Melissa Patrick

Project Personnel, All of Us Program Operational Use

2 active projects

Systemic Disease and Glaucoma (Cloned)

We have previously published a predictive model of glaucoma progression using electronic health record (EHR) data pertaining to systemic attributes from a single institution. We aim to use the All of Us dataset to 1) serve as external validation for…

Scientific Questions Being Studied

We have previously published a predictive model of glaucoma progression using electronic health record (EHR) data pertaining to systemic attributes from a single institution. We aim to use the All of Us dataset to 1) serve as external validation for this single-center model and 2) to train new models focused on predicting glaucoma progression using systemic predictors. This is important to understand whether the original findings are generalizable and provide additional knowledge about the utility of systemic predictors on a national-level dataset.

Project Purpose(s)

  • Disease Focused Research (Primary open angle glaucoma)
  • Other Purpose (This work is the result of an All of Us Research Program Demonstration Project. Demonstration Projects are efforts by the All of Us Research Program designed to meet the goal of ensuring the quality and utility of the Research Hub as a resource for accelerating precision medicine. This work has been approved, reviewed, and overseen by the All of Us Research Program Science Committee and Data and Research Center to ensure compliance with program policy. )

Scientific Approaches

We plan to primarily work with EHR data contained in All of Us for a cohort of adult participants diagnosed with primary open-angle glaucoma. We will extract data on systemic conditions and medications for this cohort, as well as physical measurements and vital signs. We will clean the data such that the format is consistent with the data from our previously published model. Then, we will use this data as an external validation of a logistic regression model derived from our prior study that was based at a single academic center. Next, we will use All of Us data to train a new set of models, using techniques such as logistic regression, random forests, and artificial neural networks. We will optimize these models using feature selection methods and class balancing procedures. By evaluating performance metrics such as area under the curve (AUC), precision, recall, and accuracy, we will assess whether we can achieve superior predictive performance when training models using All of Us.

Anticipated Findings

We anticipate that the All of Us data will validate the findings from the model, which demonstrated that blood pressure-related metrics and certain medication classes had predictive value for glaucoma progression. In addition, we anticipate that the models trained with All of Us data will outperform the model trained with single institution data due to larger sample size and greater diversity. These findings will support further investigation in understanding the relationship between systemic conditions like blood pressure with glaucoma progression.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

Collaborators:

  • Chenjie Zeng - Research Fellow, NIH

Duplicate of Demo Project - Family History in EHR & PPI Data

As a demonstration project, this study will summarize structured data elements available in the All of Us registered tier and compare to published survey results to describe data for reuse in disease specific outcomes. Specific questions include: 1. Could harnessing…

Scientific Questions Being Studied

As a demonstration project, this study will summarize structured data elements available in the All of Us registered tier and compare to published survey results to describe data for reuse in disease specific outcomes. Specific questions include:

1. Could harnessing informatics tools like predictive modeling and clinical decision support to detect and alert healthcare providers to these preventative measures significantly improve the precise care we deliver to patients?
2. How can one evaluate the availability of family medical history information within the All of Us registered tier data and characterize the structured data elements from both data sources?

Project Purpose(s)

  • Methods Development
  • 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 Approaches

We utilize the Family Medical History PPI survey to capture self-reported information but exclude participants who did not know any of their family history or who skipped every survey question. We pay particular attention to the disease/relative pairings that map to the American College of Medical Genetics and Genomics’ (ACMG) list of important diseases.

We define EHR family history information as the collection of registered tier observations with "family+history" or "FH:" anywhere in their OMOP concept name. We exclude observations of “Family social history” and remove duplicate observation and value concept pairings from the same healthcare organization regarding the same participant as these were likely due to repeated entries across multiple routine annual physical exams.

We aim to compare the data sources by summarizing the type and amount of family history information gained.

Anticipated Findings

This description of the family medical history data in the All of Us registered tier database will assist future investigators in understanding All of Us data methods and give feedback to the program on the utility of participant survey and EHR data.

We hypothesize that the survey data will provide a more complete look at family medical history due to its structured nature. Though, we are also interested in determining how much overlap there is between the PPI and EHR data. It’s plausible that the free-form nature of EHR family history information yields more detailed records. We would ultimately like to determine if a gold standard method for defining a participant’s family medical history is attainable within the All of Us registered tier data.

We anticipate facing informatics challenges because of collecting data from different sources, mapping these data to a common data model, and attempting to harness data from these sources to find the common source of truth.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Melissa Patrick - Project Personnel, All of Us Program Operational Use
  • Robert Cronin - Early Career Tenure-track Researcher, Vanderbilt University Medical Center
  • Ashley Able - Other, Vanderbilt University Medical Center
1 - 2 of 2
<
>
Request a Review of this Research Project

You can request that the All of Us Resource Access Board (RAB) review a research purpose description if you have concerns that this research project may stigmatize All of Us participants or violate the Data User Code of Conduct in some other way. To request a review, you must fill in a form, which you can access by selecting ‘request a review’ below.