Choosing a Sedative in Sepsis: What Dexmedetomidine's Reputation Actually Rests On
An intensivist wants dexmedetomidine by default for its "delirium-protective" reputation. The disagreement is about which comparison that reputation was actually built on, and whether it survives being tested against the sedative it's about to replace.
Marcus J., 58, a long-haul truck driver, was admitted two days ago with septic shock from a perforated diverticulitis, now source-controlled surgically and hemodynamically stabilized on low-dose norepinephrine. He remains intubated and requires ongoing sedation targeted to a light RASS of 0 to −2 per the unit's usual protocol. He has no prior psychiatric or cognitive history and no chronic illness of any kind before this admission, which matters here: sepsis, mechanical ventilation, and sedative exposure are each independent delirium risks, and he carries none of the baseline cognitive vulnerability that usually compounds them. His wife describes him as sharp and even-tempered, the kind of driver who does his own maintenance checks rather than trust a shop, and who has driven the same cross-country route for eleven years without a single missed delivery window.
The overnight team is choosing between propofol, the unit's default continuous sedative, and dexmedetomidine, which the attending specifically requests, describing it in rounds as "the delirium-protective one." That reputation is real but, on closer inspection, tied to a specific historical comparison — dexmedetomidine's advantage was originally demonstrated against benzodiazepines, a drug class this unit stopped using for routine sedation years ago specifically because of their own delirium risk. Whether that same advantage holds up against propofol, the comparator actually being considered tonight, is a separate and more recently tested question, and the team's working assumption — that "dexmedetomidine prevents delirium" as a general property of the drug — may not survive contact with the specific trial that tested exactly this comparison, a trial run specifically in septic, mechanically ventilated patients much like Marcus.
In the ICU, choosing a sedation strategy
I'd like dexmedetomidine as our sedative tonight. It has a real reputation for reducing delirium and improving coma-free days — that's not something I'm inventing, it goes back to the original MENDS trial, Pandharipande and colleagues, JAMA, 2007, which found dexmedetomidine outperformed lorazepam on exactly those outcomes.
I want to name something about that trial before we generalize from it: lorazepam is a benzodiazepine, a GABAergic drug we stopped using for routine sedation specifically because of its own delirium risk. The comparison that built dexmedetomidine's reputation was against exactly the drug class we already know is bad for this outcome — it doesn't tell us how dexmedetomidine compares to propofol, which is what's actually on the table tonight.
The trial that answers that direct question is MENDS2 — Hughes and colleagues, New England Journal of Medicine, 2021, 422 mechanically ventilated septic patients, dexmedetomidine versus propofol head-to-head. Days alive without delirium or coma were essentially identical — 10.7 versus 10.8 days — with no difference in ventilator-free days, 90-day mortality, or cognitive testing at six months.
So I don't think tonight's choice should be decided on the delirium-prevention question at all — MENDS2 says that's roughly a wash between these two specific drugs. What I'd actually weigh is that Marcus is likely to be extubated within the next day or two if he keeps trending this way, and dexmedetomidine doesn't suppress respiratory drive the way propofol can at deeper sedation levels, which tends to make weaning trials smoother.
That's a real, separate reason to reach for dexmedetomidine tonight — it just isn't the reason the attending originally gave, and I think it's worth being honest with the team about which rationale is actually doing the work.
Agreed: dexmedetomidine started as the continuous sedative, titrated to RASS 0 to −2, with the rationale documented explicitly as respiratory-drive preservation ahead of anticipated weaning — not as a delirium-prevention claim the team agreed the current head-to-head evidence doesn't support over propofol.
The team's reasoning converged without lasting disagreement: the intensivist accepted that MENDS2 changes what claim can honestly be made for tonight's choice, while still agreeing dexmedetomidine was the right pick for a different, real reason.