Publication

• Title: Personalized Hemodynamic Resuscitation Targeting Capillary Refill Time in Early Septic Shock: The ANDROMEDA-SHOCK-2 Randomized Clinical Trial
• Acronym: ANDROMEDA-SHOCK-2
• Year: 2025
• Journal published in: JAMA
• Citation: The ANDROMEDA-SHOCK-2 Investigators for the ANDROMEDA Research Network, Spanish Society of Anesthesiology, Reanimation and Pain Therapy (SEDAR), and Latin American Intensive Care Network (LIVEN). Personalized Hemodynamic Resuscitation Targeting Capillary Refill Time in Early Septic Shock: The ANDROMEDA-SHOCK-2 Randomized Clinical Trial. JAMA. 2025;334(22):1988-1999.

Context & Rationale

• Background

  • Early septic shock resuscitation aims to restore tissue perfusion, but indiscriminate fluids and vasoactives can be harmful (e.g., fluid overload, excessive vasoconstriction, inotrope toxicity).
  • Common resuscitation targets (e.g., lactate trends) are biologically non-specific and may lag behind real-time changes in microcirculatory perfusion.
  • Capillary refill time (CRT) is a rapid, low-cost bedside marker of peripheral perfusion and has prognostic associations in critical illness.
  • Prior interventional work suggested CRT-targeted resuscitation could reduce organ dysfunction and might reduce mortality versus lactate-targeting, but uncertainty remained due to sample size and imprecision.
• Research Question/Hypothesis

  • Whether a protocolised, haemodynamic phenotype-based resuscitation strategy targeting normal CRT (≤3 seconds) during the first 6 hours of early septic shock improves patient-centred outcomes versus usual care.
• Why This Matters

  • If beneficial, CRT could provide a globally accessible resuscitation target that responds quickly to interventions and may reduce unnecessary fluid loading.
  • A structured approach integrating arterial tone (diastolic pressure), fluid responsiveness, and cardiac function could operationalise “personalised” haemodynamic resuscitation early in shock.
  • Sepsis resuscitation remains an area where pragmatic bedside endpoints with clear biological meaning are attractive but require rigorous validation at scale.

Design & Methods

• Research Question: In adults with early septic shock, does a 6-hour CRT-targeted personalised haemodynamic resuscitation strategy improve a hierarchical composite outcome (28-day mortality, duration of vital support, and hospital length of stay) compared with usual care?
• Study Type:
  • Multicentre, international, investigator-initiated, parallel-group, randomised clinical trial (open-label).
  • 86 ICUs across 19 countries; enrolment March 2022 to April 2025; follow-up to 90 days.
  • 1:1 allocation; stratified win ratio analysis prespecified in a published protocol and statistical analysis plan¹².
• Population:
  • Setting: Adult ICU patients (with recruitment possible from ICU/ED/ward/OR pathways depending on local workflow) meeting septic shock criteria within a defined early window.
  • Key inclusion criteria: Suspected/confirmed infection; lactate ≥2 mmol/L; norepinephrine infusion required to maintain MAP >65 mm Hg after receiving ≥1000 mL of fluids; enrolment within 4 hours of septic shock onset.
  • Key exclusions (trial-level): >4 hours after onset; anticipated surgery or acute dialysis within 6 hours; life expectancy <90 days; refractory shock; do-not-resuscitate status; Child-Pugh B/C cirrhosis; severe ARDS; active bleeding; pregnancy; CRT not reliably assessable (e.g., peripheral vascular disease, hypothermia, darker skin tone, Raynaud syndrome).
• Intervention:
  • CRT measurement: Firm pressure to ventral distal phalanx using a glass slide for 10 seconds, then release; time to return of normal colour measured with a chronometer; abnormal defined as >3 seconds.
  • Protocol window: 6-hour sequential algorithm targeting CRT normalisation (≤3 seconds), with hourly reassessment when CRT normalised.
  • Tier 1 (initial haemodynamic phenotype): Pulse pressure <40 mm Hg triggered fluid-responsiveness assessment and up to two 500 mL boluses (maximum 1000 mL) in fluid-responsive patients; pulse pressure ≥40 mm Hg with diastolic arterial pressure <50 mm Hg triggered norepinephrine titration to diastolic pressure ≥50 mm Hg (1-hour test).
  • Tier 2 (if CRT remained abnormal): Basic echocardiography to assess cardiac dysfunction (with general treatment recommendations recorded); reassessment of fluid responsiveness and additional 500 mL boluses in fluid-responsive patients until CRT normalised, fluid responsiveness became negative, or safety limits reached.
  • Pressure and inotrope tests: In chronic hypertensives, a transient MAP test to 80–85 mm Hg for 1 hour (maintained if CRT normalised); dobutamine test at 5 μg/kg/min for 1 hour (continued only if CRT normalised).
  • Clinical override and rescue: Treating teams could override protocol steps if judged in the patient’s best interest; rescue therapies allowed if CRT remained abnormal at protocol completion.
• Comparison:
  • Usual care resuscitation according to local protocols and/or international guidelines; fluid-responsiveness assessment and echocardiography allowed but not mandated.
  • CRT measurement requested at baseline and 6 hours (no mandated CRT-driven algorithm); after 6 hours, management in both arms was clinician-directed.
• Blinding: Open-label (no blinding of clinicians or participants); objective outcomes (mortality) contrasted with clinician-influenced outcomes (organ support duration and length of stay).
• Statistics: Power calculation targeted a win ratio of 1.15 with α=0.05 and 90% power, requiring ~1500 participants; primary analysis used a stratified win ratio (intention-to-treat framework as prespecified) with hierarchical ordering of outcome components².
• Follow-Up Period: Primary outcome through day 28; follow-up to day 90 for mortality and selected tertiary outcomes.

