Volume 32, Issue 6 (9-2024)
JSSU 2024, 32(6): 7951-7962 |
Back to browse issues page
Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:
Kargarian-Marvasti S, Hasannezhad M, Abolghasemi J. Survival Analysis of Patients with Covid-19 Using Parametric Models in the Presence of Frailty Variable: A Prospective Cohort Study. JSSU 2024; 32 (6) :7951-7962
URL: http://jssu.ssu.ac.ir/article-1-6146-en.html
URL: http://jssu.ssu.ac.ir/article-1-6146-en.html
Abstract: (285 Views)
Introduction: This study was performed to investigate the survival analysis of patients using parametric models in the presence of fragility variables.
Methods: The data of this research belonged to a prospective cohort study of all the patients with Covid-19 in Fereydunshahr City. By conducting PCR tests on 2269 individuals suspected of Covid-19, 880 definite patients of Covid-19 were identified by census method. The death of the patients due to Covid-19 was the failure event. The response variable was the time from the onset of symptoms to the time of death (or censoring) at the end of the study. The data were analyzed through SPSS (version 16) and R software (version 4.3.2) at an error level of 0.05 and fitted with survival parametric models (considering the gamma distribution for the frailty variable) using Akaike's criterion.
Results: Based on multiple regression analysis, the risk of death in patients with a history of heart disease was 4.9 times more than of patients without heart disease (95% CI for HR=2.21-10.98, HR=4.9) and in hospitalization patients was 4.2 times more than of outpatient cases (HR=4.2, 95% CI for HR=1.74-10.24). Moreover, increasing the age showed a significant relationship with the mortality rate (95% CI for HR=1.02-1.08, HR=1.05). With the inclusion of fragility variable in the model, the variable of “cardiovascular disease” was recognized as an important risk factor in survival time of patients; while without the fragility variable, this variable was ignored. In this study, according to the Akaike's criterion, the log-normal model showed a goodness of fit with the Covid-19 data.
Conclusion: Using the fragility variable in survival regression models of patients with Covid-19, it is possible to identify the factors affecting patient mortality that it’s impossible to identify these risk factors in conventional models.
Methods: The data of this research belonged to a prospective cohort study of all the patients with Covid-19 in Fereydunshahr City. By conducting PCR tests on 2269 individuals suspected of Covid-19, 880 definite patients of Covid-19 were identified by census method. The death of the patients due to Covid-19 was the failure event. The response variable was the time from the onset of symptoms to the time of death (or censoring) at the end of the study. The data were analyzed through SPSS (version 16) and R software (version 4.3.2) at an error level of 0.05 and fitted with survival parametric models (considering the gamma distribution for the frailty variable) using Akaike's criterion.
Results: Based on multiple regression analysis, the risk of death in patients with a history of heart disease was 4.9 times more than of patients without heart disease (95% CI for HR=2.21-10.98, HR=4.9) and in hospitalization patients was 4.2 times more than of outpatient cases (HR=4.2, 95% CI for HR=1.74-10.24). Moreover, increasing the age showed a significant relationship with the mortality rate (95% CI for HR=1.02-1.08, HR=1.05). With the inclusion of fragility variable in the model, the variable of “cardiovascular disease” was recognized as an important risk factor in survival time of patients; while without the fragility variable, this variable was ignored. In this study, according to the Akaike's criterion, the log-normal model showed a goodness of fit with the Covid-19 data.
Conclusion: Using the fragility variable in survival regression models of patients with Covid-19, it is possible to identify the factors affecting patient mortality that it’s impossible to identify these risk factors in conventional models.
Type of Study: Original article |
Subject:
Epidemiology
Received: 2023/12/19 | Accepted: 2024/04/28 | Published: 2024/09/5
Received: 2023/12/19 | Accepted: 2024/04/28 | Published: 2024/09/5
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
