PAPatrick R. Alba

Patrick R. Alba

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Patrick R. Alba has a total of 160 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Impact of Cumulative Social Determinants of Health on Odds of Diabetes Incidence in US Veterans

    by Lewis J. Frey, Mulugeta Gebregziabher, Kinfe G. Bishu, Brianna Youngblood, Jihad S. Obeid, Jianlin Shi, Patrick R. Alba, Chanita Hughes Halbert

    Background: Type 2 diabetes mellitus (T2DM) is a chronic condition that has been attributed to social factors; however, the cumulative effect of social determinants of health (SDOH) on T2DM incidence is not known. Objective: The aim of the present study was to examine the association between T2DM and the cumulative burden of multiple SDOH. Design: The study is a retrospective cohort study with a baseline between 2008 and 2009 and a ten-year follow-up between 2010 and 2019. Setting: The study was conducted using data from the United States Veterans Health Administration (VHA). Participants: Out of 10,537,027 patients treated in the VHA between 2010 and 2019, 6,518,102 patients were selected who had no evidence of T2DM or Elixhauser comorbidities at baseline (2008–2009). Measurements: Over 10 years following baseline, the exposure consisted of seven types of SDOH occurring in structured data: social isolation, financial stress, employment issues, food insecurity, transportation insecurity, unstably housed, and psychosocial need. Incidence of T2DM in the ten-year follow-up window was the primary outcome. Results: Veterans with ≥3 SDOH doubled their adjusted odds of T2DM (2.07; CI: 2.05–2.09). There were significant racial differences in cumulative SDOH, with 8.8% of Black individuals having the highest burden of ≥3 SDOH compared with 3.8% of White individuals. Transportation insecurity, psychosocial need, and financial stress significantly increased the odds of T2DM across all racial and ethnic groups. Black individuals had the highest T2DM odds ratio for psychosocial need (OR = 1.58; CI: 1.56, 1.60). Limitations: The Veteran population is predominantly male, limiting generalization to the wider population. Conclusions: With each additional SDOH burden, the odds of T2DM increased, and ≥3 SDOH doubled the odds. The cumulative SDOH burden and associated disparities warrant investigation to reduce T2DM incidence.

  • Open Access

    Multimorbidity Burden in Veterans with and Without Type 2 Diabetes Mellitus: A Comparative Retrospective Cohort Study

    by Lewis J. Frey, Mulugeta Gebregziabher, Kinfe G. Bishu, Brianna Youngblood, Jihad S. Obeid, Jianlin Shi, Patrick R. Alba, Scott L. DuVall, Christopher D. Blasy, Chanita Hughes Halbert

    Background/Objectives: Multimorbidity, where patients have ≥2 comorbidities, is recognized as a major challenge for health systems worldwide, driving up morbidity and cost. The differences in multimorbidity burden between those with and without type-2 diabetes mellitus (T2DM) in the Veteran population are not well studied. This large retrospective cohort study fills the existing gap. Methods: Using a retrospective cohort of adult Veterans with and without T2DM, we examined 29 comorbidities defined by Elixhauser criteria for 10,499,394 Veterans from 1 January 2008 to 31 December 2009. We then ascertained diabetes status for 10 years of follow-up from 1 January 2010 to 31 December 2019. Multimorbidity status was categorized using the Elixhauser comorbidity index (0, 1, ≥2) and logistic regression was used to estimate the odds ratio (OR) for its association with risk of diabetes, adjusting for covariates. Results: Compared to those with zero comorbidities, the odds of having diabetes were more than doubled (2.53, CI: 2.51–2.54) for those with ≥2 comorbidities. Conclusions: The doubling of the odds of T2DM among those with more than one comorbidity is typical of Veterans with T2DM. In addition, the odds were significantly higher for Hispanics compared to other groups when adjusting for covariates. This calls for more attention to reduce the risk of T2DM through improved management and effective use of treatments informed by disparities that exist in the VHA.


  • Open Access

    Comparative Analysis of NLP Models for Automatic LOINC Document Ontology Named Entity Recognition in Clinical Note Titles

    by Annie E. Bowles, Qiwei Gan, Elizabeth Hanchrow, Scott Duvall, Patrick R. Alba, Jianlin Shi

    In order to utilize clinical notes for research studies, it is necessary to identify the most relevant notes. Mapping to the LOINC Document Ontology makes this process easier by reducing the variability of note types. We experimented with three models to automatically identify LOINC DO entities in VA note titles. The supervised BERT model performed best, but the open-source large language models (LLMs) performed well despite a lack of fine-tuning. Future work will aim to improve note classification by including additional note metadata and contents, hybridizing with rule-based approaches, testing fine-tuned LLMs, and mapping to exact LOINC codes.

  • Open Access

    Development and Evaluation of Natural Language Processing Methods for Extracting Key Melanoma Pathology Concepts

    by Johnathan C. Stanley, Mengke Hu, Cecelia J. Madison, Jianlin Shi, Tiffany Quilter, Lacy Castleton, Elizabeth E. Hanchrow, Julie A. Lynch, Wesley Y. Yu, Patrick R. Alba

    This study presents the development and evaluation of an annotation schema and rule-based natural language processing (NLP) system for extracting key melanoma pathology concepts from surgical pathology reports. Achieving high precision and recall, our system addresses melanoma’s complex staging criteria for use in downstream staging and cohort recruitment operational needs.


  • Open Access

    A Schema for Extraction of Comprehensive

    by Ivy Cerelia Valerie, Patrick R. Alba, Georgina Kennedy

    The characterization of patient smoking history is an integral part of evidence-based cancer care. We developed a schema to fully describe longitudinal history and quantified behavioral patterns for smoking consumption, including substance, modality, recency, quantity and cessation efforts, to support extraction from free-text patient histories.

