What does the Literature say?

This page summarizes landmark randomized trials in critical care medicine frequently referenced in ICU practice, board review, and evidence-based guidelines.

The Vault 🔐

Dropbox - Password Required

Dropbox is a free service that lets you bring your photos, docs, and videos anywhere and share them easily. Never email yourself a file again!

Dropbox

Landmark Critical Care Trials 📚

Table of Contents

• Albumin, Colloids & Crystalloids
• Anemia & Transfusion
• Antibiotics & Procalcitonin
• ARDS
• Cardiac Arrest
• Cardiogenic Shock
• Catheters & Hemodynamic Monitoring
• Delirium & Sedation
• ECMO
• Fever
• Heart Failure
• ICU Bundles
• Mechanical Ventilation
• Noninvasive Ventilation (NIV / HFNC)
• Nutrition
• Pancreatitis
• Renal Replacement Therapy
• Sepsis & Septic Shock
• Stress Hyperglycemia
• Stress-Related Mucosal Disease
• Stroke & Neurocritical Care
• Ventilator-Associated Pneumonia
• Trauma & Surgical ICU
• Venous Thromboembolism & Pulmonary Embolism

Albumin & Colloids & Crystalloids

SAFE , 2004.
Albumin and normal saline result in similar clinical outcomes when administered to ICU patients for volume resuscitation, except for those with traumatic brain injury (TBI GCS <7) who do worse with albumin. Randomized. 6,997 ICU pts to either albumin or saline. No difference in all-cause mortality. Nonsignificant trend towards mortality benefit in albumin use for severe sepsis; this later shown in 2013 Cochrane review to be no benefit of albumin over crystalloids for fluid resusc.

CHEST, 2012.
Blinded randomized. 7000 ICU pts. 6% hydroxyethyl starch vs 0.9% NS for fluid resusc. Colloid arm a/w 21% increased risk of RRT. No mortality benefit. Colloid arm a/w increased rates of rash, pruritus, and hepatic failure.

HES vs Ringers, 2012
Randomized 804 ICU pts w/severe sepsis to fluid resuscitation w/6% HES or RInger's acetate. HES arm had increased risk of death @90 days and more likely to need RRT.

ALBIOS, 2014.
Among patients with severe sepsis or septic shock, daily administration of albumin to maintain serum albumin ≥3 g/dL was not associated with a reduction in all-cause mortality at 28 OR 90 days when compared to no albumin. Albumin pts had better SOFA suscores and fewer vasopressors/inotropes used.

SALT-ED, 2018.
Single center n=13K NON-critically ill ER pts given balanced crystalloids (LR vs Plasma-lyte) vs isotonic crystalloid (NS). Does not reduce duration of hospitalization; however, balanced cystalloids a/w reduction in major adverse renal events.

SMART-MED, SMART-SURG, 2018.
Critically ill pts randomized to NS IVF vs "balanced" IVF (Plasma-Lyte A or Lactated Ringer's). Primary endpoint: major adverse kidney at 30 d: any death, new RRT, persistent renal dysfxn. Balanced IVF group had less new RRT, deaths, or persistent renal dysfxn at 30 days.

The SMART and SALT-ED (Vandy ER) trials are massive, multiple-crossover trials which compare saline versus balanced crystalloids among critically ill and non-critical patients.
Both studies found a 1% increase in death or renal failure with the use of saline.

Anemia

TRICC, 1999.
Mortality reduction with restrictive strategy. Randomized. 838 ICU pts without evidence of active bleeding to restrictive transfusion strategy (txf to maintain Hgb >7 g/dL) vs liberal strategy (txf to maintain Hgb >=10 g/dL). Euvolemic pre-txf. Decreased rate of in-hospital mortality in restrictive strategy. Also, trend towards improved outcome in active cardiac ischemia.

ABLE, 2015.
Randomized. <8 day transfusion with fresh pRBCs (<8d shelf-life) vs standard issue (up to 42d shelf-life) in medical critically ill pts. No mortality difference at 90 days.

TRANSFUSE, 2017.
Randomized. Young vs old pRBCs. More febrile nonhemolytic txf reactions in pts receiving younger units.

Antibiotics & Procalcitonin

KUMAR, 2006.
Retrospective of med, surg, and mixed med-surg ICUs in US/Canada. 2,731 pts with septic shock. Survival to hospital discharge. Effective antimicrobial administration within the first hour of documented hypotension in septic shock is a/w increased survival to hospital discharge. 50% of septic shock pts received effective antimicrobials within 6hours of documented hypotension.

KUMAR, 2009.
Inappropriate initial antimicrobial selection in septic shock occurs 20% of patients and is a/w 5x reduction in survival.

PRORATA, 2010.
Randomized. 630 ICU pts with suspected bacterial infections to procalcitonin or control. Procal-arm: procal-guided strategy to guide docs to start, to continue, or to stop abx. Final decision to start/to stop was the doc's. Control arm: abx per current std guidelines. Results: procal-guided abx strategy to treat suspected bacterial infections led to less abx-days WITHOUT a >10% increase in 28 day- or 60 day- mortality. In other words, mortality was noninferior to those at day 28 and day 60 of the control group. (10% as margin for non-inferiority). Intention to treat trial. Benefit: reduce abx resisitance in ICU. SSC 2016: suggested to use procal to shorten duration of abx therapy with sepsis and to stop abx if limited evidence of infection in those initially thought to have sepsis.

Pro-ACT, 2017.
Randomized to review use of procalcitonin in reducing antibiotic use in LRTI. Intention to treat. Final diagnoses reviewed in ED:
Asthma exacerbation: 39.3%
COPD exacerbation: 31.9%
Acute bronchitis: 24.2%
Community acquired pneumonia: 19.9%
The provision of procalcitonin assay results, along with instructions on their interpretation, to emergency department and hospital-based clinicians did not result in less use of antibiotics than did usual care among patients with suspected lower respiratory tract infection. Limitation: Study does not address critical question – is it safe to withhold antibiotics based on a low procalcitonin level?

