Connecting directly with individuals can transform our understanding of health beyond traditional healthcare interactions, according to Leslie Oley Wilberforce, CEO of Evidation Health. By establishing trusted relationships with patients and leveraging technology to gather diverse data streams, researchers and healthcare providers can develop a more holistic view of health journeys.
"Patients are really one of the best sources of data and insights about their own health. By going directly to individuals, you can reveal the full patient experience in quite a different way."
Traditional real-world data often comes from episodic healthcare system interactions, creating significant gaps when patients aren't regularly engaging with providers. Evidation's approach bridges these gaps by creating what they call "behavior grams" – layered visualizations combining passive data collection from wearables with active patient-reported information.
The Power of Layered Data
Wilberforce demonstrated this approach through several examples, including a 65-year-old woman with type 2 diabetes. Her behavior gram combined wearable data tracking BMI, steps, aerobic activity, sleep duration, heart rate variability, and glucose levels from a continuous glucose monitor. This passive data collection was supplemented with validated questionnaires measuring depression and anxiety.
This layered approach revealed valuable insights about the patient's health journey, including her switch to a GLP-1 medication and subsequent weight loss, followed by participation in a digital health coaching program. The combined data showed corresponding decreases in glucose levels that might not be captured in traditional healthcare records.
"Some of this data, especially today with how people are accessing GLP-1s, doesn't exactly always exist in your EHR data with the prevalence of people getting GLP-1s prescribed from Noom and hims and hers or from telehealth providers," noted Wilberforce. "This is a really different method to try and pull the full picture together."
Bridging Data Sources
The approach doesn't replace traditional healthcare data but rather complements it. Wilberforce emphasized that policy and technology changes over the past decade have empowered individuals to mediate and control their health data sharing.
"Ten years ago, most patient data from the healthcare system was largely locked in the healthcare system," she said. "The really important shift over the last number of years, through policy, through technology and a number of other things, is that there's many more ways for individuals to mediate or choose who their data gets shared with."
This patient-mediated approach allows researchers to address data gaps by going back to individuals with targeted questions. It also enables personalized health interventions at precisely the right moment, potentially improving outcomes through better timing and contextualization.
Case Studies Demonstrating Value
Wilberforce shared three case studies illustrating the practical applications of this approach:
1) Measuring medication impact on alertness: A study of 500 individuals taking OTC allergy medication combined surveys, vigilance tests, and wearable data to demonstrate improved attention and reduced sleepiness on days when medication was used, with no impact on activity levels or heart rate.
2) Understanding GLP-1 medication usage: Within a weight management cohort of 140,000 people, Evidation discovered that 40% of GLP-1 users obtained prescriptions from providers other than their primary care physician – information likely missing from traditional EHR data. The study also revealed that depression scores (measured by PHQ-8) decreased while patients were on GLP-1 therapy and slightly increased after discontinuation.
3) Early detection of respiratory illness: Using machine learning algorithms analyzing wearable data patterns from 200,000 individuals, Evidation identified over 28,000 instances of early flu-like symptoms. This early detection enabled timely interventions and facilitated enrollment in a clinical trial testing therapy to reduce household flu spread.
"The challenge our partner was having there was like, this is really hard to enroll because by the time the person shows up at the provider, they've probably been sick for a few days and they're outside of the early onset window where we actually need to get them the drug," Wilberforce explained.
Building Trust and Engagement
Critical to Evidation's success is their approach to privacy and user control. The company established early principles that users control and own their data, with transparent consent processes for how data is used. This foundation of trust enables them to collect increasingly sensitive health information.
User engagement is maintained through both extrinsic motivators (rewards and gamification) and intrinsic value (providing insights and research participation opportunities). This has allowed Evidation to build a community of 5 million people across 97% of U.S. zip codes, with over a billion data points collected daily.
Five Key Takeaways
1) Direct patient connections reveal health insights not captured in traditional healthcare data – "By going directly to individuals, you can really reveal this full patient experience in quite a different way," as Wilberforce emphasized.
2) Layering multiple data sources creates a more complete picture of an individual's health journey, with passive wearable data and active surveys contextualizing each other.
3) Patient-mediated data sharing bridges critical gaps in healthcare records, particularly for treatments obtained outside traditional healthcare settings like GLP-1 medications prescribed through telehealth.
4) Early detection algorithms can identify health changes before symptoms become severe, enabling timely interventions that improve outcomes and research opportunities.
5) Building trust through transparency and user control is essential for successful data collection: "If people didn't trust us to be good stewards of their Fitbit data, they would never trust us to be good stewards of electronic health record data or any more sensitive types of data."
For more market access & RWE insights, including latest articles, interviews and more – click here