Utilizing Unstructured Data in EMRs for Clinical Trial Patient Identification:The Human-in-the-Loop Approach

The board of experts is comprised of three leading Cardiologists who are experts in their own sub-specialities within Cardiology.

Electronic Medical Records (EMRs) offer a rich source of patient data, but a significant portion remains
unstructured. Current industry data suggests that over 80% of data in EMRs is unstructured. This
means that many care providers are dealing with vast amounts of information that may contain
essential clinical and operational insights.1 Organizing this data, especially clinician notes, lab reports,
and diagnostic summaries, can yield valuable opportunities. Structuring this data can be especially
useful for outpatient cardiology clinics in identifying suitable candidates for clinical trials.

  1. Detailed Patient Profiles: Structured data provides a clearer view of patient profiles. This organized
    format can reveal important insights hidden in the unstructured data. By implementing NLP (Natural
    Language Processing and OCR (Optical Character Recognition) we are now able to structure this data.
  1. Efficient Patient Matching: Structured data can be more easily analyzed by algorithms. This speeds
    up and refines the process of matching patients to specific clinical trials.
  2. Time and Cost Savings: Manually searching through unstructured data is time-consuming and costly.
    It is estimated that identifying patients that meet both inclusion and exclusion criteria manually can
    take up to 4 hours per patient. 2
  3. Increased Data Accuracy: Structuring data significantly reduces errors and ambiguities, making
    patient-trial matches more reliable.
    The Human-in-the-Loop Advantage:
    Even after data structuring, incorporating a ‘human in the loop’ review can be
    beneficial. Algorithms are precise, but human oversight ensures a contextual
    understanding of the data. This combination results in data that’s both accurate and
    relevant, ensuring that patient allocations for clinical trials are appropriate and wellinformed.
    In conclusion, as outpatient cardiology clinics work to improve patient care, the structured utilization
    of EMR data, combined with a human-in-the-loop approach, offers a promising path forward. This
    approach not only optimizes data use but also ensures that clinical research decisions are wellinformed and patient centric.
    Written by:
    Naveed Ahmad – CEO and Co-Founder of Cardea Health
    1 https://www.accenture.com/us-en/blogs/insight-driven-health/unleashing-the-power-of-unstructured-data
    2 https://deep6.ai/resources/interview-with-clinical-research-coordinator

About Cardea Health

Cardea Health Inc. is revolutionizing how patient data can be utilized by Cardiologists. The
mission of Cardea Health is to provide real time actionable insights on a single pane of glass so
that physicians can improve the quality of care for cardiac outpatients. Cardea Health was
founded in 2021 by Mr. Naveed Ahmad, a serial entrepreneur and Dr. Andrew Yadegari, an
Interventional Cardiologist with the Scarborough Health Network. Cardea Health is located in
Toronto, Canada.

For more information or media inquiries, please contact:

Mr. Naveed Ahmad
Cardea Health/ CEO and Co-Founder
Naveed.Ahmad@cardeahealth.io