Echo Lyn Warner
· Assistant ProfessorVerifiedUniversity of Utah · Hematology & Oncology
Active 1993–2026
About
Dr. Echo Lyn Warner is a professor at the Spencer Fox Eccles School of Medicine, primarily in the Department of Pediatrics and the Division of Hematology/Oncology. As a first-generation college student from a frontier rural community, she is passionate about public health research aimed at improving wellbeing among underrepresented communities. Her research program in cancer health disparities focuses on leveraging social technology, digital health, and artificial intelligence to enhance psychosocial and health outcomes for cancer patients and caregivers, with particular emphasis on adolescents and young adults. Her current work promotes equity in online cancer communication and information seeking through community engagement and patient navigation interventions that utilize social media. Dr. Warner studied young adult cancer caregivers' use of social media for social support during her doctoral studies at the University of Utah, supported by national fellowships including the NCI Ruth L. Kirchstein National Research Service Award and the Jonas Philanthropies Jonas Scholar Fellowship. She completed a postdoctoral fellowship at the University of Arizona Cancer Center on an NCI-funded T32 program, focusing on the interpretation and spread of cancer and vaccine misinformation on social media, especially among racial and ethnic minorities. Since rejoining the University of Utah as an Assistant Professor and Associate member of the Huntsman Cancer Institute, she has been continuously funded by the National Cancer Institute to study cancer health outcomes. Her research encompasses a range of social media data collection methods and analytic procedures developed over more than a decade of experience. Dr. Warner is also Co-Director of the DELPHI Data Science Initiative at the University of Utah.
Research topics
- Internal medicine
- Medicine
- Computer Science
- Natural Language Processing
- Family medicine
- Gerontology
- Artificial Intelligence
- Immunology
- Environmental health
- Virology
- Pediatrics
- Psychiatry
- Medical physics
- Demography
Selected publications
Journal of Adolescent and Young Adult Oncology · 2026-03-30
articleWe explored how sociodemographics and health insurance literacy (HIL) are associated with financial toxicity (FT) among adolescent and young adult (AYA) cancer patients using linear regression. Participants ( N = 86; ages: 18–39) diagnosed with cancer in the past year completed baseline surveys of an intervention trial assessing HIL (range: 4–36, lower scores = worse literacy) and the COmprehensive Score for financial Toxicity (COST; range: 0–44, lower scores = higher toxicity). Policyholders had higher FT (β = −5.24, p = 0.02). Among AYAs aged 26–39, each one-point HIL increase correlated with a 0.55 increase in COST score ( p = 0.05). Improving HIL could address FT among AYAs aged 26–39, suggesting a potential target for intervention.
Psycho-Oncology · 2026-01-01
articleOpen accessBACKGROUND: Lesbian, gay, bisexual, transgender, queer, and other sexual and gender minority (LGBTQ+) individuals in the US face significant health inequities driven by structural stigma. However, the health impacts of LGBTQ+ stigma and hypervigilance among LGBTQ+ individuals who are also a cancer caregiver are understudied. METHODS: 3 years prior) were recruited for an online survey. Survey domains/measures included sociodemographics, State Equality Index, Sexual Stigma Scale (adapted), Riggle Hypervigilance Measure, as well as health outcomes (PROMIS anxiety, depression, sleep disturbance, and fatigue). Latent class analysis (LCA) was used to group participants based on LGBTQ+ stigma and hypervigilance scores. Multivariable linear regression models tested the association between LCA classes and each PROMIS measure. RESULTS: Among 332 LGBTQ+ cancer caregivers, LCA identified four distinct classes: 1. Low stigma and hypervigilance (31.0%), 2. Moderate stigma and behavioral hypervigilance (21.7%), 3. Moderate contextual hypervigilance and scanning (19.3%), and 4. High stigma and hypervigilance (28.0%). In multivariable linear regression models, class 4 was associated with a 10.43-unit higher anxiety T-score (95% CI: 7.75-13.11), a 9.61 unit higher depression T-score (95% CI: 6.67-12.54), a 9.50-unit higher sleep disturbance T-score (95% CI: 7.02-11.98), and a 12.11-unit higher fatigue T-score (95% CI: 8.88-15.35) in comparison to class 1. Similar, but lower magnitude, trends were seen across the associations of classes 2 and 3. CONCLUSIONS: The findings of this study algin with the LGBTQ+ literature and suggest that among LGBTQ+ cancer caregivers, increased levels of LGBTQ+ stigma and the resulting hypervigilance drive health inequities.
