Table 2.
Primary and secondary outcomes
| Outcome | Methylprednisolone (n = 297) | Placebo (n = 287) | Differencea (95% CI) | p valueb |
|---|---|---|---|---|
| Primary outcome | ||||
| Died on or prior to study day 60—no./total no. (%) | 47/286 (16) | 50/277 (18) | 0.89 (0.58, 1.38)c | 0.61 |
| Secondary outcomes | ||||
| In-hospital morbidity and mortality | ||||
| Vasopressor dependent shock during initial hospital stay among those who did not have vasopressor-dependent shock at randomization—no./total no. (%) | 13/274 (5) | 12/271 (4) | 1.08 (0.48, 2.4)d | 1.00 |
| ALI-ARDS during initial hospital stay among those who did not have ALI-ARDS at randomization—no./total no. (%) | 10/265 (4) | 8/241 (3) | 1.14 (0.44, 2.94)d | 1.00 |
| MODS-free days in study days 1–8—median (IQR) and no./total no. (%) | 0 (0–0) | 0 (0–0) | NE | 0.38 |
| 0 | 249/288 (86) | 252/275 (92) | ||
| 1 | 5/288 (2) | 10/275 (4) | ||
| 2 | 4/288 (1) | 4/275 (1) | ||
| 3 | 4/288 (1) | 2/275 (1) | ||
| 4 | 3/288 (1) | 2/275 (1) | ||
| 5 | 3/288 (1) | 0/275 (0) | ||
| 6 | 0/288 (0) | 1/275 (0) | ||
| 7 | 12/288 (4) | 2/275 (1) | ||
| 8 | 8/288 (3) | 2/275 (1) | ||
| MV-free days in study days 1–8—median (IQR) | 8 (4–8) | 8 (3–8) | 0 (− 0.4, 0.4) | 1.00 |
| MV-free days in study days 1–28e—median (IQR) | 28 (23–28) | 28 (21–28) | 0 (− 0.6, 0.6) | 1.00 |
| MV-free days in study days 1–8 in the MV stratum—median (IQR) | 4 (0–7) | 1 (0–6) | 3 (1.1, 4.9) | 0.66 |
| MV-free days in study days 1–28 in the MV stratume—median (IQR) | 23 (13–27) | 21 (0–26) | 2 (− 0.8, 4.8) | 1.00 |
| Duration of initial ICU stay (from day 0)f—median (IQR) | 3 (2–7) | 4 (2–7) | − 1 (− 1.7, − 0.3) | 1.00 |
| Duration of total ICU stay up to study day 28 (from day 0)f—median (IQR) | 3 (2–8) | 4 (2–8) | − 1 (− 1.7, − 0.3) | 1.00 |
| Duration of initial hospital stay (from day 0)f—median (IQR) | 7 (4–12) | 8 (4–15) | − 1 (− 2.3, 0.3) | 1.00 |
| Hospital Mortality—no./total no. (%) | 34/291 (12) | 28/281 (10) | 1.2 (0.7, 2.03)g | 1.00 |
| Post-discharge morbidity and mortality | ||||
| Cardiovascular complicationsh—no./total no. (%) | ||||
| Up to day 28 | 11/251 (4) | 12/234 (5) | 0.85 (0.37–1.96) | 1.00 |
| Up to day 60 | 18/257 (7) | 17/250 (7) | 1.03 (0.52–2.05) | 1.00 |
| Up to day 180 | 31/257 (12) | 30/253 (12) | 1.02 (0.6–1.74) | 1.00 |
| VR-12 Physical Component Scorei | ||||
| Day 28 | 37.5 ± 14 | 37.2 ± 13.8 | 0.2 (− 2.6, 3) | 1.00 |
| Day 60 | 38.8 ± 15.3 | 40.3 ± 14.5 | − 1.5 (− 4.5, 1.5) | 1.00 |
| Day 180 | 41.2 ± 14.9 | 40.8 ± 15.7 | 0.4 (− 3, 3.8) | 1.00 |
| VR-12 Mental Component Scorei | ||||
| Day 28 | 32.7 ± 9.6 | 33.9 ± 9.3 | − 1.2 (− 3.2, 0.7) | 1.00 |
| Day 60 | 32.4 ± 8.9 | 32 ± 9 | 0.4 (− 1.4, 2.2) | 1.00 |
| Day 180 | 31.5 ± 8.2 | 32.8 ± 9.2 | − 1.3 (− 3.2, 0.6) | 1.00 |
| Katz ADLi | ||||
| Day 28 | 4.99 ± 1.84 | 4.74 ± 2.07 | 0.25 (− 0.14, 0.64) | 1.00 |
| Day 60 | 5.11 ± 1.71 | 5.03 ± 1.81 | 0.08 (− 0.27, 0.43) | 1.00 |
| Day 180 | 5.12 ± 1.66 | 5.02 ± 1.85 | 0.1 (− 0.29, 0.48) | 1.00 |
| Lawton IADLi | ||||
| Day 28 | 5.39 ± 2.61 | 5.3 ± 2.72 | 0.09 (− 0.45, 0.62) | 1.00 |
| Day 60 | 5.42 ± 2.62 | 5.71 ± 2.63 | − 0.29 (− 0.82, 0.24) | 1.00 |
| Day 180 | 5.77 ± 2.57 | 5.96 ± 2.6 | − 0.19 (− 0.76, 0.38) | 1.00 |
| Any re-hospitalization within 12 monthsj—no./total no. (%) | 135/253 (53) | 120/250 (48) | 1.24 (0.87, 1.76)g | 1.00 |
| Number of re-hospitalizations within 12 months k—median (IQR) | 1 (1–3) | 2 (1–3) | − 1 (− 1.4, − 0.6) | 1.00 |
| Died by study day 180—no./total no. (%) | 59/279 (21) | 65/274 (24) | 0.86 (0.58, 1.29)g | 1.00 |
| Restricted mean survival time up to day 180—RMST (SE) | 151.5 (3.6) | 149 (3.7) | 2.5 (− 7.7, 12.6)l | 1.00 |
| Died by 1 year—no./total no. (%) | 79/260 (30) | 84/253 (33) | 0.88 (0.61, 1.27)g | 1.00 |
| Time to death (days)—no./total no. (%) | 79/297 (27) | 84/287 (29) | 0.9 (0.66, 1.22)g | 1.00 |
| Exploratory outcome | ||||
| Duration of MV up to day 28 in the participants who were on MV at randomization—median (IQR)m | 4 (1–9) | 7 (2–27) | 1.4 (1, 2) | 0.21 |
ALI-ARDS acute lung injury- acute respiratory distress syndrome; MODS multiple organ dysfunction syndrome; IQR inter quartile range; VR-12 Veterans RAND-12; ADL Activities of Daily Living; IADL instrumental activities of daily living; RMST Restricted mean survival time
