Sedation and analgesia IV: Typical sedatives

The third in a multipart series on the basics of ICU analgesia and sedation. Start with Part I: PrinciplesPart II: Opioids, and Part III: Non-opioids.

Let’s shift gears now from analgesia to sedation. (If the distinction is blurry to you, rewind to Part I.)


In many ways, propofol is the classic IV sedative, a GABA drug appropriate for continuous sedation or bolused during procedures. It has the following characteristics:

  • Fast-on, fast-off. Propofol takes effect within seconds. With short-term use, its effects wear off within a few minutes, making it very handy in patients needing frequent neuro exams. However, propofol can still fall into the trap we learned about in Part II: after prolonged infusion, it distributes into extravascular compartments and its effects no longer rapidly diminish after stopping the drip. Bear this in mind when using it for days and days.
  • A reliable, progressive dose-response reduction in consciousness… Propofol is a potent sedative/hypnotic and amnestic. With a little propofol, patients become sleepy. A little more, they become hard to arouse. A little more, their breathing slows. A little more, they become comatose. With enough propofol, you can make an epileptic ferret keep quiet, and with its easy titratability, you can usually reach your goals in a steady, controlled fashion.
  • … and in everything else… As a neurointensivist once said to me, “Almost anything will go away with enough propofol.” It easily suppresses the respiratory drive (and so cannot be used in non-intubated patients). Its slowing effects on cerebral activity makes it an effective agent for status epilepticus. It can even reduce intracranial pressure in herniating patients. The only thing it doesn’t take away? Pain. Propofol has no analgesic effects, so many patients will need an additional analgesic.
  • … including blood pressure. Cardiac output is usually one of propofol’s victims. Propofol will consistently, reliably cause vasodilation, along with a (lesser) degree of negative inotropy. It will drop the blood pressure so predictably that it can be used as a rescue drug for hypertensive emergencies. Fortunately, like its other effects, this usually follows a steady dose-response curve and can be planned for. The possible exception is the under-resuscitated patient in hypovolemic shock, who may be compensating purely by dint of increased vascular tone; in these patients, propofol can cause unexpectedly precipitous deterioration by taking away their squeeze. Be cautious pushing this stuff in very sick patients.

Propofol also has a few idiosyncrasies. Most importantly:

  • Propofol infusion syndrome (PRIS). This is a rare, unique complication with high mortality, usually seen in high-dose, prolonged propofol infusions. (It’s thought to be caused by some kind of mitochondrial defect of lipid metabolism.) It’s characterized by the otherwise-unexplained development of rhabdomyolysis, lactic acidosis, and bradycardia, often with sequelae such as renal failure and death. The only treatment is supportive—and stopping the infusion. Fear of this has led some centers to monitor screening labs (daily CKs and lactates) and limit the dose range, but this is a rare phenomenon, and the best safeguard is probably just to have it on your radar.
  • A lipid-based carrier. Propofol is characterized by its milk-white diluent, which is essentially fat. The lipid content is so high that propofol drips need to be accounted for in dietary calculations; each milliliter contains about 1 kilocalorie of energy. Other consequences of this are the need to change out infusion sets on a regular basis (to curtail bacterial growth), and the occasional incidence of hypertriglyceridemia. Some centers check daily triglyceride levels; it is a bit unclear whether high triglycerides are associated with PRIS, but as in anyone, at some point they can cause other complications like pancreatitis.
  • Occasionally it turns your urine green. This has no significance but is entertaining.

In short, propofol is not a perfect drug, but has the advantage that its pitfalls are out in the open: we know about them and they are usually manageable. This is a good thing.


Our other sedative standby is…


These are old, effective drugs. Working on similar GABA receptors to ethanol (although different ones from propofol), they offer sedation/hypnosis, reliable amnesia, and anxiolysis, but like propofol, no analgesia whatsoever. Compared to propofol, they cause relatively little respiratory depression; massive benzo overdose can cause apnea, but a much more common cause is polypharmacy, as the synergy of benzos with opioids is rather effective at causing patients to stop breathing. Likewise, their effect on hemodynamics is usually mild, although as we learned, no drug is truly stable in shock.

The trouble is that lately, as awareness has increased of the negative effects of over-sedation, delirium, and the post-intensive care syndrome, benzodiazepines have emerged as one of the leading culprits. Although any sedative can cause delirium, evidence has accumulated that benzos may be more “deliriogenic” than most of the alternatives. Since delirium is associated with poor outcomes—like mortality—many of us have started to avoid these time-honored drugs when possible. However, that doesn’t mean they have no role.

Most often, the benzo of choice is midazolam (Versed). Midazolam is the shortest-acting benzo in our typical armamentarium, taking effect within a couple minutes and lasting an hour or two. Like propofol, except even more so, it builds up in the tissues, and prolonged use will make it last longer and longer (particularly in renal failure, since it has some active metabolites). It is typically pushed for immediate effect or given by continuous infusion for ongoing sedation. Dosing is usually around 1–4mg by push—although textbooks list much higher figures when used for induction—and around 1–10mg/hr by infusion.

The other option is lorazepam (Ativan). Unfortunately, it probably has no advantages over midazolam for the critically ill. It takes longer to kick in (as much as 15 minutes), and lasts much longer, around 6 to 8 hours. Finally, lorazepam infusions use propylene glycol as a diluent, which can cause an anion gap acidosis (with elevated osmolal gap) after prolonged use. In a recovering or intermediate-level patient, it is a decent choice for seizures, a decent choice for alcohol withdrawal (especially because it can be given PO, and has no active metabolites, which may lead to cleaner clearance in liver disease), and an okay choice for mild anxiolysis at low doses. For sick people, just use midazolam. IV dosing is similar to midazolam, but about twice as potent.

Diazepam (Valium) is rarely seen in the ICU. Despite a rapid onset, it has a very long duration of effect. Its use should probably be limited to alcohol withdrawal, but once we enter that territory, a host of options (chlordiazepoxide/Librium, oxazepam/Serax, and more) appear, all of which are really a topic for another day.

Come back next time, and we’ll finish up by discussing some more unusual sedative options.