Add sedation ONLY after adequate analgesia. Target lightest effective sedation -deep sedation kills ICU patients. PADIS, 2018
Have you treated pain first? Before reaching for a sedative, ensure fentanyl (or equivalent) is optimized. If CPOT โฅ 3 โ the patient needs more analgesia, not sedation. Go to Analgesia Guide.
Sedation Agent Comparison
Agent
Dose
Onset
Best For
Avoid When
Key Side Effects
Propofol (Diprivan) (Diprivan) 1ST LINE
5โ50 mcg/kg/min IV
30โ60 sec
Most ventilated ICU patients. Rapid on/off -ideal for daily SAT, neuro checks.
Hypotensive patient needing sedation โ Ketamine (supports BP). Or low-dose propofol with vasopressor support.
Alcohol/benzo withdrawal โ Benzodiazepines (CIWA-guided). Phenobarbital for severe/refractory. Add dex as adjunct.
Severe ARDS + NMB โ Deep sedation (RASS โ3 to โ4). Propofol or midazolam + fentanyl. Ensure adequate analgesia -paralyzed patients cannot communicate pain.
Procedural sedation โ Propofol 0.5โ1 mg/kg bolus ยฑ fentanyl. Etomidate for cardioversion. Ketamine for hemodynamically unstable procedures.
Key Pearls
Light sedation saves lives. RASS โ1 to 0 reduces ICU mortality, ventilator days, and long-term cognitive impairment vs deep sedation. There is almost never a reason for RASS โ4/โ5 without NMB.
Benzos cause delirium. Midazolam and lorazepam are independently associated with โ delirium and โ mortality compared to propofol or dexmedetomidine MENDS, 2007SEDCOM, 2009. Never use as first-line ICU sedation.
Dexmedetomidine is the only sedative without respiratory depression. Can sedate non-intubated patients (post-extubation agitation, NIV patients, procedural sedation).
Daily SAT + SBT. Every day ask: can we turn off sedation? Can we do a breathing trial? The ABCDEF bundle reduces mortality and delirium. Do it every single day.
Propofol Infusion Syndrome (PRIS) - Why It Happens
Pathophysiology - mitochondrial energy failure
Big picture: propofol is a mitochondrial toxin. Mitochondrial dysfunction means cells can't make ATP. The tissues that burn the most ATP at baseline are the ones that fail first, and the syndrome's clinical features map directly onto them:
Heart (high-throughput ATP for contraction + conduction) โ cardiac failure, Brugada-pattern ECG, refractory bradycardia, asystole.
Renal tubules (ATP-hungry Naโบ/Kโบ ATPase along the nephron) โ acute kidney injury, both from direct tubular ATP failure and from pigment nephropathy as myoglobin pours in.
System-wide: when aerobic metabolism fails everywhere, lactate climbs and a new anion-gap metabolic acidosis appears. That's the lab signal that the rest of the syndrome is already in motion.
Two-pronged mitochondrial hit: propofol blocks how cells use fat for fuel (fatty-acid ฮฒ-oxidation) and disrupts the mitochondrial energy chain itself, so ATP production collapses.
The lipid vehicle adds fuel mitochondria can't burn: the 10% intralipid emulsion delivers a continuous fat load that the impaired mitochondria can't process. Triglycerides climb, plasma turns lipemic, free fatty acids accumulate.
Direct cardiac toxicity is separate from the energy failure: propofol independently blocks cardiac Naโบ channels, L-type Caยฒโบ channels, and ฮฒ-adrenergic receptors. This is why the bradycardia is refractory to atropine and catecholamines, and why the ECG develops a Brugada pattern.
What makes PRIS more likely (and why)
High dose (> 4 mg/kg/hr) and long duration (> 48 h) - the two strongest predictors. More drug for longer = more mitochondrial toxin delivered. These are also the defining criteria for "high risk" in nearly every PRIS case series.
