Development of an Integrative, Hands-on Health Data Analysis Course

Development of an Integrative, Hands-on Health Data Analysis Course

Academic Year:
2016 - 2017 (June 1, 2016 through May 31, 2017)
Funding Requested:
Project Dates:
Chair Uniqname:
Overview of the Project:
With growing access to massive amounts of data, the research paradigm has shifted from collecting data
among a few subjects to using large data sources to efficiently produce impactful research on health and
healthcare. Trainees across the health sciences are in need of access to high quality data. In our opinion, the limiting factor is not what data are available but rather the skills required to utilize them. Therefore, we propose to develop and implement a novel, hands-on data course designed to teach health science trainees how to obtain and analyze pre-existing large US health data. This course will integrate practical research skills (e.g., applied statistics, data management and programming,research question development, creation of publication-quality tables/figures) in a format that progressively builds research independence. We have carefully selected a variety of publicly available, national health data sources based on content and complexity (i.e., the data sources used in the course will increase in complexity throughout the term). We anticipate this course will lead to a shift in how students effectively learn about data management and analysis and the extent to which national US health data are used across the university.
Number of Graduate Students Affected Annually:
20 graduate students
Proposal PDF:
Additional Supporters:
<p>Karen Szemak, Nolan Hanson,</p>
Budget Administrator:
Lori Stautz,
Final Report Fields
Project Objectives:

(1) Develop multiple high-quality training datasets from national health data sources (2) Draft a course textbook to teach data analysis to health science students (3) Establish a university-wide seminar series on National Sources of Health Data (4) Pilot course “Management and Analysis of Large US Health Data” and obtain student feedback

Project Achievements:

HS807 has now been offered to UM health sciences students twice (winter 2018 and wintere 2019). This course has impacted several health sciences programs by attracting students in need of the skills to utilize large pre-existing health data. HS807 trainees thus far have included graduate students from the SPH/SI health informatics program, Health Management and Policy (SPH), Biostatistics (SPH), Health Behavior and Health Education (SPH), School of Nursing, and the Medical School. Further, both junior and senior faculty from the School of Nursing and Medical school have participated in HS807, along with several postdoctoral research fellows from the Institute for Healthcare Policy and Innovation. HS807 was recently adopted by both the School of Nursing and the SPH/SI health informatics programs as a program elective. We hope to further integrate HS807 formally into other health science programs. Student feedback from the initial implementation has been excellent and strongly supportive of continuation of the course. In addition, the course led to the establishment of the National Sources of Health Data webinar series. This a 5-session webinar open to all of UM students, faculty, and staff that includes presenters from government agencies including: the Centers for Disease Control and Prevention, the National Center for Health Statistics, the National Cancer Institute, the Health Resources and Services Administration, and the Agency for Healthcare Research and Quality. Specific training materials developed through the support of this grant include: (1) a 305-page course textbook, (2) large training datasets for 6 research assignments (based on seminal studies) that amount to 15 GBs of data, and (3) 10 programming challenge activities. The number of students impacted is 1 undergraduate, 41 graduate students, 3 junior faculty, 2 senior faculty, and 4 postdoctoral research fellows.

This course (along with the webinar series) will continue to be offered once per academic year. We will seek to further develop the course into a national MOOC and further collaborate with federal agencies who guest present in the course. Training datasets that use national health data for the course will be updated every 2-3 years.
The training materials have been distributed to trainees and colleagues via the textbook and other training materials. In addition, the webinar series has attracted individuals from across the university.