Publication

• Title: Deferring arterial catheterization in shock
• Acronym: EVERDAC
• Year: 2025
• Journal published in: The New England Journal of Medicine
• Citation: Muller G, Contou D, Ehrmann S, Martin M, Andreu P, Kamel T, et al.; EVERDAC Trial Group. Deferring arterial catheterization in shock. N Engl J Med. 2025;393:1875-1888.

Context & Rationale

• Background

  • Invasive arterial catheters are widely used in ICU shock to provide continuous arterial blood pressure (ABP) monitoring and facilitate arterial blood sampling (blood gases, lactate, and other tests).
  • Potential benefits include beat-to-beat haemodynamic monitoring and rapid detection of hypotension; potential harms include insertion-related bleeding/haematoma, thrombotic/ischaemic complications, catheter-related bloodstream infection, and procedural burden.
  • Oscillometric non-invasive blood pressure (NIBP) may be less reliable in low-perfusion states and under vasopressor therapy; the clinical consequences of measurement error (overtreatment/undertreatment) versus harms of invasive cannulation were uncertain.
  • Prior evidence was dominated by observational associations, small measurement-accuracy studies, and practice variation data; robust randomised evidence on patient-centred outcomes (especially mortality) was lacking.
  • Contemporary shock care increasingly emphasises clinical perfusion endpoints (e.g., peripheral perfusion) in addition to numerical ABP targets, raising the question of whether routine invasive ABP monitoring is necessary for most patients.³
• Research Question/Hypothesis

  • In adults admitted to ICU with shock requiring vasopressors, is a non-invasive strategy (deferring arterial catheterisation with prespecified rescue criteria) non-inferior to routine early arterial catheterisation for 28-day all-cause mortality, within a 5 percentage-point margin?
• Why This Matters

  • Arterial lines are a high-frequency invasive intervention in critical care; even small net effects could translate into large population-level impacts on complications, costs, and ICU workload.
  • Defining when invasive ABP monitoring is truly necessary informs de-implementation strategies and could reduce avoidable harms without compromising safety.
  • The trial also interrogates downstream consequences of monitoring choices (blood sampling patterns, infection risk, patient discomfort), which are rarely captured in haemodynamic trials.

Design & Methods

• Research Question: Whether deferring arterial catheterisation (non-invasive BP monitoring strategy with rescue arterial catheterisation if safety criteria were met) is non-inferior to routine early arterial catheterisation for 28-day mortality in ICU patients with shock requiring vasopressors.
• Study Type: Multicentre, open-label, pragmatic, randomised, parallel-group, non-inferiority trial; investigator-initiated; 22 ICUs in France; 1:1 allocation; randomisation stratified by centre, invasive mechanical ventilation status, and vasopressor dose category at randomisation (<0.18 vs ≥0.18 μg/kg/min norepinephrine tartrate or epinephrine).¹
• Population:
  • Setting: ICU patients with shock receiving norepinephrine tartrate or epinephrine.
  • Key inclusion criteria: Vasopressor infusion (norepinephrine tartrate or epinephrine) with at least one sign of hypoperfusion; expected vasopressor duration ≥6 hours; eligibility allowed regardless of whether an arterial catheter was already in situ at randomisation.
  • Signs of hypoperfusion (examples used in eligibility definitions): arterial lactate ≥2 mmol/L; poor skin perfusion (e.g., mottling or capillary refill time >3 seconds); oliguria (<0.5 mL/kg/h for >2 hours); altered mental status/coma (attributing to shock rather than sedation).
  • Key exclusion criteria: Adults under legal guardianship; immediate/mandatory need for invasive ABP monitoring at inclusion (e.g., extracorporeal membrane oxygenation); very high vasopressor dose at inclusion (>2.5 μg/kg/min norepinephrine tartrate or epinephrine); inability to obtain NIBP measurements (e.g., measurement failure with appropriate cuff positioning/sizing); situations where protocolled strategy could not be delivered safely.
• Intervention:
  • Noninvasive strategy: Prefer oscillometric brachial cuff NIBP monitoring rather than routine invasive ABP monitoring.
  • Arterial catheter management at baseline: If an arterial catheter was present at randomisation, removal was mandated within 1 hour (unless safety criteria required continued invasive monitoring).
  • Rescue arterial catheterisation (prespecified safety criteria): Allowed/mandated if predefined triggers occurred (e.g., very high vasopressor dose, high-risk procedures, inability to obtain reliable NIBP, or other protocolised safety thresholds).
  • Blood sampling approach: Arterial punctures and arterial blood draws were discouraged; alternatives included central venous sampling and clinical integration with pulse oximetry where feasible (protocol supported venous-based substitutes to reduce arterial sampling burden).
  • High-risk surgery exception: If an arterial catheter was inserted for a high-risk procedure, protocol specified removal after return to ICU within a short timeframe (unless ongoing criteria required continued invasive monitoring).
• Comparison:
  • Invasive strategy: Routine early insertion of an arterial catheter (radial or femoral at clinician discretion) for continuous invasive ABP monitoring.
  • Timing target: Attempted insertion within 4 hours of randomisation (median achieved 1 hour); non-invasive cuff measurement was generally not used except during insertion/replacement or where the arterial line was temporarily not functioning.
  • Discontinuation: Removal permitted when invasive monitoring was considered no longer useful (e.g., vasopressor dose reduced and clinical hypoperfusion resolved, or palliative trajectory).
• Blinding: Open-label (clinicians and patients not blinded); primary outcome (mortality) objective; subjective outcomes (e.g., discomfort) susceptible to expectation/performance effects.
• Statistics: Planned sample size 1010 patients to test non-inferiority with a 5 percentage-point absolute margin, assuming 28-day mortality ~22.5–25%, with 80% power and α=0.05; primary analysis included both intention-to-treat and per-protocol approaches, with effect reported primarily as adjusted absolute risk difference (95% CI) and non-inferiority concluded if the upper bound of the CI was <5 percentage points.¹
• Follow-Up Period: Primary endpoint at day 28; follow-up extended to day 90 for longer-term mortality and key secondary outcomes.

