A Comparative Study to Find a Suitable Model for an Improved Real-Time Monitoring of The Interventions to Contain COVID-19 Outbreak in The High Incidence States of India

Background On March 11, 2020, The World Health Organization (WHO) declared coronavirus disease (COVID-19) as a global pandemic. There emerged a need for reliable models to estimate the imminent incidence and overall assessment of the outbreak, in order to develop effective interventions and control strategies. One such vital metrics for monitoring the transmission trends over time is the time-dependent effective reproduction number (Rt). Rt is an estimate of secondary cases caused by an infected individual at a time during the outbreak, given that a certain population proportion is already infected. Misestimated Rt is particularly concerning when probing the association between the changes in transmission rate and the changes in the implemented policies. In this paper, we substantiate the implementation of the instantaneous reproduction number (Rins) method over the conventional method to estimate Rt viz case reproduction number (Rcase), by unmasking the real-time estimation ability of both methodologies using credible datasets. Materials & Methods We employed the daily incidence dataset of COVID-19 for India and high incidence states to estimate Rins and Rcase. We compared the real-time projection obtained through these methods by corroborating those states that are containing a high number of COVID19 cases and are conducting high and efficient COVID-19 testing. The Rins and Rcase were estimated using R0 and EpiEstim packages respectively in R software 4.0.0. Results Although, both the Rins and Rcase for the selected states were higher during the lockdown phases (March 25 - June 1, 2020) and subsequently stabilizes co-equally during the unlock phase (June 1- August 23, 2020), Rins demonstrated variations in accordance with the interventions while Rcase remained generalized and under- & overestimated. A larger difference in Rins and Rcase estimates were also observed for states that are conducting high testing. Conclusion Of the two methods, Rins elucidated a better real-time progression of the COVID-19 outbreak conceptually and empirically, than that of Rcase. However, we also suggest considering the assumptions corroborated in the implementations which may result in misleading conclusions in the real world.