Population Health

November 7, 2023

Initiative awards 12 early-stage pilot grants to interdisciplinary teams of UW investigators

Research project team engaged in discussionThe Population Health Initiative announced today the award of 12 Tier 1 pilot grants to teams of interdisciplinary researchers representing eight University of Washington schools and colleges as well as several community-based partners. These awards are worth approximately $400,000, which includes school, department and unit matching funds, with the projects seeking to address grand challenges in areas such as mental health, housing affordability, maternal health and HIV/AIDS.

“The range of innovative project ideas we are funding in this round of grants is inspiring,” shared Ali H. Mokdad, the UW’s chief strategy officer for population health and professor of health metrics sciences. “We are also delighted that five of the funded projects are led by PhD students or postdoctoral scholars.”

The Population Health Initiative seeks to support researchers in laying an interdisciplinary foundation for a future project to generate proof-of-concept with its Tier 1 pilot grants. The 12 funded projects for this cycle are:

Housing affordability and chronic stress in the US: Does affordability modify the effect of neighborhoods on health?

Amy J. Youngbloom, Department of Epidemiology
Stephen J. Mooney, Department of Epidemiology
Anjum Hajat, Department of Epidemiology
Isaac Rhew, Department of Psychiatry & Behavioral Sciences
Rebecca Walter, Runstad Department of Real Estate

Project abstract
In 2019, an estimated 7.8 million very low-income households (i.e., households earning less than 50% of the area median income) in the United States paid 50% or more of their income towards housing. Despite the growing breadth of the housing affordability crisis, few US-based researchers have examined the potential effects of housing cost burden on health and whether existing housing assistance programs can reduce the negative health effects of housing cost burden. Furthermore, while public health and urban planning researchers have produced ample evidence supporting the role of built environment factors in promoting health, no US-based research has yet sought to disentangle the role of housing affordability in understanding how built environment factors contribute to population health research.

The research team for this project includes faculty from the UW’s College of Built Environment’s Real Estate Department, Epidemiology, and Psychiatry & Behavioral Sciences. This cross-disciplinary team will leverage a recently linked dataset, that combines nationally representative health data, including detailed biomarker data (NHANES) and the US Department of Housing and Urban Development (HUD)’s housing assistance database. This study uses quasiexperimental methods to explore the association between housing affordability and chronic stress, and the role that national housing assistance programs may play in reducing chronic stress. Additionally, we will test whether housing cost burden and housing assistance are modifiers of the association between built environment measures and chronic stress. This research will provide preliminary data to support future research efforts to further examine this relationship.

Increasing Monolingual Spanish Speakers’ Participation in HIV/AIDS Clinical Trials: The Language Access and Justice Initiative

Alison Cardinal, School of Interdisciplinary Arts and Sciences (UW Tacoma)
Michele Andrasik, Fred Hutch Cancer Center
Pedro Goincochea, Fred Hutch Cancer Center
Rafael Gonzalez, Fred Hutch Cancer Center

Project abstract
The aim of this initial needs assessment is to learn about the barriers, limitations, and opportunities for the enrollment of Hispanic and Latinx (from now on Latinx) individuals with limited English proficiency (LEP) in HIV/AIDS clinical research conducted by the HIV/AIDS clinical research networks funded by NIAID. The inclusion of racial and ethnic minorities is required in HIV/AIDS clinical research to assure generalizability and applicability of findings to the populations most in need of the interventions.

However, participation of Latinx participants lag, making up less than 12% of study participants when they account for nearly 30% of new HIV infections in the US. Through interviews and surveys with staff and clinical researchers from the 100 clinical research sites around the country conducting research on HIV/AIDS prevention, therapeutics, and vaccines, we hope to show proof of concept of the impact of language barriers on the participation of Latinx populations in clinical research.

We propose to conduct formative research to: 1) Understand the challenges that staff and clinical researchers face in enrolling LEP Latinx participants in HIV/AIDS clinical research. 2) Investigate how clinical research sites use and employ language access practices and plans, such as translation and interpretation, to increase enrollment and retention of LEP Latinx people. 3) Evaluate the effectiveness of language access practices currently used in education, recruitment, and retention of LEP Latinx research participants in HIV/AIDS clinical research.