Concurrent catecholamine infusions (norepinephrine, epinephrine, phenylephrine) - adrenergic stimulation drives cellular ATP demand sky-high at exactly the moment damaged mitochondria can't supply it. Worse, propofol is blocking the ฮฒ-receptor, so the pressor dose keeps climbing to no effect, a vicious cycle that masks early shock.
High-dose glucocorticoids (stress-dose hydrocortisone, dexamethasone for cerebral edema, etc.) - steroids independently impair mitochondrial function and increase catabolic load. Almost always co-administered with catecholamines in the patients who get PRIS, so risk stacks.
Sepsis - septic cells already have cytopathic hypoxia (mitochondrial dysfunction with normal oxygen delivery). Propofol pushes a mitochondrial system that's already failing right over the edge.
Severe traumatic brain injury - the classic at-risk population. These patients need deep, prolonged propofol for ICP control AND are usually on catecholamines AND often steroids. PRIS was first described in this group; treat it as the prototype red-flag scenario.
Starvation, low-carb, or ketogenic state - when cells can't use glucose, they fall back on fatty-acid ฮฒ-oxidation for energy. But that's exactly the pathway propofol blocks. The backup generator is offline.
Young age - pediatric mitochondria are more vulnerable to fatty-acid overload, and propofol clearance differs in kids. PRIS was first recognized in critically ill children; still happens in adults but pediatric ICU patients are highest risk per mg of drug delivered.
Inborn fatty-acid oxidation defects (MCAD, VLCAD, carnitine deficiency) - the ฮฒ-oxidation pathway is already broken at baseline; even modest propofol doses can trigger the syndrome. Rare but a classic board point.
Clinical presentation - the diagnostic constellation
The lab signature: a new high anion-gap metabolic acidosis with elevated lactate in a patient on a propofol drip - without shock, sepsis, or another perfusion failure to explain it - is PRIS until proven otherwise. This is Type B2 (mitochondrial) lactic acidosis, not Type A (hypoperfusion). It does not respond to fluid or pressors, because perfusion is not the problem.
High anion-gap lactic acidosis - cells can't oxidize pyruvate, so it shunts to lactate. AG widens, bicarb drops. Often the first lab abnormality, before CK or TGs move.
Hyperkalemia - Kโบ pours out of dying muscle cells. Stacks arrhythmia risk on top of the existing cardiac toxicity.
Hypertriglyceridemia + lipemic plasma - the lipid emulsion accumulates because mitochondria can't process it. TGs often > 500-1000 mg/dL; the lab will note grossly lipemic samples.
Acute kidney injury - dual hit: myoglobin causes pigment nephropathy AND tubular ATP failure stops Naโบ/Kโบ ATPase. Often oliguric.
Cardiac dysfunction - new bradycardia refractory to atropine and pressors, falling BP, reduced contractility on echo. Brugada-pattern ECG develops on the way to collapse (see next).
PRIS is a clinical diagnosis - you don't need every feature. Any 2-3 of the above in a patient on propofol > 24-48 h is enough to stop the drug and start workup.
Cardiac / conduction signature - Brugada-pattern ECG is the most common finding
Type 1 Brugada-like ECG (coved ST elevation > 2 mm in V1-V3 with T-wave inversion) is the most common and most characteristic conduction finding in PRIS, and frequently appears before circulatory collapse. Mechanism: propofol's dose-dependent cardiac sodium-channel blockade unmasks a Brugada phenotype. If you see this on a propofol drip, stop propofol now - do not wait for lactate or CK to confirm.
Early: Brugada-pattern ST elevation V1-V3, new RBBB, PR/QRS prolongation.
Mid: refractory bradycardia (atropine- and catecholamine-resistant), QT prolongation.
Terminal: ventricular tachyarrhythmias, asystole.
Monitoring on a propofol drip
Send these on anyone running propofol > 4 mg/kg/hr OR > 48 h. Tighter cadence if other risk factors (catecholamines, steroids, sepsis, TBI) are stacked on top.
BMP daily - Kโบ, anion gap, bicarb, Cr. A widening AG or new acidosis is the earliest red flag.
Lactate daily - a rising lactate without obvious shock is the key warning sign that mitochondria are failing.