ARDS

AMATO, 1998.
Randomized. 53 mechanically ventilated ARDS pts to protective vs conventional ventilation strategy for 28 day endpoint. Protective strategy had 28 day survival advantage, higher rate of MV weaning, and lower rate of barotrauma. Protective ventilation strategy: PEEP on the lower inflection point of the P-V curve, TV of <6cc/kg, driving pressure of <20 cmH20 above the PEEP, permissive hypercapnia, preferential use of pressure mode.

ARDSNet, 2000. (ARMA)
Low tidal volume strategy (6 cc/kg of ideal body weight) on mechanical ventilation of patients with ARDS leads to improved mortality at Day 180 and more ventilator-free days. NNT was 11. Dx <=36h prior to enrollment: acute decline in P:F ratio <=300, bilateral pulm infiltrates on CXR with edema, PCWP <=18 no LA hTN. Protocol: low TV - If Pplateau <25 cmH20, TV increased by 1cc/kg IBW until Pplateau >=25cm H20 OR TV 6cc/kg IBW. If severely dyspneic, TV could go to 8 cc/kg IBW for Pplateau <=30 cmH20. Min TV was 4 cc/kg IBW; Min allowed arterial pH was 7.15. Standard PEEP protocol also used.

Criticisms: Ethical conerns -educational materials part of the informed consent process were inadequate. Was 12 cc/kg 	of IBW really the standard at the time?

ALVEOLI, 2004.
Prospective, RCT of ARDS pts. Compared higher PEEP with lower PEEP using standard protocol. Showed higher PEEP led to remarkable improvements in oxygentation. No improvement in survival rate.

FACTT, 2006.
Optimal fluid strategy in ARDS? Randomized. 1,000 ALI/ARDS pts to liberal (CVP 10 to 14) vs conservative (CVP <4) fluid mgmt strategy. At primary endpoint of 60 day mortality, no difference. Conservative group: more vent-free days, lower ICU-LOS.

** Also compared CVC and PAC in monitoring CVP. No difference of 60 day mortality. More reason not to use PAC. ESCAPE, PAC-MAN in 2005 previously showed higher complication rates with PACs.

EDEN, 2012.
Randomized. 1,000 pts with ARDS to enteral nutrition to either trophic feeding (10-20 mL/hr) x 6 days OR attempt to reach goal feeding rates as soon as possible. Similar oucomes: days alive and free of vent support, hospital mortality and organ-failure free days. Trophic rates of EN or targeting goal rate during first 6 days is acceptable.

ACURASYS, 2010.
Paralysis with cisatracurium for 48 hours in patients with early severe ARDS improves 90 day survival and increases ventilator-free days. Early severe ARDS: P to F ratio <=150, 1st 48 hours. Reduction of incidence of PTX. Unexpected! 4% v 11.7%

BERLIN, 2012.
Berlin Definition of ARDS: ALI & Swan removed from AECC 1994 definition.
Respiratory insult within 1 week
Bilateral opacities not explained by nodules/effusions/lung collapse,
No sign of cardiac originating pulm edema,
Mild: 200 to <=300, Moderate 100 to <=200, Severe <=100 (all have PEEP >=5; mild can have CPAP >=5).

OSCILLATE, 2013.
Randomized. 548 pts with moderate to severe early ARDS, early HFOV might increase in-hospital mortality. Criticism: high dose of sedation was given.

OSCAR, 2013.
HFOV in ARDS and did not show mortality benefit.

PROSEVA, 2013.
Randomized. 466 medical ICU pts with severe ARDS (P:F <150) at <36h post-intubation to 16hours/day intermittent proning vs standard supine positioning. Important: PROSEVA used low-tidal ventilation AND therapeutic paralysis. Did not report increase in pressure ulcers; regular hospital beds. SSC 2016 recommends proning in ARDS from sepsis and PF <150.

AMATO, 2015.
Post hoc observational analysis. Driving pressure trial. DP = Pplat - PEEP = VT / Crs. LIP to UIP.

ROSE, 2019.
Multicenter (PETAL network) randomized 1,408 moderate-to-severe ARDS to cisatricurium vs no cisatricurium. Study stopped early due to futility. No difference in primary outcome 90 day inpatient mortality. Less sedation, higher PEEP strategy compared to PROSEVA. Early NM blockade resulted in Increase in ICU-acquired weakness and serious adverse cardiovascular events. Due to disparity in results of ROSE compared to ACURASYS, continuous NMB-ade not currently recommended.

EPVent-2, 2019.
North American study randomizing 200 mechanically ventilated pts with moderate to severe ARDS (P:f ratio <=200) to Peosphageal-guided PEEP strategy vs empirical high PEEP-FiO2 strategy. Both arms had low tidal volume strategy. Primary outcome was death and vent-free days through day28. No difference in primary composite end point (death and vent free days through day 28) so trial authors do not support Pesophageal-guided PEEP titration in ARDS.

Cardiac Arrest

HACA, 2002.
Randomized. 275 pts with ROSC after witnessed cardiac arrest (VF or pulseless VT) to mild therapeutic hypothermia (32 to 34 C) or to standard normothermia. Improved 6 MONTH neuro outcomes in TH group 55% vs 39%. NNT 6 for unfav neuro outcome; NNT 7 for survival.

AHA Post-cardiac arrest care, 2010
- Comatose adults with ROSC post-OHCA VF arrest should be cooled to 32 to 34C for 12 to 24h, Class I, Level B
- Comatose adults with ROSC post-IN hospital arrest of ANY rhythm or OHCA PEA/asystole can be considered for TH, Class 2b, Level B
AHA/ACCF STEMI, 2013
- TH for comatose pts with STEMI and OHCA from VF/pulseless VT, Class I, Level B

Bernard, 2002.
Randomized. 77 pts with ROSC after out-of-hospital VF arrest. Cooled to 33C within 2h of ROSC adn for 12h vs standard normothermia. 49% hypothermia arm survived with normal neuro fxn to rehab or home.

TTM, 2013.
Randomized. 939 pts with shockable or nonshockable cardiac arrest to TTM with goal of 33C or more normothermic 36C. With mean of 8.5 months, no difference in primary outcome of all-cause mortality btw 2 groups.

Cardiogenic Shock

SHOCK, 1999.
Early revascularization in patients with cardiogenic shock due to MI led to nonsignificant trend to improved mortality at 30 days (primary endpoint) compared to early medical optimization.  At 6 months, there was improved survival in early revasc group.

IABP-SHOCK, 2012.
Random 600 pts with acute MI complicated by cardiogenic shock to IABP or no IABP. 95% had PCI with stent. No difference in outcome mortality at 90 days. 2013 guidelines downgraded indication to Class 2A for IABP for acute MI complicated by shock (RVF, V septal rupture, acute severe MR).

Catheters

3SITES, 2015.
Multicenter randomized. 3,027 pts with 3.471 catheters were studied to see what difference in rate of major catheter-related complications were for subclavian, IJ, or femoral routes. Primary outcome: CRBSI or symptomatic DVT. Results: fewer BSIs and DVTs with subclavians as compared to IJ and fem. More PTX, in subclavians. CRBSI defined as colonization of catheter tip by >= 1000 cfu/mL of same organism recovered from peripheral BC. Tunneled and abx-impregnated caths not used (may reduce CRBSI). Experienced operators. More frequent cath tip colonization with fem vs subclav; however, not clincially significant.

PAC-Man, 2006.
Randomized. 1,014 ICU pts. PAC vs no PAC. No difference in hospital mortality. High nonfatal complication PAC rate of 9.5%.

ESCAPE, 2005.
Among patients admitted with acute decompensated heart failure, pulmonary artery catheter-guided therapy does not improve survival and increases adverse events compared to clinical assessment-guided therapy.  PAC group had improvement in quality of life and impact of disease in the first few months after discharge.

Delirium & Sedation

MENDES, 2007.
Random 106 mechanically ventilated ICU pts. Either Dexmetetomidine or lorazepam for sedation. Dex group had less delirium-days and less coma-days

SEDCOM, 2009.
Random 375 mechancially ventilated ICU pts. Either dex or MIDazolam until extubation OR for max of 30 days.
No difference in primary endpoint: time at target sedation level. Dex group: 30% less delirium and nearly 2 days less on the ventilator. Consistent with MENDES, 2007. Lack of daily sedation vacation. High dex dosing, high MIDazolam dosing.

DEXCOM, 2009.
Dexmedetomidine reduced the duration but not the incidence of delirium after cardiac surgery compared with morphine.

MIND-USA, 2018.
Randomized med-surg ICU pts, n=566. Haloperidol vs ziprasidone against placebo for delirium treatment over 14d period. 90% pts were hypoactive delirious. CAM-ICU based diagnosis. No benefit in primary outcome of days alive without delirium vs placebo.

SPICE III, 2019.
Multicenter randomized n=4K mechanically ventilated (40% resp failure). Dexmedetomidine (add-on propofol PRN) vs usual care (propofol, midazolam or both). Primary outcome at 90 day mortality. Primary use of dexmedetomidine versus usual care did not improve mortality and led to higher rate of bradycardia, hypotension, and asystole.

ECMO

CESAR, 2009.
Randomized. 180 pts transferred to ECMO center to get ECMO. Of the 75% who actually received ECMO, improved 6 month survival. Improved mortality in ECMO-referred pts.

EOLIA, 2018.
ECMO in severe ARDS. Cross-over. No difference in mortality in early ECMO vs conventional mgmt in servere ARDS.

Fever

HEAT, 2015.
Acetaminophen in febrile ICU pts. Randomized. 700 pts with fevers due to suspected/documented infection to acetaminophen or placebo. Continued until resolution of fevers, ICU discharge, cessation of abx, or death. Primary outcome at 28 days was ICU-free days: similar btw both arms. APAP use a/w shorter median ICU LOS. No benefit for acetaminophen use in pts with intent to reduce ICU LOS.

Heart Failure

V-HeFT, 1986.
In the era before beta-blockers and ACE inhibitors, ISDN/hydralazine showed a trend towards improved survival among pts with systolic heart failure. Self-identified black pts had improved survival with ISDN/hydralazine leading to positive A-HeFT trial.

CONSENSUS, 1987.
Impressive mortality benefit with ACE-I in severe, class III/IV heart failure.. 40% reduction in 6 month mortality. Risk reduction over 10 years was 30%..

V-HeFT II, 1991.
Head-to-head comparison of ISDN/hydralazine vs enalapril on mortality. Enalapril had survival benefit over ISDN/hydralazine.

SOLVD, 1991.
2,569 pts with symptomatic HF NYHA II to III. Enalapril vs placebo. 16% mortality reduction with enalapril.

SOLVD, 1992. Prevention.
Enalapril prevented development of CHF in prevention arm.

SAVE, 1992.
2,231 aystompatic pts LVEF <40% randomly assigned to captopril or placebo. Captopril arm decreased mortality by 19% at 42 mos. Decreased hospitalization for HF, and decreased recurrent MI.

AIRE, 1993.
2,006 pts with clinical HF showed 27% reduction in mortality at 15 months with ramipril.

MADIT, 1996. & MUSST, 1999.
Mortality benefit with ICD if LVEF <35% and NSVT.

