The pattern and trend of COVID-19 spread in the Eastern Mediterranean Region: Curve-fitting models estimation (Preprint)
BACKGROUND Curve fitting is used to find the "best fit" line or curve for a series of data points. It is the process of specifying the model that provides the best fit to the specific curves in the dataset. OBJECTIVE This study aimed to use the Curve Estimation Procedure to assess the pattern and trend of COVID-19 spread in the countries of the Eastern Mediterranean Region (EMR) by finding the model best fit the observed COVID-19 data in each country between 20 February 2020 - 21 April 2020. METHODS The number of daily confirmed COVID-19 cases, recovered cases, and during the period 20 February 2020 - 21 April 2020 in 21 Eastern Mediterranean countries were extracted from the World Health Organization situation reports. The Curve Estimation procedure was used to produce different curve estimation regression models for the observed data in each country. A total of 11 regression models were tested and compared to find the model that best fits the COVID-19 data during the observed period for each country. RESULTS During this observed period, the total number of confirmed cases, recovered cases, and deaths in the region were 138673, 71343, and 6291, respectivly. The overall fatality rate in the region was 4.5%. The quadratic model (Y = b0 + (b1 * t) + (b2 * t**2) and the cubic model (Y = b0 + (b1 * t) + (b2 * t**2) + (b3 * t**3) follows the observed data points fairly well during the observed time period in five and nine countries, respectively. The exponential model (Y = b0 * (e**(b1 * t))), the growth model (Y = e**(b0 + (b1 * t))), and the compound model (Y = b0 * (b1**t)) were the best fit for data during the observed time period in two, three, and two countries, respectively. CONCLUSIONS The pattern of COVID-19 spread differed between countries in the EMR. This might reflect the variations in testing and implementation of public health measure for prevention and control. The best curve-fitting model was demonstrated for each country that can be used for making very short-term predictions.