| Sumario: | The concepts of incremental specificity and incremental positive predictive accuracy (PPA) have been proposed to measure the success of dual chamber cardioverter defibrillator (ICD) algorithms for tachyarrhythmia detection in improving specificity while maintaining very high sensitivity to detection of episodes of ventricular tachycardia/fibrillation (VT/VF). While dual chamber VT/VF detection algorithms differ substantially among different ICD manufacturers, they all operate as 'add-on' features to the single chamber elementary detection algorithms that are based on simple criteria of increased ventricular rate. The incremental specificity and PPA characterize the performance of the dual chamber detection operation in this 'add-on' mode, that is within a database of rhythm episodes all meet the simple rate-based criteria. A statistical model of hypothetical devices has been used to demonstrate that the concepts of incremental specificity and PPA are very dependent on the composition of the database used to evaluate a particular dual chamber ICD. Because some sinus tachycardia and supraventricular tachyarrhythmias with regular atrioventricular conduction are more easily discriminated from true VT/VF than other supraventricular tachyarrhythmias, the model shows that rather than the performance of the dual chamber detection functions, the major contributor to the incremental specificity may be the proportion between the 'easy' and 'difficult' supraventricular episodes. The algorithms used by different ICD manufacturers to detect tachyarrhythmias based on ventricular rate are known to differ substantially in the ability to differentiate true VT/VF from other tachyarrhythmias. Consequently, the databases of rhythms against which the different dual camber ICDs are tested are also different in composition of different types of supraventricular tachyarrhythmias. Therefore, the values of incremental specificity and PPA reported by different manufacturers do not have an equivalent meaning and do not offer a valid comparison of the true performance of different dual chamber ICDs.
|
|---|