Implantable cardioverter-defibrillator FDA safety advisories: Impact on patient mortality and morbidity.

BACKGROUND: A significant proportion of implantable cardioverter-defibrillators (ICDs) have been subject to Food and Drug Administration (FDA) advisories. The impact of device advisories on mortality or patient care is poorly understood. Although estimated risks of ICD generators under advisory are low, dependency on ICD therapies to prevent sudden death justifies the assessment of long-term mortality. OBJECTIVE: To test the association of FDA advisory status with long-term mortality. METHODS: The study was a retrospective, single-center review of clinical outcomes, including device malfunctions, in patients from implantation to either explant or death. Patients with ICDs first implanted at Cleveland Clinic between August 1996 and May 2004 who became subject to FDA advisories on ICD generators were identified. Mortality was determined by using the Social Security Death Index. RESULTS: In 1644 consecutive patients receiving first ICD implants, 704 (43%) became subject to an FDA advisory, of which 172 (10.5%) were class I and 532 (32.3%) were class II. ICDs were explanted before advisory notifications in 14.0% of class I and 10.1% of class II advisories. Among ICDs under advisory, 28 (4.0%) advisory-related and 15 non-advisory- related malfunctions were documented. Over a median follow-up of 70 months, 814 patients died. Kaplan-Meier 5-year survival rate was 65.6% overall, and 64.2, 61.1, and 69.3% in patients with no, class I, and class II advisories, respectively (P = .17). CONCLUSIONS: ICD advisories impacted 43% of the patients. Advisory-related malfunctions affected 4% within the combined advisory group. Based on a conservative management strategy, ICDs under advisory were not associated with increased mortality over a background of significant disease-related mortality.

Chest radiography is a poor predictor of left ventricular lead position in patients undergoing cardiac resynchronization therapy: comparison with multidetector computed tomography.

BACKGROUND: Optimal left ventricular (LV) lead position is thought to be a major predictor of response in patients undergoing cardiac resynchronization therapy (CRT). While the post-implant posterior-anterior (PA) and lateral chest X-ray (CXR) is commonly used to determine the position of the LV lead, the accuracy to which the CXR can correctly localize the LV lead is unknown. METHODS: We collected data on 47 consecutive patients (mean age 64 years, 60% men and LV ejection fraction 23%, 49% ischemic cardiomyopathy) that underwent CRT between 2004 and 2007, who had both a post-implant CXR as well as a contrast-enhanced multi-detector computed tomography (MDCT) of the chest for any reason. The positions of the LV lead on CXR and MDCT were interpreted in a blinded fashion, independent of each other. The accuracy of the CXR in localizing various LV lead positions, with MDCT as the gold standard, was recorded. RESULTS: On CXR, the LV lead tip position was as follows: basal (4%), mid-ventricular (66%), and apical (30%) and anterior (2%), lateral (34%), and posterior (64%). On MDCT, the LV tip position was: basal (28%), mid-ventricular (60%), and apical (13%) and anterior (13%), lateral (19%), and posterior (68%). Compared to the MDCT gold standard, the percentage of LV lead positions the CXR correctly classified correctly were: 100% basal, 39% mid-ventricular, and 29% apical and 0% anterior, 12% lateral, and 77% posterior. Taking both PA and lateral views into consideration, the LV lead position was misclassified by CXR in 62% cases. CONCLUSION: Using MDCT as a gold standard, the routine post-implant CXR performs very modestly in terms of accurate LV lead positioning.

Meta-analysis of diagnostic efficacy of 64-slice computed tomography in the evaluation of coronary in-stent restenosis.

We sought to conduct a meta-analysis using available studies to determine the diagnostic efficacy of 64-slice computed tomography (CT) in evaluation of in-stent restenosis (ISR). Sixty-four-slice CT allows optimal noninvasive assessment of coronary artery disease. However, a variety of artifacts limit evaluation of stented coronary segments. We included studies that used 64-slice CT for evaluation of coronary ISR. We pooled efficacy estimates across studies using random-effects models. We identified 14 studies, which included 895 patients (1,447 stents, mean diameter 3.1 mm). Of these, 1,231 (91.4%) stents were adequately assessed by 64-slice CT. Overall sensitivity was 91% (95% confidence interval [CI] 86 to 94), specificity was 91% (95% CI 89 to 92), positive predictive value (PPV) was 68% (95% CI 63 to 73), and negative predictive value (NPV) was 98% (95% CI 97 to 99). The summary receiver operating characteristic curves graph showed a symmetric area under the curve of 0.96. When nonassessable segments were included, overall sensitivity and specificity decreased to 87% (95% CI 81 to 92) and 84% (95% CI 82 to 87), with a PPV of 53% (95% CI 47 to 59) and an NPV of 97% (95% CI 96 to 98), respectively. In conclusion, 64-slice CT detects (high sensitivity and specificity) or excludes ISR (high NPV) with a high degree of confidence; however, precise quantification of ISR is not accurate (low PPV). Efficacy estimates are even lower when nonassessable segments are included. Hence, at this point, stress imaging remains the most acceptable noninvasive technique for diagnosis of ISR.