Here are some of the studies on this issues
Hagström, H., et al. Ability of Noninvasive Scoring Systems to Identify Individuals in the Population at Risk for Severe Liver Disease. Gastroenterology 158, 200–214 (2020).
Details
Retrospective study in the general population using data-linkage in Sweden
Follow up was for: 5 years and 10 years with a maximal follow up of 27 years
Aim: to estimate the ability of simple scores (FIB4, NFS, APRI, forns and BARD) to predict fatal and nonfatal liver related outcomes.
Highlights
Compared 5 scores to see which had the highest predictive ability for detecting incident cirrhosis in the community.
The ability of all the scoring systems to identify patients who will develop cirrhosis at 10-years is low in the unselected population.
The scores performed better:
- For predicting short-term events
- In men
- In people with risk factors for NALFD
Subgroup analysis was done for people with risk factors for progressive NAFLD (Type 2 diabetes, elevated ALT, hypertriglyceridaemia)
Most of the patients who developed cirrhosis after 10 years had simple scores that predicted a low risk of advanced fibrosis at baseline
The NPV of these scores is high (>99%) but the sensitivity and PPV is low
Conclusion: simple scores may be useful to predict who will develop cirrhosis or its complications in people who have risk factors for progressive NAFLD or evidence of liver disease (type 2 diabetes, elevated ALT) but should NOT be used in the general population as a screening test to detect those at risk of incident cirrhosis.
Johnson, A. L.. et al. Predicting Liver‐Related Outcomes in People With Nonalcoholic Fatty Liver Disease: The Prognostic Value of Noninvasive Fibrosis Tests. Hepatology Communications 6, 728–739 (2022).
Details
Prospective study in Australia
Evaluated the ability of non-invasive tests (NFS, FIB4, FibroScan and ELF) to predict liver related complications
Median follow-up of 50 months
Highlights
All liver related complications occurred in:
- People with type 2 diabetes
- FibroScan >9.5kPa at baseline
5 patients (26%) with a liver related event had a low risk FIB-4 at baseline. No one with a low risk NFS at baseline had any liver related events.
Of the patients for didn’t have any liver related complications: only 20% had a low risk NFS compared to 62% who had a low risk FIB-4
Conclusion: FibroScan but not simple scores were able to predict liver related complications. The FIB-4 incorrectly classified 26% of patients with a liver related event as at low risk of advanced fibrosis. In contast the NFS had a high number of false positives and was only able to reassure 20% of the people who didn’t have a liver related event.
Graupera, I. et al. Low Accuracy of FIB-4 and NAFLD Fibrosis Scores for Screening for Liver Fibrosis in the Population. Clinical Gastroenterology and Hepatology 20, 2567–2576.e6 (2022).
Details
Population-based cohort study across 5 countries in the general popuation
Evaluated the ability of simple scores (FIB-4 and NFS) to detect an elevated liver stiffness (measured with FibroScan)
Highlights
There were more false-negative simple scores (low risk simple score, elevated FibroScan) in people with risk factors for progressive liver disease (alcohol use, diabetes, obesity).
People with an elevated LSM who were missed by the FIB-4/NFS included:
- 2-4% in the general population
- 8-9% of people at risk for advanced disease
Almost a third of people with an elevated FIB-4 or NFS had a low risk FibroScan.