Table 3.

Multivariate analysis of improvement trajectory based on demographics and disease severity prior to treatment with COVID-19 convalescent plasma.

Variables	OR (95% CI)	P-value
Intensive care status
Not in ICU prior to transfusion	1.00 (Ref)
ICU admission prior to transfusion	0.47 (0.42–0.51)	<0.0001
Mechanical ventilation status
Not on ventilator prior to transfusion	1.00 (Ref)
On ventilator prior to transfusion	0.41 (0.37–0.46)	<0.0001
Sex
Females	1.00 (Ref)	–
Males	0.93 (0.86–1.00)	0.070
Age
18–34 years	1.00 (Ref)	–
35–54 years	0.83 (0.71–0.96)	0.01
55–79 years	0.48 (0.41–0.55)	<0.0001
80+ years	0.25 (0.21–0.30)	<0.0001
Odds ratio of net improvement by day 7 stratified by disease severity
Mild or moderate	1.00 (Ref)	–
Severe	0.44 (0.38–0.50)	<0.0001
Critical or end stage	0.20 (0.21–0.30)	<0.0001
Odds ratio of net improvement by day 14 stratified by disease severity
Mild or moderate	1.00 (Ref)	–
Severe	0.60 (0.47–0.76)	<0.0001
Critical or end stage	0.32 (0.25–0.42)	<0.0001
Odds ratio of net improvement by day 21 stratified by disease severity
Mild or moderate	1.00 (Ref)	–
Severe	0.66 (0.46–0.95)	0.026
Critical or end stage	0.67 (0.46–0.99)	0.05

Adjusted GEE model was used to estimate the odds of being in a higher category (net improvement) of the ordinal clinical outcomes. Initial status of severity and time interaction term were added in the model as time-varying covariate.