Table 2.

The results of logistic regression using a backward Wald elimination of variables (response: Outcome)^a

	Regression coefficient (β)	SE	P	OR	95% CI for OR (lower-upper)
Step 1
Age	−0.006	0.023	0.81	0.994	0.950-1.041
Hospitalization period	−0.005	0.259	0.98	0.995	0.599-1.651
Hydroxychloroquine (1)	5.138	2.944	0.08	170.3	0.531-5.46E+04
Ribavirin (1)	−1.854	4.555	0.68	0.157	0.000-1180.949
Lopinavir/ritonavir (1)	0.858	1.829	0.64	2.359	0.065-85.041
Intravenous antibiotics (1)	−1.212	3.085	0.69	0.298	0.001-125.876
Hydroxychloroquine (1) by age by severity (1)	0.332	758.358	1.00	1.394	0.000-NA
Lopinavir/ritonavir (1) by age by severity (1)	−0.639	766.387	1.00	0.528	0.000-NA
Ribavirin (1) by age by severity (1)	0.236	110.643	1.00	1.266	0.000-1.91E+94
Diabetes (1)	−2.310	1.750	0.19	0.099	0.003-3.063
Hypertension (1)	2.513	2.062	0.22	12.338	0.217-702.334
Chronic obstructive pulmonary disease/asthma (1)	34.177	10895.718	1.00	6.96E+14	0.000-NA
Azithromycin (1) by hydroxychloroquine (1)	−0.087	11.366	0.99	0.917	0.000-4.34E+09
Age by azithromycin (1) by hydroxychloroquine (1)	0.028	0.207	0.89	1.028	0.685-1.544
Step 13
Hydroxychloroquine (1)	4.125	0.983	<0.001	61.859	9.009-424.722
Lopinavir/ritonavir (1) by age by severity (1)	−0.081	0.020	<0.001	0.922	0.887-0.958

Complete 13 steps of logistic regression is provided in the electronic Supplementary Material [Supplementary Table 2]. OR=Odds ratio; CI=Confidence interval; SE=Standard deviation; NA=Not applicable