Medical Device Daily Executive Editor

When you implant an implantable device and something goes wrong, what’s (who’s) to blame — the implanted device or the physician doing the implanting?

A new registry report by the Heart Rhythm Society (HRS; Washington), points to the second of these possibilities and, thus, the need for increased training in this sector.

First-year data from the National ICD registry, just published in HeartRhythm, the journal of HRS, reports few complications last year with implantable cardioverter defibrillators (ICDs), but that a number of implanting physicians had no formal training to do the procedure.

The first-year data showed that 15% of the 3,249 physicians lacked this training.

The National ICD Registry is being touted as the nation’s first comprehensive database of detailed information about patients implanted with the devices. It found that complications at the time of device implantation and prior to hospital discharge occurred at a 3.6% rate in the more than 108,000 ICD implantations at 1,117 hospitals. Adverse events included hematomas (1.2%), lead dislodgement (1%) and death (0.02%).

But then that report only points to procedural issues, not device failures and malfunctions, according to Stephen Hammill, MD, chair of the National ICD Registry steering committee and past president of HRS.

Hammill told Medical Device Daily, “All the stuff that we report at this time deals with the time of implantation, no follow-up.” And the reporting time period is “only a few days around hospitalization,” with those reports only of procedural complications.

“We don’t have any indication related to devices or [failures of] leads.”

He added that this follow-up data will be presented in future reports from the registries, perhaps as early as the American Heart Association’s (Dallas) annual meeting in November.

Hammill said there is no umbrella requirement for training of cardiologists implanting ICDs.

“[T]he filtering point,” he said, “is when the physician applies to the hospital credentialing committee to do a procedure, any kind of procedures.” Hence, approval to do the procedures is the responsibility of local credentialing committee “to determine qualifications.”

HRS, he said, has “tried to help the credentialing committees and developed guidelines to all of the hospitals to help give them guidance.”

Offering a hypothesis, Hammill said the reason for so many implants by those without any formal training may be that hospital credentialing committees have looked at the background of older docs, “with, I suspect, a significant amount of experience with pacemaker implantation, and they’ve moved over to ICD implantation.”

Hammill said that there are a variety of ways for surgeons to get this credentialing — though not in medical school — and that HRS in 2004 “put together a different pathway for those individuals so they could gain some type of credentials.”

Hammill said that the “second version” of the registry will look at follow-up data concerning implanted patients “and we will be combining registry data with Medicare claims data and the National Death Index. This information will include complications, such as infections, “further out in hospitalizations.”

Hammill said that by the spring of 2008, the registry will be able to report “immediate complications information that we can look at to tell us if there is any relationship between training or the volume of a particular physician and their specific outcomes with regard to complications.”

Ongoing are four research protocols, he said, that look at various issues that will be further extrapolated from the registry data — “gender, race, others with physician entering patients in the registry and their following recommended implantation guidelines” — with the expectation that some of this information will be reported at the November AHA meeting. “And we’ll get a much better feeling for how often those problems are occurring,” Hammill said.

Further out additional information will include data on the performance of ICDs produced by all of the companies that make these devices.

The registry data comes from all of the U.S. hospitals that do ICD implants, totaling 1,450, and representing the procedural experiences of “nearly” 3,900 physicians. HRS said.

As Hammill told MDD last year, when the registry was officially unveiled to attendees at the HRS annual meeting, that the hospitals have little choice but to provide the information since it is required as a qualification for receiving reimbursement from the Centers for Medicare & Medicaid Services (MDD, May 10, 2007).

There has been no pushback, Hammill said. “We have all the hospitals that are implanting ICDs participating in the registry.”

And he added that “a gratifying aspect” of the effort is that “most of the hospitals have elected to enter all their patients into the registry, Medicare and non-Medicare.”

“The nice aspect of the registry is that quarterly reports go back to every hospital to compare their outcomes with national averages, plus averages of hospitals of similar size.”

He said that this helps hospitals “see if they’re having more complications, longer hospital stays or other issues.” When a hospital sees data in an “outlier” category, it can then drill down. “They can look deeper to see if that’s related to a procedure room or a particular physician” or some other issue.

In a general assessment of the effort, Hammill said that it has the goal of being “a constructive way to give feedback to these hospitals and physicians, not point fingers.”

And he said that as this follow-up data is added to future registry reports, it will become even “more useful.”

The registry was initiated in 2004 with convening of a registry working group, CMS officially adopting the effort in 2005. It was then provided a sort of public unveiling last year at the HRS annual meeting in Denver.