Publication

Abstract : Scientific communities are working toward mitigating the repercussions of the SARS-CoV-2 infectious virus and its spread toward the community. The dynamics of the spreading nature can be determined by prediction models. Various prediction models are devised and/or used to know the disease dynamics and the existing ones based on statistical models are being developed for single or multiple countries. Dynamic prediction of new variants of SARS-CoV-2 infections during the present pandemic scenario is critical and herculean and this prediction is cardinal for preparing strategic maneuvers. Many review articles commonly address the statistical models adopted, whereas the studies indicating effective models that address disease dynamics and forecast potential contagion scenarios, viz., Holt’s method, Wright’s modified Holt’s method, and un-replicated linear functional relationship model (ULFR) for new variants are found scarce. This work aims at collating the basic working philosophies of most cited COVID-19 dynamic prediction model reports by a systematic literature study. The study is reported, by formulating a search strategy that is refined in terms of geographical purview, time period according to a predefined protocol and scientific reports presented across the globe. The outcome of this review shows that the study findings are critically relevant to front-line healthcare professionals and government agencies for disease prognostication, effective risk stratification, and adopting strategic maneuvers in policies. Additionally, this works aids in identifying the lacuna in the existing and or customized prediction models so as to improvise their data-driven approaches effectively.