MASH Real-World Data: Why Fibrosis Stage Is the Variable That Matters

Some diseases are difficult to study using claims alone. MASH is one of the clearest examples.

Metabolic dysfunction-associated steatohepatitis (MASH) affects an estimated nine to 15 million U.S. adults. Until March 2024, no treatment had been approved specifically for MASH. The therapeutic landscape now includes two approved therapies, both granted accelerated approval and indicated for the same narrow population: adults with moderate to advanced liver scarring who have not progressed to cirrhosis. Cirrhosis represents the most advanced stage of liver scarring, accompanied by extensive disruption of the liver’s structure.

Within that population, one clinical variable separates treatment eligible from treatment ineligible: fibrosis stage.

Fibrosis is the accumulation of liver scarring as MASH progresses, graded from F0, indicating no scarring, to F4, indicating cirrhosis. Both approved therapies are indicated for patients with F2 to F3 fibrosis. That single clinical measure determines who is eligible for treatment, who belongs in a study cohort, what the baseline disease severity looks like, and how disease progression or treatment response gets assessed. Yet, standard claims data rarely has this vital detail, leaving researchers in the dark.

Why fibrosis stage anchors MASH research

For teams working in health economics and outcomes research, fibrosis stage is not a clinical detail. It is a critical study variable.

It defines the cohort. A treatment-eligible population estimate is a count of patients at fibrosis stage F2 or F3, not a count of all patients with a MASH diagnosis.

It measures the baseline disease severity. Measuring real-world treatment response typically means comparing stage at initiation against stage at follow-up.

It stratifies the outcomes. Progression risk and mortality both move with disease severity. In a real-world analysis of patients with F2 to F3 fibrosis, 17% went on to develop cirrhosis.

And the timing matters. Both drugs were approved on biopsy findings rather than long-term outcomes, with confirmatory trials still running. Much of the near-term evidence on how these therapies perform in practice will come from MASH real-world data.

Where fibrosis stage lives

Staging starts with the Fibrosis-4 (FIB-4) index, a score calculated from a patient’s age and three routine blood test results. Because its inputs are routine lab values, it is computable wherever those labs are available. It also carries real prognostic weight, with progression risk rising as the score climbs.

The FIB-4 result is a first pass rather than a final answer. AASLD practice guidance routes elevated scores to a second round of testing: a liver stiffness scan or a blood panel that measures markers of scarring directly.

That second round is where the documentation changes character.

A liver stiffness scan, most often FibroScan, produces a reading that is typically documented in an imaging or clinical note. Liver biopsy remains the reference standard, with the findings documented in a pathology report written as a free-text narrative. For researchers studying disease progression, the free-text nature of pathology reports can limit the questions that the patient’s clinical record can answer.

The pathway is also unevenly followed. In a 2025 analysis of a real-world liver disease population, most patients at high risk for advanced scarring had no hepatology referral on record. The clinical pathway breaks down before the second test is ever ordered.

Separately, claims carry part of the picture. ICD-10-CM includes codes for early and advanced hepatic fibrosis, which may sit alongside a MASH code. However, the code set does not capture the F0 to F4 staging system used in drug labels and clinical trials. No ICD-10-CM code distinguishes F2 from F3, and a fibrosis stage designation can be assigned only when the clinician has documented the stage and a coder has captured it.

This limitation does not reflect a clinical documentation failure. Clinicians record liver stiffness measurements or biopsy findings in notes because those results belong in the clinical record and guide subsequent care. The care is documented appropriately. It reads as a gap when we try to force clinical complexity into administrative billing codes.

Coverage criteria for these therapies require documented fibrosis stage confirmed by testing, with results submitted. The specific tests vary by health plan. None of the criteria reviewed allow a diagnosis code to substitute for a test result. The real-world value of a therapy is locked behind the clinical narrative.

The measurement is becoming an endpoint

In August 2025, the FDA agreed to review a proposal to accept liver stiffness measurement as a trial endpoint for patients with MASH without cirrhosis. The agency described the action as a first step in a longer qualification process, rather than a final decision, while recognizing its potential to advance the use of noninvasive tests in place of biopsy for MASH drug development.