AVID, 1997.
Antiarrhythmics vs ICD in sudden death? Secondary prevention of ICD placement in survivors (half with symptomatic HF and half with LV dysfxn) of sudden cardiac death or hemodynamically significant sustained VT improves survival.

DIG Trial, 1997.
6,800 pts with symptomatic systolic HF. No diff in survival btw digoxin and placebo. Survival decreased during hospitalization for pts with HF.

MADIT II, 2002.
Mortality benefit with ICD if LVEF <30% only. Mainly ICMP pts.

RALES, 1999.
1,653 pts NYHA III to IV. Spironolactone or placebo. Reduction in mortality at 24months. Gynecomastia and hyperkalemia were AE.

Val-HeFT, 2001.
Valsartan or placebo added to usual therapy of HF pts. Adding ARB as combo therapy (triple tx with ACEI and BB) increases mortality!

VMAC, 2002.
Nesiritide, IV nitro, or placebo after PAC placed in acute HF. Randomized. Observed x 24h. Nesiritide decreased mPCWP at 3h. Improved dyspnea/sx but not mortality at 30 d or 6mos. Later MAs showed worsen of renal fxn, decreased 30d survival. However, ASCEND-HF 2011 disproved this. In any case, not recommended for use.

BNP, 2002.
1,586 pts in ED with chief complaint of dyspnea. Plasma BNP of >400 pg/mL accurately predicted CHF. Levels <100 pg/mL predicted noncardiac dyspnea. 100-400 pg/mL unclear.

OPTIME-CHF, 2002.
Did not show that short term use of milrinone in acute HF would avoid hospital exac.

EPHESUS, 2003.
6,632 pts eplerenone vs placebo randomized. LV dysfunction (sCHF) post-MI. 15% reduction in mortality. Improved survivability in diabetics too.

A-HeFT, 2004.
Isosorbide dinitrate plus hydralazine improves survival and reduces hospitalization among black patients with HFrEF.
AHA/ACCF Heart Failure (2013, adapted)
: Recommendation of nitrates/hydralazine for self-described African American patients with NYHA III-IV HFrEF on OMT with ACE-inhibitors and beta blockers unless contraindicated (class I, level A)
: Suggestion of nitrates/hydralazine for all patients with current or previous symptomatic HFrEF who can't tolerate ACE- inhibitor or ARB therapy unless contraindicated (class IIa, level B)

ESCAPE, 2005.
Among patients admitted with acute decompensated heart failure, pulmonary artery catheter-guided therapy does not improve survival and increases adverse events compared to clinical assessment-guided therapy.  PAC group had improvement in quality of life and impact of disease in the first few months after discharge.

SCD-HeFT, 2005.
ICD vs amiodarone in ischemic vs nonischemic HF (EF <35%). Mortality benefit with ICD in both groups.

ICU Bundles

ABCDE, 2014.
Multicenter trial implenting bundle with Awakening & Breathing (Sedation vacation, SBT) Coordination, Delirium, Early Exercise/Mobility. Primary outcome was reduced ventilator days, and it was reached including less delirium.

Mechanical Ventilation

Yang-Tobin, 1991.
RSBI is ratio of f/VT (RR over tidal volume). >105 accurately predicts weaning failure. <105 breaths/min/L was associated with weaning success. In other words: high negative predictive value.

Supine Nosocomial Pneumonia, 1999.
Supine body position is risk factor for nosocomial pneumonia. Semirecumbent position reduces frequency and risk of nosocomial pneumonia, in pts who receive enteral nutrition especially. Risk increased by long duration MV and decreased consciousness. Independent RF: supine and enteral nutrition.

Kress Sedation Vacation Trial, 2000.
Randomized. 128 medical ICU pts on MV. Daily interruption arm vs control. Sedation vacation arm a/w decreased duration of MV by ~3 days and a/w decreased time in ICU by 3.5 days. No mortality benefit.

Awakening & Breathing Controlled (ABC) Trial, 2008.
"Wake up and breathe" protocol that followed daily sedation holiday with spontaneous breathing trials (SBT). Results: reduced time on ventilator, reduced days in ICU, and reduced days in hospital.

No Sedation in ICU Patients, 2010.
Randomized. 428 Danish pts on MV. No sedation vs propofol sedation (with midazolam) with daily interruption until awake. Both groups were treated with morphine for analgesia. Primary outcome was number of vent free days at 28 day mark. Intention to treat trial. Results: reduced time on the ventilator, reduced days in the ICU, reduced days in the hospital wehn compared to sedation holiday protocol.

TracMan, 2013.
Early vs late (perc or surg) tracheostomy placement in ICU pts. Randomized with 30 day mortality endpoint and intention to treat analysis. Early placement (<4 days of ICU admission) was not associated with improvement in 30 day mortality. Also, ability of doctors to predict who required extended vent time was limited. Late defined as on or after day 10.

Noninvasive Positive Pressure Ventilation

Cochrane, 2004.
Benefit of NIPPV as first line intervention as adjunct therapy to usual medical care in acute resp failure due to AECOPD. Consider early in tx course and before severe acidosis. NIPPV in this setting decreases likelihood of endotracheal intubation, decreases treatment failure, and decreases mortality. Important: Many RCTs including meta-analyses have found NIPPV benefit with severe COPD exac, not with mild.

Winck, 2006.
Systematic review and meta-analysis. CPAP and NIPPV in acute cardiogenic pulmonary edema is supported. BOTH decrease need for endotracheal intubation and BOTH decrease mortality as compared to standard medical therapy. No increased acute myocardial infarction risk. Important: Conflicting studies on mortality.

3CPO, 2008.
Randomized. 1069 pts to standard O2 therapy, CPAP, or NIPPV. No significant difference in 7day mortality or intubation rate btw standard O2 and NIPPV or CPAP and NIPPV. In pts with acute cardiogenic pulmonary edema, NIPPV induces more rapid improvement in respiratory distress AND metabolic disturbance than does standard O2 therapy. No effect on short term mortality (7 days).

