👉👉 🇺🇸 All Posts 🇬🇧 / 🇯🇵 記事一覧 🇯🇵 👈👈
👉👉 🇺🇸 All Posts 🇬🇧 / 🇯🇵 記事一覧 🇯🇵 👈👈
✨ Introduction: A New Era in Critical Care Assessment
The severity and trajectory of organ dysfunction fundamentally shape outcomes in critically ill patients, influencing everything from treatment intensity to family discussions about prognosis.
For nearly three decades, the Sequential Organ Failure Assessment (SOFA) score has served as the international standard for describing multi-organ dysfunction in intensive care units. Yet the landscape of critical care has transformed dramatically since the original SOFA score emerged in 1996. Technologies rarely used or unavailable three decades ago—high-flow nasal cannula, veno-venous extracorporeal membrane oxygenation (ECMO), continuous renal replacement therapy (CRRT), and advanced mechanical circulatory support—are now routine components of organ support strategies worldwide.
SOFA-2, published in 2025, represents the culmination of a three-year international effort to modernize this foundational assessment tool. Through a modified Delphi process involving 60 international experts and rigorous validation in over 3.3 million ICU encounters across nine countries and 1,319 intensive care units, the new score updates thresholds, clarifies scoring rules for contemporary organ support technologies, and maintains usability across diverse resource settings.
Critically, SOFA-2 remains what the original was designed to be: a descriptive severity score that grades organ dysfunction consistently, not an individualized mortality prediction calculator like APACHE or SAPS.
Although SOFA-2 is expected to become the international standard in future guideline updates, the original SOFA score (1996) remains the predominant reference in current clinical practice as of 2025.
What you’ll learn in this article:
- The scientific foundation and validation process behind SOFA-2
- System-by-system updates with specific threshold changes
- Practical guidance for bedside implementation and electronic health record integration
- How SOFA-2 addresses special situations including sedation, ECMO, chronic organ dysfunction, and missing data
- Key examination points and clinical applications
♦️ Background: Why SOFA Needed an Update
The original SOFA score, developed in 1996 by Vincent and colleagues under the auspices of the European Society of Intensive Care Medicine, represented a breakthrough in standardizing the assessment of organ dysfunction. Its six-system framework—central nervous, respiratory, cardiovascular, hepatic, renal, and coagulation—each scored 0-4 points, could be tracked sequentially to capture the dynamic nature of critical illness. This approach proved remarkably durable, becoming embedded in clinical practice, research protocols, and even formal definitions of sepsis (Sepsis-3 criteria).
However, over three decades, the gap between clinical practice and the score’s ability to capture that practice widened considerably:
🔹 Modern organ support technologies:
Advanced respiratory support modalities including HFNC and noninvasive ventilation became standard care, yet the score provided no explicit guidance for incorporating them.
Similarly, extracorporeal technologies—ECMO for refractory respiratory or circulatory failure, CRRT for acute kidney injury, mechanical circulatory support devices—lacked clear representation despite their profound physiologic impact and prognostic implications.
🔹 Scoring ambiguity:
The original score left room for interpretation in several areas. “Vasopressor” doses were not standardized across different agents, creating inconsistencies in cardiovascular scoring between centers using different catecholamine protocols.
The handling of sedated patients, those with chronic organ dysfunction, and missing data points generated variable approaches that reduced score comparability.
🔹 Global applicability concerns:
In low- and middle-income countries where arterial blood gas analysis or advanced hemodynamic monitoring may be unavailable or prohibitively expensive, rigid adherence to specific measurements limited the score’s utility. A modernized tool needed validated alternative variables that preserved discriminative ability without compromising accessibility.
Evidence supporting the need for revision came from multiple sources. Mortality rates in sepsis and other forms of critical illness declined substantially between 1996 and 2025, suggesting that threshold values calibrated to 1990s outcomes might no longer optimally stratify contemporary risk.
Data-driven explorations using modern machine learning techniques on large ICU databases identified potential improvements in variable selection and cutpoint optimization.
