Developing a Reconciliation Module to Improve the Accuracy of EHR Allergy Lists

Sachin Vallamkonda, BA
Department of Medicine
Division of General Internal Medicine and Primary Care
Poster Overview

PURPOSE: Patient allergy information in the electronic health record (EHR) is often inaccurately documented, haphazardly stored, and not updated timely. We aim to develop an automated allergy reconciliation tool that allows clinicians to accurately and easily reconcile these discrepancies, enhancing patient safety.

METHODS: The reconciliation tool will first identify relevant area within the EHR where patient allergy information is stored. This includes allergy lists, clinical notes, allergy tests, and drug- allergy alerts that clinicians receive. The tool will then identify areas of improvement (e.g., missing or inaccurate information) automatically and continuously. We utilized natural language processing to process information that is entered in as free-text data. We also mapped the different names used to identify allergens and allergy information in order to address duplicate documentation.

RESULTS: Our preliminary results based on 2018 data showed that 43% of all active allergies had free-text in the comment field. Furthermore, 20,000 patients had duplicate allergens on their allergy list. Such duplicate entries can be reconciled to improve accuracy of patients’ health records and reduce providers’ allergy alerts.

CONCLUSION: An allergy reconciliation module would greatly improve allergy information accuracy and documentation quality. This is pertinent for patient safety and provider workflow.

Scientific Abstract

RATIONALE: Patient allergy information in the electronic health record (EHR) is often inaccurately documented, haphazardly stored, and not updated timely. We aim to develop an automated allergy reconciliation tool that allows clinicians to accurately and easily reconcile these discrepancies, enhancing patient safety.

METHODS: The reconciliation tool will first identify relevant allergy information areas within the EHR (including allergy lists and free-text comments, clinical notes, challenge tests, allergy lab tests, flowsheets, and drug-allergy alerts). The tool will then identify areas of improvement (e.g., missing or inaccurate information) automatically and continuously. We utilized natural language processing to process free-text data and mapped various versions of allergy information (e.g., different names) to address duplicate allergen documentation.

RESULTS: Our preliminary results based on 2018 data showed that 43% of all active allergies had free-text in the comment field, and 11.9% had a reaction in the comment field that could have been coded. Furthermore, 20,000 patients had duplicate allergens on their allergy list. Such duplicate entries can be reconciled to improve accuracy of patients’ health records and reduce providers’ allergy alerts.

CONCLUSION: An allergy reconciliation module would greatly improve allergy information accuracy and documentation quality. This is pertinent for patient safety and provider workflow.

Clinical Implications
Patient allergy information in the electronic health record is often inaccurately documented, haphazardly stored, and not updated timely. An automated allergy reconciliation tool would allow clinicians to accurately and easily improve allergy documentation’s accuracy and quality and bolster patient safety.
Research Areas
Authors
Sachin Vallamkonda
Principal Investigator
Dr. Li Zhou

Explore Other Posters

Leave a Reply

Your email address will not be published. Required fields are marked *