Expanding ICD Codes May Aid Rare Disease Patients, Research

Many rare diseases don't have an ICD code, and the National Center for Advancing Translational Sciences (NCATS) at NIH is exploring ways to best expand those codes to benefit rare disease patients and research, said Director Joni Rutter, PhD, during Rare Disease Day at NIH on Tuesday.

"So many rare diseases are an n of 10, or 100, or less than 500 people, so we lack the ICD codes that we use for common diseases," Rutter told MedPage Today in an interview. "Because we don't have those ICD codes within our electronic health system, those rare disease patients become invisible."

Having a diagnostic code has implications for what the rare disease community calls the "diagnostic odyssey" and clinical care, but it also has implications for research, particularly in terms of being able to understand the prevalence and economic burden of a disease.

The rare disease community has turned its attention to ICD codes in recent years, with the EveryLife Foundation releasing its "ICD Code Roadmap" in 2021, in partnership with patient advocacy groups. In a video, the foundation explains that without a diagnostic code, insurers may deny care, physicians may struggle to communicate, and research into the disease may be hindered.

Getting an ICD code alone "wouldn't create therapies or a cure, but it would be a key component in making the work towards it possible," the video states. "You can't work on something you can't see."

It concludes that patient advocates can use the roadmap to determine whether pursuing a specific ICD code is the right strategy for their community.

Some rare diseases, such as cystic fibrosis and sickle cell anemia, already have ICD codes, Rutter said, and increasing the number of rare diseases with ICD codes could be one strategy to improving care.

But there's also interest in figuring out ways to better understand symptomatology measured via current ICD codes, particularly to aid in shortening the diagnostic odyssey, she noted.

"There might be broad codes like 'seizure disorder,' 'stunted growth,' or 'failure to thrive,' -- basket codes that are often used to describe the symptoms of rare diseases. When you get a constellation of these different kinds of symptoms, you might start to narrow down that this child probably has a rare disease, maybe we should do a sequence on that child," Rutter said.

"Doing it in that symptomatic or phenotypic way might get us to a diagnosis faster," she added. "Getting treatments [to patients] earlier and faster is the goal."

Eric W.K. Sid, MD, MHA, program officer within the NCATS Division of Rare Diseases Research Innovation, told MedPage Today that the agency's IDeaS pilot program attempted to quantify the number of patients and direct medical costs for 14 rare diseases within four different health system databases. It revealed that rare disease patients were difficult to quantify using ICD coding, resulting in under-counting, and that diagnostic journeys were prolonged.

From that project, the agency developed its Shortening the Diagnostic Odyssey program, which is funding three projects that are exploring three different approaches to speeding up the diagnostic timeline for rare diseases.

"These [projects] are looking at how to pull this into some type of algorithm so that when you're seeing patients who've been seen by multiple specialists, or have hit several different milestones from a clinical perspective that would be concerning for a rare genetic disease, that they would then get an expedited genetic evaluation," Sid said.

ICD codes can be part of that algorithm, but the overall strategy is about more than just those codes, Sid noted.

Rutter proposed developing a "zebra trigger" using these methods for a physician who doesn't typically see rare disease patients.

According to Sid, another research challenge is trying to get answers about prevalence and cost burden from the data that are available. For instance, even with an ICD code, it may not be easy to assess overall prevalence because data are siloed within different health systems.

"If your research question is to capture how many patients might have this rare disease, what are the systems that they're in, and how do we start to interrogate those systems, have access to their data, and start to make those assessments?" he asked.

Several groups, including NCATS and the EveryLife Foundation, have attempted to quantify the overall healthcare burden of rare diseases in recent years. Rutter noted that the groups have come to similar conclusions, even though they used various approaches to get there -- and that includes an estimated cost of about $400 billion annually in the U.S. That's among the estimated 10,000 rare diseases affecting some 30 to 40 million people in the U.S. (and over 350 million people globally). The work has helped make the case that "rare diseases are a public health challenge," Rutter said.

Expanding ICD codes to more rare diseases can help move the field forward, she added, but "they're not necessarily sufficient for getting us to where we need to be in terms of getting an accurate diagnosis to get that patient treated as fast as possible."

"We don't need to wait to have ICD codes for everybody," she said. "We need to understand a path for how we can get there, despite the fact that we don't have ICD codes."

Comment