Perioperative Anaphylaxis: The Epinephrine Dose That Isn't the ED Dose
A minute after the antibiotic infusion starts, her airway pressures spike and her pressure collapses. The drug is right; the dose and route the emergency department would use are wrong for a patient who's already got a line in and a monitor running.
L.F., a 58-year-old woman, is thirty minutes into an elective total knee arthroplasty when the pre-incision cefazolin infusion — her fourth surgery, never a documented reaction before — is barely half infused when her peak inspiratory pressure jumps from 22 to 41 cmH2O and her blood pressure, which had been a steady 132/78 all morning, drops to 62/34 within ninety seconds. Her oxygen saturation follows it down, and a faint urticarial flush is spreading across her chest under the surgical drape.
This is IgE-mediated anaphylaxis, and epinephrine is unambiguously the correct drug — but the dose and route a bystander would give in the community, 0.3 to 0.5 mg intramuscularly, exist specifically to move drug into a patient with no other way in, fast. She already has a running IV and continuous invasive monitoring, which is a genuinely different clinical context, not just a more convenient one: small, titrated IV boluses (on the order of 10 to 100 mcg, escalated against the actual pressure response on the monitor) reach the myocardium and vasculature immediately, without the absorption delay IM dosing is built around, and without the larger fixed dose's real risk of provoking dangerous hypertension or arrhythmia in a patient whose vasculature is about to respond very differently than an outpatient's would.
Three prior uneventful cefazolin exposures make the reaction surprising to the room but not to the epidemiology: NAP6, the UK national audit that reviewed a full year of severe perioperative anaphylaxis, found antibiotics the single commonest trigger at 48 percent, ahead of neuromuscular blocking agents at 25, and sensitization by definition requires the earlier exposures her chart records as reassuring. Her pressure of 62/34 is worth reading against that audit too — NAP6's reviewers set 50 mmHg systolic as the point at which chest compressions should already have started, and criticized delay past it in more than half the arrests they examined. She is twelve points above that line, which is the margin the next few minutes are spending — and the first of those minutes went on the team's own instinct to look for an alternative explanation, a vasovagal event or an anesthetic effect, before the urticaria spreading under the drape settled it. That brief hesitation is itself a real, common feature of perioperative anaphylaxis recognition: the patient can't report symptoms, the rash is easy to miss under surgical draping, and the earliest hemodynamic signs can look like half a dozen other intraoperative problems before the pattern becomes unmistakable.
Intraoperatively, ninety seconds into the reaction
Epinephrine now, IV, titrated — not the standard 0.3 to 0.5 mg IM dose. She's already got a line and an arterial waveform in front of me; I'm giving 50 mcg boluses against the pressure response, not a fixed IM dose built for someone with no access at all.
I'd start lower than 50 and I don't think that's a quibble. NAP6 found no tachyarrhythmias at 50 mcg boluses, but its dead were disproportionately older patients on ACE inhibitors, and we don't yet know what she takes. Ten to twenty micrograms tells me the direction of her response without committing me to it. Where I'd go faster than you is the infusion — if two boluses don't move her, start it then rather than after a third, because bolus-chasing is how a reaction stays half-treated for ten minutes while everyone feels busy.
If she plateaus on an adequate infusion, vasopressin at a 2-unit bolus is the next move rather than more epinephrine — a non-adrenergic pathway for a vasoplegia that has stopped answering the adrenergic one. But I want to disagree with the premise you're both working from, which is that this is a dosing conversation. NAP6's reviewers judged fluid resuscitation insufficient in a large share of the cases they read, and the commonest management criticism they made was not the epinephrine dose. She is capillary-leaking into a third space right now. If the only thing being titrated in this room is a pressor, we will get the dose exactly right and still lose the pressure.
Agreed and executed: two titrated IV epinephrine boluses (50 and 75 mcg) restored her pressure to 104/62 within three minutes; bronchospasm resolved with the same doses. Surgery was halted, cefazolin identified as the presumed trigger and flagged in her allergy record, and she was extubated uneventfully at case end after a period of hemodynamic observation.
Vasopressin was never needed — her response to titrated epinephrine alone was adequate, which the critical care physician noted as reassuring but not something to generalize from without a genuinely refractory case to test the second-line reasoning against.