Estimating the adjusted fatality rate of COVID-19 in Iran using the fused lasso approach in distributed lag models
Distributed lag models can estimate disease fatality in long-term epidemics by adjusting the delayed effect of the number of daily cases. The fused lasso approach is a regularization method that considers the inherent order between features, such as the time order between the number of daily hospitalizations, and can remove the effect of collinearity between features by placing a penalty in the estimation of the parameters of the distributed lag model. Therefore, this study has been carried out by estimating the adjusted mortality of COVID-19 using the lasso regularization approach in distributed lag models.
The data relating to the COVID-19 disease (the number of deaths and the number of people identified daily) in Iran from the middle of February 1398 to the middle of May 1402 was prepared from the OurWorldinData database and using R4. 3.3 the distributed lag model was fitted with the fused lasso method.
The adjusted fatality rate of COVID-19 in Iran is 1.72%, with five days delay between new cases diagnosis and deaths. The highest fractions were on the first day and the fifth day after detection.
Using the fused lasso method improved the significance of the coefficients in the distributed lag model. The short time from detection to death in Iran compared to other countries is a sign of poor screening and a lack of adequate equipment and services for the emergency treatment of patients.
-
Patient’s overview of using Artificial Intelligence (AI) in health care: A Systematic Review
Zahra Koohjani, Shahrbano Nasir Shirvani, *
Razavi International Journal of Medicine, Summer 2024 -
Refractive Error and Ocular Biometric Changes in the Treatment of Diabetes Mellitus
Amir Asharlous, Abbas Riazi, Samira Janani, Sattar Rajabi*, Amir Rakhshan, Mehdi Khabazkhoob, Zahra Sadat Dibaji Forooshani,
Function and Disability Journal, Winter 2022