SOFA Score Calculator
SOFA Score Calculator
Score organ dysfunction across six systems — the score whose 2-point rise, with suspected infection, is the Sepsis-3 definition of sepsis.
SOFA Score
6 items → 0–24PaO₂/FiO₂ under 300 (2), platelets under 150 (1), bilirubin 1.2–1.9 (1), MAP under 70 (2), GCS 13–14 (1), creatinine 2.0–3.4 (2)
Scoring — six systems, 0 to 4 points each
- Respiration
- PaO₂/FiO₂ ratio, with respiratory support raising the top two grades
- Coagulation
- platelet count, ×10⁹/L
- Liver
- total bilirubin, mg/dL
- Cardiovascular
- mean arterial pressure, or vasopressor dose in µg/kg/min sustained ≥1 hour
- CNS
- Glasgow Coma Scale
- Renal
- creatinine, mg/dL, or urine output
- Sepsis-3
- suspected infection + a rise in SOFA of 2 or more from baseline (assumed 0 if no known prior dysfunction)
Worked example
PaO₂/FiO₂ under 300 (2), platelets under 150 (1), bilirubin 1.2–1.9 (1), MAP under 70 (2), GCS 13–14 (1), creatinine 2.0–3.4 (2)
2 + 1 + 1 + 2 + 1 + 2 = 9 points
Falls in the 6–9 band → moderate organ dysfunction, mortality ≈15–20%
Approximate mortality by total SOFA score
| SOFA total | Category | Approximate mortality |
|---|---|---|
| 0 – 5 | Mild | ≈10% |
| 6 – 9 | Moderate | ≈15 – 20% |
| 10 – 14 | Severe | ≈40 – 50% |
| 15 or more | Very severe | >80% |
Why the change matters more than the total
The Sequential Organ Failure Assessment score rates six organ systems — respiratory, coagulation, hepatic, cardiovascular, neurological and renal — from 0 to 4 each, and its role in the Sepsis-3 definitions is structural rather than incidental: sepsis is suspected or confirmed infection accompanied by an acute RISE in SOFA of 2 points or more from the patient’s baseline. In someone with no known chronic organ dysfunction, that baseline is assumed to be zero, so the admission score is effectively the delta already.
Two of the six items cannot be filled in from a laboratory report. The cardiovascular component is scored on mean arterial pressure and, at higher points, on the dose and choice of vasopressor; the neurological component uses the Glasgow Coma Scale. Both need bedside assessment, which is worth flagging explicitly — SOFA is often treated as a purely biochemical score because the remaining four items come from blood, but it is a clinical score that happens to lean heavily on the laboratory.
Total score correlates closely with mortality: roughly 10% below 6, 15 to 20% for 6 to 9, 40 to 50% for 10 to 14, and above 80% at 15 or more, though these figures come from mixed intensive care cohorts and shift with case mix. Vasopressor doses in the cardiovascular item are specified in micrograms per kilogram per minute, sustained for at least one hour — a brief bolus does not count, and using a shorter window will misclassify the item and inflate the score.
Frequently asked questions
How is sepsis defined using SOFA?
Suspected or confirmed infection plus an acute rise in SOFA of 2 points or more from baseline. In a patient with no known prior organ dysfunction, the baseline is assumed to be 0.
Which SOFA items are not laboratory tests?
The cardiovascular item, which uses mean arterial pressure and vasopressor dose, and the CNS item, which uses the Glasgow Coma Scale. Both require bedside assessment rather than a blood result.
What vasopressor doses does the cardiovascular item use?
Micrograms per kilogram per minute, sustained for at least one hour. A brief bolus dose does not count and using a shorter time window will overstate the score.
What does a SOFA score of 9 mean?
It falls in the 6–9 band, corresponding to moderate multi-organ dysfunction with an approximate mortality of 15 to 20% in mixed ICU cohorts. If it represents a rise of 2 or more with suspected infection, it meets the Sepsis-3 definition of sepsis.
Related calculators
References
- Vincent JL et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. Intensive Care Med. 1996;22(7):707–10.
- Singer M et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801–10.