Zworth Reading
EMMax’s EM Weekly Update
Highlight of the Week
ED Return Visits: Separating Signal from Noise
EMOttawa · Commentary / Educational (Expert opinion, LOE 4-5)
A thoughtful piece on the psychology and reality of ED bounce-backs. Dr. Sim Singh (one of my former co-residents) argues that most return visits are not errors, that using return-visit rates as a quality or competency metric is methodologically shaky, and that the emotional weight we attach to "your patient came back" is disproportionate to actual risk. It walks through why raw bounce-back rates conflate appropriate re-presentations — worsening disease, access issues, patient preference — with genuine missed diagnoses, and argues that punitive QA frameworks built on these numbers often measure noise rather than signal.
I really like some of the implications for ED practice Dr. Singh raises - recognize high risk patients for disposition, address abnormal vital signs before dispo, and make clinical reasoning visible in the chart.
Bottom line: Many if not most ED return visits are not errors, and using bounce-back rates as a quality or competency metric is not necessarily effective. They do, however, sometimes offer an opportunity to reflect and learn.
FOAM Radar
Should I treat Strep throat?Knowledge translation / Evidence review
First10EM · Synthesizes RCT and systematic review evidence (LOE 1-2 underlying)
Justin Morgenstern revisits the surprisingly uncertain evidence base for treating GAS pharyngitis. The post walks through the NNT for preventing suppurative complications (very high), the minimal symptom duration benefit (~16 hours), and the near-extinction of rheumatic fever in developed countries. It's a useful framework for the shared decision-making conversation we should probably be having more often. I summarized some of the same evidence in a previous Grand Rounds presentation.
Bottom line: Worth reading if you want ammunition for the 'maybe we don't need to treat every positive strep test' conversation. The evidence for routine antibiotics is weaker than most assume, though local rheumatic fever epidemiology matters.
Guidelines Update
ACR Appropriateness Criteria® Epigastric Pain
American College of Radiology
- CT abdomen with IV contrast is usually appropriate for suspected esophageal or gastroduodenal perforation
- Upper GI series with water-soluble contrast remains an option when CT is unavailable or contraindicated
- For uncomplicated GERD or suspected peptic ulcer disease without alarm features, imaging is usually not appropriate initially
- CT is usually appropriate for evaluating complications post-fundoplication or other reflux procedures
Mixed — GRADE methodology adapted, but many recommendations based on expert consensus where RCT evidence is lacking. The perforation imaging recommendations have stronger evidentiary support than the GERD/PUD recommendations.
Conflicts with existing guidance: No major conflicts with existing practice. This largely codifies what most EDs already do — CT for suspected perforation, restraint in imaging uncomplicated dyspepsia.
Bottom line: Useful reference document but unlikely to change ED practice. The main value is having citable guidance when declining imaging for low-risk epigastric pain or when justifying CT for suspected perforation.
Adjacent Specialties
Incidence and Outcomes of Refractory Septic Shock per Consensus Clinical Criteria: A Multicohort Retrospective StudyLOE 4 (Retrospective cohort)
Critical Care Medicine · Critical Care
In patients meeting septic shock criteria (n=15,732), what is the incidence and mortality of refractory septic shock defined as norepinephrine equivalent >0.5 µg/kg/min plus lactate >2 mmol/L?
Refractory septic shock occurred in 21.8% of septic shock patients. Mortality was 64.4% in the refractory group versus substantially lower in non-refractory patients (risk-adjusted OR 4.87 for death).
This operationalizes the recently published consensus criteria for refractory septic shock (also worth checking out) and provides the real-world incidence data. The retrospective design and reliance on EHR data are limitations, but the large sample size and multicenter design strengthen generalizability. The 4.87 OR for mortality is striking and validates that these criteria identify a genuinely high-risk population. For ED purposes, this gives us a concrete threshold to communicate prognosis and escalation needs. I will say the finding that mortality is high in patients requiring > 0.5 mcg/kg/min of norepinephrine is not groundbreaking.
Bottom line: One in five septic shock patients will meet refractory criteria, and two-thirds of those will die. Knowing this threshold (NE equivalent >0.5 plus lactate >2) helps frame goals-of-care conversations and ICU handoffs. This is prognostic information, not a treatment target.
Impact of Time to Antibiotics on In-hospital Mortality in Neutropenic Sepsis: A Prospective Multicenter Cohort StudyLOE 3 (Prospective observational cohort)
Critical Care Medicine · Critical Care / Oncology
In patients with neutropenic sepsis (n=942), does time to antibiotics (<1h vs 1-3h vs ≥3h) affect in-hospital mortality?
Delayed antibiotics (≥3h) had OR 1.50 for mortality compared to <1h; even intermediate timing (1-3h) had OR 1.26. Effect was strongest in patients with septic shock and hematologic malignancy.
This is a well-designed prospective multicenter study. Limitations include the observational design (sicker patients might get faster antibiotics, biasing toward null) and the Korean healthcare setting. The finding that the 1-3 hour group also had increased mortality compared to <1 hour is important — this isn't just about avoiding egregious delays.
Bottom line: Neutropenic sepsis is a true time-sensitive emergency. Every hour matters, and the effect is strongest in exactly the patients we worry about most — those with shock and hematologic malignancy. This reinforces aggressive door-to-antibiotic protocols for febrile neutropenia.
Major Journals Scan
Ambulance offload delays and patient outcomes: a systematic reviewLOE 1 (Systematic review of observational studies)
Emergency Medicine Journal
Why it matters to EPs: Ambulance ramping is a daily reality in most Canadian EDs and a frequent topic of media and political attention. Understanding what we actually know about patient harm from offload delays is essential for informed advocacy and resource allocation discussions.
Only 4 observational studies met inclusion criteria. Offload delays were consistently associated with longer ED length of stay. Evidence for mortality harm was inconsistent. Only one study found increased 30-day mortality, others found no difference. No studies assessed time to treatment or clinical deterioration.
This is an honest systematic review that reveals how little rigorous evidence exists on a topic we discuss constantly. For one, it's really hard to study this across health systems. Second, it tends to be the case that sicker patients often get prioritized for rapid offload, creating confounding that's difficult to fully adjust for. The consistent LOS finding is real but unsurprising. The mortality signal is weak and inconsistent. The authors appropriately call for better causal inference methods rather than overstating current evidence.
Bottom line: We have strong intuition that ambulance ramping harms patients, but the actual evidence for clinical harm is surprisingly thin. This doesn't mean ramping is safe, it means we haven't studied it well enough to quantify the harm. Useful context for policy discussions: we should fix ramping for many reasons, but citing definitive mortality data would be overstating the evidence. Also worth continuing to discuss in your shop the best way to provide care to these patients given the current reality.