The epidemiology and costs of implantable cardioverter–defibrillator therapy in Australia

The benefit of few medical interventions are as dramatic as the lifesaving effect of an ICD

Over the past three decades, implantable cardioverter–defibrillator (ICD) therapy has become the standard of care for preventing sudden cardiac death in a range of patients who are at particular risk. These patients can be broadly categorised into two groups: patients who have already been resuscitated from a sudden cardiac death event caused by a malignant ventricular arrhythmia (secondary prevention patients); and those who are deemed to be at increased risk because of the nature and severity of their cardiac diagnosis but have not had a malignant ventricular arrhythmia (primary prevention patients). Although new variants of the basic ICD platform, such as subcutaneous ICDs, are becoming available, the fundamental components of an ICD system remain constant: a lead for assessing cardiac rhythm, particularly the onset of a rapid malignant ventricular arrhythmia, and a pulse generator for rapidly delivering a direct current shock to terminate arrhythmia. In the appropriate setting, these devices can be live-saving, and the effects of few interventions in modern medicine are as compelling as the termination of ventricular fibrillation by an appropriate ICD shock.

The evidence base for ICD therapy in secondary prevention patients is robust, and patients are routinely offered ICD implantation after presenting with resuscitated cardiac arrest or haemodynamically significant ventricular arrhythmia in the absence of transient or reversible causes, such as acute myocardial infarction.1 The evidence base for primary prevention patients selected according to the presence of persistent and profound left ventricular systolic dysfunction is also solid; landmark primary prevention trials have identified a prognostic benefit for ICD therapy in patients with cardiomyopathy who have a left ventricular ejection fraction of 35% or less.2,3

However, risk stratification based solely on left ventricular systolic dysfunction is imprecise, and a significant proportion of patients receiving ICDs for primary prevention do not experience malignant ventricular arrhythmia during long term follow-up.4 Further, most sudden cardiac deaths are of patients with normal left ventricular ejection fractions. Overall, the positive predictive value ejection fraction for sudden cardiac death is only 37.9%, but it is currently the only criterion for selecting primary prevention patients.5

While the search for improved markers of sudden cardiac death risk continues, the uptake of ICD therapy for primary prevention has increased dramatically.6,7 Given the risks of late complications, including inappropriate shocks, lead failure, and sepsis, this increase has implications for patients; its costs mean there are also implications for society and health care budgets.

In this issue of the MJA, Blanch and colleagues report hospitalisation data for ICD procedures in Australia during 2002–2015.8 The analysed data, from the National Morbidity Database (NHMD), do not provide specific information about ICD procedures in octogenarians, which are increasingly common. The study results suggest that complication rates have declined, possibly because of the combined effects of advances in ICD technology and increased experience and surgical skill. However, the diagnosis-related groups complication code was changed in 2009 from comprising severe and catastrophic complications of ICD procedures to catastrophic complications alone, and this may explain some of the decline. ICD replacement, removal, and adjustment numbers also increased, reflecting increased ICD use across the period. Another finding — not exclusive to Australia6 — is that a clear majority of implants (79%) were for men.

As Blanch and her co-authors mention, there is evidence of both under- and overuse of ICD therapy in the population at risk, as well as of suboptimal targeting. While every patient who survives a cardiac arrest should be considered for ICD implantation, the indications are not always clear-cut for primary prevention ICD therapy. Decisions are still based on the two landmark trials published in 2002 (MADIT II)2 and 2005 (SCD-Heft).3 In the meantime there have been considerable advances in the medical management of heart failure, particularly of non-ischaemic cardiomyopathy. The results of the recent DANISH trial, which was adequately powered to detect changes in total mortality and included excellent background medical therapy (including cardiac resynchronisation), raised questions about the survival benefit of ICDs for patients with non-ischaemic cardiomyopathy over the age of 68 years.9

In the study by Blanch and colleagues, costs data were available for 2011–2014. There was a small decrease in the cost of each ICD procedure during this period, but the overall annual cost of hospitalisations increased by 7% because of rising procedure numbers, and was $445 644 566 over the 3 years. One of the major expenses is the ongoing cost of device follow-up.

As Blanch and her co-authors point out, the benefits and drawbacks of ICD implantation and — equally importantly — ICD replacement at battery depletion should be discussed with each patient and their family before making decisions. Comorbid conditions, such as renal failure, dementia and severe cardiac failure, may influence the decision to implant an ICD, particularly in older patients. Additionally, cardiac resynchronisation by biventricular pacing may be a preferable alternative for some patients.

Blanch and colleagues are to be congratulated for this diligent effort to ascertain the contemporary epidemiology and cost of ICD implantation in Australia. It is to be hoped that their report will stimulate further research into the current use and cost-effectiveness of this expensive but very effective therapeutic advance.