Key Results

This trial was not stopped early. Recruitment proceeded to the planned sample size, with no interim stopping for efficacy or harm reported.

• Noninvasive strategy achieved major separation in exposure to arterial catheterisation (14.7% vs 98.2%), with later insertion when required (median 22 h vs 1 h).
• Primary outcome showed non-inferiority for 28-day mortality (adjusted risk difference −3.2 percentage points; 95% CI −8.9 to 2.5; non-inferiority P=0.006), without evidence of superiority.
• Trade-offs were explicit: fewer insertion-related haematoma/haemorrhage (1.0% vs 8.2%) and lower catheter-related bloodstream infection rate (0.6 vs 2.7 per 1000 ICU-days), but more arterial puncture attempts (742 vs 269 per 1000 ICU-days) and more reported device-related pain/discomfort (13.1% vs 9.0%).

Internal Validity

• Randomisation and Allocation: Central randomisation with allocation concealment; stratified by centre, invasive mechanical ventilation status, and vasopressor dose category at randomisation (<0.18 vs ≥0.18 μg/kg/min norepinephrine tartrate or epinephrine).¹
• Dropout / exclusions: 1010 randomised; 1006 included in intention-to-treat analysis (4 excluded after randomisation due to legal guardianship and/or withdrawal of consent).
• Performance/Detection Bias: Open-label design could influence thresholds for interventions, testing frequency, and clinician perceptions of safety; primary outcome (mortality) is objective and less prone to detection bias.
• Protocol Adherence: Invasive strategy achieved arterial catheterisation in 493/502 (98.2%), with 479/502 (95.4%) inserted within 4 hours; noninvasive strategy limited arterial catheterisation to 74/504 (14.7%).
• Timing: Median time to arterial catheter insertion was 1 h (IQR 0–2) in invasive strategy vs 22 h (IQR 6–141) in noninvasive strategy, indicating substantial timing separation.
• Baseline Characteristics: Groups were closely matched at baseline (mean age 66.3 vs 66.3 years; mean SAPS II 56 vs 56; mean SOFA 11 vs 11; mean arterial pressure 63.6 vs 63.8 mmHg; mean norepinephrine dose 0.44 vs 0.41 μg/kg/min; septic shock 70.6% vs 68.9%).
• Separation of the Variable of Interest: Arterial catheters inserted per 1000 ICU-days: 18 vs 112 (incidence ratio 0.16; 95% CI 0.11 to 0.22); arterial puncture attempts per 1000 ICU-days: 742 vs 269 (incidence ratio 2.76; 95% CI 2.41 to 3.16).
• Crossover: Rescue arterial catheterisation occurred in 74/504 (14.7%) in the noninvasive group; failure to attempt insertion or delayed insertion contributed to per-protocol deviations in the invasive group (per-protocol set: 492/504 and 490/504 across two definitions).
• Adjunctive therapy use: Central venous catheters inserted during ICU stay were similar (45 vs 53 per 1000 ICU-days; incidence ratio 0.84; 95% CI 0.66 to 1.06); blood cultures were similar (203 vs 220 per 1000 ICU-days; incidence ratio 0.92; 95% CI 0.80 to 1.07).
• Outcome Assessment: Mortality endpoints were clear and objective; catheter-related bloodstream infection was defined and quantified, but local catheter infection was inconsistently collected and non-catheter bloodstream infections were not recorded (limits full infection-burden inference).
• Statistical Rigor: Prespecified dual (intention-to-treat and per-protocol) framework appropriate for non-inferiority; additional analyses addressed differential non-adherence (including inverse probability weighting and complier-average causal effect approaches).²

