be explained. This study raises questions about (1) the understanding of the mode of spread of the virus (2) the implementation of the lockdown, including the geographic coverage and timing. It is recommended that future decisions on contact tracing and lockdown should be guided by an understanding of the disease geography.