Baseline characteristics of patients
| Characteristics | Convalescent Plasma (n=80) | Control (n=80) |
| Gender, male | 49 (61.2) | 44 (55.0) |
| Age, years | 59.0 (48.0–68.5) | 62.0 (49.5–68.0) |
| Age category# | ||
| <65 years | 55 (68.8) | 52 (65.0) |
| ≥65 to <80 years | 20 (25.0) | 27 (33.8) |
| ≥80 years | 5 (6.3) | 1 (1.3) |
| Comorbidities | ||
| Diabetes | 34 (42.5) | 29 (36.3) |
| Hypertension | 49 (61.3) | 49 (61.3) |
| Cardiovascular Disease | 19 (23.8) | 16 (20.0) |
| Chronic Pulmonary Disease | 13 (16.3) | 9 (11.3) |
| Obesity | 43 (53.8) | 38 (47.5) |
| Randomisation location | ||
| Intensive Care Unit | 53 (66.3) | 53 (66.3) |
| Medical Ward | 27 (33.8) | 27 (33.8) |
| Time from symptom onset to randomisation, days | 10.0±3.0 | 9.8±3.2 |
| Time from hospitalisation to randomisation, days¶ | 3 (2–5) | 3 (2–5) |
| Score on six-level ordinal scale | ||
| 2- hospitalisation without supplemental oxygen | 0 (0) | 1 (1.3) |
| 3 - hospitalisation plus supplemental oxygen | 18 (22.5) | 21 (26.3) |
| 4 - hospitalisation plus noninvasive ventilation or high-flow supplemental oxygen | 28 (35.0) | 24 (30.0) |
| 5 - hospitalisation plus invasive mechanical ventilation and/or extracorporeal membrane oxygenation | 34 (42.5) | 34 (42.5) |
| Vasoactive Drugs | 17 (21.3) | 14 (17.5) |
| NEWS 2 | 7 (6–10) | 7 (6–9) |
| PaO2/FiO2+ | 191 (134–246) | 167 (100–258) |
| SOFA+ | 3.5 (2–7) | 4 (2–7.8) |
| Characteristics | Convalescent Plasma (n=80) | Control (n=80) |
| Laboratorial findings at randomisation | ||
| Neutralising Antibody Titre+,§ | 960 (160–2560) | 1280 (640–2560) |
| Neutralising Antibody Titre≤80§ | 19/80 (23.8) | 8/78 (10.3) |
| White blood cell count, cells×103/μL | 7.7 (5.2–11.7) | 8.2 (6.3–11.3) |
| Neutrophil count, cells×103/μLƒ | 6.4 (4.2–8.4) | 7.1 (4.9–9.4) |
| Lymphocyte count, cells/μLƒ | 0.8 (0.6–1.2) | 0.8 (0.5–1.1) |
| Platelet count, cells×103/μL | 224.3±85.2 | 225.9±81.5 |
| C reactive protein, mg·L−1ƒ | 117.4 (60.9–203.2) | 90.6 (56.8–155.3) |
| D-dimer, μg·mL−1+,ƒ | 1.7 (0.9–4.1) | 1.3 (0.7–3.1) |
| Interleukin-6, pg·mL−1+,ƒ | 7.0 (2.9–17.3) | 3.7 (2.8–8.7) |
| Tumor necrosis factor-alpha, pg·mL−1+,ƒ | 10.9 (8.5–13.6) | 11.3 (10–14.9) |
| Medications at randomisation | ||
| Glucocorticoids## | 79 (98.8) | 79 (98.8) |
| Antibacterials | 73 (91.3) | 71 (88.8) |
Data are n (%), median (Interquartile Range) or mean±standard deviation. Abbreviations: PaO2/FiO2, arterial oxygen partial pressure/fractional inspired oxygen; NEWS, National Early Warning Score; SOFA, Sepsis-Related Organ Failure Assessment.
#p=0.15. This variable was selected for post hoc analysis by a Poisson regression model to adjust the effect of intervention;
¶ p=0.11. This variable was selected for post hoc analysis by a Poisson regression model to adjust the effect of intervention;
+Ten control (12.5%) and 10 convalescent plasma (12.5%) had PaO2/FiO2 estimated from peripheral oxygen saturation /FiO2 ratio adjusted to the positive end-expiratory pressure. One patient (1.3%) in the intervention group and one (1.3%) in the control group did not collect blood samples for neutralising antibody titre measurement. One (1.3%) patient in the intervention group did not collect D-dimer; one (1.3%) patient in the control group did not collect interleukin-6 and tumor necrosis factor-alpha at randomisation;
§p=0.002 for median neutralising antibody titres and p=0.041 for neutralising antibody titre ≤1:80. Median neutralising antibody titres was selected for post hoc analysis by a Poisson regression model to adjust the effect of intervention;
ƒLeukocyte counts and inflammatory markers: p=0.14 for median white blood cell count; p=0.17 for median neutrophil count; p=0.16 for C Reactive Protein; p=0.11 for D-dimer; p=0.046 for Interleukine-6; and p=0.20 for Tumor Necrosis Factor-alpha. These variables were selected for post hoc analysis by a Poisson regression model to adjust the effect of intervention;
##Among patients who were treated with corticosteroids, 78 (99.7%) and 77 (97.5%) received dexamethasone in plasma convalescent and control groups, respectively. Other corticosteroid drugs were used in 10 (12.7%) and 16 (20.1%) patients from intervention and control groups, respectively.