CoChat
Sign Up
Cited by 2Medical Coding and Health InformationHIV, Drug Use, Sexual RiskOpioid Use Disorder Treatment

Natural Language Processing for Identification of Hospitalized People Who Use Drugs: Cohort Study

Open access
Sato, Taisuke, Grussing, Emily D, Patel, Ruchi, Ridgway, Jessica, Suzuki, Joji, Sweigart, Benjamin, Miller, Robert, Wurcel, Alysse G·2025·JMIR AI
DOI 10.2196/63147PMC PMC12294639
Open this paper in CoChat — free accountOpen PDF

Abstract

Background: People who use drugs (PWUD) are at heightened risk of severe injection–related infections. Current research relies on billing codes to identify PWUD—a methodology with suboptimal accuracy that may underestimate the economic, racial, and ethnic diversity of hospitalized PWUD. Objective: The goal of this study is to examine the impact of natural language processing (NLP) on enhancing identification of PWUD in electronic medical records, with a specific focus on determining improved systems of identifying populations who may previously been missed, including people who have low income or those from racially and ethnically minoritized populations. Methods: Health informatics specialists assisted in querying a cohort of likely PWUD hospital admissions at Tufts Medical Center between 2020‐2022 using the following criteria: (1) ICD-10 codes indicative of drug use, (2) positive drug toxicology results, (3) prescriptions for medications for opioid use disorder, and (4) applying NLP-detected presence of “token” keywords in the electronic medical records likely indicative of the patient being a PWUD. Hospital admissions were split into two groups: highly documented (all four criteria present) and minimally documented (NLP-only). These groups were examined to assess the impact of race, ethnicity, and social vulnerability index. With chart review as the “gold standard,” the positive predictive value was calculated. Results: The cohort included 4548 hospitalization admissions, with broad heterogeneity in how people entered the cohort and subcohorts; a total of 288 hospital admissions entered the cohort through NLP token presence alone. NLP demonstrated a 54% positive predictive value, outperforming biomarkers, prescription for medications for opioid use disorder, and ICD codes in identifying hospitalizations of PWUD. Additionally, NLP significantly enhanced these methods when integrated into the identification algorithm. The study also found that people from racially and ethnically minoritized communities and those with lower social vulnerability index were significantly more likely to have lower rates of PWUD-related documentation. Conclusions: NLP proved effective in identifying hospitalizations of PWUD, surpassing traditional methods. While further refinement is needed, NLP shows promising potential in minimizing health care disparities.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.2196/63147

Cites

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

No references could be resolved yet.

