Discussion:
Our data demonstrates an increasing trend in pediatric ICD admissions over time. These findings are in line with reported increasing rates of pediatric ICD implantation over time . Cardiomyopathy was the most common underlying diagnosis in all the years, likely due to the need for hospitalization for other comorbidities and may reflect an increasing trend in ICD utilization for primary prevention in patients with cardiomyopathy. An underlying diagnosis of primary arrhythmia demonstrated a decreasing trend in the rate of hospitalizations after 2006. Although data is not available to assess reasons for this decline, we speculate that utilization of ICDs in primary arrhythmia patients may be decreasing with evolving practices; as the knowledge of inappropriate shocks, electrical storm, and alternative medical management strategies improves among primary arrhythmia patients, ICD implantations may be decreasing. For example, it is likely that more caution and consideration is being taken before prophylactic ICD implantation among patients with Long QT type 3 and Catecholaminergic Polymorphic Ventricular Tachycardia. Patients with primary arrhythmias are also less likely to have other comorbidities that may necessitate hospitalization.
Our data suggest that among hospitalizations for young patients with ICDs, those with an underlying diagnosis of cardiomyopathy or congenital heart disease have significantly higher odds of in-hospital death when compared to those with primary arrhythmia diagnosis. While our study was not powered to address the possible trend of increased mortality seen among non-Hispanic Black race this may warrant further investigation in the future.
We noted that a diagnosis of heart failure increased over time among ICD admissions suggesting the possibility of increased utilization of ICDs in heart failure patients. Prior studies have demonstrated increased in-hospital mortality among patients admitted with heart failure who have arrhythmias and Silka et al demonstrated abnormal ventricular function to be significantly correlated with mortality among patients with ICDs. Although exploratory, our multivariable analysis found that only a diagnosis of heart failure was associated with in-hospital death, increasing the odds of mortality among patients admitted with ICDs by ten-fold (p < 0.001), independent of underlying diagnosis. Of note, the mortality rate among ICD admissions with heart failure in our cohort was 3.7%, which is lower than the previously reported mortality rate of 7.3% in all heart-failure related hospitalizations. The KID database is not designed to identify reasons for the lower mortality but one might speculate that children with heart failure in whom an ICD is implanted may be of a different risk profile than those who do not have ICDs. For example, implantation of an ICD requires the ability for the patient to tolerate not only the procedure, but anesthesia and induction, and thus may be limited to patients with less procedural risks. In addition, predicted survival of less than 1 year is a relative contraindication for ICD implantation.
In this study, we were unable to determine the exact cause of death based on diagnostic codes; however, our data suggests that in-hospital deaths are less among patients with primary arrhythmia disorders and higher among those with cardiomyopathy and congenital heart disease with the driver of death mainly related to heart failure.