Sheridan B, Griggs J, Mondim P, Ter Avest E, Lyon R. Maintenance of prehospital anaesthesia using an intermittent bolus regime in blunt trauma patients with a high GCS and hemodynamic reserve: a retrospective cohort study.
In-hospital anesthesia maintenance is almost universally continuous infusion. Prehospital maintenance after RSI has stayed bolus-based by convention — and until this paper, nobody had actually measured what that convention delivers.
Bolus administration of sedatives/analgesics post-RSI carries an inherent structural risk: dosing can be delayed or entirely omitted amid competing clinical priorities, and most services provide no protocol for bolus intervals or doses — drug choice, dose, and timing are left to individual clinician discretion. This study set out to quantify, not just theorize about, the practice effect of that discretion in a population specifically selected to need reasonable sedation: hemodynamically stable, GCS ≥9 blunt trauma patients without severe TBI.
| Design | Single-center retrospective observational cohort, UK HEMS (KSS) |
| Period | Jan 1, 2019 – Feb 28, 2023 |
| Population | 101 blunt trauma patients, age ≥12, pre-induction GCS 9–15, pre-induction SBP ≥100 mmHg, pre-induction shock index ≤1.0 (i.e., hemodynamically reserved, not exsanguinating, not severely head-injured) |
| Primary objective | Describe practice effect of intermittent-bolus maintenance dosing (midazolam, ketamine, fentanyl, rocuronium), standardized to weight and time (dose/kg/h) |
| Induction | 100% received ketamine + rocuronium; 92% also received fentanyl |
This is the part worth pulling into any teaching slide built from this paper.
| Agent | Median maintenance dose | Range (non-zero) |
|---|---|---|
| Fentanyl | 0.64 mcg/kg/h | 0.26–5.56 mcg/kg/h |
| Midazolam | 0.028 mg/kg/h | 0.005–0.157 mg/kg/h |
| Ketamine | 0 mg/kg/h (median) | 0.05–1.62 mg/kg/h |
For an 80 kg patient (this cohort's median weight), the study's own stated bottom line is that the average patient would receive only 2.2 mg of midazolam and 51.2 mcg of fentanyl in one hour — well below prior published guideline recommendations (e.g., the German Society of Anaesthesiology's 2016 out-of-hospital guidance: midazolam 3–5 mg and fentanyl 150 mcg repeated every 20 minutes). Eight patients (8%) received no maintenance sedative or analgesic at all, with anesthesia lasting 48–74 minutes — and in six of those eight cases, the chart showed no physiological reason for the omission.
Maybe clinicians were just titrating carefully to each patient. The data says otherwise.
If under-dosing reflected careful individualized titration, you'd expect maintenance dose to correlate with something — patient physiology, induction dose, or at minimum, transport duration (more time, more cumulative drug). None of that held. After Bonferroni correction across 20 comparisons, maintenance dosing showed no significant association with age, weight, shock index, GCS, heart rate, blood pressure, or induction dose. Total cumulative dose also showed no relationship with anesthesia duration for any of the four drugs (fentanyl r=-0.04, ketamine r=0.07, midazolam r=0.01, rocuronium r=0.15; all P>.19) — clinicians gave roughly the same total drug regardless of whether transport took 20 minutes or 100. The authors' own read: this pattern is harder to explain as deliberate precision dosing and easier to explain as a byproduct of cognitive load and competing priorities during a demanding, understaffed (two-clinician) prehospital resuscitation.
The source post's framing holds up well against the primary data, and if anything is conservative: an intermittent-bolus-only regimen, without standardized dosing intervals, produced doses well below published guideline recommendations in a substantial minority of patients, with no evidence the shortfall reflected deliberate clinical titration. The paper doesn't prove infusions are safer or more effective — that's explicitly flagged as an area needing further research — but it makes a strong, quantified case that unstructured bolus dosing has a real, measurable ceiling problem, particularly under the cognitive load of a two-person prehospital trauma team.