In the world of medical research, data drives discovery. But until recently, nearly 80% of data relevant to research has been hidden in clinical notes. These notes include detailed observations, symptoms, and treatment plans that are absent from the structured data fields available in most EHRs or claims data.
Unlock the potential of clinical notes with Truveta Data
What causes patients to stop taking their medications?
How do device complications vary by patient demographics?
How do symptoms like pain or fatigue predict disease progression?
None of these questions can be fully answered without the detailed insights hidden in clinical notes. With Truveta, researchers gain access to all notes generated during a patient’s care, including progress notes, nursing evaluations, procedure and operative reports, referral notes, discharge summaries, imaging reports, patient-reported outcomes, and more.
Billions of clinical notes across therapeutic areas
Truveta Data includes more than 120 million patients, 5 billion clinical notes, and 85 million imaging studies. The largest collection of clinical notes integrated with EHR data, Truveta Data is linked with SDOH, mortality, and claims data for the most complete view of the patient journey – at scale.
Achieving scale with the Truveta Language Model
Despite the promise of clinical notes for advancing research, accessing and analyzing this data has historically been fraught with challenges. The sheer volume of information contained in free-text fields within the EHR renders it impossible for humans to extract patterns at scale. Truveta has overcome these barriers with the most advanced, award-winning AI technology, the Truveta Language Model (TLM), which is used to extract the data from these notes and clean it to enable research.
Unlock access to any clinical concept of interest
With Truveta, researchers can access a continually expanding library of clinical concepts spanning diverse therapeutic areas and clinical scenarios, including severity and stage of disease, medication adherence, and disease symptoms.
Examples of condition-specific concepts available for research:
*Custom concepts can be extracted from notes and structured alongside EHR data.
Deep dive: cardiovascular and prostate cancer research
Study longitudinal view of prostate cancer patients
Clinical notes can be particularly valuable for studying oncology, where standardized coding systems often fail to consistently or precisely capture diagnoses, tumor staging, and other clinical events. For prostate cancer, the ability to access complex concepts such as medication details, family history, symptoms, surgery staging, Gleason scores can help lead to new insights, such as what can be learned from the 20% of prostate cancer patients with a family history of the disease.
The sample note below demonstrates examples of highly relevant patient information that can be extracted, structured, and made available to better understand this highly heterogeneous disease.
TLM extracts data from these notes and places them within the Truveta Data Model (TDM) for immediate researcher access within Truveta Studio. Additional granularity may include specific lab values, therapy dosage, comorbidities, additional SDOH, and more.
Advance cardiovascular research using echocardiogram data
With nearly 74 million distinct cardiac measurements across echocardiogram and cardiac catheterization reports, Truveta is the market leader in real-world data to advance cardiovascular research.
By leveraging TLM, Truveta extracts precise metrics that enable more advanced and nuanced queries. Echocardiogram data can be used for everything from tailoring treatment interventions, to comparative effectiveness, to long-term safety monitoring.
The figure below gives a visual of how this data was classified by severity.
The following figure leverages aortic stenosis severity classifications, along with TAVR procedure data, showing the difference between time to treatment based on severity.
Data extracted from clinical notes are critical for understanding factors associated with gaps in care and clinical outcomes. Using normalized echo data, Truveta Research was able to not only classify patients by severity of aortic stenosis, but also to classify their disease progression, and then study potential disparities in both the types of treatments provided and time to treatment.
A paradigm shift for research
Clinical notes represent the future of medical research. Their ability to capture the intricacies of the patient experience makes them indispensable for life sciences researchers aiming to push the boundaries of innovation, drug discovery and clinical trial development. With this more complete picture, researchers can help transform healthcare – delivering better outcomes and driving the next wave of breakthroughs in medicine.
Ready to unlock the full potential of clinical notes? Contact us today.