Non–Evidence-Based ICD Implantations in the United States

Based on clinical and randomized controlled trials, practice guidelines for implantable cardioverter-defibrillator (ICD) therapy for primary prevention were established in 2006 and 2008. A 40-day period following a myocardial infarction (MI) is mandated prior to ICD implantations, and ICD therapy is not recommended for patients with New York Heart Association (NYHA) class IV symptoms or for those recovering from coronary artery bypass graft (CABG) surgery. Primary prevention ICDs are recommended for patients with a low left ventricular ejection fraction (≤30% or ≥35%), making patients with newly diagnosed heart failure unlikely candidates for ICD therapy. Researchers recently conducted a retrospective cohort study to determine the prevalence of non–evidence-based ICD therapy, specifically for patients recovering from an acute MI or CABG surgery, those with NYHA class IV symptoms, or with newly diagnosed heart failure. Evidence-based patients are defined as those who do not fall into the 4 non–evidence-based groups. The results of the study were published in the Journal of the American Medical Association [2011;305(1):43-49]. The researchers compared the characteristics and in-hospital outcomes of non–evidence-based ICD patients with those who received an evidence-based ICD based on practice guidelines. They also considered the effects of site, physician specialty, and procedure year. Death, postprocedure complications, infection, and length of hospital stay were among examined in-hospital outcomes. Physician specialty categories included electrophysiologists, nonelectrophysiologist cardiologists, thoracic surgeons, and other (internists and surgeons). Data were gathered from the National Cardiovascular Data Registry’s (NCDR’s) ICD registry. After expanding coverage for ICD implantation in 2005, the Centers for Medicare & Medicaid Services mandated that data for all primary prevention ICD implants in Medicare beneficiaries be submitted to this national registry. Of 1448 participating hospitals, 78% report data on all ICD implants, not only for Medicare patients. Data were collected for adult patients (≥18 years) with cardiomyopathy who received primary prevention ICD implants between January 1, 2006, and June 30, 2009. Of 111,707 patients receiving ICD implants for primary prevention between the specified dates, 25,145 (22.5%) were for a non–evidence-based indication. The majority of patients in the 4 subgroups of non–evidence-based ICD implants were male and white, and the median age ranged from 64 to 68 years. Of the 25,145 patients receiving non–evidence-based implants, 15,604 patients were newly diagnosed with heart failure (62.1%), 9257 patients had experienced an MI within 40 days of the implant (26.8%), 3022 patients experienced NYHA class IV symptoms (12.0%), and 814 patients had undergone CABG surgery within 3 months of the implant (3.2%). Researchers reported that patients receiving a non–evidence-based implant were significantly older and had more comorbid disease than the patients who received an evidence-based ICD. Non–evidence-based implants also resulted in a significantly higher in-hospital death rate (0.57% [95% confidence interval (CI), 0.48%-0.66%] vs 0.18% [95% CI, 0.15%-0.2%]; P<.001). Postprocedure complications in patients receiving non–evidence-based ICD implants were significantly higher than in patients with evidence-based implants (3.23% [95% CI, 3.01%-3.45%] vs 2.41% [95% CI, 2.31%-2.51%]; P<.001). There was a wide variation in the use of non–evidence-based ICD therapy by site. At many sites, >40% of the ICD implants were non–evidence-based. There was a significantly lower rate of non–evidence-based ICD implants for electrophysiologists (20.8%) than nonelectrophysiologists (24.8%) for nonelectrophysiologist cardiologists; 36.1% for thoracic surgeons; and 24.9% for other specialties. Study limitations included narrowing consideration of complications to in-hospital events, the mandate for data submission of ICD implantation applying to Medicare patients only, and the possibility that non–evidence-based ICD therapy is underreported to assure that hospitals receive payment for the procedures. In conclusion, the researchers said that 22.5% of patients in NCDR’s ICD registry did not meet evidence-based criteria for ICD implantation. These patients demonstrated increased comorbidity and postprocedure complication and were at greater risk of in-hospital death.