Introduction:
Coronavirus disease 19 (Covid-19) is an infectious disease caused by the
Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) which was
initially identified in December 2019 before becoming a global pandemic.
Human to human spread is the identified form of transmission while exact
molecular pathways in this pathway are not fully understood [1, 2].
Respiratory diseases spread by the inhalation of droplets scattered by
the infected person. Avoidance of social distancing, failure of using
personal protective equipment, late detection of symptoms all contribute
to the rapid spread of the disease, and increase the burden in the
health system in Covid-19 pandemic [3]. The unexpected increase of
infected patients had put tremendous pressure on health systems causing
a capacity overload, the premature ending of medical supplies, and
exhausted health professionals to name a few. This sudden increase of
patients had caused also significant implications for non-Covid-19
patients such as failure of initiating proper workup and treatment of
conditions [4].
Such an unexpected increase of patients is not foreseeable for the
health system and there are several factors for this. First, modern
medicine had faced this scale of the pandemic first time where a
systematic approach is used to gather data about how the number of cases
evolves across nations. Second, countermeasure for this pandemic,
lockdowns, quarantines and curfews, are internationally applied for the
first time with no previous experience. Third, as the number of cases
increase with different increments between communities, prediction of
the total number of cases can be challenging.
Time series analysis with autoregressive integrated moving averages
(ARIMA) models was popularized by Box and Jenkins in 1970 with their
Box-Jenkins approach [5]. By using only one variable measured in
equally spaced points in time, forecasting can be made with the help of
the model build using the variable. Time series are used in statistics,
weather prediction, and econometrics to name a few. In medicine, time
series are used to predict the number of patients admitted in previous
studies [6, 7].
In this research, we explored whether ARIMA model is feasible to predict
the number of cases for Covid-19 patients. The aim is to forecast the
total number of patients in the United States of America (USA) using the
time series model and this modeling can provide health systems to
provide better health care to patients.