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?
Personalized treatment rules, or PTRs, are rules that help doctors recommend treatment based on a patient’s traits. Health systems have large amounts of data that researchers can use to study patient health and create PTRs. But the same PTR won’t work for everyone. The best way to treat older patients with heart failure, for example, might be different from the best way to treat this condition in younger patients. If a health system has a lot of data on older patients, it may be able to make a PTR for them, but that PTR might not work for younger groups.
In this study, the research team is testing methods to create PTRs that clinicians can use across groups with similar characteristics and account for differences in how much data are available about one group versus another.
How can this project help improve research methods?
Researchers can use the results to produce PTRs that use data from one patient group that can be applied to other patient groups.
What is the research team doing?
First, the research team is developing a way of creating PTRs that works across different patient groups. Second, the team is developing a way to create PTRs that can be applied to multiple patient groups, even if not all data are available for one of the groups. Third, the team is using machine learning to identify the patients to whom PTRs should be applied across groups. In machine learning, computers use data to learn how to perform tasks with little or no human input.
Research methods at a glance
Design Elements | Description |
---|---|
Goal |
Aim 1: To develop a PTR using data from a source population when the target population has outcome, treatment and covariate information but small numbers of patients Aim 2: To develop a PTR using data from a source population when the target population has only covariate information available Aim 3: To develop criteria for selecting patients in the target population to apply a PTR that was developed and internally validated using a separate source population |
Approach | Retrospective cohort study design with before-and-after measures and matched comparisons; machine learning; software development |