Differentiated service delivery for HIV: tailoring community ART delivery for people living with HIV in refugee settlements in Uganda

Robin E. Klabbers, Departments of Global Health and Emergency Medicine
Kelli N. O’Laughlin, Departments of Emergency Medicine and Global Health
Jacob Oluma, Medical Teams International
Timothy Muwonge, Makerere University

Project abstract
Linkage to care and initiation of anti-retroviral therapy (ART) is essential for successful HIV treatment, yet many people living with HIV in refugee settlements in Uganda do not engage in clinical care because of barriers such as long distances to clinic, transportation costs, and lack of social support. Community ART delivery is a differentiated care strategy designed to overcome these obstacles by moving HIV treatment out of health facilities and into the community.

Little is known regarding what optimal community ART delivery looks like for people living with HIV in refugee settlements. The proposed pilot research will leverage the infrastructure of the Head StART study, a cluster-randomized trial evaluating the effectiveness of community ART delivery for people newly diagnosed with HIV at 12 health centers in five refugee settlements in Uganda. A multilingual research assistant will conduct individual interviews with people living with HIV (N~20) and HIV peer-navigators (N~12), and will hold focus group discussions with health workers (N~32) at Head StART health centers to identify key barriers to community ART delivery engagement in this setting and characterize perspectives on the optimal design (preferences and feasibility) of community ART delivery to overcome these barriers.

Transcripts will be analyzed using rapid qualitative methods and a list will be generated of community ART delivery characteristics to inform the attributes and levels of a discrete choice experiment quantifying preferences for community ART delivery models among people living with HIV in refugee settlements in Uganda.

Developing and Disseminating Cold Injury Prevention Education

Barclay T. Stewart, Department of Surgery
Colleen Chalmers, King County Regional Homelessness Authority
Tony Machacha, King County Regional Homelessness Authority
Caitlin Orton, Department of Surgery

Project abstract
We are in the midst of a homelessness crisis. People living outside are at high-risk of cold injuries (e.g., frostbite, severe hypothermia) during winter months. Cold injuries are limb- and life-threatening and can compound the challenges of living with other physical and mental disabilities. UW Medicine Regional Burn Center (UWMRBC) has cared for growing numbers of people with cold injuries over the past few years related to increasing homelessness, insufficient mental and public health services, and rising costs of living. Nearly all cold-related injuries and deaths are preventable through education, harm reduction strategies, and housing. Homelessness advocates have called for a more robust cold injury response plan. The current King County Cold Response Plan addresses many issues (e.g., warming shelters, communication, transportation) but lacks sufficient, consumer-tested education materials for people experiencing homelessness (PEH).

We aim to fill the critical education gap by 1) developing a King County Cold Injury Prevention Toolkit (e.g., prevention education material, dissemination strategies, practical implementation tips), 2) disseminating the Toolkit with community-based strategies, and 3) establishing a partnership between UWMRBC and King County Regional Homelessness Authority (KCRHA). This project addresses two pillars of UW’s Population Health Initiative: human health and social and economic equity.

Measures of success for this project include the 1) development of a toolkit that includes PEH and community stakeholder feedback, 2) evaluation of dissemination feasibility and 3) building on the foundational partnership between UWMRBC and KCRHA to expand the project’s scope and reach across King County, Washington State and our country.

Developing a digital mental health intervention for Arab refugee adolescents and young adults in Jordan

Anna Larsen, Department of Psychiatry & Behavioral Sciences
Dror Ben-Zeev, Department of Psychiatry & Behavioral Sciences
Ahmad Y. Bawaneh, International Medical Corps – Jordan

Project abstract
More than 108.4 million people have been forcibly displaced due to life-threatening persecution, with over 76% of the world’s refugee population living in low-middle-income countries (LMICs). Refugees have significant unmet mental health needs, particularly among adolescents and young adults (AYAs; 15-29 yrs), who comprise more than half of the refugee population.

