The Right Way to Use AI in Healthcare

The Rowan Report / By Tim Rowan, Editor Emeritus

For better or worse, healthcare has begun the inevitable adoption of Artificial Intelligence. Before you consider adopting AI technology, know that there is a wrong way and a right way to use AI in healthcare. In a companion article this week, we describe the criticism insurance companies are getting for deploying AI in healthcare to harm patients. As a balance, here is a review of a product that we find to be using AI in healthcare to help both patients and Home Health Agencies.

The Problem 

Home Health referral documents from physicians or hospitals can consist of more than 100 electronically transmitted pages. Some agencies report occasional packets exceeding 1,000 pages, often in a variety of data formats. Some are standard data formats, such as a face sheet, but most are unstructured, consisting of images or narrations, sometimes in paragraphs, sometimes in incomplete sentences. Worse, patient data interoperability can be limited by unstructured data.

More often than not, most of these pages are never read. Thoroughly interpreting that much data is nearly impossible for a human. Consequently, nurses too often approach an admission evaluation visit with an incomplete picture of a patient. The result can be gaps in care or treatment, inaccurate OASIS assessments, incomplete or poorly sequenced diagnosis codes, and improper care plans. These obstacles can impact both patient outcomes and agency revenue.

One Newly Available Solution for the Right Way to use AI in Healthcare

We recently attended a product demonstration and followed it up with updated descriptions to learn details about new product developments. Over the next three months, Select Data, in full disclosure one of our sponsors, will be introducing an AI-powered suite of products that has been designed over many years of development to support clinical, data driven decision-making. One by one, it addresses the problems described above.

The new system, SmartCare, empowers clinicians to harness previously hidden insights while reducing bias and cognitive overload. It enables them to steer their decisions with enhanced precision while maintaining their pivotal role in patient care, eliminating one of the common reasons many Home Health administrators hesitate to invite AI into agency processes. It does, however, make the care team’s job easier and facilitates better decision-making.

  • AI can read those 100 to 1,000 page referral documents in minutes, where a human may require days.
  • SmartCare uses AI to synthesize relevant medical history to provide a care snapshot highlighting the key diagnosis, focus and considerations for care, and recommended OASIS clinical discipline. It highlights any areas for clarification needed from physician or admitting nurse.
  • Clinicians can search and index specific words in unstructured data, such as narratives, to instantly identify any detail of a patient’s condition in an easy-to-read interface. Nurses approach the initial OASIS visit armed with all of a referring clinician’s relevant care findings.
  • Recommendations for diagnostic codes strictly follow Medicare PDGM guidelines.

Suite of Tools

1 – RISE stand for Rapid Intake Summary & Evaluation. This component of the suite summarizes all clinical data from referral sources and your EHR. It compiles this data to provide clinically relevant diagnoses, focus of care, and recommendations for skilled disciplines. This is the part of the tool that reads referral documents and supports informed decision-making. The advantages we detected go a bit beyond the technical…


Read Full Article