How AI Identified 85,000 At-Risk Patients With Drug-Seeking Behavior

AI is changing healthcare delivery to improve outcomes for both providers and patients by increasing productivity and efficiency and allowing healthcare providers to spend more time in direct patient care.

Download our free ebook, Maximizing the Opportunities in Digital Health Tech

Discover how you can leverage AI in health tech to build a digital healthcare institution.

The WHO estimates that by 2030 there will be a shortage of healthcare workers, from physicians to nurses, to the tune of 9.9 million. AI has the potential to alleviate part of this challenge and reduce the burden on healthcare practitioners by playing a role in:

  • Care delivery
  • Clinical decision support
  • Tirage
  • Diagnostics
  • Self-care, prevention and wellness
  • Chronic care management

How AI identified drug-seeking behavior

Almost 21 million Americans have at least one addiction, yet only 10% of them receive treatment. From 1999 to 2017, more than 700,000 Americans died from overdosing on a drug. And alcohol and drug addiction cost the US economy over $600 billion every year.

Patients get addicted to drugs through pain medications, which are often over-prescribed and can lead to chronic abuse and addiction.

It is a challenge, especially for large healthcare organizations, to accurately and persistently identify at-risk patients before they become addicted and alert physicians to the risk. That’s why one of the largest integrated healthcare services companies in the U.S. worked with Cognizant’s AI to develop an AI-based solution that could help.

Cognizant created an AI-driven machine learning solution that can parse doctors’ notes entered into the health care organization’s electronic medical records (EMR) to identify potential drug-seeking behavior.

The healthcare organization also wanted to figure out a way to identify latent drug-seeking behavior to lessen the incidence of addiction and lower healthcare costs.

Cognizant used three different sources to identify the common traits of drug-seeking behavior by analyzing the patient’s diseases and conditions as recorded in EMR, the types of drugs that historically had been prescribed to the patient and the behaviors and symptoms exhibited due to each type of drug.

In addition, they looked at medical literature and discussed with physicians to determine that certain drug-seeking behaviors — not only symptoms but also a pattern of actions, descriptive phrases or questions used by patients, and related facts and circumstances — are a meaningful indicator of current addiction or the risk of future abuse.

The AI solution uses an advanced machine learning algorithm that mines not only patient behaviors and symptoms but examines the physician’s text-based notes from patient interactions. It combines phase-based extraction, rule-filtering and advanced text clustering to mine highly variable data to identify patients who demonstrate drug-seeking behavior.

Pop-up alerts in the EMR system prompt physicians to take corrective actions at the point of care, interceding with patients in real-time.

Cognizant says its solution learns continuously from its own results to verify the accuracy of its models and improve searches. The company also estimates that the tangible health and financial benefits include saving the health organization as much as $60 million by identifying 85,000 at-risk patients.

Sign up for the MIT SAP Leading Health Tech Innovation course to learn how to bring your health tech idea to life from the founding faculty chair of MIT's Mind+Hand+Heart Initiative, Prof. Rosalind Picard, who co-founded AI health techs Affectiva and Empatica.

MIT SAP Leading Health Tech Innovation is delivered as part of a collaboration with MIT School of Architecture + Planning and MIT Sloan and Esme Learning. All personal data collected on this page is primarily subject to the Esme Learning Privacy Policy.


© 2021 Esme Learning Solutions. All Right Reserved.