Key Results

This trial was not stopped early. Completed enrolment and follow-up per protocol.

Outcome	CRT-PHR	Usual care	Effect	p value / 95% CI	Notes
Primary hierarchical composite (28-day mortality → duration of vital support → hospital length of stay; win ratio)	131,131 wins (48.9%)	113,117 wins (42.2%)	Win ratio 1.16	95% CI 1.02 to 1.33; P=0.04	Higher favours CRT-PHR; ties 23,113 (8.6%)
All-cause mortality at 28 days	191/720 (26.5%)	198/745 (26.6%)	HR 0.99	95% CI 0.81 to 1.21; P=0.91	No mortality signal at day 28
Duration of vital support within 28 days (days)	Mean 5.9 ± 7.4; median 3.0 (0.0–8.7)	Mean 7.2 ± 8.3; median 4.0 (0.0–10.6)	Mean difference −1.31	95% CI −2.15 to −0.46	Component of primary composite; lower favours CRT-PHR
Vital support–free days within 28 days (days)	Mean 16.5 ± 11.3; median 23 (0–25)	Mean 15.4 ± 11.4; median 22 (0–25)	pOR 1.28	95% CI 1.06 to 1.54	Higher favours CRT-PHR; P not reported
Hospital length of stay up to 28 days (days)	Mean 15.3 ± 10.3; median 12.0 (7.0–22.0)	Mean 16.2 ± 10.5; median 13.0 (7.0–23.0)	Mean difference −0.85	95% CI −1.80 to 0.10	Component of primary composite; truncated at day 28
ICU length of stay up to 28 days (days)	Mean 7.9 ± 5.9; median 6.0 (4.0–10.0)	Mean 8.8 ± 6.3; median 6.0 (4.0–11.0)	Mean difference −1.02	95% CI −1.82 to −0.23	Tertiary outcome; truncated at day 28
All-cause mortality at 90 days	231/719 (32.1%)	247/744 (33.2%)	HR 0.93	95% CI 0.78 to 1.11	P not reported
Resuscitation fluids during first 6 hours (mL)	928 ± 684	1179 ± 846	Mean difference −251	95% CI −312 to −190	Demonstrated protocol-level separation
Inotrope use during first 6 hours	70/720 (9.7%)	27/747 (3.6%)	Difference 6.1%	95% CI 3.3 to 8.9	Greater early inotrope exposure with CRT-PHR
Protocol-related serious adverse events	Not reported	Not reported	Not reported	Not reported	Safety reporting focused on clinical outcomes rather than adjudicated protocol harms

