Back to main
StoryPatient experienceOutcome
K Health · Case Study

Designing Continuous
Care for Hypertension

An AI-supported chronic care experience that shifted hypertension management from isolated visits to ongoing care.

Patient measuring blood pressure at home
Role
Sole Product DesignerLed discovery, UX strategy, UI and prototyping across the full MVP.
Duration
3 monthsAn MVP designed for K Health and Cedars-Sinai.
Team
Cross-functionalProduct · Clinical · Data Science · Engineering
Scope
End to endDiscovery · UX · UI · Prototyping
Core shift

From reactive care to proactive, continuous care

Between visits, blood pressure often goes unmanaged. Follow-up can be inconsistent, patients miss reviews, or do not always stick to the plan they were given.

K Health and Cedars-Sinai chose hypertension as the first pilot in the existing app, with the goal of expanding continuous care to other chronic conditions.

Why hypertension
A strong first condition

~40% of U.S. adults have hypertension. Blood pressure can be measured at home and tracked regularly, making it a natural place to start.

For Cedars-Sinai
Better control, lower cost

Better blood-pressure management could mean fewer avoidable hospitalizations and lower costs.

For K Health
A new kind of engagement

The program could give patients a reason to return between visits, increase overall app use and visits, and create a model that could later be offered to other partners.

My role
I was the sole designer on the project, responsible for the experience end to end, from discovery through implementation.

I worked closely with the Product Manager and with Clinical, Data Science and Engineering to translate product, clinical and technical requirements into a clear patient and clinician experience.

Discovery & synthesisUsed research to identify design requirements, gaps and constraints.
Experience & interaction designShaped the end-to-end patient journey, onboarding and core monitoring interactions.
Product design & deliveryDesigned the detailed product flows and interactions, and carried them through implementation to ensure the final product matched the design.
Cross-functional collaborationPartnered with Product, Clinical, Data Science and Engineering from exploration through implementation.
Discovery

Understanding the care model before designing the product

01 · People

Patients and primary care physicians

I spoke with people managing hypertension and with primary care physicians to understand the experience between visits and where uncertainty appears at home.

02 · Existing solutions

Hypertension and chronic-care products

I reviewed existing products to compare approaches to logging readings, feedback and ongoing monitoring.

03 · Care model

Care model review

I reviewed the clinical flow with Product, Clinical and Data Science, identified gaps and missing definitions, and worked with the team to clarify what was needed for the patient and clinician experience.

From discovery to decisions

Three principles guided the design

01
Build a scalable home for continuous care.
02
Keep measurement capture structured and conversational.
03
Connect the patient and clinician experiences.
Platform thinking

Creating a real home for ongoing care

My Health became the home for the hypertension program, combining blood-pressure readings with lifestyle factors like nutrition and activity. I designed it so it could later support other long-term care programs.

My Health overview screenMy Health navigation and modular content screen
Swipe to explore
System flow

How the continuous-care loop worked

The program worked as a loop: invite, onboard, measure, interpret, guide and return to the next reading.

StartPush notification/home screen Eligible users receive a pushnotification to enroll.(Or once enrolled, reminders to login readings as medically indicated). Onboarding First time reading Within thenormalrange? No Severe(too high/low) Initiate visit CTA Visit waiting room Doctor visit Next reading Elevated Follow-up questions(Generative AI) User replies Personalized messageguiding the user tochange cadence and/orschedule a visit within x days. Yes Feedback:Great, your reported values arewithin the normal range, see youat the next reading! End
StartPush notification/home screen
Eligible users receive a push notification to enroll. Once enrolled, reminders to log readings as medically indicated.
↓
Onboarding
↓
First time reading
↓
Within the normal range?
Yes · Normal
Feedback: Great, your reported values are within the normal range, see you at the next reading!
↓
End
No · Elevated
Follow-up questions (Generative AI)
↓ User replies
Personalized message guiding the user to change cadence and/or schedule a visit within x days.
↓
Next reading
No · Severe (too high/low)
Initiate visit CTA
↓
Visit waiting room
↓
Doctor visit
↓
Next reading
The elevated and severe paths return to the next reading, continuing the care loop.
Enrollment

Meeting patients where they already were

Enrollment had to feel native to the existing app. Eligible patients could be invited through a push notification, while the same prompt appeared naturally inside the homepage task area, making the program visible at the right moment without creating a new destination.

Push notification inviting an eligible patient to enroll Hypertension monitoring task shown in the existing K Health home screen
Onboarding

Building confidence before the first reading

The first-time experience had to do more than explain the program. It needed to reassure patients, confirm that they had the right device and make them feel ready to begin.

Program welcome screenBlood pressure monitor guidance screenHow to measure blood pressure correctly
Swipe to explore
Core interaction

Balancing structure with conversation

I explored two ways to enter a blood-pressure reading. Free text fit K Health’s chat experience and left room for related questions through our medically validated Knowledge component. That was useful for hypertension, where diet and exercise can affect blood pressure. But open input also made it easier to drift from the task. A standard form was focused, but felt disconnected from the rest of the product. We chose a structured widget inside the conversation, with clear fields and immediate feedback after submission.

Patient reading flow screen 1Patient reading flow screen 2Patient reading flow screen 3Patient reading flow screen 4Patient reading flow screen 5
Swipe to explore
After the reading

Different readings led to different next steps

The care path adapted to each reading: normal readings received feedback and returned to the next measurement based on cadence; elevated readings triggered follow-up questions and tailored next steps; severe readings moved directly toward care. I translated that logic into clear patient-facing states, questions and actions.

Elevated readings
Elevated readings flow
Swipe to explore
Severe readings

When urgency mattered, remove uncertainty

For severe readings, the experience reduced conversation and moved directly toward care, with a clear route to an immediate visit.

Severe reading flow: move from the reading directly toward an immediate visit
Swipe to explore
Clinician workflow

Working inside the existing clinician product

The experience stayed inside the clinician product physicians already used. It covered the key patient states: not yet enrolled, with guidance on how to introduce the program; declined; and enrolled, with the relevant program data available to the physician.

Pre-enrollment widget inside the clinician productPost-enrollment insights inside the clinician product
Implemented experience

End-to-end walkthrough

The final flow brings the individual decisions together: program entry, onboarding, structured measurements within the conversation, immediate feedback and ongoing monitoring.

Outcome

The product was only onepart of sustained engagement.

In the first three months after launch, we reviewed patient conversations and compared them with engagement over time.

What we found

Patients who stayed engaged were often those who had received a clear, detailed explanation of the program from their physician.

What it meant

Engagement depended on more than the product. A strong handoff from the care team mattered, and continued reinforcement would be needed over time.

Next step

Personalize push reminders around each patient’s measurement routine and timing.

Longer-term measurement

Cedars-Sinai later shifted priorities for reasons unrelated to product performance, so we could not measure longer-term outcomes.

Continue exploring

Next case: Pediatric Body Picker.

View next case→