Research Projects Directory

Research Projects Directory

At this time, all listed projects are using data in the registered tier. The registered tier contains individual-level data from electronic health records, survey answers, and physical measurements. These data have been altered to protect participant privacy.

Note: Researcher Workbench users provide information about their research projects independently. Any views expressed in the Research Projects Directory belong to the relevant users and do not necessarily represent those of the All of Us Research Program.

Information in the Research Projects Directory is also cross-posted on AllofUs.nih.gov in compliance with the 21st Century Cures Act.

There are currently 214 active workspaces. This information was updated on 9/23/2020.

Sort By Title:

MDD_test

Project Purpose(s)

  • Population Health ...

Scientific Questions Being Studied

Initial exploratory data analysis to assess the AoU data potential hypotheses that could be explored

Scientific Approaches

Case-Control statistical analysis to compare across significant differences across various sub-cohorts in AoU based on labels that will be derived based on survey outcomes.

Anticipated Findings

To help develop better understanding of AoU datasets and workbench that will be used to inform future research studies

Demographic Categories of Interest

  • Race / Ethnicity
  • Age
  • Sex at Birth
  • Gender Identity
  • Sexual Orientation
  • Geography
  • Disability Status
  • Access to Care
  • Education Level
  • Income Level

Research Team

Owner:

  • Abhishek Pratap - Senior Researcher, Sage Bionetworks

Mental Health and Substance Use Demo Projects

Project Purpose(s)

  • Disease Focused Research (disease of mental health)
  • Population Health ...

Scientific Questions Being Studied

What are the prevalences of mental health conditions in the AoURP?

Scientific Approaches

Not available.

Anticipated Findings

AoURP data can be used to assess mental health conditions in previously under-represented populations.

Demographic Categories of Interest

  • Sex at Birth
  • Education Level
  • Income Level

Research Team

Owner:

  • Chen Yeh - Project Personnel, Northwestern University

Collaborators:

  • Kai Yin Ho - Project Personnel, Northwestern University
  • Joyce Ho - Mid-career Tenured Researcher, Northwestern University

Mental Health Demonstration Project

Project Purpose(s)

  • Disease Focused Research (generalized anxiety disorder, depressive disorder, bipolar disorder)
  • Other Purpose (“This work is a result of an All of Us Research Program Demonstration Project. The projects are efforts by the Program designed to meet the program's goal of ensuring the quality and utility of the Research Hub as a resource for accelerating discovery in science and medicine. This work was reviewed and overseen by the All of Us Research Program Science Committee and the Data and Research Center to ensure compliance with program policy, including policies for acceptable data access and use”.) ...

Scientific Questions Being Studied

As a demonstration project, this study aimed to explore the usability of the All of Us dataset and examined the prevalence of mental health conditions in the All of Us Research Program cohort. Specifically, we explored the lifetime prevalence of depressive disorder, bipolar disorder, and generalized anxiety disorder.

Our study looked prevalence rates for the above conditions in the following ways:
1. Prevalence in EHR data available by various demographic factors
2. Cohort characteristics
3. Congruency for diagnoses in EHR and self-report questionnaire
4. Among individuals who self-report as having been diagnosed with a mental health condition listed above, the percentage of individuals in treatment and associations between treatment and various demographic factors

Scientific Approaches

In this analysis, we calculated prevalence of mental health conditions by leveraging demographic information, questionnaire responses, and EHR data Specifically, we utilized the following surveys: Basics, Overall Health, Personal Medical History, and Healthcare Access PPIs. We utilized EHR data by creating a cohort of individuals with specific diagnoses code in their EHR. We referenced all relevant parent and child SNOMED codes for each mental health condition of the investigation (documented in Concept Set). Associations were calculated using Chi Square.

Anticipated Findings

We anticipated that the prevalence rates found in All of Us will be consistent with previous large scale studies, such as the National Comorbidity Survey. We found that the All of Us dataset is sensitive to detecting mood disorders and is usable for examining mental health conditions.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Chen Yeh - Project Personnel, Northwestern University

Collaborators:

  • Kai Yin Ho - Project Personnel, Northwestern University
  • Joyce Ho - Mid-career Tenured Researcher, Northwestern University

Miscellaneous

Project Purpose(s)

  • Other Purpose (Trouble-shooting, thanks for your help, Francis.) ...