2015 multicenter RCT put doubt on the Grade 2 recommendation to use NIV for acute hypoxemic resp failure in immunocompromised pts: no benefit of NIV vs O2

FLORALI, 2015.
Randomized n=310 to 3 groups in many ICUs in France. NIV, HFNC, Nasal mask. NIV had highest mortality at 25%. At 28d, less ppl needed HFNC than MV, but not statistically significant; Important: yet HFNC had more vent-free days & at 90days, HFNC group 2x as likely to survive.

It is estimated that for every 10 L/min of oxygen flow applied that 0.7 cm of H2O of PEEP is generated with the mouth closed.  If the mouth is open, this value falls to 0.35 cm H2O for each 10 L/min of flow

Nutrition

ASPEN Guidelines, 2009.
EN is the preferred route of feeding over parenteral nutrition (PN) for the critically ill patient who requires nutrition support therapy. (Grade: B)

EN feeding in ICU pts decreases infectious morbidity from pneumonia, CVC line infection, (and abdominal abscess. in abd trauma pts.)

REGANE, 2010.
Gastric residuals up to 500 cc can be tolerated

EPanIC, 2011.
Randomized. 2,312 ICU pts received early (<=48 hours) parental nutrition vs 2,328 pts received late (at or after Day 8). Insulin for normoglycemia. Late initiation of parental nutrition a/w faster recovery and fewer complications (less ICU infections/cholestasis, Less MV, less RRT). Greatest difference in pts whom EN was surgically contraindicated.

OMEGA, 2011.
Omega-3-fatty acid, gamma-linolenic acid, antioxidant supplements in ALI: outcome of probable harm. Pts had fewer vent-FREE, fewer ICU-FREE, and NONpulm organ failure-FREE days; increase in harm. No signficiant diff in mortality at 60 days.

EDEN, 2012.
Randomized. 1,000 pts with ARDS to enteral nutrition to either trophic feeding (10-20 mL/hr) x 6 days OR attempt to reach goal feeding rates as soon as possible. Similar outcomes: days alive and free of vent support, hospital mortality and organ-failure free days. Trophic rates of EN or targeting goal rate during first 6 days is acceptable. 85% had gastric tube.

Early PN Trial, 2013.
Randomized. Parenteral nutrition. Early PN reduced MV days. No 60 day mortality difference.

Reignier, 2013.
Randomized. GRV monitoring vs 250 mL trigger. Monitored for regurgitation and development of VAP. Outcome: No difference in VAP rates or outcomes.

CALORIES, 2014.
Randomized 2400 pts in the UK ICUs. to parenteral vs enteral nutrition starting 36h post-admission, continued x 5 days. NO significant difference in mortality @ 30 days, 90 days, or in infectious complications.

PERMIT, 2015.
Randomized 886 pts to permissive underfeeding (40-60% calculated caloric requirements) compared to standard enteral feeding (70-100%) x14days w/similar protein intake. Primary outcome 90 day mortality. No difference.

Taylor, 2016.
NJ tubes for feeding in selected pts with high aspiration risk. Otherwise, 3 RCTs showed no benefit of NJ over NG in severe acute pancreatitis. "Holding EN for GRV <500 cc in absence of intolerance signs should be avoided."

TARGET, 2018.
Randomized 3957 mechanically ventilated ICU pts to energy-dense (1.5 kcal/mL) vs routine (1.0 kcal/mL) EN. Dose was 1 mL/kg/hr. Primary outcome all-cause mortality at day 90. No difference.

Pancreatitis

Mier, 1997.
Early (<=48 to 72h of onset) vs late (>12 days after onset) necrosectomy in severe necrotizing pancreatitis. Randomized. 41 pts with severe necrotizing pancreatitis per Ranson's and dynamic pancreatography. Mortality rate did not reach significance. Late necrosectomy for selected cases became standard.

Renal Replacement Therapy

ATN, 2008.
Randomized. More intense RRT does not improve all-cause mortality at Day 60 in ICU ATN AKI pts, compared to less-intensive therapy. Intensive RRT: iHD or SLED, 6 tx per week, or CVVHDF. Intensive RRT associated with more hypotension and did not improve renal function nor did it improve nonrenal organ dysfunction.
: No benefit in more intensive therapy with ATN or RENAL Trials.

RENAL, 2009.
Randomized. 1,508 pts to CVVHDF 25 vs 40 mL/kg/hr. No difference in 90-day mortality.
: No benefit in more intensive therapy with ATN or RENAL Trials.

AKIKI, 2016.
Randomized. No mortality difference between early or delayed RRT in ICU (vented and/or vasopressor use) AKI pts. Early: within 6h of stage 3 AKI dx. Primary outcome overall survival at Day 60. Higher BSI in early group. ?powered.

BICAR-ICU, 2018.
Randomized 394 French ICU pts w/metabolic acidosis (pH <=7.20, PaCO2 <=45 serum HCO3 <=20) & elevated lactate 2+/SOFA 4+ to sodium bicarbonate infusion vs not. No difference in primary outcome of all-cause mortality at 28d. However, pts with AKI (AKIN 2-3) had 12% less composite of all-cause mortality.

IDEAL-ICU, 2018.
Randomized comparing early-vs. delyaed-intiiation stratgety for RRT . ICU pts w/septic shock and AKI. Excluded hyperkalemia >6.5, met acidosis <7.15, volume overload w/pulm edema refractory to diuretics. No difference in 90-day all-cause mortality. Early RRT initiation = within 12h of severe AKI diagnosis.

Sepsis & Septic Shock

Low Dose Dopamine, 2000.
Randomized. 328 ICU pts with early renal dysfunction assigned to low dose dopamine (2 micrograms/kg/min) vs placebo. Primary endpoint was peak serum Cr concentration during infusion. No clinically significant protection from renal dysfunction.

