Skip to main content

College of
Health Sciences

Department of Population Health Sciences

The Department of Population Health Sciences aims to discover novel ways to prevent disease and promote health, with a focus on the health conditions that most impact Geisinger patients, health plan members, and our communities. 

We conduct innovative research to understand risk factors of these diseases and develop strategies to improve healthcare delivery. Our interdisciplinary faculty study a range of health conditions, collaborating with Geisinger clinicians and community partners and leveraging extensive data resources, including longitudinal electronic health records and insurance claims, geocoded data on the social and physical environment of communities, and genotyped and whole-exome sequencing data.

Core focus areas include:

  • Epidemiology: Examining the extent and distribution of the health conditions that most impact our patients and communities and identifying the factors that put populations at risk for disease
  • Biostatistics: Applying advanced quantitative analysis to observational data to draw valid conclusions from research studies, application of machine learning for the development of predictive models, and linking health and genomic data to understand the genetic architecture of complex diseases
  • Health and behavioral outcomes research: Detecting, preventing and controlling chronic conditions through the study of health-related behaviors, healthcare practices and health-related quality of life
  • Environmental health: Evaluating how the surrounding built, natural and social environment impacts health and enables or constrains the influence of healthcare
  • Health economics: Conducting comparative effectiveness research to identify the healthcare interventions that offer the greatest benefit at the lowest cost

Contact Us

Colored bar.

Active grants

Community engagement


PI: Lisa Bailey-Davis, DEd, RD
Funder: The Degenstein Foundation
Dates: 12/2022 – 12/2024

This project engages community partners in 5 central Pennsylvania counties for an Upstream Planning Initiative to Improve Response to Child Maltreatment.  Community-engaged planning is focused on identifying solutions that will alter the health trajectories for patients and families affected by or at risk for child maltreatment.  
Contact us: or 866-630-0798, option 2

PaTH MOMS: Establishing a PCORnet PaTH Clinical Research Network Maternal Morbidity and Mortality Patient Stakeholder Council

Site PI: Lisa Bailey-Davis, DEd, RD
Funder: PCORI®
Dates: 7/2023 – 6/2025

The project aims to create a Council of patients with maternal health lived experiences, maternal health clinicians and researchers, and maternal health community partners guided by a shared vision to improve maternal outcomes for patients of racial and ethnic minorities and rural residency. The project will support the development of capacity building activities needed to create a Council that can effectively inform the development of an equitable and inclusive PCOR/CER research agenda in maternal health.

Environmental health

Diabetes LEAD (Location, Environmental Attributes, and Disparities) Network

Co-PIs: Annemarie G. Hirsch, PhD, MPH and Brian S. Schwartz, MD, MS
Funding Source: CDC
Project Number: 1U01DP006296
Dates: 09/2017 – 09/2022

The Geisinger-Johns Hopkins Bloomberg School of Public Health Center for Community Environment and Health (CCEH) is a site in the CDC-funded Network, a collaboration dedicated to identifying the contributions of community features and geography on type 2 diabetes risk and control. 

Diabetes prevalence and incidence in the US vary substantially by geography. The LEAD Network aims to guide policy decision-making to reduce the burden of type 2 diabetes across the US. LEAD sites conduct studies at their respective institutions as well as network-wide analyses. 

Using longitudinal electronic health record data and community data, the CCEH is studying the role that community types and community features have in onset of type 2 diabetes and type 2 diabetes control (e.g., blood pressure and kidney function) in central and northeastern Pennsylvania. 

We are evaluating a range of community features, including greenness, blue space, community socioeconomic deprivation, the food environment, urbanicity, and leisure-time physical activity resources. Using questionnaire and salivary cortisol measures we are exploring the various pathways through which community features and community perceptions may impact type 2 diabetes control, including stress, food insecurity, diet, and physical activity. 

