Nearly one-fourth of ICD implants failed to meet implantation criteria

Al-Khatib S. JAMA. 2011;305:43-49.

Issue: February 2011

More than 22% of patients in a retrospective cohort study received a non-evidence-based implantation of an implantable cardioverter defibrillator, which greatly increased the likelihood of in-hospital death and post-procedural complications.

“Our findings suggest that one excess complication occurred for every 121 non-evidence-based ICD implantations,” the researchers wrote. “Although the absolute difference in complications between the two groups is modest, these complications could have significant effects on patients’ quality of life and health care use, including length of hospital stay and costs.”

The study, recently published in the Journal of the American Medical Association, featured 111,707 patients from the National Cardiovascular Data Registry-ICD Registry. Patients who met one or more of the following criteria were classified as having received a non-evidence-based ICD implantation: if they had NYHA Class IV symptoms, an MI within 40 days before ICD implantation, CABG surgery within 3 months before ICD implantation or newly diagnosed HF at the time of ICD implantation.

Overall, 25,145 patients (22.5%) received a non-evidence-based ICD implant. Compared with patients who received an evidence-based implant, those who did not had a higher risk of in-hospital death (0.57% vs. 0.18%; P<.001) and any post-procedural complication (3.23% vs. 2.41% P<.001). Among electrophysiologists, rates of non-evidence-based implantations were significantly lower when compared with non-electrophysiologists (20.8% vs. 24.8%; P<.001). Over time, researchers reported no clear decrease in non-evidence-based ICD implants. – by Brian Ellis

ICDs are among the most effective therapies we can offer to patients with heart disease. That said, we need to recognize that resources are not infinite and that we need to use expensive therapies, like ICDs, in appropriately chosen patients. The electrophysiology community has done an outstanding job in mounting randomized clinical trials to define groups that derive benefit and groups that don’t. It is therefore a disappointment and a bit of a surprise that almost one-quarter of ICDs in this large survey were placed in patients who did not meet evidence-based guidelines. I have little doubt that some of the non-indicated implants reflect emotional responses by physicians to sick patients, but that seems to me unlikely to explain the magnitude of the issue. I was most struck by the tremendous variability by site of the problem. That kind of variability may be a clue to how best to approach the problem: How do some institutions do so well and why do some do so poorly? The next debate will no doubt center on whether we can improve practice simply by better education or whether more draconian measures will be required.

– Dan Roden, MD
Cardiology Today Editorial Board member

Disclosure: Dr. Roden reports no relevant financial disclosures with device manufacturers. He has consulted for Merck, Novartis, Sanofi-Aventis and Astellas.

Follow CardiologyToday.com on Twitter.