Scaling Care with Smarter Design
Company:
Grayce
Role:
UX research / UI Design
Date:
March 2023 - Present
Company Overview
Through personalized expert guidance and a comprehensive tech platform, Grayce assists caretakers manage a wide range of responsibilities, including navigating insurance, finding care facilities, and providing emotional support. Its services aim to reduce stress for caregivers, improve employee well-being, and enhance workplace productivity.
We were making the pivot from an employer benefits model to a payer market strategy.
While our existing market focused on qualitative measures like caregiver stress and employee feedback, insurers required much more robust data.
Appealing to this new market relied extensively on providing substantial data and analytics to assess risk, evaluate population health, and measure the effectiveness of our interventions.
We had no way to prove our value to payers because we weren’t tracking the value-based, population health metrics this market demands.
"Effective precision social care should reduce health plan spend and be able to quantify the dollar amount saved compared to randomized control groups within the same buyers population."
To meet payer requirements, we needed a system to capture and monitor the outcomes of our interventions and the member needs they addressed. This would provide actionable insights into population health, demonstrate the effectiveness of our platform, and quantify cost savings for payers.
The Process
Research and Discovery
Aligning on Payer Priorities
To kick things off, I partnered with cross-functional teams—Product, Clinical Operations, and Business Development—to uncover exactly what data matters most to payers.
We were currently categorizing member needs into 7 broad categories that did not give insights into population health trends:
By analyzing payer systems and government policy bodies, we pinpointed key SDOH (Social Determinants of Health) data standards we needed to be able to extract from our member's needs
We needed to gather this data from members in order to identify high-risk populations and quantify cost savings.
Auditing Historical Intervention Data
Our clinicians were prioritizing providing personalized care, resulting in nearly 20,000 separate entries of highly specfic, free-text interventions.
While this customization was valuable to members, it resulted in data that was difficult to aggregate and analyze at scale, meaning there was no way to assign time or money savings to Grayce's offerings.
Without a standardized way to classify the help we provided we could only quantify effectiveness on an individual basis. We needed population health outcomes.
Observing Real-World Workflows
To fully understand how care partners operated day-to-day, we conducted eight “Clinician Observation Sessions,” totaling 20 hours of in-depth shadowing. The clinical team shared their screens and walked us through real scenarios and daily tasks. These sessions revealed inefficiencies, inconsistencies, and opportunities to standardize processes, all of which informed our solution design.
The Solution
By designing a way to structure and categorize the thousands of needs our patients were reporting, and a separate system to organize the thousands of interventions our service providers were offering, we could provide payers with the data they needed to evaluate the impact of our services and make data-driven decisions.
We needed to find a way to extract thousands of needs into a system that categorized and filtered them in a way that was both usable by our internal service team and valuable to payers.
By blending manual inspection with AI-enabled classification, we reviewed tens of thousands of historical care deliverables and meticulously arranged them into specific social determinants of health (SDOH) categories.
Thousands of unique, one-off needs were consolidated into 90 categories aligned with industry-standard SDOH markers.
These needs were further grouped into high-level categories for easier search and navigation.
From 20,000 Interventions
to 5 Key Actions
Following a similar process to the one used to categorize patient needs, we did an audit of more than 20,000 interventions our clinicians provided and found that they could all fit under one of the following five actions:
Confirm (details or appointments)
Schedule ( appointments or tours)
Complete (applications, forms, or referrals)
Coordinate (setting up services)
Provide (resources, information, eligibility)
This standardization eliminated the need for free-text entries, which made it impossible to measure impact. This also relieved the burden of time-consuming UX writing for our services team.
Each action was tied
to a value metric
Implementing a system that categorized interventions into one of these five action types allowed us to attach concrete value metrics to each action.
For example, scheduling preventive care visits led to an X% reduction in hospital readmissions, while coordinating access to resources improved patient compliance rates by Y%. Connecting these metrics to specific actions within the platform allowed us to demonstrate how Grayce directly influenced health outcomes and reduced costs for payers.
Results
By designing a way to structure and categorize the thousands of needs our patients were reporting, and a separate system to organize the thousands of interventions our service providers were offering, we could provide payers with the data they needed to evaluate the impact of our services and make data-driven decisions.
This new system would aligning our members' needs with targeted interventions, delivering personalized support that resulted in measurable time and cost savings.