Triglycerides q48h - TGs > 400-500 mg/dL signal the lipid emulsion is overwhelming damaged mitochondria.
CK q24-48h - rising CK = rhabdo starting. Pair with urine myoglobin if elevated.
ECG - baseline at drip start; recheck with any new bradycardia, conduction abnormality, or hemodynamic change. Look specifically for Brugada-pattern ST elevation in V1-V3.
Treatment - stop the drug, support the organs
STOP propofol immediately the moment PRIS is suspected. This is the only intervention that reverses the mitochondrial toxicity - everything else is supportive. Do not wait for confirmatory labs. Switch sedation to dexmedetomidine, midazolam, or ketamine.
Remove the cofactors - wean catecholamines if hemodynamics allow, taper steroids, address starvation/low-carb state, treat sepsis aggressively. These amplified the toxicity in the first place.
Hemodynamic support - expect catecholamine-refractory bradycardia and hypotension (propofol has blocked the ฮฒ-receptor and Naโบ/Caยฒโบ channels). Use transvenous pacing for symptomatic bradycardia. Isoproterenol or glucagon can be tried for refractory ฮฒ-receptor failure (case-report level).
Bicarbonate for severe acidosis (pH < 7.1) as a bridge to renal replacement - the underlying problem won't resolve until propofol clears.
CRRT / hemodialysis for refractory acidosis, hyperkalemia, AKI, or severe hypertriglyceridemia - also clears propofol and lactate. Strongest case-series support among rescue interventions.
V-A ECMO for refractory cardiogenic shock - bridges the patient while the mitochondrial toxicity resolves. Case reports support survival with timely cannulation.
Mortality 30-50% once the full syndrome develops, and survival correlates directly with how fast propofol was stopped. When in doubt, switch sedation - the cost of being wrong about PRIS is much lower than the cost of being right and slow.
Special Clinical Scenarios
Alcohol / Benzodiazepine Withdrawal
Benzodiazepines are the treatment -NOT propofol or dexmedetomidine alone.
CIWA-Ar score q4โ8h -score > 8 โ treat
Lorazepam 1โ4 mg IV q1h PRN (symptom-triggered) OR diazepam 5โ20 mg PO/IV PRN
Severe/refractory (CIWA > 20): phenobarbital 130โ260 mg IV q15โ30 min until controlled
Add dexmedetomidine as adjunct for autonomic symptoms (tachycardia, HTN)
Thiamine 100 mg IV before any glucose (prevents Wernicke's)
Delirium Management
Non-pharmacologic first: reorientation, day/night cycle, early mobility, hearing aids/glasses
Dexmedetomidine -reduces delirium duration vs benzos MENDS, 2007
Midazolam 1โ2 mg IV + fentanyl 25โ50 mcg (or local only)
Cardioversion
Propofol 0.5โ1 mg/kg IV or etomidate 0.2 mg/kg IV
Paracentesis / thoracentesis
Topical lidocaine only; add midazolam if anxious
Pimp Questions
What is propofol infusion syndrome and how do you recognize it?
Rare but fatal complication of high-dose (> 4 mg/kg/hr) or prolonged propofol. Caused by impaired mitochondrial fatty acid oxidation. Features: new metabolic acidosis + elevated lactate, rhabdomyolysis, hyperkalemia, lipemic plasma, Brugada-like ECG, cardiac failure. Treatment: stop propofol immediately, switch agent, supportive care.
Why are benzodiazepines associated with worse ICU outcomes?
MENDS, 2007SEDCOM, 2009: more delirium, longer ventilator time, increased ICU stay vs propofol or dexmedetomidine. Accumulate in renal/hepatic failure. Paradoxical agitation in elderly. Reserve for withdrawal and status epilepticus only.
What makes dexmedetomidine unique compared to other sedatives?
Only ICU sedative without respiratory depression. Mechanism: ฮฑโ agonism (locus coeruleus) โ sleep-like state, patients arouse easily. Can sedate non-intubated patients, continue during vent weaning, and doesn't interfere with respiratory drive during SBT.