TABLE 3.

Effect of zinc supplementation compared with no zinc on primary and secondary outcomes

	Zinc	No zinc	HR or risk ratio1 (95% CI)	P value
Primary outcome
Fever, cough, or shortness of breath at baseline, n (%)
No	25 (27.2)	22 (24.7)
Yes	66 (71.7)	67 (75.3)
Missing	1 (1.1)	0 (0.0)
Days to resolution, median (95% CI) (N = 132)	3 (2–4)	3 (2–4)	0.94 (0.67, 1.33)	0.745
Secondary outcomes
Fever at baseline, n (%)
No	55 (59.8)	44 (49.4)
Yes	37 (40.2)	45 (50.6)
Days to resolution, median (95% CI) (N = 81)	2 (2–3)	2 (1–2)	0.89 (0.57, 1.40)	0.616
Cough at baseline, n (%)
No	38 (41.3)	33 (37.1)
Yes	54 (58.7)	56 (62.9)
Days to resolution, median (95% CI) (N = 109)	3 (2–5)	3 (2–5)	0.96 (0.66, 1.41)	0.843
Shortness of breath at baseline, n (%)
No	76 (82.6)	81 (91.0)
Yes	15 (16.3)	8 (9.0)
Missing	1 (1.1)	0 (0.0)
Days to resolution, median (95% CI) (N = 23)	2 (2–4)	2 (1–.)	0.61 (0.25, 1.52)	0.291
Days to discharge, median (95% CI) (N = 147)	6 (5–7)	6 (5–6)	0.83 (0.59, 1.15)	0.261
Need for assisted ventilation1 (N = 148), n (%)
No	68 (93.2)	73 (97.3)
Yes	3 (4.1)	1 (1.3)
Missing	2 (2.7)	1 (1.3)	1.03 (0.75, 1.42)	0.864

Estimates based on univariable Cox or Poisson regression models.

P for test of proportional hazards assumption (global test) >0.05 for all outcomes for which Cox regression models were performed.

¹Risk ratio calculated for need for assisted ventilation; HR calculated for all other outcomes.