Clinical investigators are pushing in the same direction. The authors of the progression analysis concluded that standardizing noninvasive test use, expanding access to liver stiffness scanning, and automating FIB-4 calculation would improve early detection and risk stratification.

A measurement that helps determine coverage today may anchor clinical trial endpoints tomorrow. In both settings, however, the result is written into a note.

What closed claims carry

Claims databases do exactly what they were built to do. They track the money and the longitudinal events. They can tell you which patients filled a prescription, switched therapies, have comorbidities (such as diabetes), and what the total cost of care looks like over time. If you are building a budget impact model, claims alone will give you the answer.

What they cannot tell you is why.

Claims carry time. A patient may be staged once at a specialist visit and then followed for years through their claims history. The stage tells you who the patient is, and the claims show what happened next.

As covered in how EHR data and clinical notes add depth to real-world data, the research question determines the right data source. For MASH, that means recognizing that the defining clinical variable lives entirely outside the billable record.

Questions to ask a MASH real-world data source

• Are the lab values behind FIB-4 present, and are they captured longitudinally?
• Are clinical notes included, and from which care settings?
• Are diagnostic reports available, including pathology and imaging?
• How is narrative text de-identified?
• Can the clinical record and the claims record be linked for the same patient, and who performs that linkage?

A cost-of-care study does not need a clinical notes layer. A treatment-eligible population estimate will not resolve from claims alone, no matter how many millions of patients are in the database.

The clinical reality is written in the note. The research method has to be ready to go in after it.

See what MASH research really requires

With fibrosis stage written into notes and reports rather than filed in a field, locating and normalizing that value becomes an extra step before analysis can begin. The Inovalon and OMNY Health Linked Closed Claims and EHR Dataset includes a structured fibrosis stage variable curated from the notes.

The dataset connects Inovalon’s primary source closed claims with OMNY Health’s structured EHR data, clinical notes, and diagnostic reports at the patient level, delivered already linked.

For teams that would rather have the analysis done than the data delivered, Inovalon’s Outcomes and Evidence Solutions team works from the same sources.

Explore linked datasets →

Frequently asked questions about MASH real-world data

Can closed claims alone support MASH research?

For many questions, yes. Treatment patterns, persistence, comorbidity burden, hospitalizations, and total cost of care are all well served by closed claims alone. Questions that require metrics related to disease severity, such as fibrosis stage, would require access to EHR data.

Is FIB-4 enough to determine fibrosis stage?

FIB-4 is computable from routine lab values and carries real prognostic signal. AASLD practice guidance treats the result as a first pass, with elevated scores routed to further testing. Coverage criteria differ on whether FIB-4 alone is accepted as confirmation.

Why do MASH studies need clinical notes?

Because the tests that confirm fibrosis stage produce narrative output. Liver stiffness readings are recorded in clinical and imaging notes, and biopsy findings appear in pathology reports written in prose.

What does linking closed claims and EHR data add for MASH?

It connects the clinical variables that define the MASH population with the longitudinal treatment and outcome record for the same patient in the real-world setting. Fibrosis stage from the clinical record defines the relevant population, while closed claims capture treatment, healthcare utilization, cost, and clinical events over the following years. The result is a real-world MASH dataset that can answer both sides of the question: who is eligible for treatment, and what happens after treatment?

 

Sources

All external sources are hyperlinked in text.

In order of appearance: American Liver Foundation (prevalence); Hepatology Communications, June 2026 (progression and FIB-4 prognostic gradient, cited twice); AASLD practice guidance (FIB-4 pathway, cited twice); Hepatology Communications, March 2024 (free-text pathology reports); Gastro Hep Advances, December 2025 (referral gaps); Florida Agency for Health Care Administration (coverage criteria); U.S. Food and Drug Administration (liver stiffness endpoint proposal). Internal: press release, Outcomes and Evidence Solutions page, EHR and clinical notes blog, linked datasets product page.

---

Inovalon and design® and Inovalon® are trademarks of Inovalon, Inc.