In collaboration with the other LEAD sites, the CCEH has developed harmonized approaches to measurement and analysis, allowing for multi-site investigations of modifiable community features, including the food and physical activity environment, that may mediate the association between community socioeconomic deprivation and type 2 diabetes onset.

Understanding the role of community determinants in opioid use disorder and program implementation factors influencing patient adherence to opioid agonist therapy

PI: Melissa N. Poulsen, PhD, MPH
Funder: National Institute on Drug Abuse
Grant number: K01 DA 049903-01A1
Dates: 07/01/2020 – 06/30/2025

Understanding the multilevel factors that influence risk of opioid use disorder (OUD) is critical to informing an effective response to the opioid epidemic. Although the efficacy of medications for treating OUD is proven, long-term retention in treatment remains a challenge. 

Efforts are needed to identify program implementation factors that can be strengthened to improve OUD treatment outcomes. This research seeks to generate evidence to understand novel, understudied risk factors for OUD, with implications for investment in community-level prevention interventions, and how health systems can improve treatment retention among individuals with OUD. 

The aims of this grant are twofold. The first aim centers on the social determinants that engender vulnerability to OUD, evaluating associations of community contextual factors related to socioeconomic, social, and physical conditions with OUD, and the role individual-level genetic risk factors, healthcare factors, and comorbid medical conditions play in these relations. We are utilizing electronic health record and genetic data from Geisinger patients coupled with secondary data characterizing community factors to conduct a nested case-control study. 

The second aim seeks to understand factors at the individual, interpersonal, and organizational levels that influence patient engagement in and retention to medication treatment for OUD. We are conducting a mixed methods study that includes interviews with patients and key informants to identify multilevel barriers and facilitators influencing treatment success followed by a survey of adult patients in Geisinger’s addiction medicine treatment clinics to quantify factors influencing treatment engagement and retention from a patient perspective.

Epidemiology-genetic epidemiology

Genetic Epidemiology of Sleep Apnea and Comorbidities in Biobanks
Site PI: H. Lester Kirchner, PhD (PI: Brian Cade, Brigham and Women’s Hospital)
Funder: NHLBI
Grant Number: R01 HL153805
Dates: 08/16/2021 – 07/31/2026

Sleep apnea (SA) and insomnia are the two most common sleep disorders. Both contribute individually and jointly to the risk of cardiopulmonary, metabolic, and psychiatric diseases. 

Treatments for SA and insomnia remain suboptimal despite their high prevalence. SA and insomnia are thought to be comprised of distinct subtypes, which remain poorly defined and may contribute to differing risks for health outcomes. 

Our goal is to use machine learning to apply precise phenotyping to biobanks to identify the genetic bases of SA and insomnia and discover SA and insomnia subtypes based on genetics and comorbidities in order to reduce phenotype heterogeneity, guide patient stratification and aid in the discovery of more personalized treatments. 

Our approach is to combine health care system biobank data with polysomnography (PSG) to achieve statistical power to discover genetic variants for SA and insomnia-related phenotypes and characterize their associated clinical outcomes and endophenotypes. We will use advanced natural language processing (NLP) methods to substantially improve the accuracy of SA and insomnia phenotyping. Polygenic risk scores derived from our results can be used to quantify sleep disorder risk, even among those without sleep phenotypes. 

Our specific aims are: 1) to construct advanced SA and insomnia phenotyping algorithms across diverse demographic groups and sites; 2) to identify and characterize the genetic associations with SA and insomnia; and 3) to identify and characterize distinct SA and insomnia patient subgroups based on related comorbidity profiles.

Discovery and CRISPR validation of genetic factors associated with antipsychotic-induced weight gain and cardiometabolic risk.
MPI: Anne Justice, MA, PhD
Grant Number: R01 MH 122686-01
Funder: NIMH
Dates: 02/01/2021 – 12/31/2025

Obesity and related metabolic consequences contribute substantially to accelerated aging and reduce lifespan in psychiatrically ill individuals. These adverse metabolic consequences, particularly antipsychotic-induced weight gain, are thought to be highly genetic, but our understanding of the genetic contributors and mechanisms remains limited. 

