. 2024 Aug 30;19(8):e0309413. doi: 10.1371/journal.pone.0309413

Table 2. Crude and adjusted hazard ratio of COVID-19 death according to the predictors (southeast of Brazil).

	Crude Cox model	Adjusted classic Cox model ^£	Model 1 -Adjusted Cox mixed-effects model ^§	Model 2 -Adjusted Cox mixed-effects model ^¥
*Hazard ratio*			Fixed effects	Fixed effects
Age	1.03^***	1.03^***	1.02^***	1.03^***
Elderly	2.14^***
Male	0.96^***	1.07^***	1.07^***	1.07^***
Race
White	1	1	1	1
Black	1.22^***	1.13^***	1.11^***	1.12^***
Yellow	0.92^***	0.93^*	0.94^•	0.92^*
Brown	1.00^NS	1.07^***	1.02^***	1.07^***
Indian	0.74^***	0.89^NS	0.87^NS	0.89^NS
Years of study
Zero	1	1	1	1
Five	0.83^***	0.88^***	0.87^***	0.88^***
Nine	0.69^***	0.85^***	0.84^***	0.85^***
Twelve	0.54^***	0.78^***	0.76^***	0.78^***
More than twelve	0.46^***	0.63^***	0.62^***	0.63^***
Not applicable	0.14^***	0.79^***	0.70^***	0.70^***
State
Espírito Santo	1	1	1
Minas Gerais	0.65^***	0.86^***	0.89^•
Rio de Janeiro	0.90^***	1.46^***	1.04^NS
São Paulo	0.60^***	0.76^***	0.75^***
Area
Urban	1	1	1	1
Rural	1.22^***	0.97^NS	1.00^NS	0.97^•
Peri-urban	1.18^***	0.87^**	1.00^NS	0.88^*
First symptoms year
2020	1	1	1
2021	0.98^***	1.18^***	1.21^***
2022	1.23^***	1.22^***	1.27^***
2023	0.96^**	0.97^NS	1.07^*
Immunization for COVID-19	1.06^***			0.79^***
2021		0.78^***
2022		0.87^***
2023		1.03^NS
Hospitalization	0.32^***	0.40^***	0.30^***	0.38^***
Risk factor	1.40^***	1.19^***	1.20^***	1.20^***
*Random effects—SD*
City (intercept)			0.35
State (intercept)				4.20
First symptoms year (slope)				0.16
*Fit measures*
Likelihood ratio		38195 (df = 21)^***	5638635 (df = 1199)^***	5647953 (df = 4)^***
Log-likelihood		-1390689 (df = 21)	-1384557 (df = 937)	-1390405 (df = 20)
AIC		2781419	2770987	2780850
BIC		2781623	2780082	2781045

*** = p-value < 0.0001,

** = p-value < 0.001,

* = p-value < 0.01,

^• = p-value < 0.1;

SD = Standard deviation; df = degrees of freedom; AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion

^£ The classic Cox-adjusted model included the following predictors: age, sex, race, years of education, state of residence, area, year of first symptoms, hospitalization, and risk factors. To access the estimated effect of COVID-19 immunization, separately adjusted models were used for each year that COVID-19 vaccination occurred in Brazil.

^§ The first Cox mixed-effects adjusted model included fixed effects of age, sex, race, years of study, state, area, first symptoms year, hospitalization, and risk factor; and the random effect of the municipality of residence (intercept).

^¥ The second Cox mixed-effects adjusted model included fixed effects of age, sex, race, years of study, area, hospitalization, and risk factor; and the random effects of state (intercept) and first symptoms year (slope).