SAFE Score Beats Eight Rivals at Predicting 10-Year Cirrhosis Risk in Primary Care
A new study finds the SAFE score outperforms eight other tools for flagging cirrhosis risk in steatotic liver disease patients seen in primary care.
Summary
Steatotic liver disease (SLD) is a leading driver of advanced liver disease, yet most early cases are clinically silent and easily missed. A new study in JAMA Internal Medicine compared nine risk-scoring tools for predicting cirrhosis or liver cancer over 10 years in primary care patients with SLD. The SAFE score — which combines age, BMI, diabetes status, and five blood markers — delivered the greatest net benefit for cirrhosis prediction, identifying roughly 1.9 additional at-risk individuals per 100 screened compared to doing nothing. No tool performed well for hepatocellular carcinoma screening. The findings support building structured SLD screening pathways in primary care, moving beyond the widely used but limited FIB-4 index. Researchers say SAFE can guide counseling and decisions about repeat liver assessments.
Detailed Summary
Steatotic liver disease is now one of the most common liver conditions worldwide and a major upstream cause of cirrhosis and liver cancer. Because early-stage disease is often silent, many high-risk patients go undetected while others undergo unnecessary specialist referrals. A new comparative study published in JAMA Internal Medicine addresses this gap by evaluating nine noninvasive risk scores for their ability to predict 10-year cirrhosis or hepatocellular carcinoma risk in a primary care population with SLD.
The SAFE score emerged as the clear frontrunner for cirrhosis prediction. A SAFE threshold of 29.5, corresponding to a 10-year cirrhosis risk of 2.5%, yielded a net benefit of 0.019 — meaning approximately 1.9 additional people who will develop cirrhosis are correctly flagged per 100 patients classified as at-risk. No other tool came close to this level of clinical utility in the primary care context.
The SAFE calculator incorporates seven readily available variables: age, BMI, diabetes status, aspartate aminotransferase, alanine aminotransferase, globulins, and platelets. This makes it practical and low-cost for routine use. The currently recommended FIB-4 index, while cheap and widely adopted, was flagged by the authors as insufficiently sensitive and specific, leaving both under-detection and overreferral as real problems.
For hepatocellular carcinoma screening, all nine tools showed minimal utility, highlighting a significant unmet need. An accompanying editorial by a University of California, San Francisco hepatologist noted that SLD screening pathways are still nascent compared with mature frameworks for hypertension or chronic kidney disease.
The practical implication is clear: primary care clinicians managing patients with metabolic risk factors or incidentally discovered steatosis now have stronger evidence to adopt SAFE over FIB-4 for cirrhosis risk stratification. However, the study is a comparative evaluation of existing tools rather than a randomized trial, and external validation across diverse populations is still needed before widespread guideline adoption.
Key Findings
- SAFE score outperformed eight other risk tools for predicting 10-year cirrhosis risk in SLD patients in primary care.
- A SAFE threshold of 29.5 flags a 2.5% 10-year cirrhosis risk, yielding ~1.9 extra correctly identified cases per 100 patients.
- SAFE uses seven routine measures: age, BMI, diabetes status, AST, ALT, globulins, and platelets — all low-cost and widely available.
- No existing tool showed meaningful utility for hepatocellular carcinoma screening in SLD, representing a major evidence gap.
- The widely used FIB-4 index was deemed insufficiently sensitive and specific for reliable cirrhosis risk stratification.
Methodology
This is a news report summarizing a comparative prediction study published in the peer-reviewed journal JAMA Internal Medicine, authored by researchers from Yale School of Medicine. The study evaluated nine noninvasive risk scores using a primary care cohort and applied decision-curve analysis to assess net clinical benefit; an accompanying editorial provides independent expert commentary.
Study Limitations
The article is a secondary news report and does not provide full details on study design, sample size, or the specific population studied, requiring review of the primary JAMA Internal Medicine publication. The SAFE score's performance may vary across different ethnic groups and metabolic phenotypes not represented in the study cohort. No tool evaluated showed utility for HCC screening, leaving a critical clinical gap unresolved.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: