Artificial Intelligence in Health Care: A Cure for Data Pains

By Scott Briercheck, Chief Scientist | Posted: 03/28/2018

Your organization, payors and government agencies collect an impressive amount of patient data to parse, read and make critical decisions going forward. But the larger the data set grows, the lower the performance of the systems that use it − and the greater the challenge in finding the right information due to incomplete or non-specific documentation. It’s a challenge that results in compromised patient care due to incomplete or non-specific documentation, as well as inaccurate coding that leads to denials. The result is compromised financial stability.

Conflicting goals – part of the problem?

There are three conflicting information technology goals in health care today that could be creating this conundrum:

  1. Quantity counts. Gather as much information available as possible.
  2. Pick up the speed. Keep making access faster.
  3. Get it right the first time. Ensure that the first request for information is complete and accurate.

Enormous data sets hinder performance and complicate getting the right information where it needs to be − quickly and accurately. The traditional approaches of adding more hardware and standardized data formats to try to speed up searches and improve completeness and accuracy are not the best tactics, and in many cases provide no improvement at all.

Users violate text and speech data entry standards and fail to follow templating practices, resulting in inconsistent data. Large quantities of data don’t always present obvious links to the most important data relationships, leaving providers scrambling to figure out how to identify the process/revenue cycle improvements they know are locked away in the data.

So how do you extract the most complete and accurate intelligence that is specific to patient care, patient access and patient billing from the massive quantities of data collected to make informed decisions and prevent denials – as quickly and inexpensively as possible? Artificial intelligence.

Artificial Intelligence (AI)

AI is the use of computer technology to take complex and irregular information and allow a computer to effectively reason about it through effective pattern matching and inference. This differs from the literal, structured data approach that most database-driven computer systems require today.

There are two primary categories of AI:

  1. Machine Learning AI - Machine Learning AI uses statistical methods trained by experts on large data sets to automatically identify patterns of interest.
  2. Rules Based AI - Rules Based AI uses algorithms and rules designed by humans to classify features of interest derived from expert knowledge.

Use AI to achieve success

AI is a hot topic today because it allows pattern recognition to occur quickly across massively large data sets. In today’s pay-for-performance landscape, securing both clinical and financial integration is crucial to the organization’s patient and financial outcomes. How does it benefit the health care industry?

There are four primary groups in health care who care about better, faster, more accurate health care data and results – the groups that rely on the outcomes from solid clinical documentation:

  1. Patients – accurate clinical documentation benefits those who want to get healthy faster and remain healthy longer; be informed about their care
  2. Physicians – correct clinical documentation is essential for treating patients promptly, safely, and effectively, and getting paid on time. It impacts clinical and business decisions that affect the organization’s ability to mitigate costs yet improve quality.
  3. Governments – clean clinical documentation fosters quality care, streamlines regulations and improves population health
  4. Health C-Level – accurate clinical documentation means leaders can confidently promise fast, accurate care, high patient satisfaction, and manageable costs, serve all equally well.

So, what exactly does AI do?

  • AI improves physician engagement by facilitating stronger patient case capture and documentation. It directly affects and improves the physician’s quality scores.
  • AI with Natural Language Processing (NLP), analyzes physician documentation and offers suggestions for improvement. Automated AI at point of care helps the physician to accurately document patient conditions and treatments which leads to better coding.
  • AI strengthens the effect of clinical documentation’s ability to improve organizational performance. AI expedites clinical documentation workflows by eliminating the sorting and confirmation of documentation that Clinical Documentation Specialist (CDS’s) must do, reducing the task from 10 to 15 minutes manually to seconds by AI.

AI serves a role in all aspects of health care and can impact many initiatives including population health in an important and transformative way. Embrace the opportunity to improve health care outcomes from all perspectives and ultimately transform your quality scores, reputation, bottom line, and quality of patient care by leveraging AI in health care.