RIVERS TRIAL (EGDT), 2001.
Single center unblinded randomized. 263 pts in ED. Early goal-directed therapy (Aline, CVC with continuous ScVO2 monitoring, CVP 8 to 12, MAP >65 with PRN vasopressors, ScVO2 >70% with PRN blood txf/dobutamine, UOP >0.5 cc/kg/hr ), within the first 6 hours in severe sepsis and septic shock decreases mortality. 16% absolute RR in mortality: NNT 6. SSC 2016: give >= 30 cc/kg IV crystalloid in 1st 3h. Add'l fluid guidance per cardiopulmonary/hemodynamic frequent reassessment. Target MAP 65; norepi is first line vasopressor; vasopressin up to 0.03 units/min can be used to lower norepi OR to raise MAP; epi can also be used to raise MAP; normalize lactate; IV antimicrobial within 1 hour of recognition.

PROWESS, 2001.
Activated protein C in severe sepsis seemed to improve survival but was later not confirmed in PROWESS_SHOCK.

ANNANE, 2002.
Among patients with septic shock and relative adrenal insufficiency, administration of corticosteroids reduces 28-day mortality, although this finding was not confirmed in the follow-up CORTICUS trial (probably not as sick as Annane's). Survival benefit per Annane. Hydrocortisone and fludrocortisone in pts with septic shock and relative adrenal insufficiency (cortisol rise <9 after ACTH stim). Later studies suggested infection-related harm a/w routine CCS.

CORTICUS, 2008.
Hydrocortisone in septic shock. Randomized. 499 septic shock ICU pts. Hydrocortisone vs placebo. All received ACTH stim test were classifed as responders (>9 mcg/dL rise in cortisol) or a NONresponder (<=9 mcg/dL). Results: In contrast to Annane Trial, hydrocortisone does NOT improve survival in septic shock, regardless of ACTH response. No survival benefit. Hydrocortisone reversed shock more rapidly in all subgroups. SSC 2016: if adequate fluid resusc and vasopressor therapy can restore hemodynamics, suggest against IV hydrocortisone; otherwise, 200mg IV qday.

VASST, 2008.
Randomized. 779 pts with septic shock to low dose 5-15 mcg of norepinephrine for MAP 65-75 mmHg OR vasopressin 0.03 to 0.05 units for MAP 65-75 mmHg. All cause mortality was primary outcome at 28days postinfusion. Vasopressin in septic shock. Low dose vasopressin did not reduce all-cause 28 day mortality when added to norepinephrine for patients in septic shock. Did not address vasopressin as monotherapy, Less than 20% of screened patients were included!

SOAP II, 2010.
Dopamine vs norepinephrine in shock. Randomized. 1,679 pts with mostly septic shock to norepi vs dopamine. Primary endpoint all cause mortality at 28 days. No statistical difference. However, higher significant rate of arrhythmias with dopamine. SSC 2016: norepi as first line.
A subgroup prospective of 280 pts (cardiogenic shock pts) found improved mortality with norepi over dopamine at 28d.

PROWESS-SHOCK, 2012.
Activated protein C (Drotrecogin alfa) in septic shock. Randomized. Phase III study failed to show improvement in outcome of pts with severe sepsis. Randomized. 1,697 pts with septic shock comparing APC vs placebo. No difference in mortality at 28- and 90- days.

CRISTAL, 2013.
Colloids vs crystalloids in shock. Randomized. 2,857 ICU pts with hypovolemic shock and mostly concurrent sepsis to volume resuscitation with colloids or crystalloids. Both had isotonic for maintenance. No difference in all cause mortality at 28 days; colloids had reduced all-cause mortality at 90 days. NNT 29.

SEPSISPAM, 2014.
For patients with septic shock, a goal MAP of 80-85 mmHg does not reduce all-cause mortality at 28 days when compared to a goal of 65-70 mmHg. The higher MAP goal was associated with reduction in rates of renal dysfunction for patients with a history of chronic hypertension.

TRISS, 2014.
Patients with septic shock who underwent transfusion at a Hgb threshold of 7 g/dL had similar mortality at 90 days but used 50% fewer units of blood compared with those who underwent transfusion at a Hgb threshold of 9 g/dL.

ProCESS, 2014.
EGDT-like protocol: SBP and shock index goals. No difference in mortality between interventions. There was no difference in all-cause in-hospital mortality at 60 days in early sepsis patients when comparing therapeutic strategy driven by EGDT, a novel protocolized care, or usual care. Criticisms: Usual care was essentially EGDT; underpowered; IV fluids were not administered during intervention, but early.

ARISE, 2014.
EGDT vs usual care in sepsis. Early goal-directed therapy did not reduce all-cause mortality at 90 days in severe sepsis and septic shock ER pts. Criticisms: Usual care was essentially EGDT; underpowered

ProMISe, 2015.
There was no difference in mortality at 90 days in early septic shock patients when comparing EGDT to strict therapy including vasopressors and IV fluids. Criticisms: usual care was essentially EGDT; outcome evaluator bias

HYPRESS, 2016.
Hydrocortisone vs placebo in severe sepsis. Randomized. 380 pts with severe sepsis with shock to IV hydrocortisone or placebo. Primary outcome was short-term prevalence of septic shock. At 14 days, hydrocortisone did not have benefit on development of septic shock vs placebo. No difference in 28 day or 90 day mortality OR time to development of shock. Final thoughts: CCS should be for refractory septic shock.