Conclusion on Internal Validity: Overall, internal validity appears moderate to strong given robust randomisation, objective primary outcome, and large separation in the intervention exposure; limitations relate mainly to open-label conduct, discretionary exclusions prior to randomisation, and strategy-based crossover (rescue arterial catheterisation) that is intrinsic to the pragmatic intervention.

External Validity

• Population Representativeness: Typical ICU shock population in a high-income setting (mean age 66; majority septic shock; high illness severity with SAPS II ~56 and SOFA ~11).
• Important exclusions: Patients with very high vasopressor requirements (>2.5 μg/kg/min) and those with mandatory immediate invasive monitoring (e.g., extracorporeal support) were excluded, limiting applicability to the most profound shock phenotypes.
• Applicability: Most applicable to ICUs able to implement close oscillometric NIBP monitoring and protocolised rescue thresholds; performance may differ where cuff reliability is poor (e.g., extreme obesity, severe arrhythmia) or where arterial blood gas sampling is deeply embedded in routine workflows.
• Health-system translation: Likely generalisable to comparable European/UK/Irish ICU systems; external validity may be reduced in resource-limited settings where staff-to-patient ratios limit frequent NIBP cycling and timely rescue insertion.
• Subpopulation caution: Profound vasopressor-dependent shock may not be adequately represented; post hoc signals in high-dose vasopressor strata should discourage extrapolation without confirmatory evidence.

Conclusion on External Validity: Generalisability is good for adults with moderate vasopressor-dependent shock in high-resource ICUs, but limited for extracorporeal support populations, perioperative contexts requiring mandatory invasive monitoring, and patients with profound catecholamine requirements.

Strengths & Limitations

• Strengths:
  • Large pragmatic multicentre randomised trial addressing a ubiquitous ICU intervention.
  • Clinically meaningful primary endpoint (28-day all-cause mortality) with objective ascertainment.
  • Appropriate non-inferiority framework with both intention-to-treat and per-protocol analyses.
  • Marked protocol separation in arterial catheter exposure (14.7% vs 98.2%) with detailed measurement of downstream testing behaviours.
  • Captured patient-centred harms (pain/discomfort) alongside procedural complications and infection outcomes.
• Limitations:
  • Open-label design introduces potential performance bias (testing frequency, transfusion/lab triggers, clinician thresholds for rescue insertion).
  • Non-inferiority margin (5 percentage points) may be considered large relative to observed mortality (~35%), affecting clinical tolerance for uncertainty.
  • Intervention is a strategy with rescue criteria rather than absolute avoidance; outcomes depend on timely recognition of monitoring failure and prompt arterial catheterisation when needed.
  • Most profound shock phenotypes were excluded or under-represented (e.g., very high vasopressor doses, extracorporeal support), limiting conclusions for these groups.
  • Local catheter infection was inconsistently recorded and non-catheter bloodstream infections were not captured, constraining full infection attribution.

