AI tools speed up identification of people who use drugs



the findings

An automated process that combines natural language processing and machine learning has identified people injecting drugs (PWID) into electronic health records more quickly and accurately than current methods that rely on manual record reviews.

my knowledge

Currently, PIDs are identified by International Classification of Diseases (ICD) codes specified in patients’ electronic health records by health care providers or extracted from those notes by trained human programmers who review them for billing purposes. But there is no specific ICD code for ICD, so providers and programmers must rely on a set of unspecified codes as proxies to identify PWIDs – a slow approach that can lead to inaccuracies.

method

Researchers manually reviewed 1,000 records from 2003-2014 of people admitted to Veterans Administration hospitals. Staphylococcus aureus Bacteremia, a common infection that occurs when bacteria enter openings in the skin, such as those at injection sites. They then developed and trained algorithms using natural language processing and machine learning and compared them to 11 ICD proxy sets to identify PWIDs.

Limitations of the study include potentially weak documentation by service providers. Also, the data set used is from 2003 to 2014, but the injection drug abuse epidemic has since shifted from prescription opioids and heroin to synthetic opioids like fentanyl, which the algorithm may miss because the data set in which it learned to classify does not contain many Examples of this drug. Finally, the results may not apply to other conditions because they are based entirely on data from the Veterans Administration.

Effect

The use of this AI model significantly speeds up the process of PWID identification, which may lead to improved clinical decision-making, health services research, and administrative monitoring.

Suspension

“Using natural language processing and machine learning, we can identify people who inject drugs with thousands of notes in a matter of minutes compared to the many weeks it takes a manual reviewer to do,” said lead author Dr. David Goodman. Mesa, MD, assistant professor of medicine in the division of infectious diseases at the David Geffen School of Medicine at UCLA. “This would allow health systems to identify IDUs to better allocate resources such as injection and substance abuse service programs and mental health treatment for people who use drugs.”

Authors

Other researchers in the study are Dr. Amber Tang, Dr. Matthew Bidwell Goetz, Stephen Schuptau, and Alex Bui from UCLA. Michiko Goto of the University of Iowa and Iowa City Virginia Medical Center; Babak Aryanfar of VA Greater Los Angeles Healthcare System; Sergio Vasquez of Dartmouth College; and Dr. Adam Gordon of the University of Utah Health Care System in Salt Lake City. Goodman-Meza and Goetz also have appointments with the VA Greater Los Angeles Healthcare System.

magazine

The study was published in the peer-reviewed Open Forum Infectious Diseases.

Finance

The US National Institute on Drug Abuse funded this study.



Source link

Related Posts

Precaliga