Project Summary
One of PCORI’s goals is to improve the methods that researchers use for patient-centered outcomes research. PCORI funds methods projects like this one to better understand and advance the use of research methods that improve the strength and quality of comparative effectiveness research.
This research project is in progress. PCORI will post the research findings on this page within 90 days after the results are final.
What is the project about?
In observational studies, researchers look at what happens after patients and their doctors choose treatments. In these studies, patients with different traits—for example, older or younger patients—may choose different treatments. As a result, researchers may have a hard time knowing if it was the treatment that affected patient health outcomes, or whether patient traits affected the outcomes. Statistical methods can help address this problem.
Propensity score, or PS, methods can help researchers conduct observational studies. When looking at outcomes for different treatments, researchers can use PS methods to balance treatment groups. That is, they can account for differences in traits among patients who get different treatments and improve understanding of whether effects are due to the treatment or patients’ traits. But by creating more balanced, and often smaller, treatment groups, PS methods can decrease precision, making results less certain.
Subgroup analyses show how the same treatment affects groups of patients differently. For example, a treatment might work better for younger patients than older ones. PS methods can be used to study subgroups, but standard PS methods may not work well in subgroups.
In this study, the research team is developing new methods for testing balance and precision when planning subgroup analyses. The team is also developing new PS methods for conducting subgroup analyses.
How can this project help improve research methods?
Results may help researchers get information about what treatments work best for different groups of patients that is more accurate and precise.
What is the research team doing?
First, the research team is developing methods researchers can use to assess balance and precision when designing subgroup analyses. The methods detect and solve potential problems that could arise in subgroup analyses. Second, the team is developing new PS methods that use machine learning to improve balance and precision. In machine learning, computers use data to learn how to perform different tasks with little or no human input. Third, the team is comparing the new methods to existing methods to determine which methods are best for conducting subgroup analyses. Finally, the team is testing the new methods using data on treatments for women with uterine fibroids.
Research methods at a glance
Design Elements | Description |
---|---|
Goal |
|
Approach |
Simulation, machine learning, propensity scoring, subgroup analyses |
COVID-19-Related Study
Summary
With this enhancement, the research team will apply the new methods for subgroup analyses from the original award. The team will develop criteria to assess sources of error, including
- Bias (when study results are distorted)
- Misclassification (when people, treatments, or outcomes are incorrect)
- Imprecision (when measurements are widely scattered)
The research team will account for these sources of error when estimating treatment effects. The team will apply the criteria and new methods to insurance claims data. They will look at the effects of three treatments for COVID-19 on subgroups heavily affected by COVID-19, such as African Americans or Latinx populations.
Enhancement Award Amount: $349,999