CardioExchange Archive

CardioExchange’s John Ryan interviews Michael S. Lauer, coauthor of an NEJM Perspective article titled “The Randomized Registry Trial — The Next Disruptive Technology in Clinical Research?”

Ryan: What challenges do you anticipate in introducing randomized registries in the U.S., compared with those in Scandinavia?

Lauer: The biggest challenge will be cultural. We are so used to doing things a certain way, with clinical trial operations wholly divorced from clinical care, that people will skeptical — and some, I fear, even cynical. Moving to this kind of model for conducting trials may be analogous to the move from film to digital photography, which was perceived (rightly) as a threat to certain interests. When it first appeared, digital photography was criticized by people in the mainstream market as inferior and inadequate. We hear the same kind of critiques today when some folks in the “mainstream clinical trial market” talk about using registry data, claims data, or electronic medical records data as platforms for conducting trials.

Other important challenges do exist, of course, though in the long run they will be largely technical. These include the embedding of informed consent, identifying and managing missing values, ensuring confidentiality, and checking data quality.

Ryan: Are some populations more readily studied in this manner? The TASTE and the SAFE-PCI studies are both interventional studies — is that a more ideal population?

Lauer: One of the most difficult challenges in all clinical research is the ability to ensure reliable, accurate, and comprehensive follow-up. Short-term studies such as TASTE are attractive because follow-up is short. We also want to look for populations already included in registries (e.g., those automatically enrolled in the various NCDR registries) and those for whom follow-up is “built in.” These might include enrollees in integrated health care plans or people in fee-for-service Medicare.

Ryan: How substantial are the cost differences between running a randomized registry and conducting an RCT?

Lauer: The differences are potentially huge if trialists and sponsors leverage an existing resource to find patients, collect baseline data, and conduct follow-up. TASTE incurred only $300,000 in incremental costs (i.e., the cost that might have been borne by NHLBI, had we funded it). A traditional RCT would have easily cost tens of millions of dollars by current standard U.S. practices. I would also note that a randomized registry trial is a randomized trial — it’s a subtype of an RCT.

Ryan: Do you anticipate that registries will have more open access to data than RCTs?

Lauer: Yes, they should, but there is no automatic assurance. Investigators and sponsors will need to work on this.

JOIN THE DISCUSSION

Share your thoughts about the promise of randomized registry trials and Dr. Lauer’s take on them.