On the potential viability of epidemiological models in utilitarian COVID-19 prediction

Abstract:

Traditional population based epidemiological models are being increasingly used to predict potential health hazards associated with infectious diseases and subsequently design public health intervention strategies. We argue that system dynamics methods, incorporating feedback loops, are perhaps a more comprehensive platform to simulate scenarios that could impart in-depth insight of the subtleties of disease transmission process. System dynamics model was developed for simulating COVID19 for several major megacities assimilating data from climatic processes and socio-economic factors to track the transmission of the disease in human population. Besides temperature and relative humidity, some of the critical factors investigated were rates of testing, contact density, behavioral risk reduction and population density. Sensitivity analysis results indicate that the number of cases and fatalities resulting from COVID-19 are most sensitive to testing rates followed by behavioral risk and contact density. Further, it was observed that during the transmission phase, fatalities and cases were almost insensitive to the change in climate variables suggesting a complex role of those variables. Recent studies show that COVID-19 is perhaps an air-borne disease and therefore, insensitivity to climatic variables may further raise questions on the utility of traditional epidemiological models as a reliable tool to scenario-predict the course of the disease.