Upload your resume. PhdFit's six research agents compare your background with faculty profiles, recent publications, lab focus, and outreach opportunities, then rank professors with evidence you can review.
Ask how her lab is extending interpretability methods into fairness audits for real-world AI systems.
University of Washington · Pathology
Active 1995–2026
Patrick C. Mathias, MD, PhD, is an Associate Professor and Vice Chair of Clinical Operations in the Department of Laboratory Medicine and Pathology at the University of Washington School of Medicine. He also serves as the Associate Medical Director of Informatics and the Medical Director of Airlift NW POCT. His clinical and academic focus includes clinical informatics, laboratory stewardship, cost-effectiveness of laboratory testing, and clinical informatics interventions. Dr. Mathias holds board certifications in Clinical Informatics and Clinical Pathology from the American Board of Pathology, obtained in 2018 and 2015 respectively. He has academic appointments as an Associate Professor and adjunct roles in the Department of Biomedical Informatics and Medical Education. His training includes a fellowship in Clinical Informatics at the University of Washington and residency in Clinical Pathology at the same institution. His research interests are centered on clinical informatics, laboratory stewardship, and improving laboratory testing practices to enhance patient care and operational efficiency.
Development and implementation of an AI system for clinical toxicology sign-outs
medRxiv · 2026-01-30
Abstract Background Modern natural language tools have potential to improve clinical workflows, but few have been successfully deployed in practice. Here, we present the development, deployment, and evaluation of an AI language tool for generating preliminary clinical sign-outs in a urine drug testing service. Methods Large language models (LLMs) were used to extract substance use patterns from 83,553 urine drug test interpretations. We then trained an AI model using these data to predict substance use from qualitative and quantitative urine testing results. Predicted substance use patterns were used to create preliminary clinical sign-out statements, which were then integrated into an existing clinical workflow. Pre- and post-deployment user studies were performed to evaluate model performance and user experience within this workflow. Results LLM-based extraction of substance-use patterns was 99.9% accurate, outperforming human labelling. Substance use prediction was similarly accurate, with area under the ROC curve > 0.99 across 33 drug categories. Workflow integration reduced clinical sign-out times by 65s per case (51% efficiency gain), with the greatest benefits seen for less experienced users. Conclusions AI-based interpretation of urine drug testing was fast and accurate, providing significant efficiency gains to the clinical service. This demonstrates that natural language tool integration can provide substantial clinical benefit, without comprising quality of care.
Alexander L. Greninger
University of Washington
Keith R. Jerome
University of Washington
Brian T. Cunningham
Pavitra Roychoudhury
University of Washington
Nikhil Ganesh
M.D.
University of Washington
Ph.D.
University of Washington
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
Improving laboratory stewardship through benchmarking: A focus on thyroid function tests ordering
Clinical Biochemistry · 2025-05-26 · 1 citations
2025-09-01
<h3>Context</h3> Prostate cancer accounts for 15% of all new cancer cases in the U.S. and is the most common cancer in males, with the Prostate-Specific Antigen (PSA) test serving as the primary screening method. While the PSA test is valuable for early detection, over-screening can lead to false positives, unnecessary biopsies, and potential overtreatment. On May 1, 2024, the University of Washington (UW) updated EPIC health maintenance care gap alerts to promote shared decision-making for PSA testing, recommending screening starting at age 45 for average-risk men and age 40 for high-risk groups. <h3>Objective</h3> To evaluate the impact of an EPIC health maintenance alert on PSA test utilization in UW primary care by analyzing changes in PSA testing practices over time, comparing pre- and post-implementation periods. <h3>Study Design and Analysis</h3> We performed a descriptive analysis on PSA testing frequency by age, race, clinician specialty, ordering location, PSA test order codes, and type of care. To assess changes over time, we compared the frequency of PSA tests 6 months before and after the EPIC health maintenance alert implementation. <h3>Dataset</h3> Data were extracted from the EPIC EHR and Clinisys lab information system via a departmental data warehouse. <h3>Population Studied</h3> Clinicians within the UW medical system. <h3>Intervention/Instrument</h3> EPIC health maintenance care gap alert. <h3>Results</h3> PSA testing increased across all age groups post-implementation, with a disproportionate rise in non-screening PSA tests ordered within the screening population. UW Primary Care clinicians ordered the most tests, followed by hospital clinic specialists. Testing increased among black patients aged 40-44 and those in this age range with a family history of prostate cancer. However, family history documentation is frequently captured in clinical notes rather than flowsheets, limiting data accuracy. <h3>Conclusions</h3> EPIC health maintenance alerts increased PSA testing in both average-risk and high-risk groups, along with incorrect PSA test orders for screening. Encouragingly, higher-risk groups, such as Black patients and those with a family history, showed increased testing, but better documentation of family history and provider education are necessary. Future efforts will focus on cost and comorbidity analyses and developing enhanced clinical decision support tools through collaboration with Family Medicine, Urology, Population Health, Chief Population Health Office, and Lab Medicine.
