Trial recruitment platform tapped by health care network

Northwell-Health-implements-CLiX-platform.jpg
(Image: Getty/Flamingoeimages) (Getty Images/iStockphoto)

Northwell Health taps Clinithink’s CLiX for accelerated patient identification in clinical trials and research.

As Clinithink establishes a network of health care institutions, Northwell will use the software company’s platform CLiX to identify trial candidates across its network of approximately 8m people.

Northwell Health system extends across 23 hospitals and 700 outpatient locations with operations across multiple specialties. It also is the home of The Feinstein Institute for Medical Research.

The health care system serves patients across Long Island and Westchester County, New York City, uses artificial intelligence (AI) and natural language processing (NLP) technologies to read up to 2m clinical documents in an hour.

Clinithink received investments in January 2019 from TT Capital Partners and MSD Global Health Innovation Fund to scale its NLP platform.

The CLiX platform enables identification of potential patients for clinical trials through the analysis of unstructured patient data through clinical NLP to query clinical notes in the health care system.

A Clinithink spokesperson told us, “The insights derived from this rich data provide research teams with precise evidence of potentially eligible patients and cohorts for clinical trials. De-identified, aggregate insights from the output are beneficial to life science organizations to plan and implement clinical trials and to support the market positioning of products in the development lifecycle.”

Northwell implemented the patient identification platform in a proof of concept study of chronic obstructive pulmonary disease (COPD) with a total patient cohort consideration of 939,378 with 3.3m associated documents for review.

CLiX assessed the patients against 22 points of criteria and produced 88 patients for review within 30 hours. A spokesperson for Clinithink told us that this type of review manually has historically taken weeks to months.