The goal of the DECODE lab is to disentangle how biological, clinical, behavioral, environmental, and structural factors contribute to disparities in healthcare utilization and outcomes. We leverage healthcare data, national health surveys, data linkage, and causal inference to identify the causes of health disparities in racial and/or ethnic minority populations and rural communities in the United States.
Our current research focuses on identifying barriers to preventive care and disparities in the screening and treatment of breast, cervical, and colorectal cancer.
PREV-CARE survey. The PREVentive Care AdheREnce in Diverse Populations (PREV-CARE) survey is a nationally representative, online survey of American Indian or Alaska Native, Asian, Black, Latino (English- and Spanish-speaking), Middle Eastern or North African, Native Hawaiian or Pacific Islander, White, and multiracial adults ≥35 years old conducted between June and August 2023 (N=4,111). Survey included questions on preventive care utilization, barriers to care, health literacy, and perceived disease risk. Participants provided their home ZIP code so neighborhood measures could be linked to their survey responses.
Disaggregating health differences and health disparities. Dr. Strassle recently developed a novel application of AI/ML to quantify how much differences in healthcare utilization and outcomes are due to clinical differences (allowable factors) and how much are due to disparities in healthcare access and other social determinants of health (unallowable factors), using major lower limb amputations among adults with vascular disease as an example. We are now working to further develop this approach and apply it to cancer screening and treatment.
Allostatic load, genetics, and health. Dr. Strassle is an active member of the EPIB Allostatic Load and Health Working Group, where she provides expertise on building patient cohorts and capturing health outcomes in electronic health records (EHR) and health survey data. Several studies are currently underway using data from the UK Biobank and NIH All of Us.
Immortal time and immortal time bias. Immortal time occurs when treatment starts after Time 0 in a prospective or retrospective cohort study. If unaddressed, immortal time can lead to bias exaggerating the effect of treatment, making it seem more beneficial than it is. Immortal time (and immortal time bias) are of particular concern in cancer care research, because treatment typically begins days, weeks, or months after diagnosis. Dr. Strassle has spent several years studying the impact of immortal time bias and describing causal inference methods to address immortal time (e.g., clone-censor-weights) for clinical audiences.
Dr. Paula D. Strassle is an Assistant Professor of Epidemiology at the University of Maryland. Her work focuses on identifying neighborhood, healthcare system, and individual barriers to high-quality, appropriate care for racial and/or ethnic minority populations and rural communities. Dr. Strassle specializes in leveraging high-dimensional healthcare data, data linkage, and advanced epidemiologic and statistical methods (including AI/ML) to conduct population health research that can inform clinical practice and reduce health disparities in marginalized populations. She is also interested in exploring the impact of bias in observational research and making complex epidemiologic methods accessible to junior researchers and clinical audiences.
Megan (Winner) Trangsrud, MPH is an Epidemiology PhD student at the University of Maryland. She is leading several projects on patterns of preventive care utilization and multilevel barriers to preventive care using the PREV-CARE survey. Her dissertation focuses on disparities in influenza hospitalization outcomes (inpatient mortality, 30-day readmission) using the HCUP State Inpatient Databases.
Hamoud M. Alotaibi, MPH, MCHES® is a Behavioral and Community Health PhD student at the University of Maryland. He is using the PREV-CARE survey to assess the impact of health literacy and neighborhood socioeconomic status on adherence to preventive care for his dissertation.
Masters and Undergraduate Students
- Hawa Bangoura - second year Epidemiology MPH student
- Katherine Campbell McLinden – first year Epidemiology MPH student
- Shika Marur – first year Epidemiology MPH student
- Sophia Thompson - second year Epidemiology BS+MPH student
- Madiha Rehman – third year Public Policy and Global Health (dual degree) undergraduate student
*Student or trainee
Green AL*, Le R*, Rodriquez EJ, Nápoles AM, Pérez-Stable EJ, Strassle PD. Patient-clinician gender, race, and ethnicity concordance and adherence to preventive services guidelines. J Gen Intern Med. Online ahead of print. Article link
Wilkerson MJ*, Green AL*, Forde AT, Ponce SA*, Stewart AL, Nápoles AM, Strassle PD. COVID-related discrimination and health care access among a nationally representative, diverse sample of US adults. J Racial Ethn Health Disparities. 2026; 13(1):475-486. Article link
Strassle PD, Minc SD, Kalbaugh CA, Donneyong MM, Ko JS*, McGinigle KL. Disaggregating health differences and health disparities with machine learning and observed-to-expected ratios: Application to major lower limb amputation. Epidemiol. 2025; 36(6):841-848. Article link
Gaber CE, Ghazarian AA, Strassle PD, Ribeiro TB, Salas M, Maringe C, Garcia-Albeniz X, Wyss R, Du W, Lund JL. De-mystifying the clone-censor-weight method for causal research using observational data: A primer for cancer researchers. Cancer Med. 2024; 13(23):e70461. Article link
October 2025. Spotlight on AI: UMD researchers work to improve public health. https://today.umd.edu/spotlight-on-ai-umd-researchers-work-to-improve-public-health
July 2025. UMD-led study: Unconscious bias contributes to disproportionate amputations for minority populations. https://today.umd.edu/umd-led-study-unconscious-bias-contributes-to-disproportionate-amputations-for-people-from-minority-populations
Graduate and undergraduate students interested in studying disparities in healthcare utilization and outcomes are encouraged to join the DECODE lab. We offer both volunteer and credit-based research assistantship opportunities. Opportunities to conduct your thesis or dissertation work within the DECODE lab also exist. To express your interest, please reach out to Dr. Strassle (pdstrass@umd.edu) directly.
Prospective graduate students interested in collaborating with the DECODE lab and Dr. Strassle are also encouraged to reach out.