Risk scores have become an essential tool in healthcare, helping providers, insurers, and researchers estimate the likelihood of conditions and divide patients into risk groups. These tools can drive early interventions, reduce costs, and improve patient outcomes. But what happens when validated risk scores do not yet exist for certain conditions, or when organizations want to test new approaches before committing to large-scale clinical validation? At Thryve, we explore this through investigative prototypes—demonstrations of how wearable-derived data can be mapped to known risk factors and applied at the population scale.
Today, we explore what risk scores are and why they are so important for preventive health. Moreover, we illustrate our investigative approach using two conditions of growing importance: NASH (Non-Alcoholic Steatohepatitis) and Insomnia. These examples illustrate how we map known risk factors to wearable data and apply them to large-scale populations in order to visualize potential risk groups.
Risk scores are statistical tools designed to estimate the likelihood of an individual developing a specific disease or experiencing a health condition. They are widely used in healthcare to guide clinical decision-making, identify high-risk patients for early intervention, and allocate resources more efficiently. A risk score typically aggregates multiple variables into a single predictive number that can be compared across individuals or populations. These may include demographics, family history, lifestyle behaviors, comorbidities, or biometric data from clinical exams and wearables.
Examples of validated tools include the Framingham Risk Score for predicting cardiovascular disease and the CHA₂DS₂-VASc Score for assessing stroke risk in patients with atrial fibrillation. These scores are not diagnostic instruments but predictive aids that help clinicians, insurers, and policymakers focus attention and preventive strategies where they will have the greatest impact. They also create a shared language between stakeholders, allowing consistent comparisons and benchmarks across large populations.
Healthcare organizations often struggle with speed when it comes to data-driven innovation. Building validated scores requires years of clinical research, but many stakeholders, from insurers to digital health startups, need fast, data-backed signals to test ideas, design interventions, or explore new markets. Investigative risk scores fill this gap.
By mapping known risk factors to wearable metrics, Thryve demonstrates how hypotheses can be tested rapidly against population-level data. This allows organizations to:
Importantly, risk scores matter because they enable proactive healthcare by providing an early warning system that highlights who might benefit from additional testing, closer monitoring, or lifestyle interventions before costly or severe health events occur. This investigative-first approach helps partners derisk innovation, cut infrastructure costs, and focus on what really matters: translating data into decisions.
At Thryve. We support the development of investigative risk scores. We do this to help our partners quickly explore how wearable-derived data can be applied to pressing health questions without the long timelines of traditional validation studies. To showcase this capability, we have built two case studies: one focused on NASH (Non-Alcoholic Steatohepatitis) and one on Insomnia. These are not validated clinical scores, but rather prototypes that showcase how quickly Thryve can deliver actionable insights by leveraging wearable data across tens of thousands of users. For digital health organizations, the message is clear: if you have a specific health problem, Thryve can build a working data prototype within weeks, not months.
NASH is a progressive liver disease strongly linked to obesity, diabetes, and metabolic syndrome. It is often called a silent disease, as symptoms are minimal until significant liver damage has occurred. Traditional screening methods are invasive, costly, and not scalable. This makes NASH a perfect candidate to test whether wearable-derived data can help identify risk groups earlier.
In our prototype investigation, we mapped known NASH risk factors to wearable and lifestyle data streams:
Two types of scoring were applied:
The divergence between absolute and relative scoring illustrates both the potential and the limitations of early-stage approaches. While absolute scoring aligns with clinical benchmarks, relative scoring highlights patterns within large datasets that might otherwise remain invisible. This dual perspective is invaluable for organizations exploring population health strategies.
Insomnia, on the other hand, is one of the most common sleep disorders, affecting millions worldwide, yet it remains underdiagnosed and undertreated. Wearable devices, with their ability to track sleep behavior objectively, open up new ways of identifying at-risk populations.
For our prototype insomnia risk score, we drew on known risk factors:
The results were insightful:
Although not validated, this prototype stratification highlighted clear behavioral clusters. For instance, users with consistently short sleep paired with high stress markers (low HRV) aligned with known insomnia risk groups. Such insights demonstrate how wearables can provide scalable screening tools and open the door to early interventions.
Both NASH and Insomnia cases illustrate important lessons that highlight not just technical feasibility but also strategic implications for healthcare organizations:
NASH and Insomnia are just two examples of how Thryve can prototype risk scores quickly, using wearable-derived data to explore known risk factors and visualize population-level insights. Thryve’s API is uniquely suited to powering these prototypes:
For potential partners, this means reduced infrastructure costs, faster turnaround times, and the confidence that your prototypes are built on reliable, harmonized data.
Book a demo with Thryve today and learn how we help you uncover risk groups, test ideas, and turn health data into actionable insights!