SEPSIS-3, 2016.
Sepsis should be defined as life-threatening organ dysfunction caused by a dysregulated host response to infection. For clinical operationalization, organ dysfunction can be represented by an increase in the Sequential [Sepsis-related] Organ Failure Assessment (SOFA) score of 2 points or more, which is associated with an in-hospital mortality greater than 10%.
Septic shock should be defined as a subset of sepsis in which particularly profound circulatory, cellular, and metabolic abnormalities are associated with a greater risk of mortality than with sepsis alone.
Patients with septic shock can be clinically identified by a vasopressor requirement to maintain a mean arterial pressure of 65 mm Hg or greater and serum lactate level greater than 2 mmol/L (>18 mg/dL) in the absence of hypovolemia.
NON-ICU pts - qSOFA is positive when 2 or more are present: AMS, sBP 100 or less, RR 22 or more.
ATHOS-3, 2017.
Angiotensin II vs placebo in vasodilatory shock. Randomized. 321 severe vasodilatory shock patients. Either angiotensin II or placebo. Primary outcome of 3-hour MAP response (>= 10 mmHg increase OR MAP > 75 mmHg without increase in existing vasopressors). Significant reduction in vasopressor use in pts receiving angiotensin II. SOFA scores improved at 48h. Trend toward mortality with angiotensin II.

Hydrocortisone, Vitamin C, Thiamine in SEPSIS, 2017.
Marik's single center postiitve outcome study but with low n of 47. Compared hydrocortisone, vitamin C, thiamine in pts w/severe sepsis or septic shock (procal >=2) to historic controls. Primary outcome = hospital survival (8.5% vs 40%). Infusion protocol: IV vitamin C 1.5 grams q6h x 4 days (or ICU discharge), hydrocortisone 50 mg q6h x 7 days (or ICU discharge, followed by 3d taper), IV thiamine 200 mg q12h x 4d (or ICU discharge). Highly criticized study - risk of selection bias.

ADRENAL, 2018.
3800 pts randomized. Septic shock on vasopressors and mechanical ventilation to hydrocortisone 200 mg/day continuous infusion vs placebo. NO difference at 90 days (Primary outcome death). Faster time to revse shock, shorter time to ICU discharge, shorter time to extubation, less blood transfusions.

Stress Hyperglycemia

LEUVEN SURGICAL TRIAL, 2001.
Randomized. 1,548 surgical ICU pts to intensive glycemic (goal 80 to 110 mg/dL) vs conventional glycemic control (goal 180 to 200 mg/dL). Primary outcome: ICU mortality. Resulted in lower ICU mortality in intensive glycemic control among pts in ICU for 5 days or more. Leuven Medical and NICE-SUGAR followed. Guidelines do not support intensive glycemic control. Recommended non-sepsis guidelines: range of 140 to 180 mg/dL while sepsis guidelines advocate less than 180 mg/dL.

LEUVEN MEDICAL TRIAL, 2006.
Unblinded randomized. 1,200 medical ICU pts to intensive glycemic control (goal 80 to 110 mg/dL) vs conventional glycemic control (180 to 200 mg/dL). Contrasted with surg trial 5 years prior: no significant difference in primary outcome of total death. Higher mortality in ICU arm of pts with >3day ICU LOSs. Due to disparate results, NICE-SUGAR was funded. SSC 2016 recommends <=180 mg/dL.

NICE-SUGAR, 2009.
Randomized. 6,104 med-surg ICU pts to intensive glycemic control (81 to 108 mg/dL) vs conventional glycemic control (target <=180 mg/dL). Intensive control arm had higher 90 day mortality and more hypoglycemic events. No diff in med vs surg pts, in LOS, in vent days, or in need for RRT. Target <110 mg/dL is NO longer recommended. SSC 2016: suggest target of <=180 mg/dL.

COIITTSS, 2010.
Randomized. 509 septic shock ICU pts. SOFA >=8 on hydrocortisone. 4 groups: 1. Continuous IV insulin drip with hydrocortisone alone, 2. Continuous IV insulin drip with hydrocortisone plus fludro-, 3. Conventional insulin tx with hydrocoritsone alone, 4. Conventional insulin tx with hydro- plus fludro-. Intensive insulin therapy did not improve in-hospital mortality.

Stress Related Mucosal Disease (SRMD)

COOK STUDY, 1994.
A simple decision rule predicts the risk of bleeding and allows more selective use of stress ulcer prophylaxis. Greatest RF for GIB in ICU pt: coagulopathy and MV. Clinicians were asked to hold previous usual care practice of administering broad GI prophy; clinically significant GIB only 1.5% of pts! Those at highest risk had coagulopathy and MV >48hrs from resp failure. Excluded: head trauma, high % burns, organ txp recipients, or known PUD or gastritis. Important: PPIs are a/w CDAD, Campylobacter, Salmonella, Shigella, Listeria, pneumonia, fracture, malabsorption syndromes, and medication interactions. No benefit from GI prophy when added to enteric feeds.

IHI Ventilator Bundle (2011)

Peptic ulcer disease prophylaxis
Elevation of the head of the bed
Daily sedation vacations and assessment for extubation readiness
DVT prophylaxis
Daily oral care with chlorhexidine

POP-UP, 2016.
Randomized 216 intubated med-surg ICU pts who were eligible for enteric feeds and low risk for stress ulcer development. to daily IV PPI or placebo. Daily IV PPI did not have impact on significant GIB (SRMB), CDAD, VAP. Routine use of PPIs should be reconsidered. SSC 2016: PPI prophy for sepsis/septic shock with RF for GIB.

Stroke

Allen, 1983.
Nimodipine in SAH for cerebral artery spasm. Randomized. 125 SAH pts 96h within their event. Pts were neurologically normal with ICH aneurysms. Hypothesis to decide if CCB nimodipine prevents or reduces severity of ischemic neurologic deficits from arterial spasm. Nimodipine should be given to neurologically intact SAH pts to reudce severe neuro deficit due to cerebral artery spasm.

NINDS, 1995.
tPA in ischemic stroke. In pts with ischemic CVA within 3 hours, tPA improved NIHSS scores and 3 month functional outcome. No mortality benefit.

