Forecasting the daily deaths caused by COVID-19 using ARIMA model

Pak J Pharm Sci. 2022 Jan;35(1):141-149.

Abstract

Coronavirus disease (COVID-19) pandemic has recently become a global health crisis. On the basis of this study the data reported from ten different countries on confirmed daily deaths caused by COVID-19. By fitting the linear regression models based on the data from ten countries to find the relationship between the new cases and deaths reported daily. We also used the autoregressive integrated moving average model (ARIMA) to predict the potential number of daily deaths caused by COVID-19 in these countries in the next 3 Months. The R2 value obtained for Iran (0.24) implies that 24% of daily deaths correspond to the daily cases. The R2of Pakistan 0.662 which indicates that 66.2% of daily deaths are explained by our predictor variable. In Turkey 70.2% of daily deaths are explained by daily cases and India recorded the highest number of deaths while UAE had the lowest number of deaths. Our results suggest that the pandemic is under control in China, UAE and Australia. Pakistan, Iran, Germany and Italy however, showed an upward trend in the spread of the disease, which may correlate with a high increase in death rate as the data indicated.

MeSH terms

  • Australia / epidemiology
  • COVID-19 / epidemiology
  • COVID-19 / mortality*
  • China / epidemiology
  • Europe / epidemiology
  • Forecasting
  • Humans
  • India / epidemiology
  • Iran / epidemiology
  • Linear Models
  • Models, Statistical*
  • Pakistan / epidemiology
  • Turkey / epidemiology