A group of current and former DCRI researchers contributed to a paper in JACC warning against relying on observational data to guide clinical decisions.
As the proliferation of real-world data continues, it is important to refocus on combining randomized trials with real-world data sources rather than relying on observations or common sense, write current and former members of the DCRI in a recently published paper in Journal of the American College of Cardiology.
“The critical next step is to apply randomization to real-world data on a broad scale, harnessing the power of randomization to understand treatment effects, and the power of real-world data to generate large, representative, infinitely reusable study populations,” the authors write.
The piece, which was led by former DCRI fellow Alexander Fanaroff MD, MHS, and DCRI faculty member Renato D. Lopes, MD, PhD, also includes DCRI’s Executive Director, Adrian Hernandez, MD, MHS, as a contributor, along with DCRI faculty members Manesh Patel, MD; John Alexander, MD, MHS, and Christopher Granger, MD. Former DCRI Executive Directors Robert Harrington, MD, and Robert Califf, MD, also contributed.
Problems traditionally associated with randomized controlled trials (RCTs)—namely, cost and complexity—have led some to argue that common sense and clinical observation should be used to generate evidence to support clinical decision making.
“Concerns about traditional RCT models are legitimate, but randomization remains a critical and irreplaceable tool to understand the causal relationship between treatments and outcomes,” Lopes (pictured left) said, adding that often, results from RCTs have directly contradicted conclusions made through observation.
“These ideas are especially important during the global COVID-19 pandemic,” Fanaroff (pictured right) said. “Many have expressed concerns that RCTs delay the use of promising therapies, but the past 40 years in cardiovascular medicine shows that many therapies supported by common sense or observation are either ineffective or harmful when studied in rigorous RCTs.”
The authors outline the following potential downfalls of relying on common sense and observation:
- Incomplete understanding of pathophysiology;
- Biases and unmeasured confounding in observational research; and
- Incomplete understanding of risks and benefits in complex systems.
“Moving forward, the clinical research enterprise needs to be transformed to focus on streamlined studies that are both randomized and incorporate real-world data, in accordance with guidance put forth by the FDA,” Lopes said.
RCTs are also not perfect, and they can experience problems such as inadequate dosing of the therapy being tested or selection of the incorrect study population. But well-designed RCTs are “the best current method to understand the causal relationship between an intervention and subsequent population-level outcomes for most common chronic illnesses,” the authors write.