As the 6th leading chronic disease in the U.S., allergies affect 30% of adults and 40% of children. Adverse drug reactions occur in 1 in 4 outpatients and 1 in 5 inpatients. Many allergies and adverse reactions warrant documentation in the electronic health record (EHR) allergy section to inform future medical care and prescribing. It is critical to obtain a complete and accurate allergy history for each patient and to provide clinicians with an efficient allergy-alerting clinical decision support (CDS) tool. However, the allergy modules in most existing EHRs have serious limitations in how allergies are documented and drug allergy alerts are fired. These include: frequently missing documentation of reaction mechanism and type, lack of a comprehensive terminology subset for encoding diverse reactions, insufficient tools for reconciling allergy information, and physician alert fatigue resulting from an alert override rate of greater than 90%. In this study, we will provide solutions to these challenges by addressing the following specific aims: 1) improve reaction documentation by developing a comprehensive and interactive value set; 2) develop an innovative allergy reconciliation module within the EHR; 3) redesign drug allergy alerting mechanisms; and 4) distribute our methods and tools to healthcare institutions and the research community.
Accurate documentation and management of patients? adverse reactions and allergies to medications, foods and other substances in the EHR is vital to ensuring patient safety and quality of care. In this study, we propose a system redesign that employs a suite of innovative health information technology (HIT) solutions, including an enhanced reaction value set, dynamic pick lists, natural language processing, new drug-allergy alerting mechanisms, and increased knowledge generated by domain experts and big data analytics, with an overall goal of improving health care quality and safety using HIT.
| Wong, Jenna; Horwitz, Mara Murray; Zhou, Li et al. (2018) Using machine learning to identify health outcomes from electronic health record data. Curr Epidemiol Rep 5:331-342 |
