Integration of social determinants of health into health outcomes research will allow researchers to study underlying causes of health inequities. The All of Us Research Program has the potential to be a rich source of social determinants of health data; however, recommendations for scoring and interpreting the All of Us Research Program Social Determinants of Health Survey are needed to help researchers overcome the “learning curve” and facilitate optimal usage. We created a user guide to give researchers an overview of the All of Us Research Program Social Determinants of Health Survey, recommendations for scoring and interpreting participant responses, and implementable R and Python functions.
This user guide targets registered users of the All of Us Research Hub Researcher Workbench, a cloud-based platform that supports analysis of All of Us Research Program data. This repository will be especially helpful for those currently conducting or planning to conduct analyses using All of Us Research Program Social Determinants of Health Survey data. If you are not already registered as an All of Us researcher, please check out the All of Us Research Hub (https://www.researchallofus.org/). Registration includes, 1) confirming your affiliation with an institution that has a Data Use and Registration Agreement (DURA), 2) creating an account and verifying your identity with Login.gov or ID.me, 3) completing the mandatory training focusing on responsible and ethical conduct of research, and 4) signing the Data User Code of Conduct (DUCC).
In the repository, we provide R and Python code for the 14 social determinants of health constructs evaluated as part of the All of Us Social Determinants of Health Survey. Details about the survey can be found on the All of Us Survey Explorer (https://www.researchallofus.org/data-tools/survey-explorer/). In the commented code we have provided literature-informed recommendations for scoring participant responses and interpretation of scores. Then, we provide example R and Python functions for relabeling responses and scoring constructs that can be directly implemented in Jupyter Notebook or RStudio within the Researcher Workbench. The 14 constructs are listed below,
- Neighborhood cohesion
- Neighborhood disorder
- Neighborhood environment
- Social support
- Loneliness
- Perceived everyday discrimination
- Perceived discrimination in health care settings
- Food insecurity
- Housing insecurity/instability
- Housing quality
- Perceived stress
- Daily spiritual experiences
- Religious service attendance
For more detail on this work and the constructs within the All of Us Social Determinants of Health Survey, please reference and cite this paper: Koleck, T., Dreisbach, C., Zhang, C., Grayson, S., Lor, M., Leggio, K., Conway, A., Higgins, P., & S. Bakken. (2024). Data User Guide for Social Determinants of Health Survey Data in the All of Us Research Program. Under review at the Journal of the American Medical Informatics Association.