2026-03-11
articleOpen access<sec> <title>BACKGROUND</title> Human papillomavirus (HPV) is a leading cause of preventable cancers, yet HPV vaccination rates remain well below national targets, particularly in rural and highly religious communities. Social media plays a dual role as both a platform for health education and a conduit for vaccine misinformation. In Utah, where internet use is high across geographic and religious lines, online discourse from news outlets may shape public perceptions of HPV vaccination. However, little is known about how the linguistic framing of HPV vaccine information varies across rural and urban, or more and less religious, online news ecosystems. </sec> <sec> <title>OBJECTIVE</title> This study aimed to evaluate how HPV vaccine information is framed in online news outlet posts on social media in Utah, comparing linguistic patterns across rural versus urban areas and communities with higher versus lower religiosity. </sec> <sec> <title>METHODS</title> We collected 851 Facebook posts related to HPV and HPV vaccination from 23 Utah-based news outlets published between March 2012 and March 2022. After removing duplicates (n=531 unique posts) and screening for relevance, 36 posts specifically addressing HPV or HPV vaccination were retained for analysis. Posts were coded by rurality (Rural-Urban Commuting Area codes) and religiosity (county-level religious affiliation prevalence) of the originating news outlet. Natural language processing was applied using Linguistic Inquiry and Word Count (LIWC-22) software to estimate emotional valence, including positive emotion, negative emotion, and overall emotionality. Narrative arc analysis was conducted to characterize the structural and psychological patterns of online discourse. Summary statistics and independent t tests compared linguistic features across geographic and religious contexts. </sec> <sec> <title>RESULTS</title> Only 5 of Utah’s 29 counties had news outlets that posted about HPV vaccination over the study period. Urban outlets produced more posts than rural ones (69.7% vs 30.3%), and nearly 70% of posts came from less religious areas. Posts from urban and less religious areas used significantly more positive language (mean 2.04 vs 1.28 words; P=.003), while posts from rural and more religious areas used more emotional language (mean 1.17 vs 0.63 words; P<.001) and more negative emotional language (mean 0.73 vs 0.42 words; P=.02). Threat-based language was significantly more prevalent in rural posts (P=.03), and rural posts contained significantly more references to death (P<.01) and religion (P<.01). Narrative arc analysis revealed that online HPV vaccine discourse followed a consistent structural pattern with descriptive, information-rich comment sections, fluctuating cognitive tension, and a conclusive, nonprogressive ending. </sec> <sec> <title>CONCLUSIONS</title> In Utah, rural and religious communities are exposed to HPV vaccine information that is less positive and more emotionally charged than that in urban and less religious communities. These linguistic patterns may reinforce vaccine hesitancy and contribute to persistently low HPV vaccination rates in underserved areas. Tailored public health messaging that accounts for the emotional and cultural dimensions of vaccine discourse in rural and religious communities is needed to improve HPV vaccine equity. </sec>
Association of Multi‐Level LGBTQ+ Stigma and Hypervigilance With Health Outcomes Among LGBTQ+ Cancer Caregivers
Open MIND · 2026-01-01
