Leveraging AI Within Rare Disease Care While Maintaining Humanity

Patient data can reshape research in real time with the utilization of artificial intelligence (AI) to better address evidence gaps in rare disease research, said the speakers of a panel discussion. The discussion, which was entitled “The Power of Patient Data: Leveraging AI to Address Evidence Gaps in Rare Diseases” and held at the Center for Innovation & Value Research 2025 Methods Summit on March 19 in Washington, DC, featured strategies for the implementation and refinement of research by leveraging patient data to ensure that research remains aligned with evolving patient needs and clinical goals.

Image credit: GamePixel | stock.adobe.com

The panelists began by highlighting their experience when generating evidence for rare diseases and their treatments. They acknowledged that the definition of “rare” can vary. As an example, they said that individuals may define rare diseases based on their own lived experience, whereas the FDA would look toward statistics and prior research. Regardless of the definition of rare disease, Sarita Edwards, MHA, CEO and president of The E.WE Foundation, explained that the biggest challenge is often finding the resources to understand a rare disease. She cited her own personal experience as a mother of a child with a rare disease.

“The challenge was really finding resources, [which is] something that seems so simple, yet is so difficult when you’re navigating a rare diagnosis… There’s limited information. A lot of the doctors that you encounter, they’ve learned about [rare diseases] in medical school, but they’ve never actually treated a patient with the condition, [or] that patient has died,” Edwards explained.

“One of the biggest challenges [as an organization] is helping families find the resources that they're looking for, whatever that looks like," Edwards continued. "The challenge there is, a lot of the resources that you find, when you find them, you find barriers…When you uncover a resource that, as a group, we feel can be supported to this family, once we connect them, they're still challenged with actually getting the resource and being able to utilize it. Resources and [having] access to resources and being able to actually put them into practice for their personal health stories and our personal health story have been a huge challenge for us."

Following this, Harsha K. Rajasimha, PhD, MS, CEO and founder of Jeeva Clinical Trials, Inc., explained how the US is shifting from “wisdom-based” medicine to “evidence-based,” in which patients seek evidence to support their decision-making involving therapies or treatments. In the case of rare diseases, those answers either may not be present or there may not be enough data generated. Echoing this is Simu Thomas, PhD, VP, global head of Health Economics and Outcomes Research, and global medical communications and training of Alexion AstraZeneca Rare Disease, who pointed to AI as a resource for the rare disease space.

Additionally, Thomas described how AI can be a method of streamlining patient information to better understand and better communicate with health care professionals. Patients also have reported using ChatGPT or Gemini to curate questions for health care professionals, allowing them to feel more empowered and comfortable having discussions during their visits.

“I believe—strongly believe—that [AI] has an immense opportunity for rare diseases because it can break silos. It can bring information to where it needs to be; it can be the voice of the rare disease patient when the patient is not able to voice [their experiences]. I do believe strongly in technology helping this situation,” Thomas said during the panel.

Alternatively, for health care professionals, AI can be a strong tool to improve efficiency and reduce tedious tasks, allowing them to preserve their energy and focus on more important issues. In certain cases, AI may be able to help professionals diagnose and identify rare diseases that previously went unrecognized. Early diagnosis can help treat patients earlier, potentially improving patient outcomes.

Rajasima compares AI to a “hammer looking for nails,” explaining that in health care and in evidence generation, AI can help reduce costs while improving both speed and patient centricity. It can identify errors and inconsistencies within foundational data, allowing health care professionals to determine how to address these challenges, whether it be patient recruitment, retention, or lessening burdens on both patients and health care professionals.

The panelists concluded that, while AI can be beneficial for all parties within health care, caution must be exercised. Additionally, important ethical considerations may need to be made while implementing machine learning within practice. Above all, technology can never replace the humanity and personal touch that are required throughout the treatment process.

“I don't think there is a substitute… We have to leverage AI, but not replace humans. And that is something that we consistently and conscientiously have to pursue. That's what makes us human,” Thomas concluded. “We have to [also] consider that what gives us meaningful purpose, and everyone has a different purpose, and so to keep the human in the loop is really important when we are using technology.”

REFERENCE

Edwards S, Rajasimha H, Thomas S. The Power of Patient-Data: Leveraging AI to Address Evidence Gaps in Rare Disease. Presented at: Center for Innovation & Value Research 2025 Methods Summit; Washington, DC. March 19, 2025.