Preventive use of implantable heart defibrillators in people with poor heart function

What was the aim of this review?

The aim of this Cochrane Review was to compare the benefits and harms of using versus not using implantable devices that restore heart rhythm (implantable cardioverter-defibrillators (ICDs)) in addition to medications in people with non-ischaemic heart disease.

Key message

The use of ICD in addition to medications in people with non-ischaemic heart disease decreases the rate of death compared to not using the device.

What was studied in the review?

Poor cardiac function decreases the capacity of the heart to pump enough blood to the various organ systems in the human body. This can occur due to poor blood supply to the heart muscles (ischaemic) or other causes (non-ischaemic). Either way, patients are at an increased risk of sudden death. The goal of treatment is to decrease this risk either through medications or through implantation of an ICD.

The available research shows that preventive use of ICD in the first category of patients (ischaemic) decreases the rate of death. In contrast, recent publication of a large trial showed no overall benefit in the second category of patients (non-ischaemic).

What are the main results of this review?

In this review, we synthesised the results of six major clinical trials that were identified through an extensive search to better understand the efficacy of ICDs in patients with non-ischaemic heart disease. Our analysis involved 3128 participants.

We found high-certainty evidence suggesting that the use of these devices (compared to no such use) in people with non-ischaemic heart disease decreases the risk of dying from any cause and of dying suddenly. There is moderate-certainty evidence suggesting that people on such devices probably have more complications than those not on these devices. The current evidence suggests that adding ICD therapy probably has little or no effect on quality of life when compared to not adding it, but electric shocks from ICDs worsen quality of life.

How up to date is this review?

We searched for studies that were available up to 10 October 2018.

Authors' conclusions: 

The use of ICD in addition to medical therapy in people with non-ischaemic cardiomyopathy decreases all-cause mortality and sudden cardiac deaths and probably decreases mortality from cardiovascular causes compared to medical therapy alone. Their use probably increases the risk for adverse events. However, these devices come at a high cost, and shocks from ICDs cause a deterioration in quality of life.

Read the full abstract...
Background: 

There is evidence that implantable cardioverter-defibrillator (ICD) for primary prevention in people with an ischaemic cardiomyopathy improves survival rate. The evidence supporting this intervention in people with non-ischaemic cardiomyopathy is not as definitive, with the recently published DANISH trial finding no improvement in survival rate. A systematic review of all eligible studies was needed to evaluate the benefits and harms of using ICDs for primary prevention in people with non-ischaemic cardiomyopathy.

Objectives: 

To evaluate the benefits and harms of using compared to not using ICD for primary prevention in people with non-ischaemic cardiomyopathy receiving optimal medical therapy.

Search strategy: 

We searched CENTRAL, MEDLINE, Embase, and the Web of Science Core Collection on 10 October 2018. For ongoing or unpublished clinical trials, we searched the US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov, the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP), and the ISRCTN registry. To identify economic evaluation studies, we conducted a separate search to 31 March 2015 of the NHS Economic Evaluation Database, and from March 2015 to October 2018 on MEDLINE and Embase.

Selection criteria: 

We included randomised controlled trials involving adults with chronic non-ischaemic cardiomyopathy due to a left ventricular systolic dysfunction with an ejection fraction of 35% or less (New York Heart Association (NYHA) type I-IV). Participants in the intervention arm should have received ICD in addition to optimal medical therapy, while those in the control arm received optimal medical therapy alone. We included studies with cardiac resynchronisation therapy when it was appropriately balanced in the experimental and control groups.

Data collection and analysis: 

The primary outcomes were all-cause mortality, cardiovascular mortality, sudden cardiac death, and adverse events associated with the intervention. The secondary outcomes were non-cardiovascular death, health-related quality of life, hospitalisation for heart failure, first ICD-related hospitalisation, and cost. We abstracted the log (hazard ratio) and its variance from trial reports for time-to-event survival data. We extracted the raw data necessary to calculate the risk ratio. We summarised data on quality of life and cost-effectiveness narratively. We assessed the certainty of evidence for all outcomes using GRADE.

Main results: 

We identified six eligible randomised trials with a total of 3128 participants. The use of ICD plus optimal medical therapy versus optimal medical therapy alone decreases the risk of all-cause mortality (hazard ratio (HR) 0.78, 95% confidence interval (CI) 0.66 to 0.92; participants = 3128; studies = 6; high-certainty evidence). An average of 24 patients need to be treated with ICD to prevent one additional death from any cause (number needed to treat for an additional beneficial outcome (NNTB) = 24). Individuals younger than 65 derive more benefit than individuals older than 65 (HR 0.51, 95% CI 0.29 to 0.91; participants = 348; studies = 1) (NNTB = 10). When added to medical therapy, ICDs probably decrease cardiovascular mortality compared to not adding them (risk ratio (RR) 0.75, 95% CI 0.46 to 1.21; participants = 1781; studies = 4; moderate-certainty evidence) (possibility of both plausible benefit and no effect). Implantable cardioverter-defibrillator was also found to decrease sudden cardiac deaths (HR 0.45, 95% CI 0.29 to 0.70; participants = 1677; studies = 3; high-certainty evidence). An average of 25 patients need to be treated with an ICD to prevent one additional sudden cardiac death (NNTB = 25). We found that ICDs probably increase adverse events (possibility of both plausible harm and benefit), but likely have little or no effect on non-cardiovascular mortality (RR 1.17, 95% CI 0.81 to 1.68; participants = 1781; studies = 4; moderate-certainty evidence) (possibility of both plausible benefit and no effect). Finally, using ICD therapy probably has little or no effect on quality of life, however shocks from the device cause a deterioration in quality of life. No study reported the outcome of first ICD-related hospitalisations.