articleBACKGROUND: Lesbian, gay, bisexual, transgender, queer, and other sexual and gender minority (LGBTQ+) individuals in the US face significant health inequities driven by structural stigma. However, the health impacts of LGBTQ+ stigma and hypervigilance among LGBTQ+ individuals who are also a cancer caregiver are understudied. METHODS: LGBTQ+ adults in the US who provided unpaid care to a cancer patient ( ≤ $\mathit{\le }$ 3 years prior) were recruited for an online survey. Survey domains/measures included sociodemographics, State Equality Index, Sexual Stigma Scale (adapted), Riggle Hypervigilance Measure, as well as health outcomes (PROMIS anxiety, depression, sleep disturbance, and fatigue). Latent class analysis (LCA) was used to group participants based on LGBTQ+ stigma and hypervigilance scores. Multivariable linear regression models tested the association between LCA classes and each PROMIS measure. RESULTS: Among 332 LGBTQ+ cancer caregivers, LCA identified four distinct classes: 1. Low stigma and hypervigilance (31.0%), 2. Moderate stigma and behavioral hypervigilance (21.7%), 3. Moderate contextual hypervigilance and scanning (19.3%), and 4. High stigma and hypervigilance (28.0%). In multivariable linear regression models, class 4 was associated with a 10.43-unit higher anxiety T-score (95% CI: 7.75-13.11), a 9.61 unit higher depression T-score (95% CI: 6.67-12.54), a 9.50-unit higher sleep disturbance T-score (95% CI: 7.02-11.98), and a 12.11-unit higher fatigue T-score (95% CI: 8.88-15.35) in comparison to class 1. Similar, but lower magnitude, trends were seen across the associations of classes 2 and 3. CONCLUSIONS: The findings of this study algin with the LGBTQ+ literature and suggest that among LGBTQ+ cancer caregivers, increased levels of LGBTQ+ stigma and the resulting hypervigilance drive health inequities.
JCO Oncology Practice · 2026-01-12
articleOpen accessPURPOSE: The use of artificial intelligence (AI) or automation in financial hardship (FH) interventions has the potential to increase reach and address implementation challenges. However, cancer survivor perceptions of how AI or automation could be used in FH interventions are understudied. METHODS: Eligibility for an online crowdsourcing study included being ≥18 years of age, a cancer survivor, and living in the United States. The survey asked open-ended crowdsourcing questions about how survivors thought AI/automation could have improved their experience with financial assistance and health insurance. A qualitative content analysis was conducted that consisted of two cycles of coding including the use of ChatGPT-5 to generate the preliminary codebook. RESULTS: A total of N = 198 cancer survivors participated and were on average age 50.3 years (standard deviation [SD], 14.3) and age 40.1 years at diagnosis (SD, 16.2), most commonly non-Hispanic/Latine (89.9%), White (86.9%), cisgender women (69.2%), and heterosexual (70.2%). Qualitative analysis revealed seven subcategories within the financial assistance category: (1) efficient search and personalized resource matching, (2) application support and process navigation, (3) insurance and billing support, (4) conversational and interactive tools, (5) connecting to human support, (6) emotional support, and (7) concerns and lack of applicability. Furthermore, five subcategories were identified within the health insurance support category: (1) health insurance education tools and decision support, (2) health insurance navigation, (3) system simplification or automation, (4) connecting to resources and human support, and (5) concerns and lack of applicability. CONCLUSION: Overall, cancer survivors generated a variety of ideas focused on reducing the administrative burden of seeking out financial assistance and dealing with health insurance. Findings demonstrate that cancer survivors could imagine AI or automation being used in FH interventions.
