Kavya NarayananJul 16, 2020 13:49:08 IST
A worst-case scenario prediction for coronavirus case in India developed by a team of researchers at the Indian Institute of Science in Bengaluru shows that the country could see the total number of cases rise to over66 lakh by 30 Septemberthis year at the current trend of relaxation to lockdown rules.
Theinteractive prediction modelalso accounts for the “best-case” scenario, in which the number of active/ongoing COVID-19 cases starts to fall by mid-September. The best-case model predicts a more conservative estimate of 26 lakh total cases, and 4.5 lakh ongoing cases by 30 September.
The prediction model is a six-dimensional, data-based computational model for epidemics that has been adapted to model the COVID-19 pandemic in India by Professors Sashikumaar Ganesan and Deepak Subramani at the Department of Computational and Data Sciences at IISc Bangalore.
A glance of the interactive predication model for COVID-19 cases developed by IISc computational modeling researchers. Image: Ganesan et al./IISc
The trend figures, researchers said on their website, are “business as usual” predictions, representing the pandemic evolving in different directions, and considering different degrees of relaxation to lockdown rules. It considers weekly lockdowns, one on Sundays only, and another on Sundays and Wednesdays. The Sundays-only complete lockdown scenario shows 2,40,000 active cases on top of the figures for the Sundays and Wednesdays lockdown as of 30 September.
Apart from the predictive model, researchers have also made observations about how specific measures have affected the COVID-19 infection curve in India, in national and state-specific scenarios. For instance, the recovery rate has increased since 3 May 2020 hand-in-hand with a drop in new infections. This coincided with access to better medical care and the effects of timely quarantine for infected COVID-19 patients.
The researchers claim that the model predicts region-wise and age-wise COVID-19 spread accurately. It also includes predictions of infected people by region, age of infected individuals, number of days since the start of the infection and severity, over time.
The data fed into the model, the researchers said, includes “immunity, pre-medical history, effective treatment, point-to-point movement of infected people (by air, train, etc), interactivity (spread in the community), hygiene and physical distancing”.
A six-dimensional data-based prediction model by IISc researchers to predict COVID19 cases by state. Image: Screengrab from Ganesan et al./IISc
The model has beenpre-published inarXivand is awaiting peer review.