Research works printed but not linked: 36

  1. [1]Chiosi JJ Mueller PP Chhatwal J Ciaranello AL A multimorbidity model for estimating health outcomes from the syndemic of injection drug use and associated infections in the United States BMC Health Serv Res 07 17 2023 23 1 760 doi 10.1186/s12913-023-09773-1 Medline 37461007 PMC10353126 · DOI
  2. [2]Mattson CL Tanz LJ Quinn K Kariisa M Patel P Davis NL Trends and geographic patterns in drug and synthetic opioid overdose deaths - United States, 2013-2019 MMWR Morb Mortal Wkly Rep Feb 12 2021 70 6 202 207 doi 10.15585/mmwr.mm7006a4 Medline 33571180 PMC7877587 · DOI
  3. [3]Sun J Mehta SH Astemborski J et al Mortality among people who inject drugs: a prospective cohort followed over three decades in Baltimore, MD, USA Addiction Mar 2022 117 3 646 655 doi 10.1111/add.15659 Medline 34338374 PMC10572098 · DOI
  4. [4]Hollander MAG Chang CCH Douaihy AB Hulsey E Donohue JM Racial inequity in medication treatment for opioid use disorder: exploring potential facilitators and barriers to use Drug Alcohol Depend Oct 1 2021 227 108927 doi 10.1016/j.drugalcdep.2021.108927 Medline 34358766 PMC8464525 · DOI
  5. [5]Hamdan S Smyth E Murphy ME et al Racial and ethnic disparities in HIV testing in people who use drugs admitted to a tertiary care hospital AIDS Patient Care STDS Nov 2022 36 11 425 430 doi 10.1089/apc.2022.0165 Medline 36301195 PMC9700355 · DOI
  6. [6]Westgard LK Sato T Bradford WS et al National HIV and HCV screening rates for hospitalized people who use drugs are suboptimal and heterogeneous across 11 US hospitals Open Forum Infect Dis 05 2024 11 5 ofae204 doi 10.1093/ofid/ofae204 Medline 38746950 PMC11093397 · DOI
  7. [7]Ti Lianping Ti Lianlian Leaving the hospital against medical advice among people who use illicit drugs: a systematic review Am J Public Health Dec 2015 105 12 e53 9 doi 10.2105/AJPH.2015.302885 Medline 26469651 PMC4638247 · DOI
  8. [8]Hazen A Pizzicato L Hom J Johnson C Viner KM Association between discharges against medical advice and readmission in patients treated for drug injection-related skin and soft tissue infections J Subst Abuse Treat 07 2021 126 108465 doi 10.1016/j.jsat.2021.108465 Medline 34116815 · DOI
  9. [9]Eaton EF Westfall AO McClesky B et al In-Hospital illicit drug use and patient-directed discharge: barriers to care for patients with injection-related infections Open Forum Infect Dis Mar 2020 7 3 ofaa074 doi 10.1093/ofid/ofaa074 Medline 32258203 PMC7096133 · DOI
  10. [10]Calcaterra SL Bottner R Martin M et al Management of opioid use disorder, opioid withdrawal, and opioid overdose prevention in hospitalized adults: A systematic review of existing guidelines J Hosp Med Sep 2022 17 9 679 692 doi 10.1002/jhm.12908 Medline 35880821 PMC9474657 · DOI
  11. [11]Straub L Gagne JJ Maro JC et al Evaluation of use of technologies to facilitate medical chart review Drug Saf Sep 2019 42 9 1071 1080 doi 10.1007/s40264-019-00838-x Medline 31111340 · DOI
  12. [12]Schaper E Padwa H Urada D Shoptaw S Substance use disorder patient privacy and comprehensive care in integrated health care settings Psychol Serv Feb 2016 13 1 105 109 doi 10.1037/a0037968 Medline 26845493 · DOI
  13. [13]Saczynski JS Andrade SE Harrold LR et al A systematic review of validated methods for identifying heart failure using administrative data Pharmacoepidemiol Drug Saf 01 2012 21 Suppl 1 1 129 140 doi 10.1002/pds.2313 Medline 22262599 PMC3808171 · DOI
  14. [14]Segar MW Keshvani N Rao S Fonarow GC Das SR Pandey A Race, social determinants of health, and length of stay among hospitalized patients with heart failure: an analysis from the Get With The Guidelines-Heart Failure Registry Circ: Heart Failure Nov 2022 15 11 doi 10.1161/CIRCHEARTFAILURE.121.009401 36378756 · DOI
  15. [15]Campanile Y Silverman M Sensitivity, specificity and predictive values of ICD-10 substance use codes in a cohort of substance use-related endocarditis patients Am J Drug Alcohol Abuse Sep 3 2022 48 5 538 547 doi 10.1080/00952990.2022.2047713 Medline 35579599 · DOI
  16. [16]Ball LJ Sherazi A Laczko D et al Validation of an algorithm to identify infective endocarditis in people who inject drugs Med Care Oct 2018 56 10 e70 e75 doi 10.1097/MLR.0000000000000838 Medline 29200131 · DOI
  17. [17]Zalesak M Francis K Gedeon A et al Current and future disease progression of the chronic HCV population in the United States PLoS ONE 2013 8 5 e63959 doi 10.1371/journal.pone.0063959 Medline 23704962 PMC3660594 · DOI
  18. [18]Ehrenfeld JM Gottlieb KG Beach LB Monahan SE Fabbri D Development of a natural language processing algorithm to identify and evaluate transgender patients in electronic health record systems Ethn Dis 2019 29 Suppl 2 441 450 doi 10.18865/ed.29.S2.441 Medline 31308617 PMC6604788 · DOI
  19. [19]Sims SA Snow LA Porucznik CA Surveillance of methadone-related adverse drug events using multiple public health data sources J Biomed Inform Aug 2007 40 4 382 389 doi 10.1016/j.jbi.2006.10.004 Medline 17185042 · DOI
  20. [20]Ridgway JP Uvin A Schmitt J et al Natural language processing of clinical notes to identify mental illness and substance use among people living with HIV: retrospective cohort study JMIR Med Inform Mar 10 2021 9 3 e23456 doi 10.2196/23456 Medline 33688848 PMC7991991 · DOI
