The integration of multiple real-world data sources is transforming how we understand patient outcomes, medical device safety, and healthcare delivery. By leveraging innovative digital health platforms, researchers can now gather comprehensive information directly from patients, electronic health records (EHRs), pharmacy systems, and personal digital devices in near real-time. This approach fosters a more patient-centric, efficient, and scalable way to generate meaningful evidence to inform clinical practice, regulation, and policy.
Our recent study demonstrates the feasibility of utilizing a patient-centered health-data-sharing platform to seamlessly consolidate data from multiple sources over an eight-week period, specifically in patients undergoing procedures like bariatric surgery and atrial fibrillation ablation. This methodology exemplifies the potential for digital health technologies to facilitate pragmatic research, improve patient engagement, and support regulatory decision-making by providing robust, real-world evidence.
Introduction
Medical products—ranging from pharmaceuticals to medical devices—are essential elements of modern healthcare, offering significant benefits when used appropriately. These products are regulated by agencies such as the Food and Drug Administration (FDA), which increasingly rely on real-world data (RWD) to evaluate their safety, effectiveness, and performance over the product lifecycle. As noted in the FDA’s guidance, leveraging RWD derived from diverse sources like EHRs, registries, and patient-generated data enhances postmarket surveillance efforts.
However, integrating data from disparate sources remains a significant challenge. Most RWD sources are siloed, often lacking interoperability, which hampers comprehensive analysis. For instance, a patient’s electronic health information from different healthcare systems is typically inaccessible due to incompatible systems or privacy restrictions. Similarly, registries focus on disease-specific data but usually lack longitudinal follow-up or integration with other data streams.
Initiatives like the Blue Button project have empowered patients to access their health data in digital formats, promoting individual agency. Yet, accessing data across multiple healthcare providers remains cumbersome unless unified platforms are employed. The rise of mobile health (mHealth) technologies offers a promising solution: applications and APIs can connect to various data sources, enabling aggregation and enabling more comprehensive patient monitoring and research.
Despite these advances, no prior study has successfully combined personal digital device data, patient-reported outcomes, EHRs from multiple health systems, and pharmacy records into a unified, patient-controlled data repository for pragmatic research. Achieving this integration can accelerate understanding of device safety and effectiveness, improve patient recovery tracking, and foster a more personalized approach to healthcare.
Our investigation tested the feasibility of using Hugo, a patient-centered health-data-sharing platform, to aggregate these diverse data streams. Hugo enables patients to access and share their health information, including EHRs, pharmacy data, PROMs, and data from personal digital devices, via any smartphone or tablet. The platform can connect with multiple healthcare systems, regardless of where patients receive care, and updates data in near real-time as devices are synced.
We conducted an eight-week prospective cohort study involving patients undergoing two procedures involving medical devices: bariatric surgery (including sleeve gastrectomy and gastric bypass) and catheter-based atrial fibrillation ablation. The goal was to assess how effectively Hugo could collect and integrate data from various sources, supporting clinical research and potential regulatory evaluation.
Results
Patient Demographics and Enrollment
Between January and October 2018, 77 patients were screened; 11 declined, 6 were ineligible, and 60 consented to participate—30 at each site, evenly split between the two procedures. The average age was 55.2 years, with a slight female predominance (58.3%). Enrollment times averaged around 70 minutes, with some participants completing parts remotely. The final follow-up was completed by December 2018.
Electronic Health Record Data: Collection and Validation
Hugo successfully linked to the EHR portals of Yale-New Haven Hospital and Mayo Clinic for all participants. Patients who received primary care at these institutions provided their portal credentials, or, in Mayo Clinic’s case, data were uploaded directly due to EHR vendor transitions. In total, data from 13 different EHR systems were aggregated.
Validation showed that 82% of encounters appeared in both Hugo and the original EHRs, with high concordance in encounter dates, types, and diagnoses—over 90% agreement. Discrepancies were mainly due to system upgrades or account lockouts, which Hugo now monitors to ensure data continuity.
Pharmacy Data
Among participants, 41% linked their pharmacy accounts from CVS or Walgreens, enabling aggregation of prescription data, including medication names, dosages, and refill information. Patients using other pharmacies lacked API integration at the time, limiting data completeness but still providing valuable insights into medication adherence post-procedure.
Personal Digital Device Data
Device data collection was robust, with weekly sync rates decreasing slightly over time but remaining substantial. Fitbit devices provided step counts and sleep data, with most patients syncing at least once weekly. Patients with bariatric surgery and atrial fibrillation ablation showed similar engagement rates. Additional devices—such as Withings scales for weight and Kardia Mobile for ECGs—also contributed data, enabling detailed monitoring of recovery patterns.
