Trials and guidelines on sedation in the ICU

Because the college examiners called on the candidates to "list three trials investigating the use of dexmedetomidine infusions" in Question 21 from the first paper of 2023, the irresistible urge to create a directory of all possible sedation-related trials and guidelines had felt normal and appropriate. The following selection of papers and statements puts the "AS" in "FASTHUG". Where possible, the original paper is linked as the name of the trial, and the Bottom Line or equivalent breakdown is linked as the year of publication.

Sedation trials in general

Kress - 2000 - what if sedation, but with "holidays"? n=150, single centre in the US. Duration of ventilation was markedly decreased (4.9 vs 7.3 days), also ICU LOS (6.4 vs 9.9 days). Control group were fairly oversedated, however (RASS -3 to -4). 

ABC - 2008 - what if we wake as well as wean? Protocolised simultaneous sedation and ventilation deescalation. n=338, in the US. Unsurprisingly, markedly better ventilator-free days (14.7 vs 11.6), shorter ICU LOS (9.1 vs 12.9) and even better mortality (28% vs 35%) but also higher self-extubation rate (10% vs 4%). Perhaps all those patients just needed the tube out anyway?

Strøm - 2010 - what if sedation, but only by boluses? n=140, single centre in Denmark. Shorter ICU LOS (by 9.7 days), fewer ventilated days (by 4.2 days) and no difference in self-extubation. None of the patients needed physical restraints because all had access to 1:1 nursing and "patient comforters". Summary: we don't need sedation, we just need more staff.

SLEAP - 2012 - what if sedation with "holidays", like Kress, except the control group is not totally comatose (RASS -3 to 0)? n=430 in the US and Canada. No difference in time to extubation or any other outcomes. Summary: there is no need to take a holiday from sensible light sedation.

And then, somewhere around 2012, it seems everyone stopped doing sedation trials on anything other than dexmedetomidine, which has meant that a whole separate section has become necessary.

ROSE - 2019 - paralysis and heavy sedation in ARDS, because ACURASYS? n=1006, in the US. No difference in mortality (42.5% vs. 42.8%), nor any of the secondary outcomes. Summary: we don't need that much sedation.

NONSEDA - 2020 - what if no sedation, like Strøm but multi-centre? n=736 in Scandinavian ICUs. No difference in mortality, ventilator-free days, or ICU stay. Predictably, self-extubation rate more than doubled (8.9% vs 4.0%), and 27% of the non-sedated group crossed over to sedation within the first 24 hours. Summary: we still need some sedation. 

Dexmedetomidine trials

MENDS - 2007- dex vs. lorazepam, n=106 in the US. Preposterous mortality difference (17% vs 27%) in favour of dex.

DEX-COM - 2009 - dex vs. morphine in the wealthy Eastern suburbs of Sydney; n = 306 CABG patients. Delirium incidence was the same, but duration was shorter with dex (2 vs 5 days).

MIDEX-PRODEX - 2012 - dex vs prop or midaz (actually two trials). n=500 and 501, in Europe. Shorter duration of ventilation vs. midazolam, but not vs. propofol. Otherise not inferior to the other agents.

DahLIA - 2016 - dex vs. placebo, looking at delirium and ventilator-free time over 7 days. n=71 in Australia and NZ. The dex group spent about 17 hours less on the ventilator, and their delirium resolved faster (23 vs 40 hours).

Xian Su - 2016 - dex, but in elderly post-op patients? n=700 elective surgical patients in China, only over-65s. Greatly reduced incidence of delirium (9% vs 23%) and ICU stay (20.9 vs 21.5 hrs) but basically all of the patients only stayed in the ICU overnight, so.. is this really the population that needs sophisticated delirium prevention techniques?

DESIRE - 2017 - dex plus whatever other agents, vs. sedation without dex, looking at 28-day mortality. n=201 in Japan. No difference in mortality or ventilator-free days.

Skrobik et al - 2018 - dex at low dose (02-0.7 mcg/kg/hr), overnight, to improve sleep. n=100 in Canada and USA. Risk of delirium greatly reduced (54% vs 80%), as well as time to extubation and length of ICU stay, 

SPICE-III - 2019 - dex as the sole agent, vs. all others, looking at 90-day mortality. n=4000, all over the world. No difference in mortality (31.5% vs 31.3%); and they weirdly expected a mortality difference of 4.4%. There was one extra ventilator-free day with dex.

MENDS2 - 2021 - dex vs propofol. n=438, in the US. None of the patients received much of either drug, and unsurprisingly there were no differences in any of the outcomes.

Melatonin and melatonin-likes 

Nishikimi - 2018 - ramelteon for all! n=92, in Japan. Reduced length of ICU stay (4.56 vs 5.86 days), halved rates of delirium (24.4% vs 46.5%), but small and underpowered.

Pro-MEDIC - 2022 - melatonin for all! n=847 in Australia. No difference in the prevalence of delirium, nor any other endpoint (sleep quality remained poor in more than 40% of patients and 20% still required antipsychotics).

Antipsychotics for delirium

HOPE-ICU - 2013 - what if haloperidol for all? n=142, in the UK. It did not prevent delirium (no difference in number of delirium-free days). 

Gaudry - 2017 - loxapine, a clozapine-like,  for agitation in ventilated patients; n=102 in France. It did not hasten weaning from ventilation; they just replaced one sedative with another. 

REDUCE - 2018 - what if haloperidol for all, again? n=1789 in Netherlands. No difference in mortality or incidence of delirium.

MIND-USA - 2018 - haloperidol, ziprazidone, or placebo? n=566 delirious patients in the US. No difference in any of the outcome measures. Placebo was equally effective.

AID-ICU - 2022 - what if haloperidol, but ony for established delirium? n=963, in Europe. No difference on any of the primary outcomes, but weirdly an improvement in 90-day mortality with haloperidol (36.3.% vs 43.3%)

Nonpharmacological measures for delirium

Simons - 2016 - daylight lighting to assert circadian rhythm? n=711, single Dutch centre. No difference in any of the outcomes.

Sedation guidelines

Anagesia/sedation/delirium prevention in ICU generally:

• 2021 KSCCM guidelines
• 2018 SCCM guidelines 

Anagesia/sedation/delirium prevention in specific settings:

• Hughes (2020) - for perioperative patients
• Temesgen et al (2021) - in the resource limited setting
• Donato (2021)- for COVID-19
• 2022 SCCM guidelines -for paediatrics
• Prabhakar et al (2021) - for the neurocritical care patient
• Carpenter (2021) - in the emergency department 

Monitoring

• Rasulo et al (2022) - EEG monitoring for sedation

Definitions and nomenclaturs

• Statement of Ten Societies on the nomenclature of delirium and acute encephalopathy (2020)