Sep 22 2022Reviewed by Mila Perera
Compared to the current methods that employ manual record reviews, an automated method combining natural language processing and machine learning detected individuals who inject drugs (PWID) in electronic health data.
Image Credit: Shutterstock.com/Elizabeth A.Cummings
Background
Presently, individuals who inject drugs are found with the help of the International Classification of Diseases (ICD) codes. These codes are specified by the healthcare providers in patients’ electronic health records or obtained from those notes by skilled human coders who evaluate them for billing.
However, there is no particular ICD code for injection drug use, so providers and coders must depend on a combination of non-specific codes as proxies to pinpoint PWIDs, a slow and less accurate approach.
Method
Staphylococcus aureus bacteremia is a typical illness that happens when the bacteria enters openings in the skin, such as those at injection sites. It was manually evaluated in 1,000 records from 2003 to 2014 of patients hospitalized at Veterans Administration hospitals.
The team then created and trained algorithms utilizing machine learning and natural language processing, comparing them with 11 proxy ICD code combinations to detect PWIDs.
The research faced certain limitations, like poor documentation by providers. The dataset utilized was from 2003 — 2014. The use of injection drugs changed from prescription opioids and heroin to synthetic opioids such as fentanyl, which the algorithm is likely to miss as the previous dataset did not have examples of that drug.
The observations might also not apply to other conditions, as they are completely based on the data from the Veterans Administration.
Impact
This artificial intelligence model dramatically accelerates the identification of PWIDs, which may enhance clinical judgment, health care research, and administrative surveillance.
Comment
By using natural language processing and machine learning, we could identify people who inject drugs in thousands of notes in a matter of minutes compared to several weeks that it would take a manual reviewer to do this. This would allow health systems to identify PWIDs to better allocate resources like syringe services programs and substance use and mental health treatment for people who use drugs.
Dr. David Goodman-Meza, Study Lead Author and Assistant Professor, Medicine, Division of Infectious Diseases, David Geffen School of Medicine, University of California, Los Angeles
Authors
Dr. Amber Tang, Dr. Matthew Bidwell Goetz, Steven Shoptaw, and Alex Bui of UCLA; Dr. Michihiko Goto of the University of Iowa and Iowa City VA Medical Center; Dr. Babak Aryanfar of VA Greater Los Angeles Healthcare System; Sergio Vazquez of Dartmouth College; and Dr. Adam Gordon of the University of Utah and VA Salt Lake City Health Care System also contributed to the research.
Goodman-Meza and Goetz also have appointments with VA Greater Los Angeles Healthcare System.
Funding
The study was funded by the US National Institute on Drug Abuse.
Journal Reference
Goodman-Meza, D., et al. (2022) Natural Language Processing and Machine Learning to Identify People Who Inject Drugs in Electronic Health Records. Open Forum Infectious Diseases. doi.org/10.1093/ofid/ofac471.
Source: https://www.uclahealth.org/