UnityPoint Health 2018-01-20T00:37:10+00:00



The nonprofit health system operator harnesses the AI computing power of DataSkill’s ACUMI software powered by IBM Watson.

In the healthcare industry, electronic medical records (EMR) and electronic healthcare records (EHR) are primarily quantitative, transaction-based billing systems. Historically, they haven’t always been designed to improve the quality of documentation and care.

Specifically, there are regular discrepancies between the structured EMR/EHR data and a healthcare provider’s written descriptions and comments. In some cases, doctors focus on the radio buttons and do not supply supporting detail. In other cases, doctors ignore the selections (relying on defaults and macros) and only supply written information. The poor common practice of “cut and paste” from common templates and prior entries has also increased inconsistencies.

UnityPoint Health (UPH), the nation’s 13th largest nonprofit health system and the 4th largest nondenominational health system, was experiencing this challenge. Each year, mounds of data are created from more than 5 million UPH patient visits. The organization’s 30,000 employees, including home care experts, hospital teams and more than 900 doctors and specialists, must communicate with each other successfully to serve its patients.

The Challenge.

As a typical EMR/EHR user, UPH suffered from under-qualifying risk stratification, under-billing Medicare and frustration from physicians due to burdensome, structurally intensive data entry. One UPH Medicare payer estimated that up to 40% of its billings are under-coded. In terms of billing revenues, this means the potential improvement is in the tens of millions of dollars annually.

Furthermore, UPH was unable to properly quantify its error rates and provide real-time improvements. As the organization’s data and health records increased, it could have been at risk of regulatory penalties and inaccurate coding.

The knowledge required to be an excellent coder includes ICD-9 and ICD-10 expertise, as well as clinical knowledge to understand physician notes and other patient information in order to code and bill properly. Further, interpretive knowledge for coding and billing contained in humans is subject to error, and therefore is reliant on each coder’s skill level. The backlog of patient records and claims simply cannot be efficiently addressed by expanding staff. Therefore, scaling to solve the problem requires intelligent, non-human automation software that understands and interprets clinical information to identify and classify each diagnosis accurately.

Refusing to compromise on its mission of “the best outcome for every patient, every time,” UPH searched for a proven solution.

The Solution.

After much research and consideration, UPH deployed the DataSkill ACUMI DataFind coding software, which harnesses the power of IBM Watson. The solution addresses UPH’s challenges by allowing an EMR user to continue business as usual while bringing in the intelligent automation power of Watson. By interpreting doctors’ notes in

their natural language, it allows them to tell “the patient’s story” without having to learn the intricacies of coding and reimbursement systems. Doctors can now focus on the care of the patient and more accurate treatment, while coders enjoy a newfound increase in productivity and precision.

DataSkill created the annotators we use to identify valuable facts in unstructured documents—clinician notes, consult reports, free text fields in electronic medical records—and converts them to a structured form.

– Lillian Dittrick, Director, UnityPoint Health

This ACUMI solution is transformational because it increases coding revenue for providers, while protecting them from compliance issues. This reduces the impact on doctors and drastically improves timely care to patients. With healthcare providers migrating from ICD-9 (14,000 codes) to ICD-10 (70,000+ codes), this technology becomes especially relevant.

Today, UPH recognizes increased revenue by automating and elevating the coding and billing process, thanks to accurate data quality and risk analytics. The ACUMI coding solution, leveraging Watson, enable doctors to simply use their voice instead of navigating the EMR/EHR system. This results in a greater quality of coding and, most importantly, superior patient care.

ACUMI Markets

  • Healthcare
  • Medical Coding and Billing
  • RPA – “Knowledge Worker” Automation & Analytics

ACUMI Product + IBM Solution

  • ACUMI DataFind
  • IBM Watson
  • Watson Explorer

Cognitive Science

  • NLP
  • Ontology Generation
  • Concept Extraction
  • Unstructured Data Analytics
  • Classification


  • On Premises installation of hardware and run over network
  • Virtual machine must have 300GB of available hard drive space and at least 8MB of ram.
  • Virus scanning must be turned off for the folder containing WEX data
  • Client is responsible for the process that runs daily, which adds to the SQL database any new data from the past day. This must be a fast process, not removing/replacing all data, but incrementally loading the new data.