• The statistically significant primary win ratio result (1.16; 95% CI 1.02 to 1.33; P=0.04) occurred without any detectable 28-day mortality difference (HR 0.99; 95% CI 0.81 to 1.21; P=0.91).
• The composite advantage aligned with shorter vital support duration (mean difference −1.31 days; 95% CI −2.15 to −0.46) and shorter ICU stay (mean difference −1.02 days; 95% CI −1.82 to −0.23), rather than hospital length of stay up to day 28 (mean difference −0.85 days; 95% CI −1.80 to 0.10).
• Prespecified subgroup analyses suggested larger benefit in higher-severity strata (e.g., SOFA ≥10 win ratio 1.475; 95% CI 1.166 to 1.863; lactate ≥4 win ratio 1.390; 95% CI 1.134 to 1.700; baseline CRT >3 seconds win ratio 1.186; 95% CI 1.006 to 1.416), but these remain exploratory and not multiplicity-adjusted.

Internal Validity

• Randomisation and allocation concealment: Central web-based randomisation with stratification by centre and variable block sizes; allocation concealment was structurally robust.
• Baseline comparability (selected markers): APACHE II median 19 (IQR 13–26) vs 18 (13–26); SOFA median 9 (IQR 6–11) vs 9 (6–12); lactate median 3.70 vs 3.62 mmol/L; norepinephrine dose mean 0.22 ± 0.19 vs 0.22 ± 0.19 μg/kg/min; pre-randomisation fluids median 1500 mL in both groups.
• Post-randomisation exclusions / missingness: 1501 randomised; 1467 included in the primary outcome analysis (CRT-PHR 720; usual care 747); excluded after randomisation due to consent withdrawal and ineligibility (CRT-PHR: 14 withdrew consent, 10 ineligible; usual care: 6 withdrew consent, 4 ineligible); loss to follow-up at day 28 was 14 vs 6.
• Performance/detection bias: Open-label care introduces risk of co-intervention and decision-driven outcomes (duration of organ support and discharge timing), although mortality is objective.
• Protocol adherence: In the CRT-PHR arm, protocol violations were 44 and protocol deviations 111; 262/720 (36.4%) had normal CRT throughout the 6-hour intervention window and therefore received no protocol-driven interventions; 458/720 (63.6%) entered the algorithm with abnormal CRT.
• Timing and delivery: Intervention was delivered within a prespecified early shock window (≤4 hours from onset) and executed over a 6-hour protocol period; post–6-hour care was clinician-directed in both groups.
• Separation of the variable of interest (6-hour process differences):
  • Resuscitation fluids: 928 ± 684 mL vs 1179 ± 846 mL (mean difference −251 mL; 95% CI −312 to −190).
  • Inotrope exposure: 70/720 (9.7%) vs 27/747 (3.6%) (difference 6.1%; 95% CI 3.3 to 8.9).
  • Capillary refill time: 2.80 ± 1.42 s vs 3.37 ± 1.91 s (mean difference −0.55 s; 95% CI −0.71 to −0.38).
  • Lactate: 3.16 ± 2.39 vs 3.45 ± 2.99 mmol/L (mean difference −0.30; 95% CI −0.53 to −0.08).
  • Norepinephrine dose: 0.28 ± 0.25 vs 0.24 ± 0.22 μg/kg/min (mean difference 0.04; 95% CI 0.02 to 0.05).
• Adjunctive therapies (within 6 hours): Vasopressin use was 274/684 (40.1%) vs 239/694 (34.4%) (difference 5.7%; 95% CI 0.1 to 11.3); corticosteroid use was similar (343/720 [47.6%] vs 362/747 [48.5%]).
• Outcome assessment and definitions: Mortality was objective; “duration of vital support” and “length of stay” were prespecified and truncated (day 28), but can be influenced by local practice patterns (weaning, extubation timing, discharge thresholds).
• Statistical rigour: Sample size was achieved; primary analysis used prespecified stratified win ratio; secondary outcomes were tested in hierarchical order to address multiplicity.

Conclusion on Internal Validity: Overall, internal validity appears moderate-to-strong, supported by concealed randomisation, large sample size, and demonstrable early protocol separation; it is limited chiefly by the open-label design, post-randomisation exclusions, and reliance on clinician-influenced components within the hierarchical composite.