Scientific Questions Being Studied

Trouble-shooting, thanks for your help, Francis.

Scientific Approaches

no approach is necessary for this workspace because this is for operational use only.

Anticipated Findings

Trouble-shooting, thanks for your help, Francis.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Guohai Zhou - Early Career Tenure-track Researcher, Massachusetts General Hospital

Collaborators:

  • Robert Carroll - Other, All of Us Program Operational Use
  • Francis Ratsimbazafy - Other, All of Us Program Operational Use

NIV Failure Characterization

Project Purpose(s)

  • Disease Focused Research (respiratory failure) ...

Scientific Questions Being Studied

We are attempting to identify the failure mechanisms of noninvasive ventilation therapy in patients with acute respiratory failure.

Scientific Approaches

We plan to validate an existing phenotyping method for ventilation therapy patients and generate summary statistics between the different cohorts to characterize any major differences.

Anticipated Findings

We anticipate that our approach will validate the previously created phenotyping algorithm and identify major differences between patients that successfully undergo noninvasive ventilation therapy and those that fail noninvasive therapy and require subsequent endotracheal intubation.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Patrick Essay - Graduate Trainee, University of Arizona

Collaborators:

  • Vignesh Subbian - Early Career Tenure-track Researcher, University of Arizona

Noninvasive Ventilation Failure

Project Purpose(s)

  • Disease Focused Research (respiratory failure) ...

Scientific Questions Being Studied

Questions:
We are interested in identifying differences in characteristics among acute respiratory failure patients treated with various mechanical ventilation therapies. In particular, what are the mechanisms of failure for certain types of ventilation? Identifying key differences between patients successfully treated with different ventilation devices and those that failed therapies requiring a change in treatment path will all clinicians to better match patients with the appropriate therapies.

Scientific Approaches

We will be validating a recently published phenotyping algorithm for identification of invasive and noninvasive mechanical ventilation. We will generate basic summary statistics across each phenotype and apply feature selection and subgroup analysis to identify key differences between patients successfully treated with each type of therapy.

Anticipated Findings

We anticipate key differences between ventilation phenotypes will be identifiable in physiological data and that these differences may be leveraged for characterizing patient groups in detail and modeling risks of potential ventilation failure.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Patrick Essay - Graduate Trainee, University of Arizona

NW_AOU_Informatics

Project Purpose(s)

  • Population Health ...

Scientific Questions Being Studied

We are informatics researchers using AOU data to study research data sets. We want to understand the kinds of data enclosed and the utility for data reuse research.

Scientific Approaches

Data mining, mostly.

Anticipated Findings

We will find the data reuse potential of AOU.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Nick Williams - Research Associate, NIH

Obesity analysis

Project Purpose(s)

  • Disease Focused Research (obesity) ...

Scientific Questions Being Studied

state level obesity

Scientific Approaches

Not available.

Anticipated Findings

state level obesity disparities after adjusting for socioeconomic factors

Demographic Categories of Interest

  • Sex at Birth
  • Education Level
  • Income Level

Research Team

Owner:

  • Guohai Zhou - Early Career Tenure-track Researcher, Massachusetts General Hospital

Collaborators:

  • Francis Ratsimbazafy - Other, All of Us Program Operational Use
  • Paulette Chandler - Early Career Tenure-track Researcher, Massachusetts General Hospital
  • Andrea Ramirez - Other, All of Us Program Operational Use
  • Elizabeth Karlson - Late Career Tenured Researcher, Massachusetts General Hospital
  • Cheryl Clark

obesitypaper02202020

Project Purpose(s)

  • Disease Focused Research (obesity) ...

Scientific Questions Being Studied

how does bmi differ by state, ses, race/ethnicity

Scientific Approaches

Not available.

Anticipated Findings

obesity will vary by region

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Paulette Chandler - Early Career Tenure-track Researcher, Massachusetts General Hospital

Older Adults

Project Purpose(s)

  • Population Health
  • Methods Development ...

Scientific Questions Being Studied

Exploration of traditional regression models vs. machine learning methods for large population health studies of older adults (65+).

Scientific Approaches

Application and comparison of traditional/contemporary regression models (e.g. generalized linear models) and "machine learning" methods (e.g. gradient boosting, random forest, LASSO) in supervised learning applications.