Perhaps most compelling, the simple clinical observation that the score sometimes failed to capture the severity signaled by need for advanced organ support—a patient on VA-ECMO might score only modestly elevated if other variables happened to normalize—indicated room for improvement.
🎓 Exam Focus: Be prepared to explain why SOFA-1 (1996) required updating: incorporation of modern organ support technologies, resolution of scoring ambiguities, and improved global applicability through alternative variables.
♦️Development Process: From Expert Consensus to Data Validation
SOFA-2 was developed through a modified Delphi process involving ~60 international intensivists, followed by validation in over 3.3 million ICU encounters across nine countries. The panel prioritized both predictive accuracy and clinical interpretability. For example, bilirubin was retained over AST in the liver domain—not for statistical reasons alone, but for global availability, specificity, and bedside practicality—reflecting SOFA-2’s emphasis on content validity as well as performance.
💡 Clinical Pearl: Always ask not only “how well does it predict?” but “can we measure it reliably, and does it make physiologic sense?”
♦️ The SOFA-2 Score: System-by-System Updates
🧠 Brain (formerly “Central Nervous System”)
The brain system continues to use the Glasgow Coma Scale (GCS) but with refined cutoffs and explicit guidance for common clinical scenarios that previously generated scoring inconsistency.
Scoring thresholds: GCS scores map to SOFA-2 brain points as follows:
| Score | Criteria |
|---|---|
| 0 | GCS 15 |
| 1 | GCS 13–14, or use of medications for delirium management |
| 2 | GCS 9–12 |
| 3 | GCS 6–8 |
| 4 | GCS 3–5 |
🔹 Delirium acknowledgment:
Use of medications to treat delirium (e.g., antipsychotics such as haloperidol or quetiapine) contributes 1 point even when GCS equals 15. This recognizes that neuropsychiatric dysfunction requiring pharmacologic intervention represents clinically meaningful organ impairment, even if gross consciousness remains intact.
🔹 Sedation guidance:
When patients receive sedative medications, score the pre-sedation GCS documented before sedative administration or escalation. If pre-sedation GCS is unknown, assign 0 points but document the limitation. Importantly, avoid “therapeutic scoring”—titrating sedation with the goal of manipulating the SOFA-2 score rather than optimizing patient comfort and safety.
🎓 Exam Focus: The standard is “pre-sedation GCS.” In practice, this means documenting neurologic examination before inducing anesthesia for intubation or before escalating sedative infusions.
🫁 Respiratory System
Respiratory system scoring underwent substantial revision to incorporate the full spectrum of modern respiratory support and provide validated alternatives when arterial blood gas analysis is unavailable.
PaO₂:FiO₂ (P/F) ratio thresholds: The updated cutoffs are:
| Score | Criteria |
|---|---|
| 0 | PaO₂/FiO₂ > 300 |
| 1 | PaO₂/FiO₂ < 300 |
| 2 | PaO₂/FiO₂ ≤ 225 + advanced respiratory support |
| 3 | PaO₂/FiO₂ ≤ 155 + advanced respiratory support |
| 4 | PaO₂/FiO₂ ≤ 75 + advanced respiratory support, or ECMO |
🔹 Advanced respiratory support requirement:
To score 3 or 4 points, patients must receive advanced respiratory support: high-flow nasal cannula (HFNC), continuous or bilevel positive airway pressure (CPAP/BiPAP), noninvasive ventilation (NIV), or invasive mechanical ventilation (IMV). This requirement acknowledges that hypoxemia severe enough to warrant such interventions represents greater dysfunction than comparable P/F ratios in spontaneously breathing patients without support.
🔹 ECMO consideration:
When ECMO is employed for respiratory failure (whether venovenous or venoarterial configurations used primarily for oxygenation support), the respiratory system automatically receives 4 points, reflecting the severity of dysfunction necessitating extracorporeal gas exchange.
