The ICD: Detection, Discrimination, and Therapy

The ICD divides heart rates into programmable detection zones. Each zone triggers a different level of response.

The VF zone sits at the top, typically above 200 bpm. Any rate in this range triggers immediate high-energy shock without delay. The assumption is straightforward: at rates this fast, the rhythm is almost certainly ventricular fibrillation, and every second counts.

The VT zone covers an intermediate range, typically 150 to 200 bpm. Rates in this window may be VT, but they could also be SVT or sinus tachycardia during exercise. The device allows more time here for discrimination algorithms to analyze the rhythm before committing to therapy.

An optional monitor zone can be programmed below the VT zone. The device detects and logs episodes in this range but delivers no therapy. This is useful for surveillance of slower VTs that may be hemodynamically tolerated, providing the clinical team with data without exposing the patient to unnecessary treatment.

Detection requires more than a single fast beat. The device counts the number of intervals that fall within a given zone and requires a programmable threshold before declaring an episode.

For VF, this threshold is short: typically 30 of 40 consecutive intervals (written 30/40). The goal is rapid detection with minimal delay. A few non-VF intervals mixed in will not prevent detection, because the counter uses an X-of-Y scheme that tolerates occasional slow beats in an otherwise fast rhythm.

For VT, the Number of Intervals to Detect (NID) can be programmed longer. Setting a higher NID gives self-terminating episodes time to stop on their own. Many nonsustained VT runs terminate within 10 to 15 seconds. A longer detection window avoids treating rhythms that were never going to cause harm.

The ADVANCE III and PROVIDE trials demonstrated that longer detection intervals significantly reduce the number of ICD therapies delivered without compromising safety. Programming an NID of 30/40 for the VT zone, compared to the older standard of 18/24, cut therapy delivery by roughly 30% in these studies.

Once the rate falls into the VT zone, the device runs a series of discrimination algorithms to separate true VT from SVT. These work in combination, each addressing a different aspect of the rhythm.

Rate branch compares atrial and ventricular rates. If the ventricular rate exceeds the atrial rate, the rhythm is VT by definition. In a dual-chamber device, the atrial lead provides this information directly. This single criterion, when satisfied, overrides all other discriminators.

Onset evaluates how the tachycardia began. A sudden jump from a normal rate into the detection zone favors VT. Gradual acceleration, the kind seen when a person starts running, favors sinus tachycardia. This discriminator is effective at withholding therapy during exercise but can be fooled by a VT that starts during exertion.

Stability examines the regularity of R-R intervals. Irregular intervals suggest atrial fibrillation with a rapid ventricular response. Regular intervals favor organized monomorphic VT.

QRS morphology compares the tachycardia QRS to a stored template of the patient's normal sinus QRS. If the tachycardia complex matches the sinus template, the rhythm is likely SVT conducted with the same ventricular activation sequence. If the morphology differs substantially, VT is more likely.

For monomorphic VT, the device can attempt painless termination before resorting to a shock. Anti-tachycardia pacing (ATP) delivers a rapid burst of 8 to 10 pacing stimuli at 80 to 90% of the tachycardia cycle length.

The logic follows from the mechanism of reentry. The circuit that sustains monomorphic VT has an excitable gap: a window of recovered tissue between the tail of one wavefront and the head of the next. A pacing burst enters this gap, captures the circuit, and terminates the reentry when the paced wavefront collides with the circulating wavefront.

For slower VTs (cycle length above 300ms), ATP terminates the rhythm in 80 to 90% of cases. It is painless. The patient often does not even notice.

Current programming standards deliver ATP before shock in the VT zone. Some devices also deliver ATP during capacitor charging, making a final attempt at painless termination while the capacitor prepares for shock. If the VT terminates during charging, the shock is aborted. This strategy further reduces the number of shocks patients receive.

When ATP fails or the rhythm is VF, the device delivers a defibrillation shock. Modern ICDs use a biphasic waveform delivered between the RV coil electrode and the pulse generator housing, which serves as the return electrode. Some systems include a second coil in the superior vena cava to shape the shock vector across more myocardium.

The device delivers between 20 and 40 joules, depending on programming. At implant, defibrillation threshold (DFT) testing was historically performed: VF was induced, and the device was allowed to detect and shock. Current practice often omits DFT testing in straightforward cases, because modern biphasic waveforms and active-can configurations have improved shock efficacy to the point where failure is rare.

The shock terminates VF by simultaneously depolarizing a critical mass of myocardium. When enough tissue is depolarized at once, all reentrant wavelets lose their excitable substrate and extinguish. The sinus node then resumes control of the rhythm.

Inappropriate shocks are the most common adverse event in ICD patients. A shock delivered for a rhythm that does not require therapy causes pain, psychological distress, and can provoke new arrhythmias.

The most frequent cause is SVT, especially atrial fibrillation with a rapid ventricular response, detected in the VT or VF zone. Careful programming of SVT discriminators and higher rate cutoffs reduces this risk substantially.

T-wave oversensing occurs when the device mistakes the T wave for a second R wave, effectively doubling the detected rate. The real ventricular rate may be 80 bpm, but the device counts 160 bpm and enters the VT zone. This is more common in patients with large T waves or reduced R-wave amplitude. Adjusting the sensitivity or decay settings corrects the problem.

Lead fracture produces high-frequency electrical noise that the device interprets as VF. The telltale sign is extremely short, irregular intervals (below 120ms) with abrupt onset. Lead impedance will be abnormally high.

Electromagnetic interference from external sources can also trigger false detections. Proper device programming, routine follow-up, and remote monitoring catch most of these issues before they result in inappropriate therapy.