In a multicenter RCT of 566 ICU patients with hypoactive or hyperactive delirium, haloperidol and ziprasidone did not significantly reduce days alive without delirium or coma compared to placebo (7.9 vs. 8.7 vs. 8.5 days; P = 0.26). Secondary outcomes, including survival and time to mechanical ventilation freedom, were also similar. Findings suggest haloperidol or ziprasidone do not improve delirium duration in critically ill patients with acute respiratory failure or shock.

This multicenter RCT assessed early ECMO versus conventional mechanical ventilation in 249 severe ARDS patients (PaO₂/FiO₂ <50 for >3 hours or <80 for >6 hours, or pH <7.25 with PaCO₂ >60 mmHg for >6 hours). Early ECMO did not significantly reduce 60-day mortality (35% vs. 46%, p=0.09) but reduced treatment failure (35% vs. 58%, p<0.001). ECMO group had more bleeding events (46% vs. 28%) but no ischemic strokes. Study limitations include early termination, high crossover rates, and underpowering. Findings suggest no definitive mortality benefit of early ECMO, with secondary benefits warranting further investigation.

This multicenter, double-blind RCT investigated hydrocortisone versus placebo in 3,658 mechanically ventilated patients with septic shock requiring vasopressors. There was no difference in 90-day mortality (27.9% vs. 28.8%, p=0.50). Hydrocortisone reduced shock resolution time (3 vs. 4 days, p<0.001), ICU discharge time (10 vs. 12 days, p<0.001), mechanical ventilation cessation time (6 vs. 7 days, p<0.001), and transfusion rates (37% vs. 41.7%, p=0.004). Findings support hydrocortisone use for vasopressor-dependent septic shock to expedite recovery.

In a multicenter RCT of 1241 patients with septic shock, hydrocortisone plus fludrocortisone significantly reduced 90-day mortality compared to placebo (43.0% vs. 49.1%; P = 0.03) and improved ICU discharge, hospital discharge, and 180-day mortality rates. The intervention accelerated vasopressor weaning (17 vs. 15 vasopressor-free days; P < 0.001) and organ failure resolution (14 vs. 12 organ-failure-free days; P = 0.003). Hyperglycemia was more common, but serious adverse events were similar. These findings support hydrocortisone use for septic shock requiring vasopressors.

In a multicenter RCT of 488 patients with septic shock and severe acute kidney injury, early initiation of renal replacement therapy (RRT) within 12 hours did not improve 90-day mortality compared to delayed initiation after 48 hours (58% vs. 54%; P = 0.38). Delayed RRT resulted in more RRT-free days (16 vs. 12 days; P = 0.006). Early RRT led to higher RRT use (97% vs. 62%; P < 0.001). These findings support delaying RRT unless emergent indications are present.

In a multicenter RCT of 706 patients with acute MI and cardiogenic shock, culprit-lesion-only PCI significantly reduced the composite outcome of 30-day mortality or renal replacement therapy (45.9% vs. 55.4%; RR 0.83, 95% CI 0.71–0.96, P = 0.01) and 30-day mortality alone (43.3% vs. 51.6%; RR 0.84, 95% CI 0.72–0.98, P = 0.03) compared to multivessel PCI. Renal replacement rates and secondary outcomes were similar. Culprit-lesion-only PCI with staged revascularization is supported as the preferred approach.

In a multicenter RCT of 1010 patients with moderate to severe ARDS, lung recruitment with PEEP titration by best respiratory-system compliance increased 28-day mortality compared to a low-PEEP strategy (55.3% vs. 49.3%; HR 1.20, 95% CI 1.01–1.42, P = 0.041). Secondary outcomes included higher 6-month mortality, fewer ventilator-free days, and increased risk of pneumothorax and barotrauma. Routine use of recruitment maneuvers is not supported, though subgroups may warrant further investigation.

In a multicenter RCT of 321 patients with catecholamine-resistant vasodilatory shock, angiotensin II significantly improved mean arterial pressure compared to placebo (69.9% vs. 23.4%; P < 0.001) and improved cardiovascular SOFA scores (−1.75 vs. −1.28; P = 0.01). Mortality at 28 days was not significantly different (46% vs. 54%; P = 0.12). While angiotensin II effectively raises blood pressure, further research is needed to assess its impact on patient-centered outcomes before widespread adoption.

In a multicenter RCT of 380 patients with severe sepsis but not septic shock, early hydrocortisone use did not significantly reduce progression to septic shock within 14 days compared to placebo (21.2% vs. 22.9%; P = 0.70). Mortality, secondary infections, weaning failure, and muscle weakness were similar between groups, though hyperglycemia was more common with hydrocortisone (90.9% vs. 81.5%; P = 0.009). These findings do not support routine early hydrocortisone use in severe sepsis.

In a multicenter RCT of 409 septic shock patients, early vasopressin use (titrated up to 0.06 U/min) did not improve kidney failure–free days compared to norepinephrine (median 9 vs. 13 days; difference −4 days, 95% CI −11 to 5). Vasopressin was associated with reduced renal replacement therapy (25.4% vs. 35.3%; difference −9.9%, 95% CI −19.3% to −0.6%) but no difference in mortality or serious adverse events. These findings do not support vasopressin as first-line therapy over norepinephrine, though potential benefits warrant further study.

In a multicenter RCT of 620 critically ill patients with severe acute kidney injury, early initiation of renal replacement therapy (RRT) did not improve 60-day mortality compared to delayed initiation (48.5% vs. 49.7%; P = 0.79). Nearly half of the delayed group avoided RRT entirely. Early RRT was associated with higher rates of catheter-related bloodstream infections (10% vs. 5%; P = 0.03). Findings support delaying RRT initiation unless clear indications are present.

In a multicenter RCT of 310 ICU patients with nonhypercapnic acute hypoxemic respiratory failure, high-flow oxygen therapy did not significantly reduce 28-day intubation rates compared to standard oxygen or noninvasive ventilation (38% vs. 47% and 50%; P = 0.18). However, high-flow oxygen improved ventilator-free days (24 vs. 22 vs. 19; P = 0.02) and lowered 90-day mortality (HR 2.01 vs. standard, P = 0.046; HR 2.50 vs. NIV, P = 0.006). Further large-scale studies are needed.