The most striking line in this paper is not about what works. It is that one of the options tested came out worse than doing nothing at all, in every country, while also costing more.
What dominated means
In health economics, one strategy dominates another when it is both cheaper and better. Being dominated is the reverse: you spend more and end up worse off, which makes the comparison easy in a way that most of these analyses are not.
Screening pathways built around liver biopsy were dominated by no screening across all six markets. [1] Biopsy is invasive, expensive and carries its own risk, and enough people decline it that a pathway depending on it identifies fewer patients than it costs to run — a refusal rate doing the same work as the people who turned down a drug that worked.
What the model says to do instead
Multi-step non-invasive testing — blood-based scores first, imaging for those who flag — came out cost-effective in every country modeled, with incremental ratios running from dominance up to £18,572 per quality-adjusted life year. In Germany, France, Italy and the US, at least one strategy dominated no screening outright, meaning it both saved money and produced better outcomes.
The probabilistic analysis put the chance of cost-effectiveness between 90% and 100% depending on the market, which is unusually consistent for this kind of work.
Why this reaches a buyer at all
Screening is the gate. Semaglutide’s liver indication is written around a fibrosis stage, which means somebody has to establish that stage before the prescription exists, and the test that establishes it is the subject of this paper.
So the question of whether payers cover non-invasive testing sits upstream of the question of what the drug costs — and liver disease in this population is already expensive without any of it, at eleven thousand dollars a year more than matched controls.
The authors' forward claim
They argue the value of screening will grow as drugs capable of reversing fibrosis reach the market, naming resmetirom and semaglutide.
That is theirs rather than a finding, and it is the kind of claim worth labeling: a model that assumes an effective treatment exists downstream will always make finding patients look more worthwhile. Whether the treatment earns that assumption is a separate argument that seven other studies have tried to settle.
Nothing here is a reason to request a test. It is a reason to know that whether your insurer pays for one was decided by an analysis like this.