COVID-19 and India: what next?

Purpose - The study aims to carry out predictive modeling based on publicly available COVID-19 data for the duration April 01, 2020 to June 20, 2020 pertaining to India and five of its most infected states: Maharashtra, Tamil Nadu, Delhi, Gujarat and Rajasthan. Design/methodology/approach - The study leverages the susceptible, infected, recovered and dead (SIRD) epidemiological framework for predictive modeling. The basic reproduction number R0 is derived by an exponential growth method using RStudio package R0. The differential equations reflecting the SIRD model have been solved using Python 3.7.4 on the Jupyter Notebook platform. For visualization, Python Matplotlib 3.2.1 package is used. Findings - The study offers insights on peak-date, peak number of COVID-19 infections and end-date pertaining to India and five of its states. Practical implications - The results subtly indicate toward the amount of effort required to completely eliminate the infection. It could be leveraged by the political leadership and industry doyens for economic policy planning and execution. Originality/value - The emergence of a clear picture about COVID-19 lifecycle is impossible without integrating data science algorithms and epidemiology theoretical framework. This study amalgamates these two disciplines to undertake predictive modeling based on COVID-19 data from India and five of its states. Population-specific granular and objective assessment of key parameters such as reproduction number (RO), susceptible population (S), effective contact rate (B) and case-fatality rate (s) have been used to generate a visualization of COVID-19 lifecycle pattern for a critically affected population.