If you’re like me, you’re often fascinated by the various ways that technology and artificial intelligence (AI) continue to change the landscape of how we move about the world. Tasks that were not too long ago relegated to human responsibility, are now comfortably being placed in the figurative hands of our AI counterparts. Not only are they as smart as us (maybe even smarter?), they have the ability to mine important data via natural language processing (NLP).
Not to be confused with its acronymical counterpart, neuro-linguistic programming, natural language processing (NLP) is, to simplify greatly, a field of computer science that marries artificial intelligence, with spoken and written text, with or without specific codification. NLP software has the ability to “read” content and perform analyses without play-by-play direction from a human director. It can be interwoven into a variety of disciplines, including Psychology where I was first introduced to the broad concept, but is significantly associated with computer science and artificial intelligence.
Already proving themselves to be leaders in innovation and technology within the pharmaceutical industry, TriNetX has spent the last three and a half years successfully building an unparalleled global health research network, that works comfortably as the conduit for a symbiotic relationship between healthcare organizations, innovative biopharmaceutical companies, and contract research organizations. Research-oriented healthcare organizations allow access to their patient data in a “de-identified and secure way”, where visits, treatments, conditions and lab value are aggregated across a vast network of healthcare organizations globally, to better determine suitable cohorts for clinical trials. Headquartered in Cambridge, Massachusetts, TriNetX has a global reach, with contracts in over ten countries.
How Does The TriNetX Platform Operate
Healthcare Organization A (HO-A) signs up with the TriNetX network.
HO-A allows TriNetX access to their patient data from their Electronic Medical Record (EMR) system. This system typically contains a significant amount of “structured” data (there is a specific code for a specific value).
Data is pulled together in a way that allows an end user to query against that de-identified data for their own research. TriNetX prioritizes research with clinical trials.
Pharmaceutical sponsor organizations (PSO), such as Novartis, gain access to data that allows them prime opportunity, when designing clinical trials, to ensure that the patient population actually exists and can be treated (or be identified to potentially be participants in a clinical trial).
PSOs can then, through the TriNetX platform, reach out to HO-A once they see that they have patients that fit a particular criteria, and make an invitation for HO-A to participate in a clinical trial.
HO-A, who was already contributing that patient data, has the sole ability to re-identify who those specific patients are.
As explained by David Fusari, Chief Technology Officer (CTO) at TriNetX, the company separates itself from the pack by allowing a path to be cleared between pharmaceutical companies and a patient population that would be the best fit for their clinical trials. While similar companies may work within this scope, there is a noticeable distinction between their ability to gain access to already anonymized data, where the surface can barely be scratched and not be connected to a patient. Through the TriNetX platform, however, pharmaceutical companies can actually reach out to healthcare organizations who seem to have the patient cohort, to invite them to participate in a clinical trial. Along with EMR data, TriNetX’s platform includes advanced information such as oncological and genomic reports.
With sophisticated algorithms within the NLP software, unstructured content is analyzed to garner specific data that can be used in wider cohort analysis. This software is run over a clinical note or narration, like a radiology report or discharge summary, and gathers unstructured data that, when combined with available EMR, plays a significant role in how a clinical trial is designed. Even negated facts, where a particular kind of data is not included, is accounted for in the logical rendering of the NLP software. Using the programmed intelligence of NLP, TriNetX is leading the charge in uncovering the important aspects of unstructured data that another program, utilizing a similar model, may be missing.
Building on a high of positive reviews from the beta program of this software earlier this year, it is now being rolled out as a viable capability. TriNetX has built the foundation for a program that is poised to vastly change the landscape of how recruitment and other aspects of cohort analysis is performed for clinical trials.