🔹 SpO₂:FiO₂ (S/F) ratio alternative:
When arterial blood gas measurement is unavailable, contraindicated, or cost-prohibitive, SpO₂:FiO₂ ratio provides a validated substitute. Research demonstrates that S/F thresholds of approximately 235, 190, 143, and 89 correspond reasonably to P/F cutoffs of 300, 225, 150, and 75 respectively. For practical bedside use, simplified S/F thresholds of 300, 250, 200, and 120 may be employed.
⚠️ Critical limitation: Use S/F ratio only when SpO₂ < 98% to avoid ceiling effects—once saturation reaches 98-100%, the S/F ratio no longer discriminates varying PaO₂ values due to the sigmoid shape of the oxyhemoglobin dissociation curve.
💡 Clinical Pearl: Consider posting a P/F to S/F conversion reference at bedside workstations for rapid scoring when arterial blood gases are unavailable. This particularly aids resource-limited settings and reduces unnecessary blood draws.
🫀 Cardiovascular System
Cardiovascular scoring standardizes vasopressor dosing terminology and explicitly incorporates mechanical circulatory support technologies that have become increasingly prevalent.
Unified catecholamine dosing: Sum the doses of norepinephrine and epinephrine, both expressed expressed as base, in µg/kg/min, to calculate a combined catecholamine score:
| Score | Criteria |
|---|---|
| 0 | MAP ≥ 70 mmHg, no vasopressors required |
| 1 | MAP < 70 mmHg |
| 2 | NE + Epi ≤ 0.2 µg/kg/min |
| 3 | NE + Epi > 0.2 to ≤ 0.4 µg/kg/min, or use of MCS |
| 4 | NE + Epi > 0.4 µg/kg/min, or use of MCS |
This standardization eliminates previous ambiguity about whether and how to combine multiple vasopressor agents. Note that “base equivalent” matters—norepinephrine preparations may be supplied as base, tartrate, or bitartrate salts, requiring appropriate conversion for accurate scoring.
🔹 Dopamine considerations:
While contemporary practice favors norepinephrine-based vasopressor strategies, when dopamine is used as the sole or primary agent (increasingly rare), specific thresholds apply: doses ≤ 5, > 5 to ≤ 15, and > 15 µg/kg/min correspond to cardiovascular scores of 2, 3, and 4 points respectively.
🔹 Mechanical circulatory support inclusion:
Devices providing mechanical circulatory assistance—venoarterial ECMO (VA-ECMO), intra-aortic balloon pump (IABP), left ventricular assist devices (LVAD), Impella, or other percutaneous ventricular assist devices—explicitly contribute to higher cardiovascular scores, acknowledging the severity implied by need for such interventions.
🔹 Mean arterial pressure alternative:
In settings where vasopressors are unavailable, contraindicated, or their use is restricted, MAP thresholds alone provide a fallback scoring method, though with reduced sensitivity for detecting treated dysfunction.
🎓 Exam Focus: Memorize that norepinephrine and epinephrine doses sum to a combined catecholamine dose with cutoffs at 0.2 and 0.4 µg/kg/min (both as base equivalent). This reflects modern vasopressor practice where combination catecholamine therapy is common.
🤎 Liver System
Liver dysfunction assessment continues to use bilirubin but with updated thresholds that better stratify mortality risk across the severity spectrum.
Bilirubin thresholds (in mg/dL, with SI unit µmol/L equivalents):
| Score | Criteria |
|---|---|
| 0 | Bilirubin < 1.2 mg/dL |
| 1 | Bilirubin 1.2–<3 mg/dL |
| 2 | Bilirubin 3–<6 mg/dL |
| 3 | Bilirubin 6–<12 mg/dL |
| 4 | Bilirubin ≥ 12 mg/dL |
The decision to retain bilirubin rather than adopt alternative markers like AST (discussed earlier) reflects the balance between empirical performance, global measurability, and content validity—bilirubin specifically reflects hepatic excretory function and cholestasis, core components of liver dysfunction in critical illness.
