Donate to Science & Enterprise

S&E on Mastodon

S&E on LinkedIn

S&E on Flipboard

Please share Science & Enterprise

Safety Reports Analyzed with A.I. to Reduce Medical Errors

Data and person graphic

(Gerd Altmann, Pixabay)

13 Nov. 2019. Research is underway to write algorithms for analyzing free text and other unstructured data in safety report databases to reduce medical errors. The new project joining researchers at a university and hospital system is funded by a three-year, $815,000 award from National Library of Medicine, part of National Institutes of Health.

A team led by statistics professor Srijan Sengupta at Virginia Tech in Blacksburg, with colleagues from the MedStar Health National Center for Human Factors in Healthcare in Washington, D.C., aims to fill a gap in analyzing patient safety event reports. These reports, residing in a number of databases, usually have both structured data — predefined fields with predictable types of information — but also unstructured data, usually free text narratives. Structured data lend themselves to quantitative analysis, but the free text narratives are not easily processed. As a result, this rich source of information on medical errors is often ignored.

The unstructured data often describe in detail an entry in a structured data code or category. Thus, when put together they offer a richer and more detailed story. Sengupta and colleagues plan to mine patient safety event reports with natural language processing and statistical models to identify where near-misses occur, and quantify further deterioration and severity of errors. The end deliverable is a set of algorithms offering statistical tools to identify these factors for faster and more accurate identification of conditions contributing to medical errors.

Identifying timing factors and patterns in unstructured data is a key objective of the project. “What may seem like an infrequent hazard at a hospital,” says Sengupta in a Virginia Tech statement, “may be part of a broader national trend when viewed across health care systems. Using our algorithms to effectively analyze documents from reporting systems has the potential to dramatically improve the safety and quality of care by exposing possible weaknesses in the care process.”

MedStar Health’s Human Factors Center is providing its expertise on patient safety, natural language processing, and machine learning for the project. Raj Ratwani, director of the Human Factors Center and co-investigator on the project, says tens of thousands of safety issues are reported to FDA, but the data are not adequately analyzed, allowing potentially unsafe products to continue on the market that could threaten patients’ well-being.

“This research,” notes Ratwani, “is critical to identifying patterns in the reported data and turning data into knowledge that the health care provider can then use to assess the safety of their technologies and processes and develop actions and interventions to prevent patients from being harmed by recognized hazards.”

The team plans to issue open-source software as one of its deliverables. “Releasing open-source software that will enable other practitioners in public and private health care systems to implement our methods on their own proprietary data sets,” adds Sengupta, “will be one of the most important outcomes of our research.”

More from Science & Enterprise:

*     *     *

2 comments to Safety Reports Analyzed with A.I. to Reduce Medical Errors