External Validity

• Population representativeness: Participants reflect ICU adults with early, vasopressor-dependent septic shock (lactate ≥2 mmol/L) after at least 1 L of fluids, but many common real-world scenarios were excluded (late presentation beyond 4 hours, anticipated surgery/dialysis, severe ARDS, advanced cirrhosis, active bleeding, pregnancy, do-not-resuscitate orders).
• CRT feasibility constraints: Patients in whom CRT could not be accurately assessed (e.g., peripheral vascular disease, hypothermia, darker skin tone, Raynaud syndrome) were excluded, potentially limiting applicability across diverse populations and environments.
• Setting and systems: Multi-national recruitment across 86 ICUs in 19 countries supports broad applicability to ICU settings; however, local practice patterns (organ support thresholds, discharge processes) may influence duration-based outcomes.
• Intervention complexity and resources: Translation requires clinician training in standardised CRT technique, the ability to perform fluid-responsiveness assessments, and access to basic echocardiography; the feasibility and fidelity of the full algorithm may be lower in resource-limited units or outside the ICU (e.g., ED, prehospital).

Conclusion on External Validity: Generalisability is moderate for ICU patients with early septic shock where CRT can be standardised and haemodynamic testing is feasible; it is more limited for later shock, settings without consistent echocardiography/FR tools, and populations in whom CRT is unreliable or difficult to interpret.

Strengths & Limitations

• Strengths:
  • Large, multinational pragmatic RCT (1501 randomised) in early septic shock, addressing an important bedside resuscitation question.
  • Clear biological target (CRT) with standardised measurement technique and structured haemodynamic decision pathway.
  • Demonstrated protocol-level separation (less early fluids, more targeted vasoactive/inotrope use, faster CRT normalisation).
  • Prepublication of protocol and statistical analysis plan supports transparency and interpretability¹².
• Limitations:
  • Open-label design with a primary endpoint partly driven by clinician-influenced outcomes (duration of vital support and length of stay).
  • Post-randomisation exclusions and loss to follow-up may introduce bias, particularly with imbalance across groups.
  • Magnitude of fluid-volume separation over 6 hours was modest; the intervention also increased inotrope and vasopressin exposure, with limited trial-level detail on protocol-attributable harms.
  • CRT assessability exclusions and the need for protocol training may reduce real-world applicability, particularly in darker skin tones and settings with variable ambient temperature control.

Interpretation & Why It Matters

• Clinical meaning

  • CRT-PHR improved a morbidity-weighted hierarchical composite outcome, but did not reduce 28-day mortality, suggesting any benefit is likely mediated through earlier resolution of organ support needs rather than survival.
  • The component pattern is consistent with a “faster shock recovery” signal (shorter vital support and ICU stay), while hospital length of stay up to day 28 was not clearly reduced.
• Mechanistic plausibility

  • CRT is a rapidly responsive peripheral perfusion marker; the intervention produced earlier CRT improvement alongside modest reductions in lactate and fluid delivery and increased tailored vasoactive/inotrope use.
  • The algorithm operationalised a phenotype-based approach: arterial tone (diastolic pressure), preload responsiveness, and ventricular function were used to select therapy.
• Practice integration

  • For clinicians, ANDROMEDA-SHOCK-2 supports incorporating CRT into early reassessment, particularly when attempting to limit additional fluids after initial loading and vasopressor initiation.
  • However, the absence of mortality benefit and the reliance on duration-based outcomes mean implementation should be coupled with attention to measurement standardisation, safety monitoring, and local weaning/discharge practices.

