Blythe Adamson

Ms. Blythe Adamson

orcid

Research Focus & Metrics

Academic Career Length

30Years (1997-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 March 2020 in medRxiv
Verified
Preprint
Open access
PDF file
XMXinran Ma
Xinran Ma
LLLura Long
Lura Long
SMSharon Moon
Sharon Moon
Blythe AdamsonBlythe Adamson
Blythe Adamson
SBShrujal S. Baxi
Shrujal S. Baxi
Background and Objective: The Surveillance, Epidemiology, and End Results Program (SEER) program and the National Program of Cancer Registries (NPCR), are authoritative sources for population cancer surveillance and research in the US. An increasing number of recent oncology studies are based on the electronic health record (EHR)-derived de-identified databases created and maintained by Flatiron Health. This report describes the differences in the originating sources and data development processes, and compares baseline demographic characteristics in the cancer-specific databases from Flatiron Health, SEER, and NPCR, to facilitate interpretation of research findings based on these sources. Methods: Patients with documented care from January 1, 2011 through May 31, 2019 in a series of EHR-derived Flatiron Health de-identified databases covering multiple tumor types were included. SEER incidence data (obtained from the SEER 18 database) and NPCR incidence data (obtained from the US Cancer Statistics public use database) for malignant cases diagnosed from January 1, 2011 to December 31, 2016 were included. Comparisons of demographic variables were performed across all disease-specific databases, for all patients and for the subset diagnosed with advanced-stage disease. Results: As of May 2019, a total of 201,570 patients with 19 different cancer types were included in Flatiron Health datasets. In an overall comparison to national cancer registries, patients in the Flatiron Health databases had similar sex, age at initial diagnosis, and geographic distributions but appeared to be diagnosed with later stages of disease compared with patients in other datasets. For variables such as stage and race, Flatiron Health databases had a greater degree of incompleteness. There are variations in these trends by cancer types. Conclusions: These three databases present general similarities in demographic and geographic distribution, but there are overarching differences across the populations they cover. Differences in data sourcing (medical oncology EHRs vs cancer registries), and disparities in sampling approaches and rules of data acquisition may explain some of these divergences. Furthermore, unlike the steady information flow entered into registries, the availability of medical oncology EHR-derived information reflects the extent of involvement of medical oncology clinics at different points in the specialty management of individual diseases, resulting in inter-disease variability. These differences should be considered when interpreting study results obtained with these databases.

15 September 2023 in Frontiers in Pharmacology
Verified
Journal Article
Open access
Blythe AdamsonBlythe Adamson
Blythe Adamson
MWMichael Waskom
Michael Waskom
ABAuriane Blarre
Auriane Blarre
JKJonathan Kelly
Jonathan Kelly
KKKonstantin Krismer
Konstantin Krismer
ACAaron B. Cohen
Aaron B. Cohen
Background: As artificial intelligence (AI) continues to advance with breakthroughs in natural language processing (NLP) and machine learning (ML), such as the development of models like OpenAI’s ChatGPT, new opportunities are emerging for efficient curation of electronic health records (EHR) into real-world data (RWD) for evidence generation in oncology. Our objective is to describe the research and development of industry methods to promote transparency and explainability.Methods: We applied NLP with ML techniques to train, validate, and test the extraction of information from unstructured documents (e.g., clinician notes, radiology reports, lab reports, etc.) to output a set of structured variables required for RWD analysis. This research used a nationwide electronic health record (EHR)-derived database. Models were selected based on performance. Variables curated with an approach using ML extraction are those where the value is determined solely based on an ML model (i.e. not confirmed by abstraction), which identifies key information from visit notes and documents. These models do not predict future events or infer missing information.Results: We developed an approach using NLP and ML for extraction of clinically meaningful information from unstructured EHR documents and found high performance of output variables compared with variables curated by manually abstracted data. These extraction methods resulted in research-ready variables including initial cancer diagnosis with date, advanced/metastatic diagnosis with date, disease stage, histology, smoking status, surgery status with date, biomarker test results with dates, and oral treatments with dates.Conclusion: NLP and ML enable the extraction of retrospective clinical data in EHR with speed and scalability to help researchers learn from the experience of every person with cancer.

