Table 1.
Short-term and longer-term effects of corticosteroids on mortality
| Mortality endpoints |
Number of |
Relative risk |
Risk difference |
|||
|---|---|---|---|---|---|---|
| Studies | Patients | Effect size (95% CI) | I 2 | Effect size (95% CI) | I 2 | |
|
Randomized controlled trials
a
|
|
|
|
|
|
|
| ICU mortality |
3 |
292 |
0.55 (0.24 to 1.25) |
75% |
−0.28 (−0.53 to −0.03)* |
76% |
| Hospital mortality |
2 |
201 |
0.49 (0.12 to 2.07) |
77% |
−0.26 (−0.65 to 0.13) |
75% |
| 60-day mortality |
2 |
279 |
0.97 (0.75 to 1.26) |
0% |
−0.01 (−0.12 to 0.10) |
0% |
|
Cohort studies |
|
|
|
|
|
|
| ICU mortality |
5 |
226 |
1.05 (0.74 to 1.49) |
1% |
Not calculable |
- |
| Hospital mortality |
4 |
317 |
1.00 (0.23 to 4.34) |
75% |
Not calculable |
- |
| 60-day mortality | 2 | 264 | 1.30 (0.96 to 1.78) | 0% | Not calculable | - |
aTrials testing preventive therapy of corticosteroids for acute respiratory distress syndrome (ARDS) were excluded. Some studies provided data on both short-term and longer-term mortality. For cohort studies, only pooled relative risk was provided because adjusted results were needed for data pooling *P-value <0.05.