Controversies & Subsequent Evidence

• Primary outcome interpretation (win ratio hierarchical composite): The statistically significant win ratio (driven by morbidity components) despite no mortality signal emphasises the importance of understanding how organ-support discontinuation and discharge processes can influence “patient-centred” composites in open-label ICU trials²³.
• Open-label delivery and clinician behaviour: Protocolising early haemodynamics may alter downstream care decisions (e.g., comfort with weaning/step-down), potentially affecting organ-support duration and ICU/hospital stay—outcomes that are partly decision-mediated even when definitions are prespecified³.
• CRT measurement reliability and reproducibility: CRT is sensitive to technique and context (pressure duration, ambient temperature, vasoconstriction, pigmentation); ANDROMEDA-SHOCK-2 mandated standardised measurement and training, but real-world reproducibility—especially across diverse populations—remains a key implementation challenge¹⁴.
• Therapeutic trade-offs: CRT-PHR reduced early fluids but increased early exposure to inotropes and vasopressin; the net benefit-risk balance (including arrhythmia/ischaemia risk) is important for bedside translation, particularly where monitoring resources vary³.
• Consistency with prior trial signals: A Bayesian reanalysis of the original ANDROMEDA-SHOCK trial suggested a high posterior probability of benefit for peripheral perfusion targeting, supporting biological plausibility but underscoring why a larger confirmatory trial was needed⁵.
• Evidence synthesis: A systematic review and network meta-analysis of parameter-guided initial resuscitation suggested peripheral perfusion–guided strategies may improve outcomes compared with some alternatives, but the evidence base is small and certainty remains limited, making ANDROMEDA-SHOCK-2 highly influential in the field⁶.
• Guideline positioning: Recent haemodynamic monitoring recommendations emphasise integrating clinical perfusion assessment (including CRT) with biochemical and haemodynamic data, while sepsis guidelines continue to prioritise lactate as a marker of hypoperfusion; formal endorsement of CRT-targeted protocols will likely depend on replication and implementation evidence beyond a single large RCT⁷⁸.

Summary

• In 1501 adults with early septic shock across 86 ICUs, CRT-targeted personalised haemodynamic resuscitation over 6 hours improved a hierarchical composite of 28-day mortality, vital support duration, and hospital length of stay (win ratio 1.16; 95% CI 1.02 to 1.33; P=0.04).
• There was no evidence of reduced mortality at day 28 (26.5% vs 26.6%; HR 0.99; 95% CI 0.81 to 1.21; P=0.91) or day 90 (HR 0.93; 95% CI 0.78 to 1.11).
• Benefits aligned with shorter vital support duration (mean difference −1.31 days; 95% CI −2.15 to −0.46) and shorter ICU stay (mean difference −1.02 days; 95% CI −1.82 to −0.23), with hospital length of stay up to day 28 not clearly changed.
• Protocol separation was achieved via less early fluid administration (mean difference −251 mL; 95% CI −312 to −190) and greater early inotrope exposure (difference 6.1%; 95% CI 3.3 to 8.9), alongside faster CRT improvement.
• Subgroup findings suggested larger effects in more severe shock (e.g., SOFA ≥10; lactate ≥4), but these are exploratory and should not be used to restrict application without further confirmation.

Further Reading

Other Trials

• 2019Hernandez 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.
• 2010Jansen TC, van Bommel J, Schoonderbeek FJ, et al. Early lactate-guided therapy in intensive care unit patients: a multicenter, open-label, randomized controlled trial. Am J Respir Crit Care Med. 2010;182(6):752-761.
• 2014Yealy DM, Kellum JA, Huang DT, et al. A Randomized Trial of Protocol-Based Care for Early Septic Shock. N Engl J Med. 2014;370(18):1683-1693.
• 2014Peake SL, Delaney A, Bailey M, et al. Goal-Directed Resuscitation for Patients with Early Septic Shock. N Engl J Med. 2014;371(16):1496-1506.
• 2015Mouncey PR, Osborn TM, Power GS, et al. Trial of Early, Goal-Directed Resuscitation for Septic Shock. N Engl J Med. 2015;372(14):1301-1311.

Systematic Review & Meta Analysis

• 2024Wang H, Zhang Y, Chen D, et al. Prolonged capillary refill time and short-term mortality of critically ill patients: a systematic review and meta-analysis. Am J Emerg Med. 2024.
• 2023Yumoto T, Watanabe A, Naito H, et al. Clinical parameter-guided initial resuscitation in adult patients with sepsis or septic shock: a systematic review and network meta-analysis. Acute Med Surg. 2023;10(1):e914.
• 2023Jacquet-Lagrèze M, Pernollet MG, Kattan E, et al. Prognostic value of capillary refill time: a systematic review and meta-analysis. Crit Care. 2023.
• 2019Pan J, Peng M, Liao C, Hu X, Wang A, Li X. Relative efficacy and safety of lactate clearance-guided therapy in sepsis: a meta-analysis. Medicine (Baltimore). 2019;98(8):e14453.
• 2015Gu WJ, Zhang Z, Bakker J. Early lactate clearance-guided therapy in patients with sepsis: a meta-analysis. Intensive Care Med. 2015.

