Zworth Reading
EMMax’s EM Weekly Update
Highlight of the Week
CMAJ · Clinical overview / narrative review
Concise CMAJ clinical primer on Ebola virus disease for Canadian clinicians, prompted by the 2026 DRC Bundibugyo ebolavirus outbreak (case fatality 30-50%). Authors are Canadian ID physicians at U of T. Three ebolaviruses cause human disease, transmitted through bodily fluid contact, with incubation of 2-21 days. Despite the 'hemorrhagic fever' label, fewer than half of patients develop hemorrhagic symptoms. Early presentation is non-specific (fever, fatigue, myalgia, GI symptoms). Diagnosis is by blood PCR. Vaccines and monoclonal antibodies have significantly improved outcomes for Zaire ebolavirus, but supportive care remains the mainstay for Bundibugyo.
Very useful clinical primer aimed at the front-line clinician. The most important EP-relevant point is the non-specific early presentation. An undifferentiated febrile traveller from an outbreak region cannot be screened out by absence of bleeding.
Bottom line: Most of us will never see an Ebola case, but if we do, the cost of missing it is enormous. Two takeaways: (1) ask about travel from outbreak regions in any undifferentiated febrile illness, and (2) absence of bleeding does not rule out Ebola. Most patients don't bleed. If suspected, isolate first, then call Public Health and ID. Worth a 5-minute read as the DRC outbreak continues.
Practice-Changing EM
Vasopressors or Fluids in Early Septic Shock (ARISE FLUIDS)LOE 2 (Multicentre open-label RCT)
New England Journal of Medicine
In ED patients with septic shock and persistent hypotension after an initial ≥1 L fluid bolus, does a restricted-fluid + early-vasopressor strategy improve days alive and out of hospital at 90 days compared to liberal fluids + later vasopressors?
963 patients across 51 sites in Australia, New Zealand, and one other site. Inclusion required suspected sepsis with SBP <90 or MAP <65 despite ≥1000 mL crystalloid, lactate >2 mmol/L, on antimicrobials, within 6 hours of ED presentation. The intervention separation was meaningful. At 24 hours the vasopressor arm had received ~1100 mL less fluid (1140 vs 2248 mL), pressors started a full hour sooner (median 0.4 vs 1.4 h), and 86.5% vs 67.6% were on pressors. Despite this, the primary outcome was flat: median 76 days alive and out of hospital in both groups (difference 0.0, 95% CI -2.7 to 2.7, p=1.00). Mortality numerically favoured the fluids arm at every timepoint: 28-day mortality 12.9% (vasopressors) vs 10.0% (fluids), RR 1.29 (95% CI 0.91-1.85); 90-day mortality 16.4% vs 14.4% (RR 1.14, 95% CI 0.85-1.54). Neither difference reached statistical significance. Pulmonary edema was less common with restrictive fluids (0.6% vs 5.0%, RR 0.12).
This is the long-awaited adequately-powered trial of the restricted-fluid + early-pressor strategy in early septic shock and the result is pretty clear on the primary outcome: equivalence. Two important caveats. First, the mortality numbers numerically favour the fluids arm, not the restrictive arm. They don't reach significance and shouldn't be over-read, but they certainly do not support 'less is better.' Second, the pulmonary edema benefit was acknowledged by the authors themselves as potentially inflated by reporting bias given the open-label design (clinicians knew which arm patients were in). Other limitations worth noting: patients with other reasons for fluid restriction (e.g., known cardiac failure) were excluded; treatment separation was real but attenuated (many fluids-arm patients also got pressors. This trial appears broadly consistent with CLOVERS. Restrictive and liberal approaches produce similar outcomes in this selected population. It does not show that less fluid is better, but it does show that less fluid is not worse on a hard patient-centred outcome.
Bottom line: ARISE FLUIDS adds to a growing body of evidence (CLOVERS, CLASSIC, the broader restrictive-fluid literature) showing that rigid 30 mL/kg protocols don't improve outcomes in early septic shock. My practical takeaway is after an initial bolus, reassess. Don't reflexively chase 30 mL/kg in patients you're worried about fluid-overloading (especially cardiac/renal failure), and don't be afraid to start norepinephrine early. But also keep in mind that data do not suggest a restrictive approach is superior.
