Health Notes

A speech-enabled tablet for rural clinics that turned dense health surveys into guided, spoken interactions, boosting completion to 96% and restoring trust in Appalachian communities.

Role
Design Lead

Timeline
8 weeks (field work + iterative design)

Team
Researchers
(U. Pitt, WVU) Clinicians Developers Community liaisons

Compliance
HIPAA-compliant IRB approved

The Problem

An NIH-funded study provided underserved Appalachian families with dental care in exchange for completing health surveys. But over time, the experience broke down. The paper surveys were dense and academic, riddled with errors, and nearly impossible for participants to understand. Long clinic waits often ended with families leaving without the care they were promised. As trust eroded, participation plummeted, putting both the integrity of the data and the future of the research itself at risk.

How might we create a digital experience that enables large-scale, accurate health data collection without compromising trust, dignity, or access to care?

The answer didn’t arrive fully formed.

This project unfolded across multiple iterations, each shaped by what we observed in the field, beyond what had been scoped.

We started by:

Replacing paper surveys with a digital tablet flow
Enabling speech guidance for users with low literacy
Designing for offline data capture and fast transfer

Then, as we embedded in clinic environments, new problems surfaced:

Clinicians were hand-drawing complex family trees—causing massive delays
Patients were waiting all day and still leaving without care due to overflow emergencies

We adjusted course every time, expanding our design scope beyond UI into service delivery, trust repair, and systemic change.

Key Design Decisions

From Paper Chaos to Controlled Conversations

We transformed a 120+ question academic survey into a voice-guided tablet experience.
Participants could toggle between listening and reading, depending on comfort level to reduce friction without sacrificing autonomy.

📷 [Before/after: academic paper vs. simplified interface with voice support]
📷 [UI showing audio on/off toggle]

Digitizing Onboarding After a Field Discovery

During in-clinic observation, we discovered clinicians were spending hours manually creating three-generation family trees.
I redesigned the onboarding process to allow fast digital input that generated these trees automatically—freeing up staff and significantly reducing wait times.

📷 [Input screen for family info]
📷 [Auto-generated tree display]

Reorganizing Service Delivery to Rebuild Trust

We didn’t plan to redesign clinic ops—but field testing showed families often left without dental care because emergency cases derailed scheduling.
I collaborated with the research team to bring in overflow dental teams during survey days, ensuring no one left empty-handed.

This was a service design move rooted in user advocacy, and it paid off.

📷 [Revised clinic staffing diagram]
📷 [Participant quote or attendance stat]

The Breakthrough: Follow the Pain, Not the Plan

None of these system shifts were in the original brief.
But observing real-world pain points—and acting on them—was what allowed the design to succeed.

The biggest insight?

Designing for trust requires going beyond the screen.

Once participants saw they’d be treated with respect and receive care, participation rebounded.

Results

90% improvement in data accuracy
2× increase in participation and completion rates
Data transfer time dropped from days to hours
Clinic attendance fully recovered
Program now under consideration as a model for other rural research sites

Real Design Requires Real Proximity

This work proved that field presence isn’t optional—it’s essential.

We didn’t just ship a tablet. We rebuilt a fractured ecosystem:

From paper to digital
From waiting to care delivered
From mistrust to re-engagement

And we did it by listening—not just to users, but to the systems they’re stuck inside.