One of the predictable downsides to the digital revolution in healthcare has been the rapidly growing mountains of data providers often have to pore through, even in a single patient’s EHR.
But researchers at MIT and the Beth Israel Deaconess Medical Center may have come up with a ray of hope by tapping AI to develop a tool that enhances both data input and retrieval when it comes to EHRs.
In a recently published study, the researchers extolled the value of readily accessible, well-written clinical documentation.
“At their best,” they wrote, “cogent clinical narratives can help clinicians understand a patient’s case, function as a powerful communication method between clinicians, and serve as learning tools to improve future care practice.”
The problem, they add, is that both on the data entry and data retrieval sides of the process, EHRs are rarely able to provide such valuable narratives.
“Because structured and unstructured data can be hard to reconcile, EHRs often store and display information in separate pages or windows, and physicians have to synthesize the patient narrative by navigating across a variety of sources,” they explain. “This creates increased cognitive burden to discover unstructured information, and studies have shown that clinicians spend more time reading past notes than doing any other activity in the EHR. Further, the fragmented interfaces hinder comprehensibility and necessitate frequent task- switching.”
In an attempt to address the problem, the researchers created an AI system called MedKnowts and implemented it at the Beth Israel Deaconess Medical Center. “The AI-backed EHR system integrates the information retrieval system with a note-taking editor so that the search is efficient. The system enables clinicians to use natural language and automates the intake of structured data. Documentation is streamlined with features such as auto-population of text, proactive information retrieval, and easy parsing of long notes.”
According to the researchers, following several months of prototype usage by one physicians and four scribes across 1185 patients, the average system usability scale rating by the scribes was 83.75 out of a possible 100. Moreover, the scribes found the system easy to learn and use.
“MedKnowts can leverage existing health knowledge graphs or outside resources that clinicians use, aiding their decision-making during documentation,” the team said, noting some of the tool’s benefits. “Normalization to a standard ontology also allows notes to be translated to different audiences; medical acronyms can be automatically unravelled to layman’s terms if a patient wants to understand their note. Clinicians with specific language preferences can also personalize note templates and autocomplete functionality with the vocabulary choices that they prefer.”
Moving forward, the researchers plan, among other things, to enhance the tool to identify the portion of a patient’s health record that is most relevant for the clinician to focus on reading.
Photo by Estherrr/Getty Images