Assessment of Lockdown Effectiveness in the Wake of COVID-19 in India Using ARIMA Model

Abstract

Background: Novel coronavirus disease COVID-19 has been recognized as a global threat, and several studies are being conducted using various mathematical models to predict the probable evolution of this epidemic which are subject to potential bias. In this study we aimed to assess and compare the impact of lockdown between Punjab, Delhi and Gujarat states of India using Auto Regressive Integrated Moving Average (ARIMA) model by comparing forecasted COVID-19 data with real time data. Methods: We analyzed COVID-19 data of Indian states from index case till 17th of May, 2020. ARIMA (1,1,3) (0,0,0) model was used to forecast the possible cumulative cases till May 17, from data up to May 3rd. and compared with real time data. Collated recovery rate, case-fatality rate, test per million of states. Results: Trend of cumulative cases in Punjab was moving downward below the forecasted lower confidence limit (R2 = 0.9799) whereas, the cumulative case trend of Delhi was moving along the forecasted upper confidence limit with the forecasted data till May 3rd (R2 = 0.9971) and the trend of cumulative cases was below the forecasted upper confidence limit (R2 =0.9992) in Gujarat. Conclusions: In Gujarat and Delhi, the lockdown was not merely effective in controlling the rise in COVID-19 cases even after 56th day of lockdown whereas, Punjab state was succeeded in preventing havoc of COVID-19. In lieu of lockdown, using facemasks, improving ventilation in closed workspace settings, crowded spaces and close contact settings is more pragmatic than keeping away from others in India.