SPARCL, 2006.
80 mg of atorvastatin a day reduced overall stroke incidence and of cardiovascular events for more than 5 years. Despite small increase in incidence of hemorrhagic stroke.

ECASS-3, 2008.
Intravenous alteplase improved neurologic outcomes within 4.5 hours of stroke onset among patients with acute ischemic stroke. Despite an increased incidence of intracranial hemorrhage, there was no mortality benefit. Primary outcome: improvement by >4 points on NIHSS or complete resolution at 24hours.

AHA/ASA acute Ischemic CVA, 2013 Guidelines:
- IV tPA up to 3h after onset of ischemic stroke Class I, Level A
- Eligible pts should receive IV tPA ideally within 60 	minutes of hospital arrival, Target time: door to needle 	Class I, Level A
- IV tPA for selected pts btw 3 to 4.5 hours after onset of ischemic stroke Class I, Level B
- 0.9 mg/kg (max 90 mg total dose): 10% (0.09 mg/kg) IV bolus over 1 min, remaining 90% (0.81 mg/kg) as IV infusion over 60 min LEVEL A

INTERACT, 2008.
Early intensive BP-lowering treatment is clinically feasible, well tolerated, and seems to reduce haematoma growth in ICH
STICH, 2008.
Randomized. Emergent surgical hematoma via crani w/in 72h of ICH onset does not improve outcome vs initial med mgmt

INTERACT 2, 2013.
BP mgmt in randomized ICH pts monitored within 6hours of onset of ICH. Lowering sBP to 140 within 1 hour for next 24h showed proof of concept for early BP lowering vs 180 mmHg.

ATACH
4 arm protocol of dose escalation of IV nicardipine for BP control in ICH pts.

ATACH-2
Aggressive sBP treatment to <120 mmHg did not improve neurologic outcomes but a/w 2x renal complications.

Ventilator-Associated Pneumonia

Heyland, 2008.
Multicenter RCT. 739 pts w/suspected VAP randomized to receive meropenem with or without ciprofloxacin. No difference in primary endpoint of mortality. Subgroup of Pseudomonas, Acinetobacter, or MDR Gram neg bacilli, combo tx more likely to provide effective coverage.

PneumA, 2003.
Double blinded randomized. 401 ICU pts with VAP. Intention to treat trial. Equivalent 28-day mortality mortality with VAP at 8 days vs 15 days of abx. However: lactose, nonfermenting GNRs (Pseudomonas aeruginosa) should be treated for 15 days.

Trauma / Surgical ICU

CRASH-2, 2010.
Randomized. 20, 211 trauma adult pts at risk for significant hemorrhage. Tranexamic acid within 8h of traumatic injury vs placebo.. TXA a/w reduced 28-day mortality including bleeding-related mortality when TXA given within 3 hours of injury. No difference in requirement for blood txfs or vascular occlusive events. TXA is antifibrinolytic derivative of lysine.

PROPPR, 2015.
Randomized trial of severe trauma adult pts. 24h and 30-d mortality was studied with NO difference. 1:1:1 vs 1:1:2 (plasma, plts, RBCs). HOWEVER, 1:1:1 arm had improved survival with less exsanguination at 3hours.

VTE

PIOPED, 1990.
Low-probability and high-probability V/Q scans aided in ruling-out or diagnosing of PE when concordant with clinical assessment. An intermediate-probability V/Q scan is not of help in establishing diagnosis of PE. 30% had normal EKG.

MOPPETT, 2013.
Low-dose tPA plus anticoagulation reduced incidence of PH or outcome of recurrent PE vs anticoag alone.
Nonblinded single center randomized. 121 pts with moderate PE to low dose tPA or control. All received LMWH or UFH including warfarin. Submassive defined: HDS with RV dysfunction/strain.

PEITHO, 2014.
Randomized. 1,005 pts with submassive PE (HDS with RV strain and elev troponins) to UFH with fibrinolytic tenecteplase or placebo. At 7 days, tenecteplase a/w significant decrease in primary endpoint of all-cause mortality or hemodynamic DEcompensation. tPA group: major extracranial bleeding at 7 days.

PEERLESS, 2025. (Circulation)
Large-bore mechanical thrombectomy (LBMT; FlowTriever) vs catheter-directed thrombolysis (CDT) for
intermediate-risk pulmonary embolism showed superiority on a hierarchical composite “win ratio” outcome,
driven by less clinical deterioration/bailout and markedly less postprocedural ICU use, with no difference in
mortality, intracranial hemorrhage, or major bleeding.

Prospective multicenter randomized controlled trial. 550 intermediate-risk PE patients with RV dilatation and
additional clinical risk factors randomized 1:1 to LBMT vs CDT.

Primary end point: hierarchical win-ratio composite assessed at hospital discharge or 7 days after procedure:
(1) all-cause mortality, (2) intracranial hemorrhage, (3) major bleeding, (4) clinical deterioration and/or
escalation to bailout therapy, (5) postprocedural ICU admission and ICU length of stay.

Results:

• Primary end point favored LBMT: win ratio 5.01 (95% CI 3.68–6.97); P<0.001
• Clinical deterioration and/or bailout: 1.8% vs 5.4%; P=0.04
• Postprocedural ICU use: admissions 41.6% vs 98.6%; ICU stays >24 h 19.3% vs 64.5% (overall ICU use P<0.001)
• No significant differences in mortality, intracranial hemorrhage, or major bleeding
• 30-day mortality: 0.4% vs 0.8%; P=0.62
• Secondary win ratio (first 4 components only): 1.34 (95% CI 0.78–2.35); P=0.30

Reference: Jaber WA, et al. Circulation. 2025 Feb 4;151(5):260–273. DOI: 10.1161/CIRCULATIONAHA.124.072364.
NCT05111613.