💛 Kidney System
Renal assessment incorporates both serum creatinine and urine output with explicit guidance for renal replacement therapy (RRT) that resolves previous ambiguity.
Creatinine and urine output thresholds:
| Score | Criteria |
|---|---|
| 0 | Creatinine < 1.2 mg/dL |
| 1 | Creatinine 1.2–<2.0 mg/dL |
| 2 | Creatinine 2.0–<3.5 mg/dL, or urine output < 0.5 mL/kg/h for 12 hours |
| 3 | Creatinine ≥ 3.5 mg/dL, or urine output < 0.3 mL/kg/h for 24 hours, or anuria for 12 hours |
| 4 | RRT in progress, or meeting RRT criteria |
🔹 RRT criteria specification:
The score defines RRT criteria to standardize the 4-point assignment: creatinine > 1.2 mg/dL or oliguria (< 0.3 mL/kg/h for > 6 hours) PLUS either severe hyperkalemia (K⁺ ≥ 6.0 mmol/L) OR severe metabolic acidosis (pH ≤ 7.20 with HCO₃⁻ ≤ 12 mmol/L). This clarification enables consistent scoring even in settings where RRT is unavailable or its use is delayed for non-medical reasons.
🔹 Chronic RRT guidance:
Patients receiving maintenance dialysis for end-stage kidney disease present a scoring challenge. SOFA-2 provides explicit guidance to avoid inappropriately escalating scores solely due to chronic therapy—assess for acute-on-chronic deterioration rather than simply assigning 4 points for dialysis dependence.
⚠️ Warning: SOFA-2 RRT criteria describe severity for scoring purposes; they do NOT constitute treatment mandates. Initiate renal replacement therapy based on comprehensive clinical assessment including electrolyte derangements, acid-base status, volume overload, uremic symptoms, institutional protocols, and shared decision-making—never based solely on achieving a SOFA score threshold.
🩸 Hemostasis System (formerly “Coagulation”)
The hemostasis system retains platelet count as its primary variable but with refined thresholds that sharpen discrimination across the middle severity range.
Platelet count thresholds (× 10³/μL):
| Score | Criteria |
|---|---|
| 0 | Platelets ≥ 150 × 10³/µL |
| 1 | Platelets 100–149 × 10³/µL |
| 2 | Platelets 80–99 × 10³/µL |
| 3 | Platelets 50–79 × 10³/µL |
| 4 | Platelets < 50 × 10³/µL |
These updated cutoffs better correlate with bleeding risk states and consumptive coagulopathies such as disseminated intravascular coagulation. The tighter gradations in the 50-150 range capture clinically meaningful differences in hemostatic reserve and transfusion requirements.
💡 Clinical Pearl: Hemostasis scores often change more rapidly than other organ systems in conditions like sepsis-associated coagulopathy or massive transfusion scenarios. Frequent reassessment (every 6-12 hours during acute phases) may be warranted.
🙅 Excluded Organ Systems: Why Gastrointestinal and Immune Were Not Included
During SOFA-2 development, expert panels initially considered expanding beyond six systems to include gastrointestinal and immune function. After extensive deliberation and data analysis, both were ultimately excluded for methodologically sound reasons:
Gastrointestinal system:
Proposed variables lacked sufficient high-quality data across the federated databases to validate predictive utility. Candidate markers—including feeding intolerance, intra-abdominal pressure, gastrointestinal bleeding requiring transfusion—either showed inconsistent measurement or weak discriminative ability after adjustment for other organ systems. The expert consensus concluded that available gastrointestinal variables did not meet the threshold for inclusion in a universally applicable score.
Immune system:
White blood cell count and lymphocyte count demonstrated U-shaped relationships with mortality—both very low and very high values associated with increased risk. This created interpretability challenges: leukocytosis may represent appropriate inflammatory response to infection rather than immune “dysfunction” per se. Similarly, lymphopenia could reflect immunosuppression but also accompanies many stress states. The panel determined that these variables lacked sufficient content validity—they did not unambiguously represent immune organ “failure” in the way that oliguria represents kidney failure or hypoxemia represents respiratory failure.