UNC Libraries · 2026-01-29
articleOpen accessBACKGROUND: Lesbian, gay, bisexual, transgender, queer, and other sexual and gender minority (LGBTQ+) individuals in the US face significant health inequities driven by structural stigma. However, the health impacts of LGBTQ+ stigma and hypervigilance among LGBTQ+ individuals who are also a cancer caregiver are understudied. METHODS: LGBTQ+ adults in the US who provided unpaid care to a cancer patient ( ≤ $\mathit{\le }$ 3 years prior) were recruited for an online survey. Survey domains/measures included sociodemographics, State Equality Index, Sexual Stigma Scale (adapted), Riggle Hypervigilance Measure, as well as health outcomes (PROMIS anxiety, depression, sleep disturbance, and fatigue). Latent class analysis (LCA) was used to group participants based on LGBTQ+ stigma and hypervigilance scores. Multivariable linear regression models tested the association between LCA classes and each PROMIS measure. RESULTS: Among 332 LGBTQ+ cancer caregivers, LCA identified four distinct classes: 1. Low stigma and hypervigilance (31.0%), 2. Moderate stigma and behavioral hypervigilance (21.7%), 3. Moderate contextual hypervigilance and scanning (19.3%), and 4. High stigma and hypervigilance (28.0%). In multivariable linear regression models, class 4 was associated with a 10.43-unit higher anxiety T-score (95% CI: 7.75-13.11), a 9.61 unit higher depression T-score (95% CI: 6.67-12.54), a 9.50-unit higher sleep disturbance T-score (95% CI: 7.02-11.98), and a 12.11-unit higher fatigue T-score (95% CI: 8.88-15.35) in comparison to class 1. Similar, but lower magnitude, trends were seen across the associations of classes 2 and 3. CONCLUSIONS: The findings of this study algin with the LGBTQ+ literature and suggest that among LGBTQ+ cancer caregivers, increased levels of LGBTQ+ stigma and the resulting hypervigilance drive health inequities.
Accuracy and inclusiveness of health insurance information in generative artificial intelligence.
Journal of Clinical Oncology · 2025-05-28
articlee13723 Background: GenerativeAI (genAI) has the potential to revolutionize how cancer patients seek information. As more cancer survivors look to genAI for guidance on health insurance topics, it is critical to evaluate the accuracy and inclusiveness of genAI generated responses to questions about health insurance. Methods: Prompts were constructed based on the content of an NCI-funded health insurance literacy patient navigation intervention (CHAT-S). In CHAT-S, participants meet with a navigator over four 30-minute sessions to cover information on health insurance terms/processes, specifics of their health insurance, healthcare laws, information about appeals, tips for budgeting, and financial resources. Topics from these sessions were converted into 13 unique AI prompts (e.g., what is preauthorization, how do I file an appeal, etc.). Each prompt was put into Microsoft Copilot. A codebook was applied to the genAI responses to systematically evaluate accuracy and inclusiveness on a scale of 0-2 (0 = no meaningful difference, 1 = appropriate slight difference, 2 = meaningful difference). To assess accuracy, genAI content was compared against the CHAT-S intervention content and evaluated for incorrect information, lack of assertive language, and missing context. To assess inclusiveness, content was evaluated for dehumanizing language and Flesch reading ease. Results: Across all 13 genAI responses, context was consistently lacking across all genAI responses (Mean: 2, Standard Deviation: 0). While every response included appropriate information, there was important information included in the booklet that was absent in the genAI response. For example, when defining a deductible, genAI did not include that it resets annually. Overall, the main content presented in CHAT-S and the genAI responses were consistent with each other (1.2, 0.44) When content differed, it was in specificity. For example, the booklet provided names and contact information for financial resources, whereas genAI linked to resource databases where a survivor could find additional resources independently. Assertiveness was appropriate across all genAI responses (1.2, 0.44). The language used by genAI was inclusive in sentiment; however, not in reading level. The average Flesch reading ease across responses was more challenging than recommended for health educational materials (Mean 11 th grade; recommended 6 th grade). Notably, genAI responses did not produce any inaccurate information. Conclusions: GenAI has potential to guide and inform cancer survivors on health insurance topics. In this explorative study, genAI responses included helpful steps to guide patients in understanding and using their health insurance. While we used broad prompts about insurance, future studies should evaluate the ability of genAI to generate tailored recommendations based on individuals’ specific scenarios.
Journal of Adolescent and Young Adult Oncology · 2025-02-10
articleOpen accessWe investigated insurance coverage among adolescents and young adults (AYA) with cancer before and during the COVID-19 pandemic. AYAs diagnosed with cancer 15–39 years of age were identified using Utah Cancer Registry records and linked with University of Utah electronic health records. Poisson models calculated incidence rate ratios (IRRs) of health insurance coverage during pre-pandemic (11/4/2017–3/5/2020; n = 2,140) and pandemic (3/6/2020–7/6/2022; n = 1,894) periods. Prior to the pandemic, insurance gaps were higher (pre-pandemic = 16.40%, pandemic = 13.73%; IRR = 0.84, 95%CI = 0.71–0.98); more AYAs had continuous public insurance during the pandemic (pre-pandemic = 8.60%, pandemic = 10.98%; IRR = 1.28, 95%CI = 1.05–1.56). Research is needed on the durability of pandemic relief programs on insurance coverage among AYA cancer survivors.
Supportive Care in Cancer · 2025-06-14 · 1 citations
articleOpen accessSenior authorUNC Libraries · 2025-03-21
articleOpen accessBACKGROUND: Cancer survivors frequently experience cancer-related financial burdens. The extent to which Lesbian, Gay, Bisexual, Transgender, Queer, Plus (LGBTQ+) populations experience cancer-related cost-coping behaviors such as crowdfunding is largely unknown, owing to a lack of sexual orientation and gender identity data collection and social stigma. Web-scraping has previously been used to evaluate inequities in online crowdfunding, but these methods alone do not adequately engage populations facing inequities. OBJECTIVE: We describe the methodological process of integrating technology-based and community-engaged methods to explore the financial burden of cancer among LGBTQ+ individuals via online crowdfunding. METHODS: To center the LGBTQ+ community, we followed community engagement guidelines by forming a study advisory board (SAB) of LGBTQ+ cancer survivors, caregivers, and professionals who were involved in every step of the research. SAB member engagement was tracked through quarterly SAB meeting attendance and an engagement survey. We then used web-scraping methods to extract a data set of online crowdfunding campaigns. The study team followed an integrated technology-based and community-engaged process to develop and refine term dictionaries for analyses. Term dictionaries were developed and refined in order to identify crowdfunding campaigns that were cancer- and LGBTQ+-related. RESULTS: Advisory board engagement was high according to metrics of meeting attendance, meeting participation, and anonymous board feedback. In collaboration with the SAB, the term dictionaries were iteratively edited and refined. The LGBTQ+ term dictionary was developed by the study team, while the cancer term dictionary was refined from an existing dictionary. The advisory board and analytic team members manually coded against the term dictionary and performed quality checks until high confidence in correct classification was achieved using pairwise agreement. Through each phase of manual coding and quality checks, the advisory board identified more misclassified campaigns than the analytic team alone. When refining the LGBTQ+ term dictionary, the analytic team identified 11.8% misclassification while the SAB identified 20.7% misclassification. Once each term dictionary was finalized, the LGBTQ+ term dictionary resulted in a 95% pairwise agreement, while the cancer term dictionary resulted in an 89.2% pairwise agreement. CONCLUSIONS: The classification tools developed by integrating community-engaged and technology-based methods were more accurate because of the equity-based approach of centering LGBTQ+ voices and their lived experiences. This exemplar suggests integrating community-engaged and technology-based methods to study inequities is highly feasible and has applications beyond LGBTQ+ financial burden research.
Recent grants
Frequent coauthors
- 256 shared
Anne C. Kirchhoff
- 119 shared
Deanna Kepka
- 89 shared
Austin R. Waters
University of Utah
- 71 shared
Douglas Fair
University of Utah
- 70 shared
Karen Kuhlthau
Walsh University
- 69 shared
Heydon K. Kaddas
- 68 shared
Kevin C. Oeffinger
- 68 shared
Karen Donelan
Brandeis University
Education
Ph.D., Public Health
University of Utah
Other, Public Health
University of Utah
B.A., Biology
University of Utah
Awards & honors
- National Cancer Institute's Ruth L. Kirchstein National Rese…
- Jonas Philanthropies Jonas Scholar Fellowship
- two-year fellowship from the Jonas Center for Nursing Excell…
- NCI-funded T32 in Cancer Prevention and Control Health Dispa…
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Echo Lyn Warner
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup