Trial characteristics. Data are number (%) of trials unless statement otherwise
| Characteristic | All trials (n=356) | Published trials without preprints (n=181) | Trials only available as preprints (n=101) | Difference between trials only available as preprints and published trials without preprints (%; 95% CI) | P value | Trials first posted as preprints and subsequently published (n=74) | Difference between trials first posted as preprints and subsequently published and published trials without preprints (%; 95% CI) | P value |
| Trial registered | 319 (89.6) | 149 (82.3) | 97 (96.0) | 13.72 (7.0 to 20.5) | <0.001 | 73 (98.6) | 16.3 (10.2 to 22.5) | <0.001 |
| Study status | ||||||||
| Completed | 268 (75.3) | 142 (78.5) | 75 (74.3) | −4.2 (−14.6 to 6.2) | 0.44 | 51 (68.9) | −9.5 (−21.7 to 2.6) | 0.12 |
| Interim results | 20 (5.6) | 5 (2.8) | 10 (9.9) | 7.1 (0.8 to 13.4) | 0.03 | 5 (6.8) | 4.0 (−2.2 to 10.2) | 0.21 |
| Terminated early for benefit | 2 (0.6) | 0 | 2 (2.0) | 2.0 (−0.7 to 4.7) | 0.15 | 0 | 0 (0%) | 1.00 |
| Terminated early owing to feasibility | 66 (18.5) | 34 (18.8) | 14 (13.9) | −4.9 (−13.7 to 3.9) | 0.28 | 18 (24.3) | 5.5 (−5.8 to 16.9) | 0.34 |
| Type of treatment | ||||||||
| Drug treatment | 293 (82.3) | 157 (86.7) | 77 (76.2%) | −10.5 (−20.2 to −0.8) | 0.03 | 59 (79.7) | −7.0 (−17.4 to 3.4) | 0.19 |
| Prophylaxis | 17 (4.8) | 7 (3.9) | 8 (7.9) | 4.0 (−1.9 to 10.0) | 0.18 | 2 (2.7) | −1.2 (−5.8 to 3.5) | 0.64 |
| Antiviral antibodies and cellular treatments | 46 (12.9) | 17 (9.4) | 16 (15.8) | 6.5 (−1.8 to 14.7) | 0.13 | 13 (17.6) | 8.2 (−1.5 to 17.8) | 0.10 |
| Inpatient* | 252 (74.3) | 129 (72.3) | 65 (69.9) | −4.3 (−15.6 to 7.1) | 0.47 | 58 (80.6) | 6.4 (−4.8 to 17.6) | 0.27 |
| Severe/critical disease* | 88 (26.0) | 46 (26.4) | 26 (28.0) | 1.5 (−9.7 to 12.8) | 0.80 | 16 (22.2) | −4.2 (−15.8 to 7.4) | 0.49 |
| No of centres | ||||||||
| Single centre | 151 (42.4) | 86 (47.5) | 39 (38.6) | −8.9 (−20.9 to 3.1) | 0.15 | 26 (35.1) | −12.4 (−25.5 to 0.7) | 0.06 |
| Multicenter | 177 (49.7) | 85 (47.0) | 48 (47.5) | 0.6 (−12.0 to 12.7) | 0.93 | 44 (59.5) | 12.5 (−0.8 to 25.8) | 0.07 |
| Median (IQR) number of participants | 106 (60-266) | 101 (60-208) | 163 (64-412) | — | 0.01 | 100 (60 to 249) | — | 0.54 |
| Primary outcome—significant | 151 (42.4) | 74 (40.9) | 49 (48.5) | 7.6 (−4.5 to 19.7) | 0.22 | 28 (37.8) | −3.1 (−16.2 to 10.1) | 0.66 |
| Any secondary outcome(s)—significant | 186 (52.2) | 90 (49.7) | 53 (52.5) | 2.8 (−9.4 to 14.9) | 0.67 | 43 (58.1) | 8.4 (−5.0 to 21.8) | 0.22 |
| Reasons for high risk of bias | ||||||||
| Randomisation | 122 (34.3) | 67 (37.0) | 34 (33.7) | −3.4 (−15.0 to 8.2) | 0.58 | 19 (25.7) | −11.3 (−23.5 to 0.9) | 0.07 |
| Deviations from intended intervention | 214 (60.1) | 111 (61.3) | 56 (55.5) | −5.9 (−17.9 to 6.1) | 0.34 | 47 (63.5) | 2.2 (−10.9 to 15.3) | 0.76 |
| Missing outcome data | 19 (5.3) | 11 (6.1) | 5 (5.0) | −1.1 (−6.6 to 4.4) | 0.70 | 3 (4.1) | −2.0 (−7.7 to 3.7) | 0.50 |
| Measurement of the outcome | 7 (2) | 6 (3.3) | 1 (1.0) | −2.3 (−5.6 to 0.9) | 0.16 | 0 | −3.3 (−5.9 to −0.7) | 0.01 |
| Selective reporting | 22 (6.2) | 9 (5.0) | 8 (7.9) | 3.0 (−3.2 to 9.1) | 0.35 | 5 (6.8) | 1.8 (−4.8 to 8.3) | 0.60 |
| Overall high risk of bias | 228 (64) | 117 (64.6) | 61 (60.4) | −4.2 (−16.1 to 7.6) | 0.49 | 50 (67.6) | 2.9 (−9.8 to 15.7) | 0.67 |
| Funding† | ||||||||
| Industry | 79 (22.2) | 32 (17.7) | 30 (29.7) | 12.0 (1.5 to 22.5) | 0.02 | 17 (23) | 5.3 (−5.8 to 16.4) | 0.35 |
| Government | 126 (35.4) | 47 (26.0) | 37 (36.6) | 10.7 (−0.7 to 22.0) | 0.07 | 42 (56.8) | 30.8 (17.8 to 43.8) | <0.001 |
| Institution | 128 (36.0) | 69 (38.1) | 32 (31.7) | −6.4 (−17.9 to 5.1) | 0.28 | 27 (36.5) | −1.6 (−14.7 to 11.4) | 0.82 |
| Not-for-profit | 51 (14.3) | 24 (13.3) | 12 (11.9) | −1.4 (−9.4 to 6.6) | 0.75 | 15 (20.3) | 7.0 (−3.4 to 17.4) | 0.19 |
| Not reported | 23 (6.5) | 12 (6.6) | 10 (9.9) | 3.3 (−3.6 to 10.1) | 0.36 | 1 (1.4) | −5.3 (−9.8 to −0.8) | 0.02 |
| No funding | 49 (13.8) | 33 (18.2) | 10 (9.9) | −8.3 (−16.4 to −0.2) | 0.04 | 6 (8.1) | −10.1 (−18.5 to −1.7) | 0.02 |
CI=confidence interval; IQR=interquartile range.
*Estimates only include trials addressing treatment.
†Trials might be classified in more than one category.