🩺 THIS WEEK IN EM
Prehospital Whole Blood: The First Big RCT Just Came Back Negative — Here's What That Does and Doesn't Mean
The enthusiasm for whole blood in trauma has been real and growing. The logic made sense: instead of giving separate units of red cells, plasma, and platelets with varying ratios, why not use whole blood — the natural balanced product — and avoid the complexity? The problem was that the enthusiasm ran well ahead of randomized outcome data. The SWiFT trial is the first large RCT to try to close that gap.
SWiFT enrolled 942 trauma patients across ten air ambulance services in England who required prehospital transfusion. Patients were randomized to receive up to two units of whole blood or up to two units of red cells and two units of plasma. The primary outcome was a composite of death or massive transfusion within 24 hours. The result was flat: 48.7% of the whole blood group versus 47.7% of the standard component group hit the primary outcome. No difference. Mortality was the same. Blood product use in the first 24 hours in hospital was the same. The whole blood group actually had a higher rate of abnormal prothrombin time.
Before this gets read as "whole blood is finished," two structural problems with SWiFT are worth understanding. First, the trial only gave two units of blood prehospital. For a patient who ends up needing a massive transfusion, two units is a rounding error compared to what they'll receive in the trauma bay and ICU. Both groups quickly converged on essentially identical products in hospital, which made detecting a meaningful difference almost mathematically impossible. Second, this was a civilian trauma population with predominantly blunt mechanism — not the military hemorrhage context where most of the whole blood observational enthusiasm originated.
The honest read is that SWiFT tells you prehospital whole blood at this dose in this population did not produce a measurable win. It does not tell you that whole blood has no future. Larger trials, higher doses, different populations, and in-hospital use still need to be studied. But it does tell you that anyone treating this as settled — either direction — is ahead of the data.
Bottom line: The SWiFT trial, the first large RCT of prehospital whole blood in civilian trauma, found no difference in death or massive transfusion within 24 hours compared to component therapy — but the two-unit dose and rapid convergence of in-hospital products make this a limited test of the hypothesis rather than a definitive answer on whole blood's role in trauma resuscitation.
Smith JE, Cardigan R, Sanderson E, et al; SWiFT Trial Group. Prehospital Whole Blood in Traumatic Hemorrhage — a Randomized Controlled Trial. N Engl J Med. 2026 Mar 17.
Suspected PE in Pregnancy: The Imaging Reflex Is Costing More Than It Buys
Pulmonary embolism in pregnancy is one of the most anxiety-provoking diagnostic situations in emergency medicine. The disease is real, the miss is catastrophic, and the standard protective instinct is to image early and often. The problem is that routine CTPA in a pregnant patient carries real fetal radiation exposure, and the D-dimer — normally a useful filter — is almost universally elevated in pregnancy by the third trimester, which strips it of most of its discriminatory value under traditional cutoffs. The result is that many EDs have effectively abandoned D-dimer as a useful tool in this population and moved straight to cross-sectional imaging. The evidence says that is not the right call.
The pregnancy-adapted YEARS algorithm was developed to bring structured risk stratification back into this workup. The three YEARS criteria — clinical signs of DVT, hemoptysis, and PE as the most likely diagnosis — are applied alongside fixed D-dimer thresholds that do not vary by trimester. If none of the three YEARS criteria are present, D-dimer can rule out PE at a threshold of less than 1,000 μg/L. If one or more criteria are present, the threshold drops to less than 500 μg/L. A prospective management study published in NEJM validated this approach in 498 pregnant patients across multiple European centers, and only one patient — with a popliteal DVT, not PE — had a thromboembolic event at three-month follow-up. CTPA was avoided in 39% of the cohort.
The leg ultrasound piece is also underused. Many pregnant patients with suspected PE present with leg symptoms as well. A negative compression ultrasound with no YEARS criteria and a D-dimer below threshold is a very low-risk result that doesn't automatically need CTPA to close the loop — though ultrasound-positive patients with suspected DVT still need anticoagulation decisions and further imaging.
The practical gap is implementation. US data suggests that outside of academic centers, pregnant patients with suspected PE still get routed to CTPA at very high rates without structured risk stratification. If your department doesn't have a pathway for this population, the pregnancy-adapted YEARS algorithm is a reasonable framework to build from — it's prospectively validated, the failure rate is low, and it avoids unnecessary fetal radiation exposure in a substantial portion of patients who would have been imaged reflexively.
Bottom line: The pregnancy-adapted YEARS algorithm using fixed D-dimer thresholds (not trimester-adjusted) safely avoided CTPA in nearly 40% of pregnant patients with suspected PE in prospective validation, with a three-month thromboembolic failure rate near zero — a structured approach that outperforms the reflex of imaging everyone.
van der Pol LM, Tromeur C, Bistervels IM, et al; Artemis Study Investigators. Pregnancy-Adapted YEARS Algorithm for Diagnosis of Suspected Pulmonary Embolism. N Engl J Med. 2019;380(12):1139-1149.
Video EMS Dispatch: Seeing the Scene Changes the Call More Than You'd Think
Most EMS dispatch decisions are made on voice alone. A caller describes what they're seeing, a dispatcher interprets it, and a response priority gets assigned — all without anyone actually looking at the patient or the scene. That gap between what the caller says and what's actually happening has been a known problem in EMS for decades. Video dispatch is one proposed solution, and the evidence on what it actually changes is starting to accumulate.
The core finding across multiple studies is straightforward: when dispatchers can see the scene via video, they change their assessment and response decisions at a meaningful rate. A large observational study by Linderoth and colleagues found that live video from callers' smartphones altered the assessed criticality and changed the dispatched response in more than a quarter of cases. A 2026 randomized simulation study published in BMC Emergency Medicine showed that video calls improved dispatch code accuracy — 41% correct versus 25% in audio-only — though the difference did not reach statistical significance, and video calls took about 40% longer. A 2026 observational study from Victoria, Australia examining video triage for low- to medium-acuity calls found that video changed triage and transport decisions in a clinically meaningful proportion of cases, with lights-and-sirens transport to the ED remaining stable, suggesting the main effect was appropriate de-escalation rather than dangerous under-triage.
The tradeoffs are real. Longer call duration, technical failures, dispatcher cognitive load, and questions about consent and recording retention all complicate implementation. But the more important takeaway for emergency physicians is what video dispatch implies about the data you're working with when a patient arrives. The acuity call that came in as a "minor fall" was assigned that priority based on an audio-only description from a stressed bystander. Video systems are beginning to show that this process misclassifies cases in both directions — some high-acuity patients get under-triaged, some low-acuity patients get over-triaged — and the pattern of miscategorization matters when you're planning resource allocation and staffing.
This isn't an argument for EDs to lobby for video dispatch systems. It's an argument for understanding what EMS triage decisions are actually based on, and for not treating the prehospital priority code as more reliable than it is when it conflicts with what's in front of you.
Bottom line: Multiple studies show that video-assisted EMS dispatch changes triage assessments and transport decisions in a substantial minority of cases compared to audio-only, with the primary benefit appearing to be appropriate de-escalation of lower-acuity calls — a finding that also serves as a reminder of how much uncertainty exists in any prehospital priority assignment made without visual information.
Linderoth G, Lippert F, Østergaard D, et al. Live video from bystanders' smartphones to medical dispatchers in real emergencies. BMC Emerg Med. 2021;21(1):101. doi:10.1186/s12873-021-00493-5. // Nehme E, Flaus C, Clarke S, et al. Video triage by emergency medical service secondary triage clinicians in Victoria, Australia. J Telemed Telecare. 2026. doi:10.1177/1357633X251383395.
🔭 Next Week
IO vs IV access in cardiac arrest: three large RCTs and updated AHA guidelines that should change how quickly you reach for the drill
AI-assisted chest X-ray reads in the ED: whether having a second opinion from an algorithm actually improves decisions or just creates new noise
Intranasal vs subcutaneous ketamine for acute traumatic pain: what an Annals RCT says about which non-IV route gets you there faster with fewer side effects
The Hallway Consult is built for EM clinicians who want the useful version of the literature. Forward it to a colleague if it helped.
— The Hallway Consult team