FibroScan is a highly accurate tool for detecting cirrhosis, with studies showing it correctly classifies severe scarring 85% to 94% of the time. It’s less precise at identifying moderate fibrosis, where accuracy drops noticeably. How much you can trust your result depends on what stage of liver disease is being evaluated, what’s causing it, and a few physical factors like body weight and whether you ate before the scan.
How Accuracy Is Measured
Researchers compare FibroScan results against liver biopsy, which is considered the gold standard for staging fibrosis. The key metric is called the AUROC score, a number between 0 and 1 where 1.0 means perfect accuracy and 0.5 means the test is no better than flipping a coin. Anything above 0.80 is generally considered good, and above 0.90 is excellent.
Accuracy for Detecting Cirrhosis
FibroScan performs best at the extremes: ruling out serious scarring or confirming cirrhosis. A large multicenter study in patients with chronic hepatitis B or C found an overall AUROC of 0.89 to 0.90 for predicting cirrhosis, significantly outperforming blood-based biomarkers (which scored between 0.77 and 0.86). Across multiple studies, FibroScan correctly classifies cirrhosis versus non-cirrhosis in roughly 85% to 94% of cases.
This is where the test earns its reputation. If your FibroScan score is very high (typically above 12 to 14 kPa, depending on the underlying liver condition), cirrhosis is likely. If it’s very low, cirrhosis can be confidently ruled out.
Accuracy for Moderate Fibrosis
The picture gets murkier in the middle stages. For detecting significant fibrosis (stage F2 and above, meaning scarring has progressed beyond mild), AUROC scores drop to the 0.72 to 0.78 range in chronic viral hepatitis. A meta-analysis in hepatitis B patients found an AUROC of 0.794 for significant fibrosis. That’s a meaningful gap compared to cirrhosis detection.
In practical terms, this means FibroScan can struggle to distinguish between mild and moderate scarring. If your result lands in an intermediate zone, your doctor may recommend follow-up testing or monitoring rather than drawing firm conclusions from a single scan. The test is better thought of as a screening tool at these stages than a definitive diagnosis.
Accuracy for Liver Fat
Modern FibroScan machines also measure liver fat using a feature called CAP (Controlled Attenuation Parameter). A large meta-analysis of patients with non-alcoholic fatty liver disease found CAP detects any meaningful fat buildup with 84% sensitivity and 86% specificity. That’s solid performance for a quick, noninvasive scan.
However, accuracy declines when trying to distinguish between moderate and severe fat accumulation. For moderate-or-higher steatosis, sensitivity drops to 79% and specificity falls to 64%. So while FibroScan can reliably tell you whether your liver has excess fat, it’s less reliable at grading exactly how much.
How Body Weight Affects Results
This is one of the biggest practical limitations. In a study of over 8,200 patients, the standard probe (called the M probe) failed to produce reliable measurements in 14.9% of normal-weight patients. That failure rate jumped to 49.5% in obese patients (BMI of 30 or higher). Nearly half the time, the standard probe simply couldn’t get a usable reading through thicker tissue.
A larger probe (the XL probe) was developed specifically for this problem. It recovers usable results in about 60% to 80% of patients who fail with the standard probe, with success rates of 80% in normal-weight patients, 68% in overweight patients, and roughly 60% in obese patients. If you have a higher BMI, ask whether the facility uses the XL probe. Without it, there’s a real chance the scan won’t produce a result at all.
Why Eating Before the Scan Matters
Food intake temporarily increases blood flow to the liver, which stiffens the tissue and inflates your reading. In a study of 100 healthy volunteers, eating an 800-calorie meal raised liver stiffness from an average of 4.80 kPa to 5.74 kPa. That’s roughly a 20% increase, enough to push a borderline result into a higher fibrosis category.
Current guidelines recommend fasting for at least two hours before a FibroScan, though overnight fasting is preferred. If you ate shortly before your appointment and your result seems higher than expected, it’s worth mentioning to your doctor. A repeat scan under proper fasting conditions may give a more accurate baseline.
How It Compares to Other Tests
FibroScan consistently outperforms blood-based fibrosis scores, which use combinations of routine lab work to estimate scarring. In head-to-head comparisons against biopsy, FibroScan’s AUROC scores run 0.03 to 0.13 points higher than biomarker panels for cirrhosis detection.
MRI-based elastography (MRE) is more accurate still. A meta-analysis of hepatitis B patients found MRE achieved an AUROC of 0.981 for significant fibrosis compared to FibroScan’s 0.794, and 0.972 versus 0.905 for cirrhosis. MRE is considerably more expensive and less widely available, though, so FibroScan remains the most common first-line test. Many clinicians reserve MRE for cases where FibroScan results are inconclusive or when precise staging matters for treatment decisions.
What Your Result Actually Tells You
A FibroScan gives you a number in kilopascals (kPa), where higher values mean stiffer liver tissue. The cutoff values used to define each fibrosis stage vary depending on the underlying condition. Hepatitis C, hepatitis B, and fatty liver disease each use slightly different thresholds, so the same kPa reading can mean different things in different contexts. Your doctor interprets the number in light of your specific diagnosis, not from a universal chart.
A single FibroScan is a snapshot. Results can fluctuate based on inflammation, recent alcohol use, food intake, and even the time of day. Trending multiple scans over time gives a much more reliable picture of whether your liver is stable, improving, or getting worse. If your first result is borderline or unexpected, a repeat scan in a few months is a reasonable next step rather than jumping to conclusions from one reading.