Scribd
Upload
84 views23 pages

ADEpedia 2.0-Integration of Normalized Adverse Drug Events (ADEs)

2013 Summit on Clinical Research Informatics

Uploaded by

AMIA
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
84 views23 pages

ADEpedia 2.0-Integration of Normalized Adverse Drug Events (ADEs)

2013 Summit on Clinical Research Informatics

Uploaded by

AMIA
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 23

ADEpedia 2.

0: Integration of Normalized Adverse Drug Events (ADEs) Knowledge from the UMLS
Guoqian Jiang, MD, PhD Mayo Clinic
2013 AMIA Summit on Clinical Research Informatics March 22, 2013

2013 MFMER | slide-1

Acknowledgements
Co-authors Hongfang Liu, PhD Harold R. Solbrig Christopher G. Chute, MD. Dr. PH This work was supported in part by the SHARP Area 4: Secondary Use of EHR Data (90TR000201)

2013 MFMER | slide-2

Introduction
Adverse drug events (ADEs) are a well-recognized cause of patient morbidity and increased health care costs in the United States. Traditionally, spontaneous reporting is used as the main source for drug safety surveillance. Notably, the US Food and Drug Administration (FDA) uses an adverse event reporting system (AERS) to monitor for new adverse events and medication errors that might occur with all approved drug and therapeutic biologic products
2013 MFMER | slide-3

Introduction
To facilitate the ADE reporting and detection, there is emerging interest in secondary use of clinical data from the electronic medical records (EMRs). A fundamental challenge is that the community lacks a publicly-available, standardized ADE knowledge base that encodes known ADE information for drug surveillance.

2013 MFMER | slide-4

Introduction
The known ADE information is very useful for avoiding over-alerting of ADE signals detected by data mining algorithms, and reducing pharmacovigilance study noise levels, so that only novel signals are considered for further exploration. ADEpedia: a standardized knowledge base of ADEs for drug safety surveillance FDA Structured Product Labeling (SPL), FDA Adverse Event Reporting System (AERS) and the Unified Medical Language System (UMLS) http://adepedia.org

2013 MFMER | slide-5

Introduction
The Unified Medical Language System (UMLS), developed by the National Library of Medicine (NLM), intends to promote creation of more effective and interoperable biomedical information system and services [12]. A systematic review and organization of known ADE knowledge from the UMLS would be a good starting point to facilitate the integration of known ADE knowledge and ultimately form a comprehensive ADE knowledge base.

2013 MFMER | slide-6

Co-occurrence
Co-occurrence is one of essential statistics that has been widely used in ADE text mining and knowledge acquisition from biomedical and clinical documents. Wright et al. used co-occurrence-based association rule mining technique to identify a large number of clinical accurate associations that may be useful for identifying probable gaps in the problem list. Liu Y, et al demonstrated an approach that statistically significant co-occurrence of drug-disease mentions in the clinical notes can be used to detect ADE signals .

2013 MFMER | slide-7

The objectives of this study


To extract and normalize ADE knowledge from the UMLS and integrate the ADE knowledge with the ADEpedia; and To enrich the knowledge base with the drugdisorder co-occurrence data from a large-scale EMR system.

2013 MFMER | slide-8

Knowledge Resources for ADE Integration


1) UMLS 2011AB: We downloaded and installed the version of UMLS 2011AB. 2) Semantic groups: We used 2 UMLS semantic groups Chemicals & Drugs (CHEM) and Disorders (DISO), and their corresponding semantic types. 3) ADE data from the ADEpedia. In the knowledge base, the medication data are represented by RxNorm codes and the ADE data are represented by SNOMED CT and MedDRA codes.

2013 MFMER | slide-9

Electronic Medical Records (EMRs) Data Sources


We used the UMLS CUI co-occurrence data that were extracted from a 51-million-document corpus of clinical text covering Mayo Clinic clinical notes between 1/1/2001-12/31/2010 in a previous study. The clinical notes were originally retrieved from the Mayos Enterprise Data Trust (EDT), which is a comprehensive snapshot of Mayo Clinics service areas and includes structured data, unstructured text, and Clinical Notes Indexing (CNI)-produced annotations.
2013 MFMER | slide-10

Methods
Profiled the Drug-Disorder Pairs and Relationships in the UMLS Chemicals & Drugs Disorders Enriched the UMLS ADE data with the EMR cooccurrence statistics Evaluated the Utility of the UMLS ADE Data

2013 MFMER | slide-11

Profiling Results of ADE Knowledge in the UMLS


We extracted 266,832 concept pairs between the two concept sets, covering 14,256 (1.69%) concepts from the group Chemicals & Drugs and 19,006 (3.53%) concepts from the group Disorders. We profiled the contribution of source vocabularies to the concept pairs. In total, there are 47 source vocabularies having the contribution.

2013 MFMER | slide-12

Distribution of drug-disorder concept pairs contributed by source vocabularies (top 20)


NDFRT SNOMEDCT SCTSPA MTH SNMI MEDLINEPLUS MSHSWE NCI MSHFIN MSHFRE MSHGER MSHCZE MSHRUS MSHPOL MSHITA MEDCIN MSHSCR MSH CCPSS MSHSPA

70.4% 7.3% 7.3% 3.1% 2.1% 0.7% 0.7% 0.7% 0.6% 0.6% 0.6% 0.6% 0.6% 0.6% 0.6% 0.5% 0.4% 0.4% 0.3% 0.3% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0%

0.0%

2013 MFMER | slide-13

Profiling results of relationships between drugs and disorders

2013 MFMER | slide-14

Co-occurrence Enrichment Using the EMRs


Drug-disorder concept pairs identified from UMLS Out of 266,832 concept pairs, 71,626 concept pairs co-occurred in the EMRs, accounting for 26.8%. Drug-disorder concept pairs from the ADEpedia Of 299,476 drug-disorder concept pairs, 100,577 concept pairs co-occurred in the EMRs, accounting for 33.6%.

2013 MFMER | slide-15

Utility Evaluation and Case Study


We identified 2,164 CUIs overlapping across the two data sets (ADEpedia vs. UMLS), accounting for 56.5% of RxNorm Drugs in the ADEpedia. The 2,164 CUIs linked with 66,058 drug-disorder pairs from the UMLS. The result indicated that the UMLS is a useful resource from the perspective of its integration with the ADEpedia. As a case study, we selected a drug called Digoxin, which is indicated for the treatment of congestive heart failure.

2013 MFMER | slide-16

ADE knowledge extracted from the UMLS for the drug Digoxin|C0012265

2013 MFMER | slide-17

Discussion
The profiling results of the UMLS clearly demonstrated that the drug-disorder associations in the UMLS are largely underspecified. Only 1.69% (14,256) concepts from the group Chemicals & Drugs and 3.53% (19,006) concepts from the group Disorders in the UMLS had drug-disorder associations asserted.

2013 MFMER | slide-18

Discussion
From the perspective of ADE detection application, we classified the asserted relationships between the drug-disorder pairs into 4 categories. This kind of categorization would provide aggregation capability for the knowledge source and improve its utility for the purpose of drug surveillance.

2013 MFMER | slide-19

Discussion
We enriched the knowledge base with the statistical co-occurrence information extracted from the EMRs. The co-occurrence information extracted from a 51-million-document EMR system will be useful for validating ADE detection algorithm across clinical institutions and across text corpora. In future, we will explore the characteristics of the UMLS term occurrences, focusing on the ADE detection use case.
2013 MFMER | slide-20

Summary
The UMLS is a useful source for providing standardized ADE knowledge relevant to indications, contraindications and adverse effects, and complementary to the ADE data from drug product labels. The statistics enrichment from EMRs would potentially enable the meaningful use of ADE data for drug safety surveillance. The EMR-enriched ADE dataset is available for download at the ADEpedia website (http://adepedia.org).

2013 MFMER | slide-21

Application Layer Service Layer

ADE Detection / Drug Surveillance Applications Semantic Web RESTful Services (JSON, XML, RDF, TTL, etc.)

Knowledge Base Layer

Standardized ADE Knowledge Base (RDF Store)

Data Layer

ADE Data From SPLs

ADE
Knowledge From UMLS

ADE
Knowledge From EMRs

Other ADE Resources (AERS)


2013 MFMER | slide-22

Questions & Discussion

2013 MFMER | slide-23

You might also like