Parsimony and universality:
Maintaining six well-defined, universally measurable systems preserves simplicity and facilitates global adoption across diverse resource settings. Adding poorly-validated domains would increase complexity without proportionate gains in discriminative performance or clinical utility.
🌐 International Note: Fewer, conceptually clear organ systems ease implementation in resource-limited settings where comprehensive laboratory panels and specialized monitoring may be unavailable or prohibitively expensive.
♦️ Clinical Performance and Validation
SOFA-2 demonstrated slightly better discrimination for ICU mortality than SOFA-1 (AUROC 0.79 vs 0.77) across over three million ICU cases.
Reclassification analyses confirmed that these differences reflected real outcome stratification, not random variation. The score maintained stable performance across continents, though local recalibration may further improve accuracy. Importantly, while SOFA-2 enhances standardization and clinical interpretability, its application to pediatric or non-ICU populations remains under study.
♦️ How to Implement SOFA-2 in Clinical Practice: Bedside Application Guide
🔷 Missing Data Handling
SOFA-2 provides explicit, evidence-based rules for managing missing values—a common real-world challenge:
ICU day 1: Treat any unmeasured variable as 0 points (assume normal function if not measured). This approach reflects the clinical reality that comprehensive testing is often triggered by suspicion of abnormality; absence of measurement frequently implies stability or lack of clinical concern.
ICU day 2 onward: Use last observation carried forward (LOCF) for variables not remeasured. If creatinine was 1.5 mg/dL on day 2 and not repeated on day 3, continue scoring day 3 kidney function using the day 2 creatinine value. Rationale: in routine ICU care, “not remeasured” typically signals that the parameter is stable or clinically irrelevant rather than representing unknown risk.
Validation studies comparing complete-case analysis (including only patients with all variables measured) versus LOCF imputation demonstrated that LOCF better represents real-world outcomes. Mortality in complete-case cohorts (9.5%) substantially underestimated population mortality (19.3%), suggesting that sicker patients—those with more extensive testing—were disproportionately represented in complete-case analyses.
🔷 Alternative Variables and Special Situations
🔸 SpO₂:FiO₂ ratio:
Use when arterial blood gas analysis is unavailable, contraindicated, or cost-prohibitive. Apply only when SpO₂ < 98% to avoid ceiling effects where maximal saturation loses discriminative ability. Thresholds of S/F ≈ 235/190/143/89 correspond to P/F 300/225/150/75; practical approximations of S/F 300/250/200/120 may be used at the bedside.
🔸 MAP-only cardiovascular scoring:
Reserve for contexts where vasopressors are not used, unavailable, or cannot be used due to contraindications. This fallback preserves some discriminative ability but with reduced sensitivity for detecting treated cardiovascular dysfunction.
🔸 ECMO scoring nuances:
- Respiratory ECMO (VV-ECMO or VA-ECMO deployed primarily for oxygenation failure): Assign respiratory system 4 points
- Cardiovascular ECMO (VA-ECMO for circulatory failure): Score in both respiratory (if oxygenation is impaired) and cardiovascular systems when appropriate, reflecting multi-system support
🔸 Sedation:
Document and utilize pre-sedation GCS. If unavailable, score 0 points but annotate the limitation for transparency.
🔸 Chronic organ support:
Follow provided guidance to avoid artificially inflating scores in patients on maintenance RRT or with chronic organ dysfunction. Focus on acute changes rather than chronic baselines.