aDifference between treatment group is expressed as: odds ratio for binary variables; difference in medians for MV-free days up to day 8 and day 28, duration of ICU stay, total ICU days up to day 28, duration of initial hospital stay, and number of re-hospitalizations within 12 months; difference in means for other continuous variables; hazards ratio for time to event variables. NE: The difference in median and the 95% CI could not be estimated, because majority of the observed values were at the one end of the distribution
bThe p values for secondary and exploratory outcomes are adjusted for multiplicity by Bonferroni correction (for 12 in-hospital outcomes and 21 post-discharge outcomes). The widths of the confidence intervals for the treatment differences in secondary and exploratory outcomes have not been adjusted for multiplicity and therefore cannot be used to infer treatment effects
cThe adjusted odds ratio for 60-day mortality is 0.90 (95% CI 0.57–1.40; p = 0.63) when adjusted for site and mechanical ventilation status at randomization, and is 0.87 (95% CI, 0.53–1.42; p = 0.58) when adjusted for site, mechanical ventilation status at randomization, age, APACHE III score, CCI, bacteremia, use of anti-inflammatory medications at baseline, and use of macrolide at baseline. The unadjusted absolute risk difference in the 60-day mortality is − 2% (95% CI − 8 to 5%)
dThe unadjusted absolute risk difference is 0% (95% CI − 3 to 4%) for vasopressor-dependent shock, and 0% (95% CI − 3 to 4%) for ALI-ARDS
eMV-free days from days 1 to 28 was calculated in 558 participants (287 in the methylprednisolone group and 271 in the placebo group). In the MV stratum, MV-free days from days 1 to 28 was calculated in 181 participants (92 in the methylprednisolone group, 89 in the placebo group)
fThe number of participants for whom the outcome was calculated in the methylprednisolone and placebo groups was, respectively, duration of initial ICU stay: 295 and 281; duration of total ICU stay up to day 28: 291 and 280; duration of initial hospital stay: 291 and 281
gThe unadjusted methylprednisolone versus placebo absolute risk difference is 2% (95% CI − 3 to 7%) for hospital mortality, 5% (95% CI − 3 to 14%) for re-hospitalization within 12 months, − 3% (95% CI − 10 to 4%) for 180-day mortality, − 3% (95% CI − 11 to 5%) for 1-year mortality, and -3% (95% CI − 10 to 5%) for mortality over the study period
hCardiovascular complications included acute myocardial infarction, serious arrhythmias, new congestive heart failure or acute worsening of long-term congestive heart failure, and cardo-respiratory arrest
iThe number of participants for whom the VR-12 Physical Component Score (PCS) and Mental Component Score (MCS) was calculated at day 28, 60, and 180 was: 197, 201, and 166 in the methylprednisolone group and 184, 177, and 148 in the placebo group. The number of participants for whom the Katz Activities of Daily Living (ADL) score was calculated at day 28, 60, and 180 was: 203, 203, and 172 in the methylprednisolone group and 187, 180, and 150 in the placebo group. The number of participants for whom the Lawton Instrumental Activities of Daily Living (IADL) score was calculated at day 28, 60, and 180 was: 202, 201, and 169 in the methylprednisolone group and 187, 180, and 149 in the placebo group
jAmong the 510 participants (257 in methylprednisolone group, 253 in placebo group) who were discharged alive from initial hospitalization. Seven of the 510 participants had missing re-hospitalization data (4 in methylprednisolone group, 3 in placebo group) and were excluded from analysis
kAmong the 255 patients (135 in methylprednisolone group, 120 in placebo group) who had at least one re-hospitalization within 12 months
lThe estimated difference in RMST up to study day 180, adjusted for MV status at randomization in an RMST regression using pseudovalue method, is 2.3 (95% CI − 7.8 to 12.4), and p value adjusted for multiplicity = 1.00
mParticipants who died on MV on or prior to day 28 were censored on day 29. 185 participants were included in this analysis (91 in the methylprednisolone group and 94 in the placebo group)