Observational Studies

• 2024Hernandez G, et al. Effects of peripheral perfusion-targeted fluid challenges and vasopressor tests in septic shock: a proof-of-concept observational study. Ann Intensive Care. 2024;14(1):49.
• 2022Contreras M, et al. Changes in capillary refill time after fluid resuscitation in septic shock: association with outcomes. J Clin Monit Comput. 2022.
• 2017Lara B, Enberg L, Ortega M, et al. Capillary refill time during septic shock resuscitation: a prospective observational study. PLoS One. 2017;12(10):e0188548.
• 2014Ait-Oufella H, Bige N, Boelle PY, et al. Capillary refill time exploration during septic shock. Intensive Care Med. 2014;40(7):958-964.
• 2009Lima A, Jansen TC, van Bommel J, Ince C, Bakker J. The prognostic value of the capillary refill time test in critically ill patients: a prospective multicenter study. Crit Care Med. 2009;37(3):934-939.

Guidelines

• 2025Monnet X, et al. Circulatory shock and hemodynamic monitoring: recommendations from the European Society of Intensive Care Medicine. Intensive Care Med. 2025.
• 2025Shime N, et al. Japanese clinical practice guidelines for management of sepsis and septic shock 2024 (J-SSCG 2024). Acute Med Surg. 2025.
• 2025Brunkhorst FM, et al. S3 guideline update: Sepsis – prevention, diagnosis, therapy and aftercare. Med Klin Intensivmed Notfmed. 2025.
• 2021Evans 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(11):1181-1247.
• 2016National Institute for Health and Care Excellence. Sepsis: recognition, diagnosis and early management (NG51). NICE guideline. 2016 (updated).

Notes

• Further Reading links open to DOI landing pages where available; guideline links may point to official guideline landing pages when no DOI exists.

Overall Takeaway

ANDROMEDA-SHOCK-2 is the largest randomised evaluation of CRT-targeted, phenotype-based haemodynamic resuscitation in early septic shock. It demonstrated a modest but statistically significant improvement in a hierarchical composite outcome driven by shorter organ support and ICU stay, without evidence of reduced mortality. The trial strengthens the case for integrating bedside peripheral perfusion assessment into early resuscitation while highlighting ongoing debate about composite endpoints, open-label care effects, and the practical reproducibility of CRT measurement at scale.

Bibliography

• 1.Kattan E, Bakker J, Estenssoro E, et al. Hemodynamic phenotype-based, capillary refill time-targeted resuscitation in early septic shock: study protocol for the ANDROMEDA-SHOCK-2 randomized clinical trial. Rev Bras Ter Intensiva. 2022;34(1):96-106.
• 2.Orozco N, García-Gallardo G, Cavalcanti AB, et al. Statistical analysis plan for hemodynamic phenotype-based, capillary refill time-targeted resuscitation in early septic shock: the ANDROMEDA-SHOCK-2 randomized clinical trial. Crit Care Sci. 2025;37:e20250140.
• 3.Machado FR, Semler MW. Capillary refill time in sepsis—searching for the holy grail. JAMA. 2025;334(22):1983-1985.
• 4.Hernandez G, Teboul JL. Monitoring capillary refill time in septic shock: current evidence and future directions. Intensive Care Med. 2024;50(4):580-582.
• 5.Zampieri FG, Damiani LP, Bakker J, et al. Effects of a resuscitation strategy targeting peripheral perfusion status versus serum lactate levels among patients with septic shock: a Bayesian reanalysis of the ANDROMEDA-SHOCK trial. Am J Respir Crit Care Med. 2020;201(4):423-429.
• 6.Yumoto T, Watanabe A, Naito H, et al. Clinical parameter-guided initial resuscitation in adult patients with sepsis or septic shock: a systematic review and network meta-analysis. Acute Med Surg. 2023;10(1):e914.
• 7.Monnet X, et al. Circulatory shock and hemodynamic monitoring: recommendations from the European Society of Intensive Care Medicine. Intensive Care Med. 2025.
• 8.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(11):1181-1247.