Clinical Data Insights

March 2, 2020

Factors Involved in Real World Data Comparisons

The clinical community is eager to have more information about clinical outcomes in patients treated with mechanical circulatory support devices. In this interview, George Vetrovec, MD, discusses how members of the community can more confidently make real world data comparisons. Dr. Vetrovec is a Master of the American College of Cardiology, a Master of SCAI, and Emeritus Professor of Medicine and Cardiology at Virginia Commonwealth University.

“There are good things and there are risks to it,” Dr. Vetrovec responds to a question about the use of paired data, such as Optum or Nationwide Inpatient Sample (NIS) data, for assessing clinical outcomes. “The good thing is, and I think the clinicians really appreciate this, is it’s real world data. On the other hand, when you start doing comparisons, you have to be really very careful that you’re choosing similar groups of patients to compare.” Dr. Vetrovec emphasizes the need to understand whether the differences in outcomes are a result of therapy or whether they are due to other characteristics in the patient population.

Dr. Vetrovec explains that a propensity analysis can help raise the quality of the data analysis. “You take a series of risk factors that are common to these groups and pick a population that got one therapy or the other and make it comparative based on a number of risk factors, so that they look like similar patients. Making sure that there aren’t any obvious biases in that makes a real difference in terms of being sure you get a really important dataset that you can have confidence in analyzing.”

Dr. Bilazarian then asks Dr. Vetrovec about comparing datasets over long periods of time while best practice understandings are dramatically changing. For example, what is the impact of the evolution of best practices such as Impella pre-PCI and the use of right heart cath guidance in the cardiogenic shock patient population?

Dr. Vetrovec expresses concern about taking 10 years of data and trying to make assessments when so much is changing over that time. “One of the challenges over my career in cardiology,” responds Dr. Vetrovec, “is things have been changing so rapidly. It’s very hard to get a sample that’s not in some way biased just by the change in how we manage patients.” Vetrovec explains that one of the advantages of a large database is that it provides a rich data source over a much shorter period of time.

Related Content:

NPS-490
NPS-610

Subscribe or join the conversation by following us on Twitter: @ProtectedPCI


To learn more about the Impella® platform of heart pumps, including important risk and safety information associated with the use of the devices, please visit: www.protectedpci.com/indications-use-safety-information/

Subscribe to the Digital Community

Receive weekly updates including case review videos, tips and tricks, specialists opinions, and resources and downloads regarding Protected PCI and Cardiogenic Shock.

Subscribe to the Digital Community

Receive weekly updates including case review videos, tips and tricks, specialists opinions, and resources and downloads regarding Protected PCI and Cardiogenic Shock.