The discovery of genetic variation influencing antipsychotic-induced weight gain has the potential to identify the pathways leading to poor outcomes, suggesting avenues for alternative treatments with less metabolic impact. 

Our study aims to identify and functionally validate key genetic determinants of antipsychotic-induced increases in adiposity and cardiometabolic risk across. We are implementing population-based genetics, existing biospecimen with linked clinical data including precisely measured adiposity and insulin sensitivity, and advanced molecular tools to identify and functionally validate key genetic determinants of antipsychotic-induced weight gain and cardiometabolic risk across the age-span. 

This approach leverages 1) existing population-level data from large biobanking initiatives and epidemiological studies with genetic and relevant phenotypic data, 2) existing clinical and biospecimen data from NIH funded randomized clinical trials characterizing the metabolic effects of antipsychotics in children, adults and older adults with direct and precise measures of body fat, and 3) CRISPR based in vitro drug exposure, followed by cellular functional assays to characterize molecular mechanisms impacted by antipsychotic medications. 

This combined set of molecular techniques will allow us to build on known genetic associations, while discovering new genes and genetic variants that are associated with the greatest risk for treatment-related fat gain in psychiatrically ill patients.

Integrative approaches to identifying function and clinical significance of adiposity susceptibility genes
MPI: Anne Justice, MA, PhD
Grant Number: R01 DK 122503
Funder: NIDDK
Dates: 09/22/2020 – 07/31/2025

Large scale genetic studies have laid the foundation for later investigations into disease progression and public health applications, including for obesity. While hundreds of genetic associations with obesity have been identified, further work is needed to narrow in on the causal factors contributing to these associations. 

This study leverages collaborations in the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE), TransOMICs for Precision Medicine (TOPMed) Program, the Genome Sequencing Project (GSP), and the EHR database from the Geisinger MyCODE Community Health Initiative study (MyCode) to narrow in on genes underlying GWAS signals, perform clinical characterization, and conduct in vitro functional studies to characterize the molecular underpinnings and biological mechanisms of obesity-risk loci. 

To accomplish these goals, we are 1) integrating multiple genomic, methylomic, transcriptomic, and metabolomic data resources from the TOPMed Program to identify causal genes/variants underlying known obesity-risk loci, 2) evaluating the clinical significance these genes/variants in the clinical phenome of the MyCode patient population to develop a more comprehensive picture of obesity-related disease etiology, and 3) confirming the functional role of these targets through experiments in relevant human cell lines. 

Our approach will substantially move the field away from tag variants and loci to causal variants, genes, and mechanisms. We anticipate that this work will generate fundamental and important insights into the underlying etiology of obesity and ultimately point the way forward towards prevention and treatment.

Integrated trajectory data and machine learning approaches for precision obesity medicine
PI: Anne Justice, MA, PhD
Grant Number: SAP-4100085740
Funder: Pennsylvania Department of Health
Dates: 06/01/2020 – 05/31/2024

The overall goal of this project is to develop a novel integrated framework including statistical and machine learning approaches for characterizing patterns and subtypes of longitudinal body mass index (BMI) change in adults, identifying its genetic antecedents and its influence on downstream influence on type 2 diabetes (T2D) and related comorbidities. 

The findings from this investigation will improve early detection of disease risk and reveal subgroups of individuals at higher risk for disease spectrum to be targeted for early intervention and monitoring, ultimately moving precision health forward. 

Specifically, this project is 1) identifying subtypes of body mass change by characterizing common mathematical features of BMI trajectories in adults and machine learning approaches for identifying distinct trajectory patterns for BMI; 2) identifying significant associations between BMI change subtypes and genetic variants and T2D related comorbidities; and 3) developing group-specific polygenic risk score models for associated T2D outcomes. 

