Abstract: In this survival study, the range on the days of observation is from January 2020 to December 2020 consisting of the patients diagnosed with COVID-19. An accelerated failure time (AFT) model is used to identify covariates associated with recovery time (days from result of test to death/recovery of patients). AFT models with five different distributions (exponential, log-normal, Weibull, generalized gamma, and log-logistic) are generated. Akaike’s information criterion (AIC) is used to identify the most suitable model. The total number of patients used in this study is 66142 and is broken into 2116 events and 64026 censored patients. This study shows that generalized gamma having the lowest AIC value made the best fit of the model. The covariates used in determining the factors associated to the recovery of patient are age, sex, admitted and quarantined. The model shows that when patients are being quarantined, the recovery time of patients increases.
|
Keywords and phrases: accelerated failure time model, Akaike’s information criterion, probability distributions, COVID-19, survival study.
Received: June 3, 2022; Accepted: July 11, 2022; Published: July 19, 2022
How to cite this article: Ebni A. Jal-Usman, Azman A. Nads and Danilo G. Langamin, Accelerated failure time (AFT) model: determining the factors associated in the recovery of patients diagnosed with COVID-19, Advances and Applications in Statistics 79 (2022), 1-9. http://dx.doi.org/10.17654/0972361722056
This Open Access Article is Licensed under Creative Commons Attribution 4.0 International License
References:
[1] M. J. Brewe and A. Butler, Model Selection and the Cult of AIC, Biomathematics and Statistics Scotland, 2017. [2] D. Collett, Modelling Survival Data in Medical Research, Chapman & Hall/CRC, 2nd ed., 1993. [3] D. M. Diez, Survival Analysis in R, July 2021. [4] E. T. Lee and J. Wenyuwang, Statistical Methods for Survival Data Analysis, 3rd ed., Wiley-Interscience, NJ, 2003. [5] X. Liu, Survival analysis models and applications, Uniformed Services University of the Health Sciences and Walter Reed National Military Medical Center, USA, 1st ed., 2012. [6] Y. A. Mustefa and D. Chen, Accelerated failure-time model with weighted least-squares estimation: application on survival of HIV positives, Archives of Public Health 79 (2021), Article number 88. https://doi.org/10.1186/s13690-021-00617-0. [7] R. Saikia and M. P. Barman, A review on accelerated failure time models, Int. J. Stat. Syst. 12(2) (2017), 311-322. [8] G. Thiruvengadam, R. Ramanujam and L. Marappa, Modeling the recovery time of patients with coronavirus disease 2019 using an accelerated failure time model, 2021. https://doi.org/10.1177/03000605211040263. [9] L. J. Wei, The accelerated failure time model: a useful alternative to the Cox regression model in survival analysis, Stat. Med. 11 (1992), 1871-1879. [10] R. Womack, Rutgers University Libraries, 2016 (libguides.rutgers.edu/data_R) or (github.com/ryandata/Survival), code rev. 2016-12-23.
|