Beyond ICD-10 codes: Finding rare disease patients in real-world data

When rare diseases lack distinct diagnosis codes, life sciences teams must rely on rigorous, clinically informed analytics to uncover hard-to-find patients and transform imperfect data into decision-ready insights.

Rare diseases are gaining increased attention as diagnostic techniques improve and novel therapies enter the market. But accurately finding these patients using claims data is a challenge when a rare condition lacks its own diagnosis code.

Duchenne muscular dystrophy (DMD) is a prime example: it’s coded under a broader muscular dystrophy classification, making it hard to separate from related conditions like Becker muscular dystrophy (BMD).

Why does this matter? Inaccurate patient identification can distort market forecasts, misalign clinical and commercial programs, and prevent therapies from reaching the patients they’re designed to help.

For life science companies, solving this problem requires more than the right data. It takes a partner who understands the strengths and limits of that data, applies sound methodology, and knows the disease landscape inside and out.

The challenge of rare disease identification

Finding rare disease patients in real-world data is often complicated by:

• Misdiagnosis and delayed or missed diagnoses
• Shared or non-specific diagnosis codes
• Coding inconsistencies across systems
• Overlapping clinical presentations

Even within structured healthcare data, accurately isolating a rare disease cohort can be difficult when patients share codes with related but clinically distinct conditions.

In the case of DMD, patients are coded under a broader muscular dystrophy category that also includes Becker muscular dystrophy (BMD). While clinically related, DMD is generally more severe, diagnosed earlier in life, and follows a distinct progression pattern.

In short, DMD and BMD patients have different characteristics that aren’t captured by the ICD-10 code, making them hard to find.

These clinical distinctions translate into meaningful strategic differences for life science companies, from the types of providers managing care (pediatric neuromuscular centers versus adult neurology) to payer considerations, therapy eligibility windows, and treatment timelines.

Without deeper analytics, distinguishing these populations and aligning strategy appropriately becomes challenging.

Why this matters: Accurate patient identification ensures clinical trials target the intended population for recruitment, therapies reach the patients most likely to benefit, and commercial efforts are directed toward the right providers and treatment centers.

Why domain expertise matters in rare diseases

Even the best data isn’t enough on its own. Translating real-world data into useful insights requires expertise: recognizing the limits of claims data, understanding the nuances of patient journeys, and applying the right methodology, including modeling and validation approaches. It’s this combination of data and applied expertise that determines whether insights are just directional or decision-grade.

In rare conditions, that expertise matters even more. Rare conditions will always be a little like finding a needle in haystacks. Thoughtful analysis of the data by that right partner will let you drastically shrink the number of haystacks. This leads to things like better site selection, faster trial recruitment, and linking patients with novel therapies sooner.

Finding DMD patients without a unique ICD-10 code

In a recent project, our team focused on isolating a defensible DMD patient population from a broader muscular dystrophy cohort.

When an exclusive diagnosis code isn’t an option, our advanced analytics teams use a variety of techniques to find the right patient population. This can include various machine learning and predictive analytics techniques or, like in the case of DMD, clinically aligned indicators and patient demographics.

The outcome was a high-confidence DMD patient cohort that can be used to examine the patient journey as a whole, surface treatment pattern insights, and inform both commercial and clinical strategies.

Reaching key treating sites for a rare disease

With a validated patient cohort in place, we identified the top organizations in the U.S. treating DMD patients by volume. These institutions represent significant centers of expertise and referral hubs within the DMD landscape:

Top U.S. treating sites for DMD based on patient volume

Fig 1. Leading U.S. care sites treating Duchenne muscular dystrophy (DMD), ranked by patient volume derived from a validated cohort within Definitive Healthcare’s Atlas All-Payor Claims dataset. Data current as of February 2026.

Why this matters: Once the patient population is clear, strategy gets clearer, too. When your rare disease population is defined correctly, every downstream decision improves, from spotting true referral hubs and prioritizing trial sites with sufficient patient density to aligning field deployment with actual treatment volume rather than assumed prevalence.

How Definitive Healthcare supports rare disease identification

In rare diseases, precision is critical. When diagnosis codes obscure clinical nuance, rigorous methodology grounded in real-world clinical patterns becomes the difference between directional insight and decisions you can stake your strategy on.

Duchenne muscular dystrophy is just one example of how the right data and analytics can surface defensible insights in the rare disease space. As rare disease pipelines expand, competitive advantage won’t come from having more data. It will come from interpreting imperfect data more intelligently.

To learn how Definitive Healthcare can support your rare disease strategy, connect with our team or book a demo to explore our data and analytics capabilities.