Example of trial* reporting different outcome data between preprint51 and publication,2 by outcome and treatment group
| Outcome | Preprint51 | Publication2 |
| Mortality | Corticosteroids: 454/2104 (21.6) Standard care: 1065/4321 (24.6) | Corticosteroids: 482/2104 (22.9) Standard care: 1110/4321 (25.7) |
| Mechanical ventilation | Corticosteroids: 92/1780 (5.2) Standard care: 258/3638 (7.1) | Corticosteroids: 110/1780 (6.2) Standard care: 298/3638 (8.2) |
Data are events/total number (%) of participants in each treatment group.
*The RECOVERY publication on dexamethasone treatment for covid-19 reported different results for mortality and mechanical ventilation—likely not because of an error in the preprint but because of events accumulating in the trial from when the preprint was posted to when it was published.2 51