Interpretation & Why It Matters

• Clinical implication

  A noninvasive-first strategy with prespecified rescue arterial catheterisation can be implemented in ICU shock without increasing 28-day mortality, while materially reducing arterial catheter use (14.7% vs 98.2%) and some catheter-related harms.
• Net harms/benefits

  Benefits clustered around invasiveness and infection/bleeding (haematoma/haemorrhage 1.0% vs 8.2%; arterial catheter-related bloodstream infections 0.6 vs 2.7 per 1000 ICU-days), whereas the noninvasive strategy increased arterial puncture attempts (742 vs 269 per 1000 ICU-days) and slightly increased device-related discomfort (13.1% vs 9.0%).
• How it reframes monitoring

  The trial supports the view that invasive ABP monitoring is not automatically required as a default “bundle” component for most vasopressor-treated shock, provided clinicians accept a strategy that prioritises clinical perfusion assessment and uses invasive monitoring selectively.

Controversies & Subsequent Evidence

• Non-inferiority margin: A 5 percentage-point margin may be viewed as clinically permissive given baseline mortality ~35%, meaning small harms could be masked while still meeting non-inferiority.
• Strategy versus device effect: The intervention is a monitoring strategy (defer with rescue) rather than a pure comparison of invasive versus noninvasive measurement accuracy; efficacy and safety rely on trigger thresholds, clinician vigilance, and workflow.
• Profound shock uncertainty: In a post hoc subgroup with very high vasopressor dose at randomisation (>1.0 μg/kg/min norepinephrine tartrate or epinephrine), mortality was numerically higher in the noninvasive strategy group (54.0% vs 41.0%; adjusted risk difference 13.0 percentage points; 95% CI −3.0 to 29.0; P=0.11), leaving residual uncertainty in the sickest patients.
• Downstream burden shift: Avoiding arterial catheters reduced invasive line days but increased arterial puncture attempts; whether this “burden shift” is acceptable may depend on local staffing, ultrasound availability, and pain management practices.
• Interpretation of discomfort outcomes: Reported device-related discomfort may reflect repeated cuff cycling and puncture attempts; mitigation strategies (e.g., optimised cuff protocols and ultrasound-guided selective arterial access) may influence real-world acceptability.⁴
• Post-publication synthesis: High-quality post-2025 meta-analyses and guideline updates incorporating this trial were not identified in the available sources.

Summary

• EVERDAC randomised 1010 ICU patients with shock on vasopressors to a noninvasive-first strategy versus routine early arterial catheterisation.
• Noninvasive strategy was non-inferior for 28-day mortality (34.3% vs 36.9%; adjusted risk difference −3.2 percentage points; 95% CI −8.9 to 2.5; non-inferiority P=0.006).
• Arterial catheter exposure was markedly reduced (14.7% vs 98.2%), with later insertion when needed (median 22 h vs 1 h).
• Invasive-catheter harms and infection signals favoured the noninvasive strategy (haematoma/haemorrhage 1.0% vs 8.2%; arterial catheter-related bloodstream infections 0.6 vs 2.7 per 1000 ICU-days).
• Trade-offs included increased arterial puncture attempts (742 vs 269 per 1000 ICU-days) and more device-related discomfort (13.1% vs 9.0%).

Further Reading

Other Trials

• Hernández G, Ospina-Tascón GA, Damiani LP, et al. Effect of a resuscitation strategy targeting peripheral perfusion status vs serum lactate levels on 28-day mortality among patients with septic shock: the ANDROMEDA-SHOCK randomized clinical trial. JAMA. 2019;321(7):654-664.
• Asfar P, Meziani F, Hamel JF, et al. High versus low blood-pressure target in patients with septic shock. N Engl J Med. 2014;370:1583-1593.
• Rivers EP, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. 2001;345(19):1368-1377.
• Yealy DM, Kellum JA, Huang DT, et al. A randomized trial of protocol-based care for early septic shock. N Engl J Med. 2014;370:1683-1693.
• Peake SL, Delaney A, Bailey M, et al. Goal-directed resuscitation for patients with early septic shock. N Engl J Med. 2014;371:1496-1506.

