Table 2. Primary and Secondary Outcomes (Modified Intention-to-Treat Population).*.
| Outcome | Hydroxychloroquine plus Azithromycin (N=172) |
Hydroxychloroquine (N=159) |
Control (N=173) |
Effect Estimate (95% CI) | ||
|---|---|---|---|---|---|---|
| Hydroxychloroquine plus Azithromycin vs. Control |
Hydroxychloroquine vs. Control |
Hydroxychloroquine plus Azithromycin vs. Hydroxychloroquine |
||||
| Primary outcome: seven-level ordinal outcome at 15 days† | ||||||
| Median score (IQR) | 1 (1–2) | 1 (1–2) | 1 (1–2) | 0.99 (0.57 to 1.73)‡ | 1.21 (0.69 to 2.11)‡ | 0.82 (0.47 to 1.43)‡ |
| Distribution — no. (%) | ||||||
| 1: Not hospitalized with no limitations on activities | 118 (68.6) | 102 (64.2) | 117 (67.6) | |||
| 2: Not hospitalized but with limitations on activities | 22 (12.8) | 27 (17.0) | 29 (16.8) | |||
| 3: Hospitalized, not receiving supplemental oxygen | 15 (8.7) | 12 (7.5) | 8 (4.6) | |||
| 4: Hospitalized, receiving supplemental oxygen | 5 (2.9) | 6 (3.8) | 5 (2.9) | |||
| 5: Hospitalized, receiving noninvasive ventilation or high-flow nasal cannula | 0 | 2 (1.3) | 2 (1.2) | |||
| 6: Hospitalized, receiving mechanical ventilation | 9 (5.2) | 5 (3.1) | 7 (4.0) | |||
| 7: Death | 3 (1.7) | 5 (3.1) | 5 (2.9) | |||
| Secondary outcomes | ||||||
| Six-level ordinal outcome at 7 days§ | ||||||
| Median score (IQR) | 2 (1–3) | 2 (1–2) | 2 (1–3) | 0.81 (0.54 to 1.22) | 0.92 (0.61 to 1.38) | 0.89 (0.58 to 1.34) |
| Distribution — no./total no. (%) | ||||||
| 1: Not hospitalized | 84 (48.8) | 67 (42.1) | 75 (43.4) | |||
| 2: Hospitalized, not receiving supplemental oxygen | 38 (22.1) | 53 (33.3) | 45 (26.0) | |||
| 3: Hospitalized, receiving supplemental oxygen | 31 (18.0) | 25 (15.7) | 38 (22.0) | |||
| 4: Hospitalized, receiving noninvasive ventilation or high-flow nasal cannula | 3 (1.7) | 2 (1.3) | 4 (2.3) | |||
| 5: Hospitalized, receiving mechanical ventilation | 15 (8.7) | 10 (6.3) | 9 (5.2) | |||
| 6: Death | 1 (0.6) | 2 (1.3) | 2 (1.2) | |||
| No. of days free from respiratory support within 15 days¶ | 11.1±4.9 | 11.2±4.9 | 11.1±4.9 | 0.1 (−0.7– to 0.9) | −0.2 (−1.1 to 0.6) | 0.3 (−0.6 to 1.1) |
| Use of high-flow nasal cannula or noninvasive ventilation within 15 days — no. (%) | 16 (9.3) | 17 (10.7) | 16 (9.2) | 1.10 (0.60 to 2.03) | 1.19 (0.65 to 2.21) | 0.92 (0.50 to 1.70) |
| Use of mechanical ventilation within 15 days — no. (%) | 19 (11.0) | 12 (7.5) | 12 (6.9) | 1.77 (0.81 to 3.87) | 1.15 (0.49 to 2.70) | 1.54 (0.71 to 3.35) |
| Duration of hospital stay — days‖ | 10.3±8.4 | 9.6±6.5 | 9.5±7.2 | 0.9 (−0.3 to 2.1) | 0.2 (−1.0 to 1.3) | 0.7 (−0.6 to 1.9) |
| In-hospital death — no. (%)‖** | 5 (2.9) | 7 (4.4) | 6 (3.5) | 0.64 (0.18 to 2.21) | 1.47 (0.48 to 4.53) | 0.43 (0.13 to 1.45) |
| Thromboembolic complications within 15 days — no. (%) | 2 (1.2) | 3 (1.9) | 2 (1.2) | 0.89 (0.31 to 2.54) | 1.39 (0.53 to 3.65) | 0.64 (0.24 to 1.68) |
| Acute kidney injury within 15 days — no. (%) | 6 (3.5) | 4 (2.5) | 5 (2.9) | 1.18 (0.44 to 3.20) | 0.88 (0.29 to 2.63) | 1.35 (0.47 to 3.84) |
Plus–minus values are means ±SD. The modified intention-to-treat population included patients who had undergone randomization and who had a confirmed diagnosis of coronavirus disease 2019 (Covid-19). Effect estimates are mean differences (for the outcomes of the number of days free from respiratory support within 15 days and the duration of hospital stay), hazard ratios (for the outcome of in-hospital death), or odds ratios (for all other outcomes). Odds ratios for the ordinal outcome at 7 days and at 15 days that are lower than 1 indicate a benefit for treatment groups as compared with control. We used the Lipsitz method to test proportionality of odds with a P value of 0.15 (see the Additional Details on Statistical Analyses section in the Supplementary Appendix). The widths of the confidence intervals for the secondary outcomes were not adjusted for multiple comparisons, so the intervals should not be used to infer definitive treatment effects.
The ordinal outcome as assessed at 15 days was evaluated on the seven-level scale. Odds ratios lower than 1 indicate treatment benefit.
P=1.00 for all comparisons. P values for the three two-by-two comparisons for the primary outcome were adjusted with the use of Bonferroni correction for multiple comparisons.
The ordinal outcome as assessed at 7 days was evaluated on a six-level scale, in which levels 1 and 2 from the seven-level scale were combined as level 1 (no hospitalization). Data were available for 171 patients assigned to receive hydroxychloroquine plus azithromycin, for 157 in the hydroxychloroquine-alone group, and for 173 in the control group. Effects were calculated after multiple imputation. Odds ratio values lower than 1 indicate treatment benefit.
Data were available for 169 patients assigned to receive hydroxychloroquine plus azithromycin, 157 patients in the hydroxychloroquine-alone group, and 171 patients in the control group. Effects were calculated after multiple imputation.
As of June 4, 2020, a total of 21 patients were still in the hospital (range of duration of follow-up, 22 to 49 days): 8 patients assigned to receive hydroxychloroquine plus azithromycin, 5 patients in the hydroxychloroquine-alone group, and 8 in the control group. These patients were considered to be discharged alive.
A total of 18 patients died in the hospital. Hazard ratios are shown for this analysis. The causes of death were as follows: among patients assigned to receive hydroxychloroquine plus azithromycin, 5 patients died (the cause of death was Covid-19–related acute respiratory failure or septic shock in all patients); among patients assigned to receive hydroxychloroquine, 7 patients died (the cause of death was Covid-19–related acute respiratory failure or septic shock in 6 patients and abdominal-wall hemorrhage with shock in 1); and among patients assigned to the control group, 6 patients died (the cause of death was Covid-19–related acute respiratory failure or septic shock in 5 and myocardial infarction in 1).