As the rates of forced migration continue to rise, governmental and non-profit entities have placed significant focus on developing scalable solutions for refugees with unmet mental health needs. These efforts are particularly salient in Jordan, which hosts the second-largest number of refugees per capita worldwide due to a continuous flow of displaced people into their country over several decades. Digital psychological interventions delivered using mobile devices (mHealth for Mental Health) have the potential to significantly expand access to low-cost and culturally tailored treatment for vulnerable refugee AYAs who otherwise would struggle to access care.

This project will address the urgent mental health disparities experienced among Arab refugee AYAs living in Jordan by developing a mHealth mental health intervention using a community-participatory approach. Through active collaboration with a large community-based NGO in Jordan, this project aims to lay the foundation for developing a mHealth intervention prototype that is highly acceptable and feasible to the context of the lived experiences of forcibly displaced AYAs living in an LMIC setting. We propose conducting a user-centered mixed-methods needs assessment using a community-participatory approach to guide our proof-of-concept and future grant applications for our novel mHealth intervention for Arab refugee AYAs living in Jordan.

Characterization of urinary glyphosate levels in the US population by environmental and social factors

Lianne Sheppard, Departments of Environmental & Occupational Health Sciences and Biostatistics
Anjum Hajat, Department of Epidemiology
Magali Blanco, Department of Environmental & Occupational Health Sciences
Amanda Gassett, Department of Environmental & Occupational Health Sciences
Cynthia Curl, Boise State University
Carly Hyland, University of California Berkeley
Edward Kasner, Department of Environmental & Occupational Health Sciences
Anne Reiderer, Department of Environmental & Occupational Health Sciences

Project abstract
Glyphosate (the active ingredient in RoundUp®) is the most widely used herbicide worldwide and has been recently classified as a probable human carcinogen by the International Agency for Research on Cancer. Still, studies have only recently begun to characterize glyphosate levels in large human populations. Research is needed to determine how environmental and social factors impact glyphosate exposures in the general US population.

The recent release of urinary glyphosate measurements from the National Health and Nutrition Examination Surveys (NHANES) make it possible to address these questions. We will use 2013-2018 NHANES data from ~7,000 participants ages 3+ on urinary glyphosate levels, residential urbanicity, diet, socioeconomic status (SES) and race/ethnicity. Moreover, we will link glyphosate data to residential proximity to cultivated agricultural land, an Area Deprivation Index developed from socioeconomic indicators, and regional glyphosate use rates. We will assess whether glyphosate levels are associated with agricultural land proximity, urbanicity, or the consumption of foods with high glyphosate residue levels. We will investigate whether race/ethnicity, SES, or other environmental factors (e.g., regional glyphosate use rates) modify the observed associations.

This work will provide pilot data for our planned future exposure science studies, which will characterize sources, routes, and pathways of exposure among particularly vulnerable populations including pregnant women. It will also provide strongly needed glyphosate characterization in the general population and support future exposure and epidemiologic studies on this widely-used pesticide. This work is well-aligned with two population health pillars: human health and social and economic equity.

Estimating True Demand of Acute Pediatric Mental Health Services in Washington State

Shan Liu, Department of Industrial & Systems Engineering
Lawrence S. Wissow, Seattle Children’s Hospital
Alysha Thompson, Seattle Children’s Hospital

Project abstract
Pediatric mental health represents the most vulnerable patients. Mental disorders such as ADHD, anxiety, depression, and behavior disorders can cause serious harm to children’s and youth’s well-being and lifetime developmental prospects. The pediatric mental health crisis is worsening in the U.S. The CDC estimated that about 4.4% to 9.8% of American children aged 3-17 years was diagnosed with at least one mental disorder in 2016-2019, and the prevalence of functionally impairing symptoms is 3 to 4 times higher in several groups. Access to mental health services is limited and demand is much greater than supply in Washington state (WA).

An ideal continuum of care aims for patients to receive care in least restrictive settings. Few implemented studies help planners decide on the optimal location and size of mental health facilities due to lack of accurate understanding of the unmet demand geographically. We aim to determine the current usage of acute pediatric mental health services, estimate the true demand for these services and project the growth of true demand for the next 40 years in WA using diverse data sources.

Touching on the pillars of human health and social and economic equity, results from this research can build a foundation for an integrative set of systems engineering models to design a statewide pediatric mental health service continuum through capacity building, prioritizing the most effective services to reduce mental health burden and ensuring healthcare providers and state policy makers to have evidence-based tools to design an optimal continuum of care system.

