Clinical trials provide valuable information to health care providers on the average effect of treatment through the comparison of treated and control groups that are balanced (similar) on important variables. While this comparison yields insights on the average treatment effect, results may not apply equally to all patients. Certain subgroups may have better or worse outcomes based on factors such as genetics, demographics or underlying health conditions.
The current standard for evaluating subgroups is to consider one variable at a time, for example, males versus females. These methods do not allow for analysis of multiple variables such as Hispanic females over age 65. Though methods exist that allow for multiple variables, they may not identify subgroups that are balanced on important variables. For example, such methods may show that a treatment looks promising in individuals over age 65. However, closer examination shows that treated individuals within this age group had higher socioeconomic status, raising the question of whether the improved outcomes were due to treatment or socioeconomic status.
To help address this gap, researchers at Wake Forest University School of Medicine have been approved for a $1 million award by the Patient-Centered Outcomes Research Institute (PCORI) for a methodology study.
“Using a combination of matching and machine learning, we will evaluate a new method capable of considering multiple variables to identify subgroups balanced on important variables,” said Joseph Rigdon, Ph.D., assistant professor of biostatistics and data science at Wake Forest University School of Medicine.
Rigdon initially developed the method with colleagues at Stanford University and plans to apply the method to three studies, representing a range of conditions and treatments:
- The Systolic Blood Pressure Intervention Trial (SPRINT), led by Wake Forest University School of Medicine, showed intensive management of high blood pressure, below a commonly recommended blood pressure target, significantly reduces rates of cardiovascular disease.
- SEARCH for Diabetes in Youth, the largest surveillance effort of diabetes among youth under the age of 20 conducted in the U.S. to date. Wake Forest University School of Medicine served as the coordinating center for this study.
- Diabetes in Children, Adolescents, and Young Adults (DiCAYA) Network study, which aims to modernize surveillance efforts and monitor trends, is led by NYU Long Island School of Medicine and NYU Grossman School of Medicine.
“A key component of the study is producing a visual representation of the findings to help providers and patients engage in shared decision-making before selecting a treatment,” Rigdon said. “The visual will make it clear who will benefit most and least.”
The research team also plans to create a Stakeholder Advisory Committee that includes patients, parents and guardians, patient advocates, doctors, and researchers to guide strategic decisions such as how best to present findings.
This study is among the latest methodology studies that PCORI has funded to address gaps in comparative clinical effectiveness research (CER) methods. These studies provide results that guide researchers in planning future studies and improve the strength and quality of evidence generated by CER.
The award has been approved pending the completion of a business and programmatic review by PCORI staff and issuance of a formal award contract.
Rigdon said the method will not only help providers and patients but will also be useful to other researchers.
“With this funding support, we will create software to make this method widely available and easily adapted for other research projects,” Rigdon said.
PCORI is an independent, nonprofit organization authorized by Congress with a mission to fund patient-centered comparative clinical effectiveness research that provides patients, their caregivers and clinicians with the evidence-based information they need to make better-informed health and healthcare decisions.
Media contact: Myra Wright, mgwright@wakehealth.edu