Same patient, different SOFA: SOFA-2 versus the original organ failure score

In more than 134,000 adult ICU admissions scored both ways, SOFA-2 moved individual patients between severity categories, and patients with the same total under one version still differed in ICU mortality under the other. This episode covers what SOFA-2 changed, the mirror-image matching analyses, the Chinese and Korean validation cohorts, and what each discipline must chart for the new score.

In this episode

* SOFA-2 was built from a modified Delphi process and ICU data from 1,319 ICUs across 9 countries, and it discriminated ICU mortality slightly better than the original score, an area under the curve of 0.79 against 0.77.

* In 134,528 Japanese ICU admissions, mean day-one totals were 5.64 for the original score and 5.81 for SOFA-2, but the brain, respiratory and cardiovascular rows mostly shifted upward and the kidney row moved in both directions.

* Among patients matched on SOFA-2, ICU mortality was 9.0 percent when the original score was at least 2 points higher, 5.7 percent when the scores roughly agreed, and 5.0 percent when SOFA-2 was at least 2 points higher; matched on the original score, the pattern reversed, 3.4 percent, 5.0 percent and 9.5 percent.

* Much of the disagreement ran through organ support: among patients matched on SOFA-2, norepinephrine in half of the patients whose original score ran higher against 4.8 percent of those whose SOFA-2 ran higher, and kidney replacement therapy at 3.7 percent against 7.4 percent.

* Sepsis-3 still defines organ dysfunction with the original score, and no study here measured how many patients would be reclassified as septic, so name the version every time and rebaseline before comparing trends.

Chapters

(00:00) Cold open

(00:58) What they did

(02:56) Key numbers

(05:04) What it does not show

(06:42) Bedside takeaways by discipline

(07:51) Teaching pearl

(08:01) Close

Read the full article: https://iccn.substack.com/p/same-patient-different-sofa?utm_source=podcast&utm_medium=audio&utm_campaign=q4_paid

Sources

* Ono S, Uchino S, Kumaido Y, et al. Patient-level reclassification and discordance between SOFA and SOFA-2: a multicenter ICU study. Crit Care. 2026;30:425. doi:10.1186/s13054-026-06241-0

* Ranzani OT, Singer M, Salluh JIF, et al. Development and Validation of the Sequential Organ Failure Assessment (SOFA)-2 Score. JAMA. 2025;334(23):2090-2103. doi:10.1001/jama.2025.20516

* Ahn S, Kim SM, Park JH, et al. Validation of the updated sequential organ failure assessment (SOFA)-2 score in emergency department septic shock. J Intensive Care. 2026;14:92. doi:10.1186/s40560-026-00921-3

* Liufu R, Cao M, Shi XQ, et al. External validation of SOFA 2 score: a retrospective Chinese cohort study. Ann Intensive Care. 2026;16:100107. doi:10.1016/j.aicoj.2026.100107

* Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801-810. doi:10.1001/jama.2016.0287

ICCN publishes one article a day for ICU clinicians. $8 a month or $80 a year, with a 7-day free trial.

This episode is for educational purposes for healthcare professionals. It is not medical advice, a clinical guideline or a standard of care. Clinical decisions must be individualized by qualified clinicians.

Hosted by Javier Amador-Castaneda, BHS, RRT, FCCM. Voices rendered with AI text-to-speech from a script written and reviewed by ICCN.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit iccn.substack.com/subscribe