= Emerging. More than 5 years before clinical availability. (26.55%)
= Expected to be clinically available in 1 to 4 years. (39.66%)
= Clinically available now. (33.79%)
MSACL 2020 US : Mays

MSACL 2020 US Abstract

Topic: Data Science

Podium Presentation in Room 3 on Thursday at 9:20 (Chair: Lindsay Bazydlo)

An Interactive Dashboard for Toxicology Results Using the R Shiny Package

Alex Mays (Presenter)
University of Washington

Presenter Bio(s): I am a fourth year resident and current chief resident in anatomic and clinical pathology at the University of Washington. During my time in the departments of Laboratory Medicine and Pathology, my research work has focused predominantly on informatics-based solutions to issues commonly arising in the operation of a clinical laboratory, with special emphasis on clinical chemistry. During residency I also spent two years studying anatomic pathology, where I developed an interest in novel tissue imaging methods. In my last year of residency I am focusing predominantly on data warehouses, application development, and bioinformatics pipelines. After residency, I will begin a fellowship in clinical informatics at Massachusetts General Hospital.

Authors: Alex Mays (1), Patrick Mathias (1)
(1) University of Washington



The epidemiology of substance abuse is often locally specific and subject to rapid change. Retrospective review and real time monitoring of drug testing results in emergency department patients allows laboratories, treating clinicians, and public health authorities to identify new trends in substance abuse as they occur. Our institution screens for drugs of abuse using a standard screen consisting of immunoassays for eleven common drugs of abuse and a comprehensive drug screen that includes a GC/MS assay able to detect 213 other compounds.


The primary objective was to create an interactive dashboard for the display and analysis of toxicology results at our institution.


We queried our laboratory information system (LIS) for all standard and comprehensive drug screens. Data from January 2018 onwards is contained in a departmental data warehouse, and data prior to this was queried directly from the LIS. Using the free open source statistical programming language R and the visualization packages shiny, flexdashboard, and plotly, we developed a browser-based dashboard for interactive display of summarized historical clinical results in a secure, deidentified manner.


For the standard screen, our institution performed 60,459 unique tests representing 42,670 unique patients. Within this population, there were 25,385 unique tests for the GC/MS assay representing 20,649 patients. The data spanned from April 1st, 2004 to the current date. Over the study period, 180 unique compounds were detected using the GC/MS assay. Both immunoassay and GC/MS results identified marked changes over time in positivity rates for certain drugs, including a positive rate for methamphetamine of less than 10% in 2004 to 35% of all screens in 2019, and an increase in cannabinoids detected from approximately 20% in 2004 to 48% in 2019 thus far.


Data visualization dashboards for clinical laboratory results represent a flexible method for real-time laboratory analytics. This dashboard allows investigation of trends over time, by site, drug, and patient demographic variables including age and sex. Clinicians and laboratorians using the tool can track local trends of drug use in a visual, accessible format not given by LIS queries. Additional issues and considerations for deploying clinical applications include the necessity of structured data formatting, a prior existing infrastructure for data storage and retrieval, and appropriate measures to secure and anonymize sensitive patient data. Future work includes the deployment of this tool as a continuously updated web-based dashboard using data derived from a departmental data warehouse.

Financial Disclosure

Board Memberno
IP Royaltyno

Planning to mention or discuss specific products or technology of the company(ies) listed above: