Programs | Center for Clinical Trials Innovation (CCTI)

Center for Clinical Trials Innovation (CCTI)

About

Led by Founding Director Dr. Michael Harhay, the Center for Clinical Trials Innovation (CCTI) serves as a dynamic research, training, and collaboration hub that improves health through methodological innovations in clinical trials.

This program, an evolution of the PAIR Center’s Clinical Trials Outcomes and Methods Lab that launched in 2021, directly informs the conduct and analysis of all trials at the PAIR Center and has broad applications for the research community at large.

Picture of two people's hands typing on computers side by side.

What We Do

Novel pragmatic and cluster-randomized trial designs

Evaluating the effectiveness of interventions via pragmatic trials (trials conducted in real-world settings) is one of the most important methods by which health systems may learn. Further, cluster-randomized trials, or CRTs, that randomize treatments by different groups of patients, rather than individual patients, are particularly well suited for pragmatic studies, as they allow for natural implementation of new treatments within routine care delivery at the hospital, clinic, or unit level.

At CCTI, we develop and refine novel pragmatic and cluster-randomized designs and analytic methods that support the research mission of learning health systems, driving innovations in stepped-wedge and crossover CRTs and non-inferiority and equivalence trial designs.

Bayesian analysis approaches

Most randomized clinical trials fail to find statistically significant differences between treatments or management strategies when using a 0.5 p-value threshold, as is typical in conventional frequentist statistics. However, declaring that an intervention does or does not work based on this standard dramatically reduces the ability of a trial to change practice.

CCTI has been on the forefront developing Bayesian methods as an alternative for the design and analysis of clinical research. These methods portray results by showing the probability that an intervention provides one or more benefits of a certain size. Particularly when interventions are devoid of harms, even small benefits of interventions can be important. This work on Bayesian approaches enables investigators at the PAIR Center and beyond to produce results that are more likely to improve clinical care.

Patient-centered outcome measures

Patient-centered outcome measures, such as quality-of-life scales or hospital length of stay, are essential for evaluating new interventions in ways that take into account patients’ goals, values, and preferences for their health care. However, missing patient outcomes is a common challenge due to gaps in administrative data or research participants who are unable to contribute responses. Without this key information, study analyses can become biased.

For this reason, CCTI works to develop new patient-centered outcome measures and provide guidance for their interpretation. For example, within trials of seriously ill patients, data missing due to death makes it difficult to measure how the treatment affected patient-centered outcomes. Work from CCTI focuses on new methods of analysis that can address these challenges. Further, the team takes into account the perspectives of diverse stakeholders to ensure that new outcome measures are able to record what matters most to patients and their families.