Data collection

We deploy online platforms and uses proprietary tools to simplify the collection of verbatim patient information about their problems, how the problems are affecting their daily functioning, treatments, and preferences.   

 

Data analytics

From the verbatim data, we compose a standardized and informative problem list that is date-stamped and linked to the medications that patients say they are taking. We then apply systematic and intelligent data analytics to create a patient narrative that is clear, coherent and accurate. We then deploy NLP and machine learning (ML) techniques to analyze patient input about their illness and quality of life in ways that facilitate patient, clinician, and researcher use of these vital data.

 

Clinical curation and patient validation

Our clinical research consultants review, collectively and independently, the verbatim problems and functional consequences that patients report in order to better categorize NLP terms for the disease of interest. The clinical curation process contextualizes the verbatim reports, increases the granularity of term assignment by about 50%, and informs machine learning to develop a clinically meaningful ontology.  The clinically curated and contextualized NLP output can be applied to the creation of cross-sectional and longitudinal patient-reported natural histories of disease, refinement of clinical outcome assessments (COAs), and the development of patient-reported endpoints for clinical trials.