Identifying adverse drug reactions from free-text electronic hospital health record notes

Arthur Wasylewicz; Britt van de Burgt; Aniek Weterings; Naomi Jessurun; Erik Korsten; Toine Egberts; Arthur Bouwman; Marieke Kerskes; René Grouls; Carolien van der Linden

doi:10.1111/bcp.15068

Identifying adverse drug reactions from free-text electronic hospital health record notes

Arthur Wasylewicz, Britt van de Burgt, Aniek Weterings, Naomi Jessurun, Erik Korsten, Toine Egberts, Arthur Bouwman, Marieke Kerskes, René Grouls, Carolien van der Linden

Research output: Contribution to journal › Article › Academic › peer-review

Abstract

BACKGROUND: Adverse drug reactions (ADRs) are estimated to be the fifth cause of hospital death. Up to 50% are potentially preventable and a significant number are recurrent (reADRs). Clinical decision support systems have been used to prevent reADRs using structured reporting concerning the patient's ADR experience, which in current clinical practice is poorly performed. Identifying ADRs directly from free text in electronic health records (EHRs) could circumvent this.

AIM: To develop strategies to identify ADRs from free-text notes in electronic hospital health records.

METHODS: In stage I, the EHRs of 10 patients were reviewed to establish strategies for identifying ADRs. In stage II, complete EHR histories of 45 patients were reviewed for ADRs and compared to the strategies programmed into a rule-based model. ADRs were classified using MedDRA and included in the study if the Naranjo causality score was ≥1. Seriousness was assessed using the European Medicine Agency's important medical event list.

RESULTS: In stage I, two main search strategies were identified: keywords indicating an ADR and specific prepositions followed by medication names. In stage II, the EHRs contained a median of 7.4 (range 0.01-18) years of medical history covering over 35 000 notes. A total of 318 unique ADRs were identified of which 63 were potentially serious and 179 (sensitivity 57%) were identified by the rule. The method falsely identified 377 ADRs (positive predictive value 32%). However, it also identified an additional eight ADRs.

CONCLUSION: Two key strategies were developed to identify ADRs from hospital EHRs using free-text notes. The results appear promising and warrant further study.

Original language	English
Pages (from-to)	1235-1245
Number of pages	11
Journal	British Journal of Clinical Pharmacology
Volume	88
Issue number	3
Early online date	1 Sept 2021
DOIs	https://doi.org/10.1111/bcp.15068
Publication status	Published - Mar 2022

Bibliographical note

Funding Information:
We would like to thank Wilma Compagner and Paul de Clerq for their contribution in programming and helping to implement the rule-based model.

Publisher Copyright:
© 2021 The Authors. British Journal of Clinical Pharmacology published by John Wiley & Sons Ltd on behalf of British Pharmacological Society.

Keywords

adverse drug event
adverse drug reaction
clinical decision support
clinical decision support system
drug allergy
free-text
natural language processing
text-mining

Access to Document

10.1111/bcp.15068Licence: CC BY-NC-ND

Brit J Clinical Pharma - 2021 - Wasylewicz - Identifying adverse drug reactions from free‐text electronic hospital healthFinal published version, 1.41 MBLicence: CC BY-NC-ND

Cite this

@article{5c7b8ed3db9b4e7bb3a8b00298d16068,

title = "Identifying adverse drug reactions from free-text electronic hospital health record notes",

abstract = "BACKGROUND: Adverse drug reactions (ADRs) are estimated to be the fifth cause of hospital death. Up to 50% are potentially preventable and a significant number are recurrent (reADRs). Clinical decision support systems have been used to prevent reADRs using structured reporting concerning the patient's ADR experience, which in current clinical practice is poorly performed. Identifying ADRs directly from free text in electronic health records (EHRs) could circumvent this.AIM: To develop strategies to identify ADRs from free-text notes in electronic hospital health records.METHODS: In stage I, the EHRs of 10 patients were reviewed to establish strategies for identifying ADRs. In stage II, complete EHR histories of 45 patients were reviewed for ADRs and compared to the strategies programmed into a rule-based model. ADRs were classified using MedDRA and included in the study if the Naranjo causality score was ≥1. Seriousness was assessed using the European Medicine Agency's important medical event list.RESULTS: In stage I, two main search strategies were identified: keywords indicating an ADR and specific prepositions followed by medication names. In stage II, the EHRs contained a median of 7.4 (range 0.01-18) years of medical history covering over 35 000 notes. A total of 318 unique ADRs were identified of which 63 were potentially serious and 179 (sensitivity 57%) were identified by the rule. The method falsely identified 377 ADRs (positive predictive value 32%). However, it also identified an additional eight ADRs.CONCLUSION: Two key strategies were developed to identify ADRs from hospital EHRs using free-text notes. The results appear promising and warrant further study.",

keywords = "adverse drug event, adverse drug reaction, clinical decision support, clinical decision support system, drug allergy, free-text, natural language processing, text-mining",

author = "Arthur Wasylewicz and {van de Burgt}, Britt and Aniek Weterings and Naomi Jessurun and Erik Korsten and Toine Egberts and Arthur Bouwman and Marieke Kerskes and Ren{\'e} Grouls and {van der Linden}, Carolien",

note = "Funding Information: We would like to thank Wilma Compagner and Paul de Clerq for their contribution in programming and helping to implement the rule-based model. Publisher Copyright: {\textcopyright} 2021 The Authors. British Journal of Clinical Pharmacology published by John Wiley & Sons Ltd on behalf of British Pharmacological Society.",

year = "2022",

month = mar,

doi = "10.1111/bcp.15068",

language = "English",

volume = "88",

pages = "1235--1245",

journal = "British Journal of Clinical Pharmacology",

issn = "0306-5251",

publisher = "Wiley-Blackwell Publishing Ltd",

number = "3",

}

TY - JOUR

T1 - Identifying adverse drug reactions from free-text electronic hospital health record notes

AU - Wasylewicz, Arthur

AU - van de Burgt, Britt

AU - Weterings, Aniek

AU - Jessurun, Naomi

AU - Korsten, Erik

AU - Egberts, Toine

AU - Bouwman, Arthur

AU - Kerskes, Marieke

AU - Grouls, René

AU - van der Linden, Carolien

N1 - Funding Information: We would like to thank Wilma Compagner and Paul de Clerq for their contribution in programming and helping to implement the rule-based model. Publisher Copyright: © 2021 The Authors. British Journal of Clinical Pharmacology published by John Wiley & Sons Ltd on behalf of British Pharmacological Society.

PY - 2022/3

Y1 - 2022/3

N2 - BACKGROUND: Adverse drug reactions (ADRs) are estimated to be the fifth cause of hospital death. Up to 50% are potentially preventable and a significant number are recurrent (reADRs). Clinical decision support systems have been used to prevent reADRs using structured reporting concerning the patient's ADR experience, which in current clinical practice is poorly performed. Identifying ADRs directly from free text in electronic health records (EHRs) could circumvent this.AIM: To develop strategies to identify ADRs from free-text notes in electronic hospital health records.METHODS: In stage I, the EHRs of 10 patients were reviewed to establish strategies for identifying ADRs. In stage II, complete EHR histories of 45 patients were reviewed for ADRs and compared to the strategies programmed into a rule-based model. ADRs were classified using MedDRA and included in the study if the Naranjo causality score was ≥1. Seriousness was assessed using the European Medicine Agency's important medical event list.RESULTS: In stage I, two main search strategies were identified: keywords indicating an ADR and specific prepositions followed by medication names. In stage II, the EHRs contained a median of 7.4 (range 0.01-18) years of medical history covering over 35 000 notes. A total of 318 unique ADRs were identified of which 63 were potentially serious and 179 (sensitivity 57%) were identified by the rule. The method falsely identified 377 ADRs (positive predictive value 32%). However, it also identified an additional eight ADRs.CONCLUSION: Two key strategies were developed to identify ADRs from hospital EHRs using free-text notes. The results appear promising and warrant further study.

AB - BACKGROUND: Adverse drug reactions (ADRs) are estimated to be the fifth cause of hospital death. Up to 50% are potentially preventable and a significant number are recurrent (reADRs). Clinical decision support systems have been used to prevent reADRs using structured reporting concerning the patient's ADR experience, which in current clinical practice is poorly performed. Identifying ADRs directly from free text in electronic health records (EHRs) could circumvent this.AIM: To develop strategies to identify ADRs from free-text notes in electronic hospital health records.METHODS: In stage I, the EHRs of 10 patients were reviewed to establish strategies for identifying ADRs. In stage II, complete EHR histories of 45 patients were reviewed for ADRs and compared to the strategies programmed into a rule-based model. ADRs were classified using MedDRA and included in the study if the Naranjo causality score was ≥1. Seriousness was assessed using the European Medicine Agency's important medical event list.RESULTS: In stage I, two main search strategies were identified: keywords indicating an ADR and specific prepositions followed by medication names. In stage II, the EHRs contained a median of 7.4 (range 0.01-18) years of medical history covering over 35 000 notes. A total of 318 unique ADRs were identified of which 63 were potentially serious and 179 (sensitivity 57%) were identified by the rule. The method falsely identified 377 ADRs (positive predictive value 32%). However, it also identified an additional eight ADRs.CONCLUSION: Two key strategies were developed to identify ADRs from hospital EHRs using free-text notes. The results appear promising and warrant further study.

KW - adverse drug event

KW - adverse drug reaction

KW - clinical decision support

KW - clinical decision support system

KW - drug allergy

KW - free-text

KW - natural language processing

KW - text-mining

UR - http://www.scopus.com/inward/record.url?scp=85116924884&partnerID=8YFLogxK

U2 - 10.1111/bcp.15068

DO - 10.1111/bcp.15068

M3 - Article

C2 - 34468999

SN - 0306-5251

VL - 88

SP - 1235

EP - 1245

JO - British Journal of Clinical Pharmacology

JF - British Journal of Clinical Pharmacology

IS - 3

ER -

Identifying adverse drug reactions from free-text electronic hospital health record notes

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this