The Lines That Held Us: Assessing Racial and Socioeconomic Disparities in SARS-CoV-2 Testing
UNC Libraries · 2025-05-14
BACKGROUND: Racial disparities in SARS-CoV-2 prevalence are apparent. Race is a sociocultural construct, necessitating investigation into how sociocultural factors contribute. METHODS: This cross-sectional study linked laboratory data of adult patients between February 29 and May 15, 2020 with socio-demographics variables from the 2018 American Community Survey (ACS). Medical sites included healthcare organizations in Michigan, New York, North Carolina, California, Florida, Pennsylvania, and Washington. Race was treated as a proxy for racism and not biological essentialism. Laboratory data included patient age, sex, race, ethnicity, test result, test location, and residential ZIP code. ACS data included economic and educational variables contributing to an SES Index, population density, proportion Medicaid, and racial composition for corresponding ZIP code. Associations between race/socioeconomic variables and test results were examined using odds ratios (OR). RESULTS: Of 126 452 patients [mean (SD) age 51.9 (18.4) years; 52 747 (41.7%) men; 68 856 (54.5%) White and 27 805 (22.0%) Black], 18 905 (15.0%) tested positive. Of positive tests, 5238 (SD 27.7%) were White and 7223 (SD 38.2%) were Black. Black race increased the odds of a positive test; this finding was consistent across sites [OR 2.11 (95% CI 1.95-2.29)]. When subset by race, higher SES increased the odds of a positive test for White patients [OR 1.10 (95% CI 1.05-1.16)] but decreased the odds for Black patients [OR 0.92 (95% CI 0.86-0.99)]. Black patients, but not White patients, who tested positive overwhelmingly resided in more densely populated areas. CONCLUSIONS: Black race was associated with SARS-CoV-2 positivity and the relationship between SES and test positivity differed by race, suggesting the impact of socioeconomic status on test positivity is race-specific.
Use of large-language models to generate automated drug screen interpretations
Journal of Pathology Informatics · 2025-11-01
PLoS Pathogens · 2024-03-26 · 10 citations
SARS-CoV-2 transmission is largely driven by heterogeneous dynamics at a local scale, leaving local health departments to design interventions with limited information. We analyzed SARS-CoV-2 genomes sampled between February 2020 and March 2022 jointly with epidemiological and cell phone mobility data to investigate fine scale spatiotemporal SARS-CoV-2 transmission dynamics in King County, Washington, a diverse, metropolitan US county. We applied an approximate structured coalescent approach to model transmission within and between North King County and South King County alongside the rate of outside introductions into the county. Our phylodynamic analyses reveal that following stay-at-home orders, the epidemic trajectories of North and South King County began to diverge. We find that South King County consistently had more reported and estimated cases, COVID-19 hospitalizations, and longer persistence of local viral transmission when compared to North King County, where viral importations from outside drove a larger proportion of new cases. Using mobility and demographic data, we also find that South King County experienced a more modest and less sustained reduction in mobility following stay-at-home orders than North King County, while also bearing more socioeconomic inequities that might contribute to a disproportionate burden of SARS-CoV-2 transmission. Overall, our findings suggest a role for local-scale phylodynamics in understanding the heterogeneous transmission landscape.
Transfusion · 2024-04-22
BACKGROUND: With the widespread adoption of Blood Establishment Computer Systems and other Blood Collection and Transfusion Service (BCTS) clinical information systems (CIS), electronic blood donor, product, and patient data are now routinely required for clinical, regulatory, operational, and quality needs. That data are often not readily accessible for such secondary use within CIS databases, particularly for applications with significant data availability requirements such as machine learning and artificial intelligence. Data replication provides one avenue by which CIS data can be made more readily available. STUDY DESIGN AND METHODS: Members of the AABB's Information Systems Committee along with institutional information technology colleagues provided a multi-institutional viewpoint on data replication through the lens of BCTS specific use cases. Case studies of informatics offerings leveraging such technologies were also elicited. RESULTS: Six distinct use cases describe the potential role of data replication including the creation of data warehouses for frontline laboratory staff. Specific BCTS examples for each use case are presented to highlight the value of data replication, including visualization of critical inventory (O red blood cells, HLA-compatible platelets) and utilization analytics for patient blood management. Two case studies describe the approach to implement such technologies to (1) optimize staffing via laboratory workload reporting and (2) improve access to blood via antigen-negative blood product location services. DISCUSSION: Data replication and warehousing can empower BCTS analytic offerings not otherwise natively available through one's CIS to improve patient care and laboratory operations.