Community-Centered Service Design: Communal Approaches Toward Maternal Health Equity for Black Birthing People

Leslie Coney, Department of Human Centered Design & Engineering
Julie Kientz, Department of Human Centered Design & Engineering
Monica McLemore, Department of Child, Family, and Population Health Nursing
Jazmin Williams, BLKBRY

Project abstract
Black women are 3-4x more likely to die during childbirth, and over 60% of these deaths are preventable. The overmedicalization of Black birth and obstetric racism have negatively impacted Black birthing people for centuries, thus causing a growing interest in out of hospital care and support during the pregnancy and postpartum period for Black birthing people. We are interested in how Black birthing people access and use community maternal health resources (CMHR) during pregnancy and postpartum.

A CMHR can be defined as prenatal and/or postpartum support in the form of a product, service, or knowledge sold, administered, and/or shared by a community member or organization. One particular type of CMHR we would like to focus on is community-based organizations (CBOs) providing perinatal services primarily to Black birthing people. The focus on CBOs is intentional and sensitized to the historical, contemporary and cultural accounts of medical racism and institutional harm toward Black bodies.

We plan to work closely with BLKBRY, a Black owned, CBO providing perinatal services and doula care for Black and Indigenous folks in Washington State. Leveraging our multidisciplinary backgrounds, we will apply service design approaches to co-design the delivery of seamless, phenomenal and culturally-relevant services for Black birthing people. We plan to build the foundation for the implementation and evaluation of innovations co-created through this new partnership.

A Qualitative Examination of Evictions Caused by Default Judgments in Washington State

Will von Geldern, Evans School of Public Policy & Governance
Karin Martin, Evans School of Public Policy & Governance
Rachel Fyall, Evans School of Public Policy & Governance
Will Beck, Tacomaprobono Housing Justice Project

Project abstract
Housing evictions pose a substantial risk to population health and the wellbeing of low-income households. Because of long-standing racial and socioeconomic inequities in American society – and rental housing markets specifically – this risk disproportionately affects low-income communities, female-led households, and people of color. Our pilot study aims to inform policies which could affect one salient aspect of evictions: default judgments. Default judgments occur when evicted tenants do not participate in eviction proceedings and are thus unable to dispute or delay their eviction. When a tenant fails to respond to a civil summons and complaint or misses their court hearing, the court sides with the landlord in response to the defendant’s non-compliance with the process. While default judgment rates are not systematically measured in most places, recent measurements in King County have found that approximately half of all tenants do not appear at their court hearings.

Through semi-structured interviews with approximately 30 tenants who have missed their court hearings, this study aims to elucidate tenants’ perceptions of the circumstances leading up to court absences. Evidence generated by this research could be especially impactful in Washington State, where the state now provides court-appointed legal representation to tenants who do participate in their eviction proceedings. In addition to generating information about the consequential but poorly understood subject of default judgments, this study is intended to lay the groundwork for future research which employs an experimental design to increase tenants’ participation in their eviction proceedings.

Improving Prediction of Psychiatric Outcomes in Youth Using High-Dimensional Genetic and Phenotypic Data

Katherine T. Foster, Departments of Psychology and Global Health
Jennifer Forsyth, Department of Psychology
Matthew P. Conomos, Department of Biostatistics
Daniel J. O. Roche, University of Maryland

Project abstract
The pressing need to identify youth most vulnerable to psychiatric problems before they onset has spurred the search for risk biomarkers to support population-scale testing early in development. The success of precision risk assessment will hinge on identification of risk markers alone or in combination that are accessible, low-cost, and maximally predictive of future psychopathology. Large-scale genetic studies have identified common variants robustly associated with psychiatric outcomes, signaling their utility when integrated with other phenotypic risk markers (e.g., psychological and social behavior). Yet, research infrequently integrates multimodal data to select risk markers differentiating which symptoms will appear when for who, along what trajectory of illness, and following which early warning signs.

The primary aim of this pilot is to incorporate a data collection pipeline for baseline genotyping into ongoing high dimensional clinical assessments of children/families (n=400) – the PEARL(Y) study (PI: Foster). The long-term scientific aims are to optimize clinical assessment prediction by identifying which biopsychosocial risk markers account for the most longitudinal variation in a wide variety of psychopathology outcomes. PEARL(Y) applies accelerated longitudinal cohort design (i.e., initial ages 10-25) featuring >90-day, biannual bursts of smartphone self-report of daily experiences (i.e., ecological momentary assessment, EMA) and digital phenotyping (e.g., passive sensing of physiology, location, etc.). Resultant multimodal data will be analyzed using machine learning techniques (e.g., elastic net regression and other approaches, Foster & Beltz, 2018, 2021) that select which putative risks (e.g., baseline genetic biomarkers vs. daily variation in individual psychosocial phenotypes) with the strongest predictive association with clinical outcomes.

Developing Novel and Fair Machine Learning Strategies for Glioblastoma Segmentation in Sub-Saharan Africa Patient Population

Mehmet Kurt, Department of Mechanical Engineering
Jacob Ruzevick, Department of of Neurological Surgery

Project abstract
Glioblastomas pose a substantial global health challenge with rising incidence and mortality rates. Despite advancements in treatment, the prognosis remains grim, particularly in low and middle-income regions like Sub-Saharan Africa (SSA). Recent research has highlighted the significance of surgical resection, combined with chemotherapy, in extending patient survival. However, accurate tumor segmentation from MRI scans, which is crucial for surgical planning, remains a challenge due to intricate tumor structures.

Machine learning (ML) models have emerged as a promising tool for glioblastoma segmentation and diagnosis, with the Brain Tumor Segmentation (BraTS) Challenge, a challenge to assess state-of-the-art ML methods used for brain tumor segmentation, serving as a key resource. In 2023, the BraTS Challenge introduced BraTSAfrica, focusing on SSA patients with lower-quality MRI scans. Training ML models on this dataset is complicated due to limited samples and poor image quality.

This proposal outlines a novel domain adaptation approach to improve glioma segmentation in BraTS-Africa MRI data. The primary goal is to attain comparable segmentation performance in the challenging BraTS-Africa dataset, addressing the resolution- and quantity-related disparities. The hypothesis is that applying domain adaptation techniques from the main BraTS to BraTS-Africa datasets can significantly enhance tumor segmentation accuracy, contributing to algorithmic fairness in brain tumor segmentation research. This project seeks to bridge the gap in healthcare disparities by leveraging ML to enhance the diagnosis and treatment of glioblastomas in underserved regions, ultimately striving for improved patient outcomes.

Community Codesign to Integrate Low-Barrier, Culturally and Linguistically Responsive Mental Health Care into Community-Based Social Services

Lesley Steinman, Department of Health Systems and Population Health
Najma Mohammed, Neighborhood House

Project abstract
The COVID-19 pandemic brought again to the forefront the devastating impact of mental health conditions like depression, anxiety, stress, and isolation on individual and community health and well-being. Though effective interventions exist, 1 in 10 U.S. adults do not receive effective care; access is worse for linguistically diverse, people of color living in poverty. Barriers to care include stigma from people and providers; a shortage of trained, culturally and
linguistically appropriate workforce; and a fragmented, clinical mental health care system with poor coverage and long waiting lists.

Integrating mental health care into community-based social service organizations (CBOs) may improve access to low barrier, culturally relevant mental health care and address upstream social drivers of mental health like food insecurity, poverty, discrimination, and unsafe housing. Furthermore, addressing mental health within CBOs can better support an essential workforce with high turnover and burnout.

Our project will build on our community academic partnership to co-design mental health programming for CBOs by convening staff and community members to adapt multiple mental health screeners and programs for community-defined evidence. We will focus on the Somali community based on our co-PI CBO’s recent assessment that identified both mental health need and strong community connections. We will use and disseminate a community-engaged, co-design process using anti-racist implementation science to share decision-making power and center community voices and experiences. Learnings will guide a proof-of-concept R34 to deliver and evaluate the mental health intervention on participant, provider, and organizational outcomes with BIPOC and LEP communities with Neighborhood House and other CBOs.

More information about the Population Health Initiative pilot grant program, tiering and upcoming deadlines can be found by visiting our funding page.