This project will develop a novel integrated framework with statistical and machine learning approaches for characterizing patterns and subtypes of longitudinal body mass index change, identifying its genetic antecedents and its influence on downstream influence on type 2 diabetes and related comorbidities. The long-term benefits of this project include new improved methods for early detection of high-risk individuals that may benefit from weight loss intervention or weight gain monitoring.

Assessing the Burden of Diabetes by Type in Children, Adolescents, and Young Adults (DiCAYA).
PI: Annemarie G. Hirsch, PhD, MPH
Funding agency: CDC
Project number: U18DP006509-01-00
Dates: 09/2020 – 09/2025

Geisinger is part of the CDC-funded network DiCAYA ( with the goal to advance diabetes surveillance efforts of youth and young adults through the use of large-volume electronic health record data across eight US-based centers. 

Diabetes prevalence and incidence in the US is increasing in youth and young adults and younger onset is associated with greater risk of morbidity and mortality. Traditional survey-based surveillance is limited in its ability to distinguish diabetes types and achieve the scalability needed to generate valid estimates for uncommon diseases such as pediatric diabetes. Moreover, surveys and conventional disease registries are costly and can be inefficient. EHRs provide a novel approach to conducting timely surveillance. 

The DiCAYA network is spread across the US, with half the centers conducting surveillance in youth (0-17 years), and the other half conducting surveillance in young adults (18-45 years), and a coordinating center. Geisinger is one of the sites studying young adults.

The DiCAYA network has three primary aims: 1) develop EHR-based computable phenotype algorithms for the accurate identification of incident and prevalent type 1 and type 2 diabetes in youth and young adults; 2) estimate the incidence and prevalence of diabetes in these age groups between 2018 and 2024, by type, sex, race/ethnicity, and geography; and 3) evaluate the DiCAYA network’s EHR surveillance of diabetes on key surveillance attributes (e.g., data quality and accuracy). 

DiCAYA’s methodology will translate to diverse settings because its centers include a mix of health systems, including a membership-based health system, and geographically-based centers. DiCAYA’s work will inform public health and health services strategies, improve national assessments of the burden of diabetes, and advance nascent EHR-based surveillance methods for other illnesses.

Chronic Rhinosinusitis Integrative Studies Program (CRISP)
Site PI: Annemarie G. Hirsch, PhD, MPH (PI: Robert P. Schleimer, PhD; Project Leader: Brian S. Schwartz, MD, MS)
Funder: National Institute of Allergy and Infectious Diseases (NIAID)
CRISP1: first six years of funding (with no cost extension): U19 AI106683, 2013 to 2019
CRISP2: competitive renewal: P01 AI145818, 2019 to 2024

CRISP1: In CRISP1 we studied a large sample of subjects, representative of the general population in 38 counties in central and northeastern Pennsylvania, across the full spectrum of CRS. We conducted the first U.S. longitudinal study of nasal and sinus symptoms lasting three months, following 7847 subjects over 16 months with six questionnaires assessing symptoms, comorbidities, risk factors, and seasonal exacerbations. 

We then conducted the first study of sinus inflammation by CT scan in a general population, on 646 subjects selected from baseline questionnaire respondents enriched for sinus symptoms. In parallel studies, using electronic health record (EHR) diagnoses, we conducted longitudinal analysis to identify incident diagnoses that predated and occurred after CRS diagnosis. CRISP1 was a critical advance in the epidemiologic understanding of CRS; prior studies were limited to tertiary care, representing only the most severe patients; did not objectively measure sinus inflammation; and/or lacked the longitudinal data needed to assess the temporal relations between CRS and other diseases.

We made several important and novel observations. CRS identified with symptoms was common in the general population at 12% and quite disabling. We found heterogeneity in symptom combinations and stability over time and identified risk factors for these symptom patterns. We identified five symptom clusters in those with chronic sinonasal symptoms, with three predicting workplace absence, lost productivity, and/or sinus opacification. 

Analysis of sinus opacification suggested that current methods of measuring opacification may need to be revised; we used statistical methods to identify localized and diffuse patterns of opacification. There were prominent sex differences in degree and location of sinus opacification and also in symptom patterns. Our EHR-based and longitudinal questionnaire studies both provided preliminary evidence that patients with CRS without nasal polyps are at greater risk of other airway diseases. 

Studies are now needed that can identify the presence of inflammation earlier than sinus CT; using novel approaches to identification of CRS and new phenotypes; with additional longitudinal information; to address risk of long-term outcomes; and to further evaluate key sex differences.

CRISP2: We have four primary specific aims in CRISP2. First, we want to develop new approaches to CRS phenotyping: We hypothesize that patterns of longitudinal symptoms, longitudinal medical history, and CT scan findings can be used to identify new CRS phenotypes that will be associated with 1) natural history; 2) quality of life; 3) health care utilization; and 4) workplace productivity. 

Second, we want to evaluate if CRS endotypes, defined with nasal lavage fluid biomarkers identifying type 1, type 2, type 3, and mixed patterns of inflammation, epithelial-mesenchymal transition, and fibrin deposition, are associated with phenotype, natural history, clinical outcomes, quality of life, utilization, and workplace productivity. 
Third, we will evaluate CRS as antecedent risk for other diseases: we hypothesize that newly identified phenotypes and endotypes will be associated with increased risk for development of asthma and bronchiectasis. 

Finally, we will assess sex differences: we hypothesize that all specific aims will show sex differences. CRISP1 found that different diagnostic standards for CRS by sex are needed and CRS in women may be sub-optimally managed.

Central Pennsylvania Rural Birth Cohort
PI: Lisa Bailey-Davis, DEd, RD
Dates: 2022 – present

This is a collaborative prospective longitudinal pilot study between Geisinger and The Pennsylvania State University (Penn State) to determine what strategies are most successful: 1) in building and retaining a cohort of families from rural communities in Central Pennsylvania with recruitment beginning in pregnancy, infant/toddler age, and preschool age using a cohort sequential design; 2) for collecting Geisinger clinical electronic health record (EHR) and remote biobehavioral measurements (Penn State) to better characterize synergistic factors associated with obesity and substance use in this population; and 3) for identifying points for future intervention, treatment, prevention, and policy efforts to reduce health disparities in maternal-child morbidity and promote positive family processes. 

Child and Adolescent Trend (CAT) Data Registry 
PI: Lisa Bailey-Davis, DEd, RD 
Dates: 2022 – present

The Childhood and Adolescent Trend (CAT) data registry includes a subset of the Geisinger electronic health record (EHR) and collects data on 542,211 Geisinger patients with a pediatric BMI measured with a same day height and weight at age 20 years or younger between 1982 and 2022 (refreshed annually or on-demand).  

In August 2022, the CAT cohort had a median age of 18 years (min 0 years, max 45 years, mean: 18.32 years) with 5.1 million height/weight measurements (median age @ measurement 7years; mean age @ measurement 7.95 years; median of 5 measurements and mean of 9.4 measurements). 

The CAT registry is a collection of EHR data (demographics, vitals, encounters with diagnosis, problem list diagnosis, labs, social history, medication history, and referrals) as well as custom data including adult BMI, CDC and WHO BMI percentiles, BMI Extended, Family Nutrition and Physical Activity (FNPA) and Early Healthy Lifestyles (EHL) behavior and home environment data, food insecurity, and youth Blood Pressure Percentiles. 

Health economics

Surveillance for Outcomes of Genomic Medicine Policies
Site-PI: Jing Hao, PhD, MD, MS, MPH (PI: Chris Lu, Harvard Pilgrim)
Grant number: 1R35HG011285-01
Dates: 09/01/2020-06/30/2025

There are currently no surveillance structures in place to examine outcomes associated with policies for genomic tests. Insurance coverage policies and levels of patient cost-sharing are major factors that influence patient access to new genomic tests by determining who gets tested, screened, and ultimately treated based on genomic information. 

These coverage determinations can have significant impact on the health of our patients and their access to potentially lifesaving medical care. Currently, we know little about the effects of policies for genomic tests despite the increasing implementation of such policies. Monitoring the consequences of policies is an important public health issue. 

This project seeks to advance the field of genomic medicine and society by developing and validating analytical methods for efficient, rigorous evaluation of policies that impact access to genomic tests and associated outcomes to inform healthcare and policy decisions.

PCORnet PaTH Network

A PaTH Toward a Learning Health System
Site PI: H. Lester Kirchner, PhD (PI: Kathleen McTigue, University of Pittsburgh)
Funder: PCORI
Grant Number: CDRN-1306-04912
Dates: 01/01/2022 – 12/31/2024

PCORI funded the development of PCORnet®, the National Patient-Centered Clinical Research Network, to make it easier and more efficient to conduct research. PCORnet is made up of Partner Networks that harness the power of large amounts of health data and patient partnerships. 

Clinical Data Research Networks (CDRNs) are one type of network supported by PCORI. CDRNs consist of two or more healthcare systems, including hospitals, integrated delivery systems, and federally qualified health centers. Each CDRN transforms data gathered from routine patient care across its participating health systems to a consistent format, the Common Data Model (CDM), to enable rapid response to research-related questions.

PCORI funded PaTH Towards a Learning Health System’s (PaTH, participation in PCORnet from 2015 to 2024. Led by the University of Pittsburgh, PaTH comprises seven health systems in the Mid-Atlantic and midwestern regions, including Geisinger. The network uses electronic health records (EHRs) across its partners to conduct research that may ultimately help clinicians and patients make more informed health decisions.

Pragmatic trials

ENCIRCLE Trail: PatiENt-Clinic-Community Integration to PRevent Obesity among Rural PresChooL ChildrEn Trial
PI: Lisa Bailey-Davis, DEd, RD
Funder: PCORI
Grant Number: CER-2019C1-16040
Dates: 10/2018 - 5/2025

This project compares three types of well-child visits for preventing obesity among preschool-aged children in rural areas. Participants are children aged 2-5 years and their families who receive an annual well-child visit at Geisinger. Learn more about the study here.  

Contact us: or 866-207-9289, option 2

Reversing Metabolic Syndrome: Eat, Love, Move: A multi-site Randomized Controlled Trial
PI: Lisa Bailey-Davis, DEd, RD
Funder: William G. McGowan Charitable Fund
Grant Number: CER-2019C1-16040
Dates: 10/2018 - 9/2024

This is a multi-site randomized controlled trial of a lifestyle intervention designed to reverse metabolic syndrome. The project enrolled 600 persons with metabolic syndrome at 5 sites across the US including Geisinger.

Contact us: or 866-630-0798, option 2

Site PI: Lisa Bailey-Davis, DEd, RD
Funder: National Institutes of Health
Grant Number: IUG3HL162971
Dates: 5/2022 – 3/2029

The Early Intervention to Promote Cardiovascular Health of Mothers and Children multicenter trial aims to test the effectiveness of an intervention designed to promote cardiovascular health and address cardiovascular disparities in both mothers and children (0-5 years old).  ENRICH is funded by the National Heart, Lung and Blood Institute in collaboration with several federal partners. Geisinger is part of the Penn State Clinical Center, one of seven clinical centers in the trial. 

Contact us: or 570-214-1424

PI: Lisa Bailey-Davis, DEd, RD
Dates: 2013 – present

The goal of this prospective, longitudinal cohort study is to determine if the pediatric care redesign with evidence-based screening (using the Family Nutrition and Physical Activity, FNPA, tool), patient-centered education, negotiated decision-making and health information technology to reverse childhood obesity clinical improvement initiative is improving the quality of care by examining process and outcome measures associated with parent-child exposure to PREVENT. 

FNPA is collected as a patient-reported outcome measure at annual well-child visits for children aged 2 to 9 years (2013-2019), 2 to 12 years (2020-present) and 2-17 years (beginning 2024).  A study-developed tool, Early Healthy Lifestyles (EHL) is used to collect parenting practices and child eating, sleeping, and activity data from parents/caregivers aged 0-26 months at routine well-child visits (2013-present). Findings from this longitudinal study are used to continuously improve the value of and quality of care for the primary prevention of pediatric obesity and related comorbidities. Our team collaborates with National FNPA Leaders ( to facilitate dissemination and implementation of the tool to clinical sites globally.

Transplant Candidacy for Patients with Stage 4-5 CKD or dialysis-dependent ESKD
PI: Alex Chang, MD, MS
Funder: Novo Nordisk ISS\

As severe obesity is one of the most common barriers to eligibility for kidney transplantation in the U.S., this placebo-controlled, randomized controlled trial examines whether subcutaneous semaglutide can improve kidney transplant candidacy for patients with advanced CKD.

Precision medicine

Title: Developing and Validating EHR-Integrated Readmission Risk Prediction Models for Hospitalized Patients with Diabetes
Site PI: H. Lester Kirchner, PhD (PI: Daniel Rubin, Temple University)
Funder: NIDDK
Grant Number: 5R01DK122073

Dates: 08/21/2020 – 06/30/2025

Hospitalized patients with diabetes are at higher risk of readmission within 30 days than patients without diabetes, and >1 million readmissions occur among diabetes patients in the US annually. Certain interventions can reduce readmission risk, but applying these interventions widely is cost prohibitive. One approach for improving the efficiency of interventions that reduce readmission risk is to target high-risk patients. We previously published a model, the Diabetes Early Readmission Risk Indicator (DERRI), that predicts the risk of all-cause 30-day readmission of patients with diabetes. The DERRI, however, has modest predictive accuracy (C-statistic 0.63- 0.69), and requires manual data input. 

Recently, we demonstrated that adding variables to the DERRI substantially improves predictive accuracy (DERRIplus, C-statistic 0.82). However, using this larger model to predict readmission risk based on manual input of data would be too labor intensive for clinical settings. Indeed, most readmission risk prediction models are limited by the trade-off between accuracy and ease of use; lack of translation to a tool that integrates with clinical workflow; modest accuracy; lack of validation; and dependence on data only available after hospital discharge. 

The objectives of the current proposal are: 1) To develop more accurate all-cause unplanned 30-day readmission risk prediction models using electronic health record (EHR) data of patients with diabetes (eDERRI); 2) To translate the models to an automated, EHR-based tool that predicts % readmission risk of hospitalized patients; and 3) To prospectively validate the eDERRI models and tool. To develop the models, we will leverage data from the PaTH Clinical Data Research Network (CDRN), a multi-center, 40-plus hospital member of the National Patient-Centered Clinical Research Network (PCORnet).

Rational Integration of Polygenic Risk Scores (RIPS)
MPI: Josh Peterson, Vanderbilt University; Jing Hao, Geisinger; David Veenstra, University of Washington
Grant number: 1 R01 HG012262-01
Dates: 03/10/2021 – 02/28/2027

There has been extraordinary growth in new techniques to predict common, complex disease based on polygenic risk scores (PRS). Without an understanding grounded in evidence, it is unlikely that the clinical use of PRS will propagate from highly specialized applications and environments to become adopted more broadly and provide greater benefit to the US population. 

Critical challenges include: 1) understanding the impact of clinical PRS for multiple diseases on long-term patient outcomes, 2) identifying risk thresholds for return of results that optimize patient outcomes and provide cost-effective care, 3) understanding how PRS performance across diverse populations may affect existing disparities and subsequent patient outcomes. 

We propose to address these challenges using decision analytic modeling and by building on our extensive work in this area to create a novel framework capable of assessing PRSs in the context of monogenic and clinical risks. We have already created clinical-economic models to project lifetime clinical impact and cost-effectiveness for population-level genomic screening with return of monogenic disease risks associated with three CDC Tier 1 conditions: hereditary breast and ovarian cancer, Lynch syndrome, and familial hyperlipidemia. 

As part of this proposal, titled Rational Integration of Polygenic Risk Scores (RIPS), we will create models to assess the clinical outcomes and economic value of population screening using PRS in real world settings and applied to large and diverse populations. 

The Aims of the proposal include 1) to evaluate published and real-world evidence on the clinical value of adding PRS to inform comprehensive genomic risk assessment; 2) to understand the impact of PRS performance and return risk thresholds on incremental clinical benefit and cost effectiveness for breast cancer, atherosclerotic cardiovascular disease, and colorectal cancer, and 3) to develop research priorities for the equitable development and implementation of PRS across underserved and underrepresented populations.


Lisa Bailey-Davis, DEd, RD

Lisa Bailey Davis
Associate Professor
Associate Director, Center for Obesity and Metabolic Research  
Women’s health; child and adolescent health; nutrition; obesity; cardiometabolic disease; substance use; integrated care; rural health disparities

Gemme Campbell-Salome, PhD

Campbell Salome
Assistant Professor, Secondary Appointment 
Health communication; precision medicine; mixed methods; family communication 

Alexander Chang, MD, MS

Alex Chang
Associate Professor
Faculty member, Center for Kidney Health Research 
Kidney and cardiometabolic disease, genetics

Yasser El-Manzalawy, PhD

Yasser El manzalawy
Assistant Professor
Biomedical data science; natural language processing; chronic diseases; critical and perioperative care

Jamie A. Green, MD, MS

Rilow Jamie Green
Associate Professor
Faculty member, Center for Kidney Health Research

Jing Hao, PhD, MD

Jing Hao
Associate Professor 
Health services research; health economics; cost-effectiveness analysis; precision medicine; inherited diseases screening and prevention

Annemarie G. Hirsch, PhD, MPH

Annemarie Hirsch
Associate Professor
Director, Center for Community Environment and Health 
Electronic health record-based epidemiology; social and environmental determinants of health; disease surveillance; diabetes; airways diseases; Lyme disease

Yirui (Iris) Hu, PhD

Yirui Iris Hu
Assistant Professor 
Risk prediction; machine learning; causal inference; cluster randomized trials; stepped-wedge disease; meta-analyses

Bobbie Johannes, PhD, MPH

Bobbie Johannes
Assistant Professor 
Access to primary care/preventative care; rural health; social determinants of health

Anne Justice, PhD

Anne Justice
Associate Professor 
Genetic Epidemiology; multinomics; obesity; cardiometabolic disease; sleep; health disparities

H. Lester Kirchner, PhD

Lester Kirchner
Professor and Chair 
Biostatistics; predictive modeling; comparative effectiveness trials; electronic health records

Thomas Morland, MD

Thomas Morland
Assistant Professor 
Therapeutic lifestyle intervention for cardiometabolic syndrome; implementation and evaluation of predictive modeling of cardiovascular events

Melissa N. Poulsen, PhD, MPH

Melissa Poulsen
Assistant Professor
Faculty member; Center for Community Environment and Health 
Multilevel influences on health and health behavior; environmental epidemiology; implementation science; qualitative research; substance use disorders

Katrina M. Romagnoli, MLIS, MS, PhD

Photo not available.

Assistant Professor

Brian S. Schwartz, MD, MS

Brian Schwartz
Co-Director, Center for Community Environment and Health 
Environmental epidemiology; airways disease; community contributors to chronic and chronic episodic diseases; diabetes; built environment; unconventional natural gas development; community natural environments and health

Erin Van Enkevort, PhD

Erin Vanenkevort
Assistant Professor 
Mixed methods; psychosocial factors; health behaviors; health and wellbeing; patient/caregiver relationships

Staff scientists

Postdoctoral fellow
  • Jazmine Gabriel, PhD
Content from General Links with modal content