Designing Ethical Health Tech for Rare Diseases

Published: 12 May 2025

Reading Time: 3 minutes

In rare disease research, data is often framed as the great equalizer. the asset that will finally unlock answers for the millions living with poorly understood conditions. But data without context isn’t just ineffective, it’s dangerous.

Artificial intelligence, wearable technologies, and decentralized trial platforms promise to accelerate discovery. Yet when designed without lived experience or structural awareness, these tools risk widening disparities and eroding trust.

As data justice advocate Virginia Eubanks writes, “Predictive models and automated decisions reflect the biases of the people who build them. Data is never neutral.”¹

For the rare disease community, where populations are small, marginalized, and often invisible to dominant systems, the ethical design of health tech is not a side issue. It’s the foundation of equity.

The Opportunity and the Risk

In theory, technology can solve major rare disease barriers:

AI models can identify drug repurposing opportunities at scale
EHR integrations can flag misdiagnosed or overlooked symptoms
Wearables can generate longitudinal data outside clinical settings

But in practice, these innovations often replicate the very gaps they claim to close.

Algorithms trained on majority datasets overlook minority phenotypes
Data collected from privileged populations skews research priorities
Platforms built without patient collaboration ignore accessibility and consent

In one 2020 study published in Science, researchers found that a widely used clinical algorithm in U.S. hospitals underestimated the health needs of Black patients by nearly 50%—not due to intent, but because historical data was biased.²

The lesson is clear: technological promise without ethical grounding becomes harm at scale.

What Ethical Health Tech Looks Like

For rare disease innovators, ethical design means embedding contextual intelligence—the insights of patients, community partners, and equity practitioners—into every phase of development.

Key principles include:

1. Co-Design with Lived Experts

The National Academy of Medicine recommends that health technology teams include patients “not as advisors, but as design partners with decision-making power.”³ In rare disease, where patient expertise is often deeper than clinical literature, this is especially urgent.

2. Data Sovereignty and Transparency

Patients must have control over how their data is used, shared, and monetized. This includes clear consent models and shared ownership where feasible.

3. Bias Auditing and Inclusive Datasets

Tools should be tested not only for accuracy but for equity. As AI ethicist Timnit Gebru has argued, “You don’t get fairness by accident. You get fairness by design.”⁴

4. Access and Accessibility

Platforms must account for disability, digital literacy, language, and affordability, not as afterthoughts but as core features.

A Case Example: EveryCure

EveryCure, a nonprofit focused on unlocking the full potential of FDA-approved drugs, is applying AI to identify repurposing opportunities for rare and understudied diseases. What makes their approach ethical is not just the algorithm; it’s the ecosystem around it.⁵

They collaborate with patient organizations to prioritize conditions
They make their database open-access to foster transparency
They advocate for regulatory alignment and public-sector partnerships

This combination of data plus context ensures that technological innovation serves those most often left behind.

Elevate Impact’s Approach

At Elevate Impact, we’ve worked with biopharma firms, hospitals, and public health entities to build ethics-first design strategies for rare disease innovation. Our model integrates:

Component	Description
Equity Impact Assessment	A pre-launch tool to evaluate whether health tech products align with DEI goals and patient needs
Lived Experience Advisory Boards	Not just patient testimonials, but embedded decision-makers at every design stage
Cross-Sector Review Panels	Input from ethicists, regulators, clinicians, and data scientists to ensure inclusive design frameworks
Training for Tech Teams	Hands-on workshops on bias in AI, equitable data sourcing, and community accountability

Conclusion

Data is not destiny. It is only as valuable as the context, consent, and community that shape it.

In rare disease research, where each data point may represent a life at the edge of visibility, ethical design is non-negotiable. Healthcare leaders, funders, and technologists must treat equity not as a downstream benefit but as an upstream design principle.

Because when tech is built with the wrong assumptions, it doesn’t just miss the mark. It risks missing the people entirely.

References:

Eubanks, V. (2018). Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor. St. Martin’s Press.
Obermeyer, Z., et al. (2019). “Dissecting racial bias in an algorithm used to manage the health of populations.” Science, 366(6464), 447–453.
National Academy of Medicine. (2022). Patient Perspectives in Digital Health Innovation. The National Academies Press.
Gebru, T. (2020). “Race and Gender Bias in AI.” MIT Technology Review.
EveryCure. (2023). Unlocking the Full Potential of Approved Medicines. https://everycure.org