Systematic Review & Meta Analysis

• Tang L, Wang F, Li Y, et al. Ultrasound guidance for radial artery catheterization: a systematic review and meta-analysis of randomized controlled trials. PLoS One. 2014;9(1):e111527.
• Cavallaro F, Sandroni C, Marano C, et al. Diagnostic accuracy of passive leg raising for prediction of fluid responsiveness in adults: systematic review and meta-analysis of clinical studies. Intensive Care Med. 2010;36:1475-1483.
• van Walraven C, Naylor CD. Do we know what inappropriate laboratory utilization is? A systematic review of laboratory clinical audits. JAMA. 1998;280(6):550-558.
• Scheer B, Perel A, Pfeiffer UJ. Clinical review: complications and risk factors of peripheral arterial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine. Crit Care. 2002;6(3):199-204. (DOI: Not reported)
• Lakhal K, Ehrmann S, Boulain T. Noninvasive blood pressure monitoring in the critically ill: time to abandon the arterial catheter? Chest. 2018;153:1023-1039. (DOI: Not reported)

Observational Studies

• Martin C, Saux P, Papazian L, Gouin F. Long-term arterial cannulation in ICU patients using the radial artery or dorsalis pedis artery. Chest. 2001;119(3):901-906.
• Low LL, Harrington GR, Stoltzfus DP. The effect of arterial lines on blood-drawing practices and costs in intensive care units. Chest. 1995;108(1):216-219.
• Corwin HL, Parsonnet KC, Gettinger A. RBC transfusion in the ICU. Is there a reason? Chest. 1995;108(3):767-771.
• Gershengorn HB, Wunsch H, Scales DC, Zarychanski R, Rubenfeld G, Garland A. Association between arterial catheter use and hospital mortality in intensive care units. JAMA Intern Med. 2014;174(11):1746-1754. (DOI: Not reported)
• Lucet J-C, Bouadma L, Zahar J-R, et al. Infectious risk associated with arterial catheters compared with central venous catheters. Crit Care Med. 2010;38(4):1030-1035. (DOI: Not reported)

Guidelines

• Evans L, Rhodes A, Alhazzani W, et al. Surviving sepsis campaign: international guidelines for management of sepsis and septic shock 2021. Intensive Care Med. 2021;47:1181-1247.
• Rhodes A, Evans LE, Alhazzani W, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock: 2016. Intensive Care Med. 2017;43(3):304-377.
• Mermel LA, Allon M, Bouza E, et al. Clinical practice guidelines for the diagnosis and management of intravascular catheter-related infection: 2009 update by the Infectious Diseases Society of America. Clin Infect Dis. 2009;49(1):1-45.
• Antonelli M, Levy M, Andrews PJD, et al. Hemodynamic monitoring in shock and implications for management: International Consensus Conference, Paris, France, 27-28 April 2006. Intensive Care Med. 2007;33:575-590. (DOI: Not reported)
• Lévy B, Bastien O, Karim B, et al. Experts' recommendations for the management of adult patients with cardiogenic shock. Ann Intensive Care. 2015;5(1):52. (DOI: Not reported)

Notes

• Where a DOI is labelled “Not reported”, it was not available in the provided sources; no DOI has been inferred.

Overall Takeaway

EVERDAC is a landmark de-implementation trial in haemodynamic monitoring: it shows that, for many ICU patients with vasopressor-treated shock, routine early arterial catheterisation is not required to preserve survival when a structured noninvasive-first strategy with rescue criteria is used. Its practical impact lies in quantifying the trade-offs—fewer invasive-catheter complications and lower catheter-related infection rates, balanced against increased arterial puncture attempts and modestly higher monitoring-related discomfort—thereby enabling more selective, patient-tailored arterial line use.

Bibliography

• Muller G, Kamel T, Contou D, et al. Early versus differed arterial catheterisation in critically ill patients with acute circulatory failure: a multicentre, open-label, pragmatic, randomised, non-inferiority controlled trial: the EVERDAC protocol. BMJ Open. 2021;11(9):e044719.
• Dodd M, Carpenter J, Thompson JA, Williamson E, Fielding K, Elbourne D. Assessing efficacy in non-inferiority trials with non-adherence to interventions: Are intention-to-treat and per-protocol analyses fit for purpose? Stat Med. 2024;43(12):2314-2331.
• Hernández G, Ospina-Tascón GA, Damiani LP, et al. Effect of a resuscitation strategy targeting peripheral perfusion status vs serum lactate levels on 28-day mortality among patients with septic shock: the ANDROMEDA-SHOCK randomized clinical trial. JAMA. 2019;321(7):654-664.
• Tang L, Wang F, Li Y, et al. Ultrasound guidance for radial artery catheterization: a systematic review and meta-analysis of randomized controlled trials. PLoS One. 2014;9(1):e111527.