Emerging infectious diseases · 2023-01-04 · 15 citations
Genomic data provides useful information for public health practice, particularly when combined with epidemiologic data. However, sampling bias is a concern because inferences from nonrandom data can be misleading. In March 2021, the Washington State Department of Health, USA, partnered with submitting and sequencing laboratories to establish sentinel surveillance for SARS-CoV-2 genomic data. We analyzed available genomic and epidemiologic data during presentinel and sentinel periods to assess representativeness and timeliness of availability. Genomic data during the presentinel period was largely unrepresentative of all COVID-19 cases. Data available during the sentinel period improved representativeness for age, death from COVID-19, outbreak association, long-term care facility-affiliated status, and geographic coverage; timeliness of data availability and captured viral diversity also improved. Hospitalized cases were underrepresented, indicating a need to increase inpatient sampling. Our analysis emphasizes the need to understand and quantify sampling bias in phylogenetic studies and continue evaluation and improvement of public health surveillance systems.
Utility of Vizient Clinical Data Base as a benchmarking tool for laboratory stewardship programs
American Journal of Clinical Pathology · 2023-11-01
Abstract Establishing an effective laboratory stewardship program across healthcare systems to improve patient care has been of growing interest. Despite general guidelines from various organizations, clear definitions for these laboratory-based quality measures, and associated benchmarks, are still lacking. Hence, we utilized Vizient® Clinical Data Base (CDB) as a tool to benchmark our institution against U.S. News and World Reports’ Honor Roll (USNW) healthcare organizations 2021-2022. We focused on inpatient laboratory testing with specific targets including mean laboratory resource units used per patient per day to obtain an overview of our institution’s overall performance, daily basic metabolic panel and CBC test ordering patterns across various service lines, and C-reactive protein (CRP) to erythrocyte sedimentation rate (ESR) order ratios compared to USNW hospitals. As one of the five institutions with the highest mean laboratory resource units used per case per day (12.7) from January 1, 2021-January 31, 2022, we investigated this further to identify tests driving this result. Of all the laboratory tests ordered, blood gas was identified at the 99th percentile compared to peers. Stratifying by service, we found cardiac surgery (17%), general medicine (15%), pulmonary/critical care (11%), neonatology (10%) and transplant (9%) services at our institution were the major contributors to blood gas orders. Investigating daily laboratory test orders (BMP, CBC) showed that our institution was at the 46th and 71st percentile overall, respectively. Further CBC analysis showed that general medicine (28%) and oncology (11%) services were the major contributors for the test volume and highest total resource units used per quarter in 2021. Assessing CRP to ESR ratio depicted that our institution’s ratio of 2.2 was lower compared to USNW hospitals’ average of 2.9, meaning ESR is likely overutilized. In conclusion, Vizient CDB can be used as a resource for benchmarking laboratory utilization patterns. It can be helpful in developing road maps and prioritizing interventions for lab stewardship programs. This study identified a general overuse of laboratory resources at the hospital compared to peers, and identified some intervention targets including blood gas testing and ESR.
On Developing Better Practices for Reproducible Data Analysis in Laboratory Medicine
The Journal of Applied Laboratory Medicine · 2023-01-04 · 1 citations
Journal Article On Developing Better Practices for Reproducible Data Analysis in Laboratory Medicine Get access Patrick C Mathias Patrick C Mathias Department of Laboratory Medicine and Pathology, University of Washington School of Medicine, Seattle, WADepartment of Biomedical Informatics and Medical Education, University of Washington School of Medicine, Seattle, WA Address correspondence to this author at: Department of Laboratory Medicine and Pathology, University of Washington School of Medicine, 1959 NE Pacific St., Box 357110, Seattle, WA 98195, USA. E-mail pcm10@uw.edu. https://orcid.org/0000-0003-4266-561X Search for other works by this author on: Oxford Academic Google Scholar The Journal of Applied Laboratory Medicine, Volume 8, Issue 1, January 2023, Pages 229–231, https://doi.org/10.1093/jalm/jfac097 Published: 04 January 2023 Article history Received: 06 July 2022 Accepted: 28 September 2022 Published: 04 January 2023
Trevor Bedford
Fred Hutch Cancer Center
Ian D. Block
University of Fribourg