Important EM Papers
Association of Rescue Breathing With Outcomes in Adult Suffocation-Related Cardiac ArrestLOE 4 (Retrospective cohort with IPTW)
Resuscitation
In adults with witnessed suffocation-related OHCA, does bystander compression-plus-ventilation CPR (CCV-CPR) improve favorable neurological outcome compared to compression-only CPR (CC-CPR)?
All-Japan Utstein Registry, 2005-2023, 76,774 patients. After IPTW adjustment, CCV-CPR was not significantly different from CC-CPR for favorable neurologic outcome at 1 month (aOR 1.01, 95% CI 0.78-1.32). Counterintuitively, the No-CPR group was associated with higher favorable neuro outcomes than CC-CPR (aOR 1.20). The authors attribute this to unmeasured confounding and selection bias.
Several real limitations. The No-CPR > CC-CPR finding is almost certainly selection bias. The No-CPR group likely includes more witnessed-arrest-with-rapid-EMS-response scenarios and other situations where bystanders deferred to professional responders. IPTW can't fix what isn't measured (scene circumstances, EMS response time, perceived viability). The suffocation category is also heterogeneous (choking, hanging, FBAO, smothering, different pathophysiology). All that said, the interesting signal here is the CCV vs CC comparison: in a population where biological reasoning says ventilation should matter most (hypoxic arrest), adding ventilation didn't help.
Bottom line: Don't take this as clear evidence to skip ventilation in suffocation arrest. The No-CPR comparison is too confounded. But the CCV-vs-CC signal is interesting and consistent with the broader trajectory of CPR evidence: chest compressions remain the high-value bystander intervention, and the marginal value of adding ventilation may be smaller than we used to assume even in arrests we think should be ventilation-dependent.
FOAM Radar
PulmCrit: Is oseltamivir assassinating ICU patients?Critical appraisal / commentary
EMCrit · LOE 2 (adaptive platform RCT — preliminary, full paper pending)
Farkas analyzes the REMAP-CAP oseltamivir arm, which reported a 98.3% Bayesian probability that oseltamivir increases mortality in critically ill influenza patients. The arm was stopped for harm. Farkas walks through methodological concerns with the trial design. The analysis is worth reading for the methodology discussion alone (adaptive platform interpretation, Bayesian probability framing, what stopping rules can and cannot tell us).
Bottom line: Some real methodological issues with this trial that Farkas does a good job unpacking. His ultimate take is that limitations aside, this is the best evidence we currently have for critically ill flu patients, and for now stopping reflexive oseltamivir in this population is probably the right move. Less clear how to extrapolate to less sick populations. Full paper not yet published. Either way, the signal is consistent with the 2025 JAMA Network Open meta-analysis and the prior Cochrane review showing limited benefit from oseltamivir.
Adjacent Specialties
Heart Failure Occurring in the Perinatal Period: A Scientific Statement From the American Heart AssociationLOE 5 (Scientific Statement / Expert Consensus)
Circulation · Cardiology / Obstetrics
Pregnant and postpartum patients with new or preexisting heart failure — what are the diagnostic challenges, treatment standards, and research gaps?
The AHA acknowledges that perinatal HF remains poorly defined, underdiagnosed, and undertreated. Physiologic pregnancy symptoms overlap significantly with HF symptoms, leading to diagnostic delays. The statement calls for standardized definitions, better screening tools, and inclusion of pregnant patients in HF trials. Notably, the document addresses BNP/NT-proBNP interpretation: levels are mildly elevated in normal pregnancy (more so in the first trimester), and an expert ACC Cardio-Obstetrics panel has proposed pregnancy-specific thresholds, though robust data are lacking.
For EPs, the key message is that dyspnea, edema, and fatigue in pregnant or postpartum patients deserve a lower threshold for cardiac workup than we typically apply.
Bottom line: Pregnant and postpartum patients with 'normal pregnancy symptoms' can have HF. May be worth having a lower threshold for workup. On BNP: there is a role, but interpret with caution. NT-proBNP is mildly elevated in normal pregnancy (especially first trimester), so positive predictive value is lower than in the non-pregnant population. Negative predictive value remains reasonable (BNP <100, NT-proBNP <128), and in true acute HF the values are usually markedly elevated. Real uncertainty remains. Use BNP as one data point alongside echo and clinical assessment, not in isolation.