The IMPACT-AFib trial lays the groundwork for future large-scale trials to be conducted using electronic claims data.
Although the intervention tested as part of IMPACT-AFib was found not to be effective, the design and implementation of the trial showcased the effective use of a new electronic data system to conduct a large-scale pragmatic trial.
The trial, the findings of which were presented by the DCRI’s Sean Pokorney, MD, MBA, at the virtual European Society of Cardiology Congress 2020, was the first to use the U.S. Food and Drug Administration’s (FDA) FDA-Catalyst System and Sentinel System network of electronic health record (EHRs) and insurance claims data from a diverse group of data partners. The trial enrolled 80,000 patients with atrial fibrillation who were at high risk of stroke and not taking medication.
IMPACT-AFib was a collaborative effort among the DCRI, the Clinical Trials Transformation Initiative (CTTI), the Department of Population Medicine at Harvard Medical School and Harvard Pilgrim Health Care Institute, the FDA, and five data partners (Aetna, Harvard Pilgrim HealthCare, HealthCore Anthem, Humana, and Optum). The study team also received input from a patient representative during the study.
At least one third of patients with atrial fibrillation and risk of stroke are not treated with anticoagulation, resulting in large numbers of preventable strokes. This study was designed to address that problem by determining whether mailing educational information on oral anticoagulants and their role in stroke prevention to both patients with atrial fibrillation and their providers would support increased adoption of these drugs. Study participants were identified using administrative medical claims and pharmacy dispensing data from the Sentinel System.
“We hypothesized that the educational information would help patients initiate discussions about oral anticoagulants with their providers, thereby increasing prescription rates for these important drugs,” said Pokorney, co-principal investigator of the study. “Now that we’ve completed the study, we know that we need to find other methods to help bolster stroke prevention.”
Despite the lack of impact from the educational intervention, investigators said IMPACT-AFib opens the door for new ways to conduct studies using claims data and other types of real-world data.
“The evidence generated from the trial showed that a single educational mailing about the recommended treatment, as compared to a control group, wasn’t effective in this context,” said DCRI’s Christopher Granger, MD, co-principal investigator of the study. “And while the intervention did not increase the proportion of patients with at least one oral anticoagulant dispensed, the study itself is a success. We conducted an extensive collaboration to conduct the first-ever trial using a decentralized database of claims data—that, in and of itself, sets a foundation for future clinical trials in atrial fibrillation and other diseases.”
CTTI, for which the DCRI acts as a coordinating center, was a key contributor to IMPACT-AFib’s success, conducting extensive pre-work within the Catalyst system to prove the viability of conducting a large-scale trial using the system.
“The study is a successful proof of concept of embedding a randomized clinical trial into a claims system, while confirming in a large scale experiment that the use of educational interventional approaches in medicine might be limited,” said Pamela Tenaerts, MD, MBA, executive director at CTTI. “We believe that all future clinical trials should maximally leverage available clinical and nonclinical data to minimize collection of necessary trial specific data, and the IMPACT-AFib trial is a fantastic example of that; it serves as a strong model for future research.”