  21. [21]Bui DDA Zeng-Treitler Q Learning regular expressions for clinical text classification J Am Med Inform Assoc 2014 21 5 850 857 doi 10.1136/amiajnl-2013-002411 Medline 24578357 PMC4147608 · DOI
  22. [22]Almudaifer AI Covington W Hairston J et al Multi-task transfer learning for the prediction of entity modifiers in clinical text: application to opioid use disorder case detection J Biomed Semantics Jun 7 2024 15 1 11 doi 10.1186/s13326-024-00311-4 Medline 38849884 PMC11157899 · DOI
  23. [23]Goodman-Meza D Tang A Aryanfar B et al Natural language processing and machine learning to identify people who inject drugs in electronic health records Open Forum Infect Dis Sep 2022 9 9 ofac471 doi 10.1093/ofid/ofac471 Medline 36168546 PMC9511274 · DOI
  24. [24]Bartholomew TS Tookes HE Spencer EC Feaster DJ Application of machine learning algorithms for localized syringe services program policy implementation – Florida, 2017 Ann Med Dec 31 2022 54 1 2137 2150 doi 10.1080/07853890.2022.2105391 35900201 PMC9341345 · DOI
  25. [25]Cresta Morgado P Carusso M Alonso Alemany L Acion L Practical foundations of machine learning for addiction research. Part I. Methods and techniques Am J Drug Alcohol Abuse 05 4 2022 48 3 260 271 doi 10.1080/00952990.2021.1995739 Medline 35389305 · DOI
  26. [26]Rivero-Juárez A Guijo-Rubio D Tellez F et al Using machine learning methods to determine a typology of patients with HIV-HCV infection to be treated with antivirals PLoS ONE 2020 15 1 e0227188 doi 10.1371/journal.pone.0227188 Medline 31923277 PMC6953863 · DOI
  27. [27]D Grussing E Pickard B Khalid A et al Implementation of a bundle to improve HIV testing during hospitalization for people who inject drugs Implement Res Pract 2023 4 26334895231203410 doi 10.1177/26334895231203410 Medline 37936964 PMC10548809 · DOI
  28. [28]Wurcel AG Yu S Burke D et al Implementation of a patient-provider agreement to improve healthcare delivery for patients with substance use disorder in the inpatient setting J Patient Saf Dec 1 2021 17 8 e1827 e1832 doi 10.1097/PTS.0000000000000721 Medline 32398540 PMC7785299 · DOI
  29. [29]Reed JR Jordan AE Perlman DC Smith DJ Hagan H The HCV care continuum among people who use drugs: protocol for a systematic review and meta-analysis Syst Rev 07 11 2016 5 1 110 doi 10.1186/s13643-016-0293-6 Medline 27401499 PMC4940695 · DOI
  30. [30]Hoffman KA Ponce Terashima J McCarty D Opioid use disorder and treatment: challenges and opportunities BMC Health Serv Res Nov 25 2019 19 1 884 doi 10.1186/s12913-019-4751-4 Medline 31767011 PMC6876068 · DOI
  31. [31]McGrew KM Homco JB Garwe T et al Validity of International Classification of Diseases codes in identifying illicit drug use target conditions using medical record data as a reference standard: a systematic review Drug Alcohol Depend Mar 1 2020 208 107825 doi 10.1016/j.drugalcdep.2019.107825 Medline 31982637 PMC9533471 · DOI
  32. [33]Matsuzaka S Knapp M Anti-racism and substance use treatment: addiction does not discriminate, but do we? J Ethn Subst Abuse 2020 19 4 567 593 doi 10.1080/15332640.2018.1548323 Medline 30642230 · DOI
  33. [34]Acevedo A Panas L Garnick D et al Disparities in the treatment of substance use disorders: does where you live matter? J Behav Health Serv Res Oct 2018 45 4 533 549 doi 10.1007/s11414-018-9586-y Medline 29435862 PMC6087681 · DOI
  34. [35]Thakarar K Weinstein ZM Walley AY Optimising health and safety of people who inject drugs during transition from acute to outpatient care: narrative review with clinical checklist Postgrad Med J Jun 2016 92 1088 356 363 doi 10.1136/postgradmedj-2015-133720 Medline 27004476 PMC4967553 · DOI
  35. [36]Zubiago J Murphy M Guardado R Daudelin D Patil D Wurcel A Increased HIV testing in people who use drugs hospitalized in the first wave of the COVID-19 pandemic J Subst Abuse Treat 05 2021 124 108266 doi 10.1016/j.jsat.2020.108266 Medline 33771274 PMC7831875 · DOI
  36. [37]Sundaram G Sato T Goodman-Meza D et al Perspectives on benefits and risks of creation of an “injection drug use” billing code J Subst Use Addict Treat Sep 2024 164 209392 doi 10.1016/j.josat.2024.209392 Medline 38735482 PMC13112322 · DOI

Cited by

Sorted by citations

Authors
Sato, T., Grussing, E. D., Patel, R., Ridgway, J., Suzuki, J., Sweigart, B., Miller, R., & Wurcel, A. G.
Affiliations
  • Tufts Medical Center (US)
  • University of Chicago (US)
  • Brigham and Women's Hospital (US)
Corresponding author
Wurcel, Alysse G
ORCID
Venue
JMIR AI
Publisher
JMIR Publications Inc.
Published
2025-07-18
Volume
4
Pages
e63147-e63147
ISSN
2817-1705 (ISSN-L)
DOI
10.2196/63147
PMC
PMC12294639
OpenAlex
W4408527265
Cited by
2
References
0
Citations by year
2026: 2
FWCI
1.57
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • JMIR AI (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Health Sciences › Health Professions › Health Information Management › Medical Coding and Health Information
  • Health Sciences › Medicine › Epidemiology › HIV, Drug Use, Sexual Risk
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Opioid Use Disorder Treatment
Keywords
drug users, natural language processing, electronic health records, opioid use disorder, drug toxicology, ICD-10 codes, Social Vulnerability Index, racial and ethnic minorities, health care disparities, injection-related infections, positive predictive value

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review