  • Patterns of Outpatient Urinalysis Testing and the Detection of Microscopic Hematuria

    by Richard S. Matulewicz, Samuel A. Gold, Fady Baky, Andrew Nicholson, Eric Wahlstedt, Patrick Alba, Bernard H. Bochner, Harry W. Herr, David S. Goldfarb, Julie A. Lynch, et al.


  • Open Access

    Evaluating Clinical Note Deidentification Tools and Transformer Transferability between Public and Private Data from the US Department of Veterans Affairs

    by Hannah Eyre, Qiwei Gan, Mengke G Hu, Annie Bowles, Johnathan Stanley, Jianlin Shi, Scott L. DuVall, Patrick R. Alba

    1 Background: Deidentification of clinical notes is critical for enabling secondary use of electronic health records (EHRs) while maintaining patient privacy. Existing deidentification tools vary in methodology, with rule-based, hybrid, and transformer-based approaches offering different trade-offs in performance. However, most evaluations of these tools are conducted on publicly available datasets, limiting their generalizability to private healthcare systems such as the U.S. Department of Veterans Affairs (VA). Methods: We introduce an annotated corpus of 1,000 VA clinical notes and evaluated the performance of four open-source deidentification tools (Physionet Deid, Philter, CliniDeID, and Stanford Deid) using precision, recall, and F1-score. Additionally, we fine-tuned BioClinicalBERT using VA and I2B2 training data to assess the transferability of transformer models across public and private datasets. Results: Transformer-based models outperformed rule-based tools when identifying names and locations, while rule-based systems demonstrated higher recall for more structured entities such as dates. CliniDeID showed the highest performance among external tools, but no tool achieved optimal balance across all PHI categories. Transformer models trained on both public and VA datasets performed best, though challenges remained in handling VA-specific terminology and annotations. Discussion: These findings underscore the limitations of applying publicly trained models to private healthcare systems without adaptation. Differences in text structure, clinical terminology, and annotation standards impact deidentification effectiveness. While transformer models offer superior flexibility, computational cost and privacy concerns may hinder large-scale adoption. Conclusion: The results emphasize the need for institution-specific evaluation and fine-tuning of deidentification models. Future work should explore large language model and hybrid approaches that integrate rule-based and transformer-based methods to enhance adaptability and accuracy in diverse clinical settings.

  • Open Access

    Incorporating Structured and Unstructured Data Sources to Identify and Characterize Hereditary Cancer Testing Among Veterans With Metastatic Castration-Resistant Prostate Cancer

    by Danielle Candelieri-Surette, Anna Hung, Fatai Y. Agiri, Mengke Hu, Elizabeth E. Hanchrow, Kyung Min Lee, Nai-Chung N. Chang, Ming Yin, Jeffrey W. Shevach, Weiyan Li, et al.

    PURPOSE: This study introduces an integrated approach using structured and unstructured data from an electronic health record to identify and characterize patient utilization of hereditary cancer genetic testing among patients with metastatic castration-resistant prostate cancer (mCRPC). Secondary objectives were to describe factors associated with the receipt of testing. METHODS: This retrospective cohort study included a cohort of Veterans diagnosed with mCRPC from January 2016 to December 2021. Receipt of genetic testing was identified using structured and unstructured data. Time to testing, age at testing, and testing rate were analyzed. Sociodemographic and clinical factors associated with receipt of hereditary cancer genetic testing were identified including race, marital status, rurality, Charlson comorbidity index (CCI), and genetic counseling. RESULTS: Among 9,703 Veterans with mCRPC who did not decline testing, 16% received genetic testing, with nearly half of the tests occurring in 2020-2021. Factors positively associated with genetic testing included receipt of genetic counseling (adjusted odds ratio [aOR], 11.07 [95% CI, 3.66 to 33.51]), enrollment in clinical trial (aOR, 7.42 [95% CI, 5.59 to 9.84]), and treatment at a Prostate Cancer Foundation-Veterans Affairs Center of Excellence (aOR, 1.43 [95% CI, 1.04 to 1.95]). Negative associations included older age (aOR, 0.95 [95% CI, 0.93 to 0.97]) and severe CCI score (aOR, 0.82 [95% CI, 0.71 to 0.94]). Trends revealed that time to testing decreased per diagnosis year while median age at testing increased per year. CONCLUSION: Although testing rates are still suboptimal, they have increased steadily since 2016. Educating Veterans about the benefits of genetic testing may further improve testing rates.

  • Open Access

    An NLP Framework for the Extraction of Concept Measurements from Radiology and Pathology Notes

    by Annie Bowles, Cris Perez, Anil Vachani, Jennifer Steltz, Brent Rose, Alex K. Bryant, Hannah Eyre, Scott L. DuVall, Julie A. Lynch, Patrick R. Alba

    Natural language processing (NLP) tools can automate the identification of cancer patients eligible for specific pathways. We developed and validated a cancer agnostic, rules-based NLP framework to extract the dimensions and measurements of several concepts from pathology and radiology reports. This framework was then efficiently and cost-effectively deployed to identify patients eligible for breast, lung, and prostate cancers clinical pathways.


  • Open Access

    Toward Real-World Reproducibility: Verifying Value Sets for Clinical Research

    by Scott L. DuVall, Craig G. Parker, Amanda R. Shields, Patrick R. Alba, Julie A. Lynch, Michael E. Matheny, Aaron W.C. Kamauu

    Standardized operational definitions are an important tool to improve reproducibility of research using secondary real-world healthcare data. This approach was leveraged for studies evaluating the effectiveness of AZD7442 as COVID-19 pre-exposure prophylaxis across multiple healthcare systems. Value sets were defined, grouped, and mapped. Results of this exercise were reviewed and recorded. Value sets were updated to reflect findings.