Table 4. Multivariable analysis for the probability of hospitalization.

Variables	OR	95% CI	P-value
Age (for each year)	1.033	1.023–1.044	<0.001
Hypertension (yes vs. no)	1.418	1.076–1.869	0.013
Cardiovascular disease (yes vs. no)	1.648	1.077–2.523	0.021
Recent high dose of steroids (yes vs. no)	1.866	1.229–2.832	0.003
Time of symptoms (for each day)	0.983	0.971–0.996	0.008
Fever (yes vs. no)	1.649	1.290–2.108	<0.001
Dyspnea (yes vs. no)	3.673	2.803–4.812	<0.001
Nasal congestion (yes vs. no)	0.588	0.410–0.845	0.004
Headache (yes vs. no)	0.525	0.402–0.685	<0.001
Arthralgia (yes vs. no)	0.250	0.117–0.533	<0.001
Diarrhea (yes vs. no)	1.783	1.367–2.325	<0.001
Anosmia (yes vs. no)	0.519	0.387–0.696	<0.001
Nausea and vomiting (yes vs. no)	2.284	1.427–3.655	0.001

AKI, acute kidney injury; eGFR, estimated glomerular filtration rate.

Footnote: Variables included in the modeling: age, type of donor (living vs. deceased); hypertension, diabetes, previous cardiovascular disease, and neoplasia as comorbidities; recent use of high dose of steroids and recent use of Thymoglobulin; time of the start of COVID-19 symptoms; fever, chills, cough, dyspnea, runny nose, headache, nasal congestion, sore throat, myalgia, arthralgia, diarrhea, nausea and vomiting, anosmia, and ageusia as COVID-19 symptoms.

Values with missing values < 10%: recent use of high dose of steroids and recent use of Thymoglobulin. For both the mode value was imputed, i.e., no use.

Variables excluded from the modeling:

    a) number of missing were higher than 10%: eGFR (10.9%) and AKI (37.4%)

    b) collinearity: hypoxemia (colinear with dyspnea).