CRT Utilization in ICD Patients: Insights from the NCDR ICD Registry

Interview | Lucas Marzec, MD

In the 1980s and 1990s, the regional variations for the use of interventions such as percutaneous coronary intervention (PCI) also varied tremendously across the country. The advent of registries had helped fine-tune some of these variations. CardioSource WorldNews caught up with Lucas Marzec, MD, a fellow in cardiac electrophysiology at the University of Colorado School of Medicine, to speak on the topic. This interview was conducted at AHA 2015 in Orlando.

CSWN: We’re starting to get more control in terms of who’s doing what. That’s what you were looking at?

Lucas Marzec: That’s correct. It’s well known that there’s variation and a number of strategies to improve outcomes related to cardiovascular disease. You mentioned PCI. Implantable defibrillators are another where there exists a lot of regional variation and hospital variation in terms of what proportion of patients are implanted with defibrillators. Before now, it was unknown whether or not the same variation existed with regard to cardiac resynchronization therapy (CRT).

There certainly are guidelines to guide CRT implantation and guidelines for ICD implantation. What did you analyze and what did you find?

We looked at the NCDR ICD Registry™ for patients who had a guideline-based indication for CRT either by the 2008 guidelines or the focused update, which was just published in 2012. We looked to see what proportion of those patients received CRT and whether or not the proportion of patients who did receive CRT varied by hospital. We found that there was a significant amount of variation in the percentage of patients receiving CRT by hospital. Anywhere from 100% of eligible patients were implanted with CRT, down to 23% at some hospitals.

If you’ve got a hospital serving a poor community that may not have the necessary insurance, you can understand some variations. Did you adjust for things like that to see if a hospital is essentially serving the same population as another hospital somewhere else?

We did. Insurance status is one factor that may play a role in whether or not a patient is implanted with a CRT device or any device, though those sorts of factors are often affect the decision of whether or not to implant any device, not necessarily a CRT device.

Other patient factors, may include race, patient sex, and other demographic variables. We adjusted for the identified patient provider and [any] hospital-level factors that we could within the NCDR. Even adjusting for those factors, there still remained a significant amount of hospital variation in the rate of CRT use.


That’s a good question. We don’t know. It may be related to technology. Certainly, electrophysiology (EP) or EP-trained providers had a higher rate of CRT use compared to those not trained in EP, and so in a center where there isn’t EP expertise, that may have played a role. But we found a number of factors are associated with the use of CRT, and so further work needs to be done to understand why that variation exists. We can then approve rates of CRT use in eligible patients who may benefit from the therapy.

There’s a new JACC EP journal. Maybe we should get more people to read that so that they understand more about what is supposed to be done.

Certainly I think publishing these findings is important for having a good understanding of defining the problem and use of CRT. In some centers, as few as 23% of people were receiving CRT devices. Having a better understanding of why that is happening is really important to improving the rate of use of the therapy.

In the past, there’s actually some indication that women do better with CRT. You adjusted for that in your analysis. Did you also look specifically at women and then compare based on gender? Are they getting as much CRT? Are they getting less?

As you mentioned, there’s some suggestion that women have a preferential benefit to CRT over men. We did look at patient sex as a risk factor or a predictor of CRT use, and contrary to what you would expect, we did not find a difference in the rate of CRT use between men and women. So that was not something that seemed to predict CRT use either at the patient level in our entire cohort or across hospitals.

On the one hand it’s good, because there’s equality. On the other hand, if there are data showing that maybe they do get some extra benefit, maybe those numbers could be improved a bit.

Correct, although it’s hard to know. We can’t say for certain. You might imagine the lack of a gender disparity in some cases may reflect poor practice in the sense that people who should get the therapy aren’t getting it. However, you want to see disparity, and we just don’t know enough, based on our analysis, to suggest what role gender is playing in the variation.

< Back to Listings