DNV and CareCom launch CareIndexing: a new company shaping up disorderly clinical data with artificial intelligence
Healthcare professionals create and update hundreds of thousands of patient notes, diagnostic reports, lab analyses, and other clinical records every day. Yet 80% of the data included in them is unstructured. Information about clinical diagnoses is reported using different expressions, terms, acronyms and even languages, making records impossible to consolidate, analyse and report on at scale without significant manual legwork.
A new company formed by independent assurance and risk management provider DNV and CareCom, a leading provider of healthcare interoperability solutions, is using natural language processing based on machine- and deep learning models to transform disparate clinical data into meaningful information.
“The healthcare sector is one of the most data-intensive industries in the world. But it is also losing out on a huge opportunity to turn fragmented clinical data into insight. CareIndexing’s machine learning solutions aim to minimize the time healthcare institutions spend on manual indexing of terminology, helping the sector to spend more time on patient care and innovation,” said Jacob Boye Hansen, CEO and Founder, CareCom.
CareIndexing provides healthcare trusts, care providers, software vendors and research institutions with essential technology to scan free text and other unstructured data in clinical documents. The company’s solutions extract meaning from this disparate data by structuring it under internationally recognized SNOMED terminology standards and codes. This results in more efficient administration and undiscovered clinical insights:
• New health record insights: CareIndexing rapidly searches through large quantities of electronic health records to identify information about specific diseases and clinical procedures. The solution also scans individual patient records, instantly picking out information about key issues from free text for quick insight into a patient’s history.
• Better patient outcomes: Disorderly clinical information structured by CareIndexing can be easily inserted into data models that identify patients at risk of disease and provide predictive analytics for more accurate diagnoses.
• Quicker, more accurate billing and reporting: CareIndexing makes it easy to retrieve and report on large volumes of medial data, helping to efficiently analyse the performance of an institution’s clinical departments and services, and improve institutions’ billing accuracy.
“The digital transformation of healthcare needs to be accelerated. This requires the removal of barriers to accessing, sharing and analysing data across the sector’s value. CareIndexing helps to solve these issues by turning disparate clinical terminology across a vast number of records into structured formats that can be easily processed by IT systems. DNV is proud to team up with CareCom to rapidly scale this technology in the market,” said Stephen McAdam, Managing Director, Digital Health, DNV.
CareIndexing was developed by CareCom in collaboration with Aalborg University in Denmark. The solution has been tested against large datasets and compared with metrics published in scientific journals to validate its accuracy and performance.
DNV has invested in forming CareIndexing as a new company in partnership with CareCom through its corporate venture capital unit, DNV Ventures. The investment forms part of DNV’s strategic expansion into the digital health market.