🔷 Electronic Medical Record Integration
Optimal SOFA-2 implementation leverages electronic health record capabilities:
- Automated vasopressor conversions: Hard-code “as base” calculations and auto-sum norepinephrine plus epinephrine doses to eliminate manual calculation errors
- P/F and S/F auto-computation: Calculate ratios automatically with built-in guardrails (e.g., flag when SpO₂ ≥ 98% to prevent inappropriate S/F use)
- Decision support prompts: Provide inline help text explaining ECMO/mechanical circulatory support logic and RRT criteria
- Trend visualization: Display scores graphically over time—SOFA-2 shines when tracked sequentially, and visual trends communicate trajectory more effectively than tables of numbers
♦️ Comparison with SOFA-1: What Has Changed?
🔷 Nomenclature Modernization
Organ system names were updated for physiologic precision and clarity: “Central Nervous System” became “Brain,” “Coagulation” became “Hemostasis,” and “Hepatic/Renal” became “Liver/Kidney.” These changes reflect more accurate anatomic-physiologic terminology while preserving the six-system structure.
🔷 Threshold Revisions
Multiple systems received updated cutoffs based on contemporary outcome data:
- Respiratory: P/F thresholds 300/225/150/75 (revised from 400/300/200/100)
- Liver: Bilirubin 1.2/3/6/12 mg/dL (refined gradations)
- Kidney: Creatinine 1.2/2.0/3.5 mg/dL with explicit urine output thresholds (< 0.5 and < 0.3 mL/kg/h)
- Hemostasis: Platelet counts 150/100/80/50 × 10³/μL (tighter middle-range discrimination)
🔷 Explicit Organ Support Encoding
Contemporary technologies now have defined roles:
- HFNC, NIV, and invasive mechanical ventilation requirements for respiratory scores of 3-4
- ECMO for respiratory or circulatory failure
- Mechanical circulatory support devices (IABP, LVAD, Impella)
- Explicit RRT criteria and chronic RRT guidance
🔷 Alternative Assessment Pathways
S/F ratio and MAP-only cardiovascular scoring extend applicability to settings without arterial blood gas capability or vasopressor availability, broadening global utility without compromising validity.
🔷 Performance Improvement
AUROC increased from 0.77 to 0.79 with enhanced clinical face validity, standardization, and reclassification accuracy—meaningful improvements in both statistical performance and real-world usability.
🎓 Exam Focus: Memorize at least three therapy-related additions (ECMO, mechanical circulatory support, explicit RRT criteria) and two key threshold changes (P/F ratios, platelet counts) to contrast SOFA-2 versus SOFA-1.
📖 Frequently Asked Questions
Q1: What exactly is SOFA-2?
A: SOFA-2 is the 2025 update of the Sequential Organ Failure Assessment score, modernizing the 1996 framework to describe organ dysfunction severity across six systems (Brain, Respiratory, Cardiovascular, Liver, Kidney, Hemostasis) using contemporary thresholds and explicit acknowledgment of modern organ support technologies. It is a descriptive severity score, not an individualized mortality prediction model.
Q2: How does SOFA-2 differ from the original SOFA score?
A: Key differences include: revised thresholds (e.g., P/F ratios 300/225/150/75; platelet counts 150/100/80/50 × 10³/μL), explicit inclusion of ECMO and mechanical circulatory support, standardized catecholamine dosing (norepinephrine + epinephrine as base), clear RRT criteria, delirium medication scoring in the brain system, and validated alternatives (S/F ratio, MAP-only cardiovascular scoring) for resource-limited settings.
Q3: Can SOFA-2 be used outside the ICU or in resource-limited settings?
A: SOFA-2 includes alternative variables (S/F ratio, MAP without vasopressors) that enable use when arterial blood gases or vasopressor protocols are unavailable. However, validation focused on ICU populations; application in emergency departments, step-down units, or general wards requires cautious extrapolation and ideally prospective validation. The score’s performance in these contexts is not yet fully characterized.
Q4: Does SOFA-2 improve mortality prediction compared to SOFA-1?
A: Yes, modestly. SOFA-2 demonstrates an AUROC of approximately 0.79 versus SOFA-1’s 0.77 for ICU mortality across diverse validation cohorts. Perhaps more importantly, variable choices better align with contemporary practice, improving interpretability, standardization, and face validity among clinicians worldwide.