Analysis revealed expected trends: step counts increased after procedures, weight decreased significantly in bariatric patients, and heart rhythm data aligned with clinical expectations. These digital biomarkers offer real-time insights into patient recovery trajectories.
Patient-Reported Outcomes (PROMs)
PROM completion rates were high, with over 80% of questionnaires completed across the cohort. Patients reported on pain, appetite, palpitations, and other symptoms at scheduled intervals. Most completed questions thoroughly, with median response times of under 30 seconds per survey. Data demonstrated typical recovery patterns: pain decreased over time, appetite gradually increased post-bariatric surgery, and palpitations declined after ablation.
Patterns of Recovery
Digital device data revealed that patients generally regained activity levels within days after procedures. Weight loss in bariatric patients was pronounced over the study period. PROMs reflected expected symptom improvements, confirming the platform’s ability to monitor recovery dynamically.
Discussion
This study confirms that a patient-centered health-data-sharing platform like Hugo can effectively aggregate diverse data sources—including EHRs, pharmacy records, PROMs, and personal digital device data—in near real-time. Such integration enhances the understanding of patient recovery, device performance, and treatment safety, offering a pragmatic model for future research and regulation.
The FDA’s increasing reliance on real-world evidence underscores the importance of comprehensive data aggregation. Initiatives like the National Evaluation System for Health Technology aim to support this goal by facilitating data sharing and validation across multiple sources. Removing data silos and improving interoperability are critical steps toward more accurate assessments of medical device and drug performance.
Digital health advances, including API updates, mobile reminders, and sensor technology, will further improve data collection quality and participant engagement. Connecting personal devices directly to cloud platforms reduces user burden, increases data fidelity, and enables continuous monitoring—pivotal for evaluating complex procedures and long-term outcomes.
Our patient-centric approach empowers individuals to access and share their health data, aligning with efforts like Medicare’s Blue Button 2.0, which provides millions of beneficiaries with access to their claims data. As health systems expand data availability, the potential for comprehensive, patient-driven research grows, ultimately enhancing clinical care and regulatory oversight.
Limitations include the small sample size and short follow-up, but the study’s success suggests scalability. Challenges like sync failures, data validation, and ensuring regulatory-grade data quality remain, but ongoing technological improvements and validation strategies will address these issues.
In conclusion, integrating multiple real-world data sources via a patient-focused platform is feasible and promising. It offers an innovative pathway for evaluating medical products, understanding patient recovery, and advancing personalized medicine in a regulatory environment increasingly supportive of real-world evidence.
Methods
Study Population and Enrollment
Patients scheduled for bariatric surgery or atrial fibrillation ablation at Yale-New Haven Hospital and Mayo Clinic were recruited. Eligibility required adults over 18, English fluency, a compatible device, and an email or willingness to create one. Patients received detailed information, provided informed consent, and were enrolled in Hugo. Enrollment targeted 15 patients per procedure per site, with follow-up extending eight weeks post-procedure. The study was approved by institutional review boards and registered at ClinicalTrials.gov.
Data Sources and Collection
Hugo linked to participating health system portals, including Epic and Cerner EHRs, to access clinical data. Patients provided credentials for their portal accounts. For Mayo Clinic, data were uploaded directly due to EHR vendor changes. Pharmacy data from CVS and Walgreens were linked via portal accounts, capturing medication details. Personal digital devices—Fitbit Charge 2, Withings scales, and Kardia Mobile—were provided, set up, and linked to Hugo accounts. Patients were instructed to sync devices weekly, with reminders sent via email and later via SMS to improve adherence.
PROMs were delivered through Hugo’s secure platform, with questionnaires tailored to procedures. Post-procedure recovery PROMs focused on pain and appetite or palpitations, administered twice weekly. Disease-specific PROMs, including NIH PROMIS® and Cardiff questionnaires, were sent at predefined intervals.
Data Validation and Analysis
Collected data were validated by comparing encounter details, diagnoses, and medication records with original EHRs, ensuring high concordance. Discrepancies were analyzed and addressed through platform updates. Descriptive statistics summarized demographics, engagement rates, and clinical patterns. Digital biomarker trajectories were plotted over time, and response rates for PROMs were assessed. Statistical comparisons between groups were performed using chi-square tests, with significance set at p<0.05.
Future Directions
Advances like API enhancements, device connectivity, and patient engagement strategies will further improve data completeness and quality. Integration of device identifiers into EHRs, better handling of missing data, and validation against regulatory standards are ongoing priorities to translate this feasibility into scalable, regulatory-grade evidence generation.
References and supplementary materials are available upon request.