- Design
- Retrospective cohort comparing nine clinical risk scores with decision-curve analysis
- Population
- 853,131 US veterans with imaging-confirmed steatotic liver disease, no viral hepatitis
- Primary outcome
- Cirrhosis and hepatocellular carcinoma within 10 years
- Effect
- Cirrhosis 3.96% and carcinoma 0.35% at 10 years; SAFE net benefit 0.019 at 2.5% cirrhosis risk threshold
This cohort study used national US Veterans Affairs records of 853,131 adults with imaging-confirmed steatotic liver disease and no viral hepatitis. Nine risk scores were calculated at the first imaging that showed steatosis, and patients were followed for cirrhosis and hepatocellular carcinoma. Over 10 years, 3.96% developed cirrhosis and 0.35% developed carcinoma.
FIB-4, APRI and SAFE had the best discrimination for cirrhosis, and SAFE, Tate and FIB-4 for carcinoma. On decision-curve analysis SAFE had the greatest net benefit for cirrhosis: using a score of 29.5 (10-year cirrhosis risk 2.5%) gave 1.9 extra true cases identified per 100 people classified at risk. For carcinoma, no risk score showed useful net benefit for deciding on screening in people without cirrhosis.
For a general physician, this supports using a simple score such as FIB-4 to decide who needs further fibrosis assessment or repeat screening, not to start liver cancer surveillance. The cohort was 93% male veterans.
- FIB-4, APRI and SAFE gave the best discrimination for later cirrhosis in steatotic liver disease.
- SAFE had the greatest net benefit, but FIB-4 performed well and is simple to calculate from routine tests.
- Scores added little to decisions about hepatocellular carcinoma screening in non-cirrhotic liver disease.
- Consider repeating fibrosis scoring over time rather than relying on a single result.
- The cohort was 93% male veterans, so check applicability to women and to Indian patients.
Why it matters
Most people with fatty liver never develop advanced disease, so who to re-test matters more than who to label.
Don't overread it
Outcomes came from diagnosis codes in a mostly male veteran population, and nothing here shows that score-guided screening improves outcomes.
The statistics, in plain English
Net benefit weighs true positives against false positives at a chosen risk threshold. A net benefit of 0.019 means about 1.9 additional people who will develop cirrhosis are identified per 100 classified at risk, compared with treating nobody as at risk, without a penalty for false alarms beyond that threshold.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for internal medicine, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free