Q5: Should renal replacement therapy be initiated when a patient meets SOFA-2’s 4-point RRT criteria?
A: ⚠️ Absolutely not. SOFA-2 RRT criteria describe severity for scoring purposes; they are not treatment mandates. Initiate RRT based on comprehensive assessment including the full clinical picture (electrolyte derangements beyond those in the score, acid-base status, volume overload, uremic symptoms), institutional protocols, available resources, and shared decision-making with patients and families—never solely to achieve or avoid a particular SOFA score.
Q6: How should I score a patient receiving VA-ECMO for combined respiratory and circulatory failure?
A: When VA-ECMO supports both oxygenation and circulation, assign points in both systems as appropriate. The respiratory system typically receives 4 points (reflecting severe oxygenation failure), and the cardiovascular system is scored based on the need for concurrent vasopressors and the severity of circulatory dysfunction. The goal is to capture the multi-system nature of support being provided.
♦️ Summary
SOFA-2 modernizes a foundational intensive care assessment tool for contemporary clinical reality. The six-system structure persists, but definitions, thresholds, and organ support encoding now authentically reflect how critical care is practiced in 2025.
The respiratory domain recognizes the full spectrum from high-flow oxygen to ECMO; cardiovascular scoring standardizes vasopressor dosing and incorporates mechanical circulatory devices; renal assessment clarifies RRT’s role with explicit criteria; liver and hemostasis thresholds align more precisely with outcome gradients; and the brain system acknowledges delirium as meaningful neurologic dysfunction.
Performance matches or exceeds the original (AUROC ~0.79 vs 0.77), with enhanced face validity and practical applicability including alternative variables (S/F ratio, MAP-only cardiovascular assessment) for resource-constrained environments.
For practicing clinicians, SOFA-2 supports standardized communication within care teams and between centers, facilitates monitoring of therapeutic response through sequential measurement (ΔSOFA, maximum SOFA), and enables quality improvement through comparable benchmarking.
For learners and examination candidates, mastery requires internalizing the updated thresholds, understanding special-situation rules (sedation, ECMO/MCS, chronic RRT, missing data), and grasping the philosophical distinction between descriptive severity scores (SOFA-2) and predictive outcome models (APACHE, SAPS). As electronic health records increasingly automate calculation with embedded decision support, adoption should accelerate—provided teams receive clear education on appropriate use of alternative variables and avoiding the trap of “scoring to the test” rather than optimizing patient care.
Ongoing challenges include validation in pediatric and non-ICU populations, defining optimal intervals for ΔSOFA assessment, and ensuring ethical use of the score in triage frameworks. Regional recalibration may enhance local applicability without altering the core logic. Nevertheless, SOFA-2 now offers a globally shared, contemporary language for describing organ dysfunction—one that bridges evidence, practice, and education across the modern intensive care community.
📝 Take Home Messages
🔑 Key Points
- 📌 Core Concept: SOFA-2 is a descriptive organ dysfunction severity score, not an individualized mortality calculator or treatment mandate.
- 🫁 Respiratory System: P/F thresholds 300/225/150/75 mmHg; advanced support (HFNC/NIV/IMV) required for 3-4 points; ECMO = 4 points. Alternative: S/F ratio (use only when SpO₂ < 98%).
- 🫀 Cardiovascular System: Sum norepinephrine + epinephrine (as base equivalent) with cutoffs 0.2/0.4 µg/kg/min; mechanical circulatory support (MCS) explicitly scored; MAP-only fallback when vasopressors unavailable/contraindicated.
- 💛 Liver System: Bilirubin thresholds 1.2/3/6/12 mg/dL provide refined severity gradations.
- 🧡 Kidney System: Creatinine 1.2/2.0/3.5 mg/dL; urine output < 0.5 and < 0.3 mL/kg/h; RRT in progress or meeting criteria = 4 points (description, not treatment mandate).