5 January 2022 in medRxiv
Verified
Preprint
Open access
PDF file
Blythe AdamsonBlythe Adamson
Blythe Adamson
RSRobby Sikka
Robby Sikka
AWAnne L. Wyllie
Anne L. Wyllie
PPPrem Premsrirut
Prem Premsrirut
The performance of Covid-19 diagnostic tests must continue to be reassessed with new variants of concern. The objective of this study was to describe the discordance in saliva SARS-CoV-2 PCR and nasal rapid antigen test results during the early infectious period. We identified a high-risk occupational case cohort of 30 individuals with daily testing during an Omicron outbreak in December 2021. Based on viral load and transmissions confirmed through epidemiological investigation, most Omicron cases were infectious for several days before being detectable by rapid antigen tests.

20 March 2023 in Cancers
Verified
Journal Article
Open access
PDF file
CBCorey Benedum
Corey Benedum
ASArjun Sondhi
Arjun Sondhi
EFErin Fidyk
Erin Fidyk
ACAaron B. Cohen
Aaron B. Cohen
SNSheila Nemeth
Sheila Nemeth
Selen BozkurtSelen Bozkurt
Selen Bozkurt
Meaningful real-world evidence (RWE) generation requires unstructured data found in electronic health records (EHRs) which are often missing from administrative claims; however, obtaining relevant data from unstructured EHR sources is resource-intensive. In response, researchers are using natural language processing (NLP) with machine learning (ML) techniques (i.e., ML extraction) to extract real-world data (RWD) at scale. This study assessed the quality and fitness-for-use of EHR-derived oncology data curated using NLP with ML as compared to the reference standard of expert abstraction. Using a sample of 186,313 patients with lung cancer from a nationwide EHR-derived de-identified database, we performed a series of replication analyses demonstrating some common analyses conducted in retrospective observational research with complex EHR-derived data to generate evidence. Eligible patients were selected into biomarker- and treatment-defined cohorts, first with expert-abstracted then with ML-extracted data. We utilized the biomarker- and treatment-defined cohorts to perform analyses related to biomarker-associated survival and treatment comparative effectiveness, respectively. Across all analyses, the results differed by less than 8% between the data curation methods, and similar conclusions were reached. These results highlight that high-performance ML-extracted variables trained on expert-abstracted data can achieve similar results as when using abstracted data, unlocking the ability to perform oncology research at scale.

16 August 2022 in npj Digital Medicine
Verified
Journal Article
Open access
PDF file
LCLorinda Coombs
Lorinda Coombs
AOAbigail Orlando
Abigail Orlando
XWXiaoliang Wang
Xiaoliang Wang
PSPooja Shaw
Pooja Shaw
ARAlexander S. Rich
Alexander S. Rich
KMKathi Mooney
Kathi Mooney
We present a general framework for developing a machine learning (ML) tool that supports clinician assessment of patient risk using electronic health record-derived real-world data and apply the framework to a quality improvement use case in an oncology setting to identify patients at risk for a near-term (60 day) emergency department (ED) visit who could potentially be eligible for a home-based acute care program. Framework steps include defining clinical quality improvement goals, model development and validation, bias assessment, retrospective and prospective validation, and deployment in clinical workflow. In the retrospective analysis for the use case, 8% of patient encounters were associated with a high risk (pre-defined as predicted probability ≥20%) for a near-term ED visit by the patient. Positive predictive value (PPV) and negative predictive value (NPV) for future ED events was 26% and 91%, respectively. Odds ratio (OR) of ED visit (high- vs. low-risk) was 3.5 (95% CI: 3.4–3.5). The model appeared to be calibrated across racial, gender, and ethnic groups. In the prospective analysis, 10% of patients were classified as high risk, 76% of whom were confirmed by clinicians as eligible for home-based acute care. PPV and NPV for future ED events was 22% and 95%, respectively. OR of ED visit (high- vs. low-risk) was 5.4 (95% CI: 2.6–11.0). The proposed framework for an ML-based tool that supports clinician assessment of patient risk is a stepwise development approach; we successfully applied the framework to an ED visit risk prediction use case.

1 July 2021 in The American Journal of Managed Care
Verified
Journal Article
Closed access
Blythe AdamsonBlythe Adamson
Blythe Adamson
ACAaron B. Cohen
Aaron B. Cohen
CGCary P. Gross
Cary P. Gross
MEMelissa Estevez
Melissa Estevez
FMFnp-Bc Kelly Magee
Fnp-Bc Kelly Magee
ADAmy J. Davidoff
Amy J. Davidoff
Medicaid expansion was associated with a reduction in the racial disparity in timely treatment of patients with advanced cancer in the United States.