Anticipated Findings

There is an active debate in the literature as to whether (potentially) increased predictive ability of machine learning techniques is worth the additional, potential black-box complexity vs. traditional regression models. A particular thread of this discussion examines the potential for unintended bias in machine learning methods.

Demographic Categories of Interest

  • Age

Research Team

Owner:

  • John Boscardin - Other, University of California, San Francisco

One

Project Purpose(s)

  • Ethical, Legal, and Social Implications (ELSI) ...

Scientific Questions Being Studied

I am interested in exploring the race and ethnicity distribution of AoU participants. It is important that AoU participants represent the diversity of the United States.

Scientific Approaches

We will use basic statistical methods to compare race/ethnicity and socioeconomic status of AoU participants of different US regions.

Anticipated Findings

By understanding trends in research participation, we can develop ways to promote a more equitable distribution of the risks and benefits of health research

Demographic Categories of Interest

  • Race / Ethnicity
  • Income Level

Research Team

Owner:

  • Susan Passmore - Other, University of Wisconsin, Madison

One

Project Purpose(s)

  • Ethical, Legal, and Social Implications (ELSI) ...

Scientific Questions Being Studied

I am interested in exploring the race and ethnicity distribution of AoU participants. It is important that AoU participants represent the diversity of the United States.

Scientific Approaches

We will use basic statistical methods to compare race/ethnicity and socioeconomic status of AoU participants of different US regions.

Anticipated Findings

By understanding trends in research participation, we can develop ways to promote a more equitable distribution of the risks and benefits of health research

Demographic Categories of Interest

  • Race / Ethnicity
  • Income Level

Research Team

Owner:

  • Susan Passmore - Other, University of Wisconsin, Madison

Opioid Use in Cancer Patients

Project Purpose(s)

  • Disease Focused Research (cancer)
  • Population Health ...

Scientific Questions Being Studied

Is there a difference between how much patients with cancer use opioids depending on ethnicity and environmental setting

Scientific Approaches

Not available.

Anticipated Findings

That there is a racial disparity among opioid use as evidenced by how much they are prescribed.

Demographic Categories of Interest

  • Sex at Birth
  • Geography
  • Access to Care
  • Education Level
  • Income Level

Research Team

Owner:

  • Toluwalase (Lasė) Ajayi - Early Career Tenure-track Researcher, Scripps Research

Collaborators:

  • Francis Ratsimbazafy - Other, All of Us Program Operational Use

Orgs_NCD

Project Purpose(s)

  • Methods Development ...

Scientific Questions Being Studied

A noncommunicable disease (NCD) is traditionally thought of as a disease that is not spread from human to human, such as heart disease or cancer. Over the past few decades however, researchers have found pathogenic organisms (POs) that either increase risk for or directly cause what would be traditionally thought of as an NCD, such as certain human papillomavirus strains causing cervical cancer or Epstein-Barr virus increasing the risk of developing multiple sclerosis. In this study we are exploring if there are still unknown associations between POs and human disease. Previously, utilizing a different cohort, we identified many new associations. However, we would like to verify these results by seeing which associations replicate on the All of Us data. Finding these relationships between POs and human disease would highlight the importance of vaccination, as preventing the infection could also protect or reduce a person’s risk of developing other diseases later in life.

Scientific Approaches

We plan to test models developed using a different cohort on the All of Us data.
Datasets:
- Clinical diagnoses for many different diseases
- Laboratory measurements of antibody titers for different pathogenic organisms
- Sociodemographic data to adjust for possible confounding.

Research Methods:
We will be using statistical analysis to look for associations between pathogenic organisms and disease diagnoses. Our previously built model uses a logistic regression to calculate this association.
Tools:
For the most part we will be using custom R and Python code.

Anticipated Findings

If we can replicate our results found in the previously analyzed cohort, we will have identified new links between certain pathogenic organisms and human diseases. Previously, similar results describing how certain human papillomavirus strains cause most cervical cancers helped spur on the development of a vaccine against those strains and after widescale rollout of that vaccine we are starting to see significant drops in the rate of cervical precancers. We would hope our results could encourage similar developments.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Mike Lape - Research Assistant, Cincinnati Children's Hospital Medical Center