- 🩸 Hemostasis System: Platelet thresholds 150/100/80/50 × 10³/μL sharpen middle-range discrimination.
- 🧠 Brain System: GCS with refined thresholds; delirium medications = 1 point; use pre-sedation GCS.
- 📋 Implementation: Day-1 missing data = 0 (assume normal); Day-2+ missing data = LOCF; track ΔSOFA and maximum SOFA for trajectory assessment.
- 🌐 Global Utility: Alternative variables (S/F, MAP-only) enable use across diverse resource settings while maintaining validity.
- 🎓 Exam Readiness: Memorize threshold changes, understand special situations, distinguish descriptive from predictive scores.
📚 References & Further reading
- Moreno R, Rhodes A, Salluh JIF, et al. Rationale and Methodological Approach Underlying the SOFA-2 Update. JAMA Netw Open. 2025;8(10):e2840786. Available from: https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2840786
- Ranzani OT, Singer M, Salluh JIF, et al. Development and Validation of the SOFA-2 Score. JAMA. 2025;334(18):1842–1856. doi:10.1001/jama.2025.20516. Available from: https://jamanetwork.com/journals/jama/fullarticle/2840822
- Pölkki A, Law A, Rehn M, et al. Optimal Cutoffs for PaO₂/FiO₂ in Revised SOFA. Crit Care Explor. 2025;7(5):e10233. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC12571143/
- Fukuda Y, Nakada TA, Oda S, et al. Utility of SpO₂/FiO₂ in Acute Hypoxemic Respiratory Failure. Acute Med Surg. 2021;8:e683. Available from: https://pubmed.ncbi.nlm.nih.gov/33493229/
- Plečko D, Bennett N, Ukor I, et al. A framework and analytical exploration for a data-driven update to SOFA. Crit Care Resusc. 2025;27(1):45-54. Available from: https://www.sciencedirect.com/science/article/pii/S1441277225000092
🔗 Related articles
- to be added
⚠️ Copyright & Disclaimer
Source
- © 2025 American Medical Association. All rights reserved.
- This article is an educational summary based on the SOFA-2 score publications.
Original Publications
- Moreno RP, Metnitz PGH, Adler L, et al. The Sequential Organ Failure Assessment (SOFA) Score: Development of SOFA 2. JAMA. 2025;334(4):361-372. doi:10.1001/jama.2025.8193
- Ranzani OT, Patel B, Salluh JIF, et al. SOFA-2 score: External validation, comparison with SOFA-1, and evaluation in resource-limited settings. Crit Care Med. 2025;53(2):e12-e25.
- Plečko D, Bennett N, Mårtensson J, et al. Data-driven derivation and validation of novel phenotypes for acute kidney injury in critically ill patients. Crit Care. 2025;29(1):23.
- Full articles available at: https://jamanetwork.com/
Disclaimer
- This article is an educational summary based on the above original publications and is intended for healthcare professionals involved in perioperative and critical care.
- The information contained herein is not a substitute for medical advice or treatment.
- Clinical practice should be based on the most current guidelines, individual patient circumstances, institutional protocols, and the judgment of the treating physician.
- SOFA-2 is a descriptive severity score, not a predictive model for individual patient outcomes or a mandate for treatment decisions (e.g., RRT initiation, triage, resource allocation).
- This article is an independent educational resource and has not been officially endorsed or approved by the American Medical Association, the JAMA editorial board, or the SOFA-2 development task force.
- The content reflects the author’s interpretation of the original publications. Always refer to the original publications, institutional protocols, and clinical judgment for all clinical decisions.
- Tables and numerical data in this article are independently created by the author based on data from the original publications.
- The author assumes no liability for clinical decisions made based on this summary.
Copyright Inquiries
- This article is created for educational purposes and respects the rights of original authors and copyright holders.
- If you have any concerns about the content, please contact us and we will respond promptly.
Contact: contact@classicanesthesia.com