Follow-up Duration, Competing Risk and Lifespan-Gain From ICDs
Follow-up Duration, Competing Risk and Lifespan-Gain From ICDs
Background In at-risk patients with left ventricular dysfunction, implantable cardioverter defibrillators (ICDs) prolong life. Implantable cardioverter defibrillators are increasingly implanted for primary prevention and therefore into lower risk patients. Trial data have demonstrated the benefit of these devices but does not provide an estimate of potential lifespan-gain over longer time periods, e.g. a patient's lifespan.
Methods Using data from landmark ICD trials, lifespan-gain was plotted against baseline annual mortality in the individual trials. Lifespan-gain was then extrapolated to a time-horizon of >20 years while adjusting for increasing 'competing' risk from ageing and non-sudden cardiac death (pump failure).
Results At 3 years, directly observed lifespan-gain was strongly dependent on baseline event rate (r = 0.94, P < 0.001). However, projecting beyond the duration of the trial, lifespan-gain increases rapidly and non-linearly with time. At 3 years, it averages 1.7 months, but by 10 years up to 9-fold more. Lifespan-gain over time horizons >20 years were greatest in lower risk patients (~5 life-years for 5% baseline mortality, ~2 life-years for 15% baseline mortality). Increased competing risks significantly reduce lifespan-gain from ICD implantation.
Conclusion While high-risk patients may show the greatest short-term gain, the dramatic growth of lifespan-gain over time means that it is the lower risk patients, e.g. primary prevention ICD implantation, who gain the most life-years over their lifetime. Benefit is underestimated when only trial data are assessed as trials can only maintain randomization over limited periods. Lifespan-gain may be further increased through advances in ICD device programming.
Implantable cardioverter defibrillators (ICDs) reduce the risk of sudden arrhythmic death in clinical trials of patients with chronic heart failure who have, or have not already experienced a life-threatening arrhythmia. Despite this, many patients who meet the guidelines for implantation do not receive an ICD, and conversely ICD implantation often falls outside evidence-based recommendations. Early studies of ICDs focused on patients at high risk of sudden cardiac death with a high event rate, so that relatively short follow-up could demonstrate a significant mortality effect. However, >70% of ICDs are now implanted for primary prevention and these patients are often at the lower end of the risk spectrum.
High-risk patients obtain the clearest early benefit, but as many patients with heart failure die for reasons other than arrhythmia, e.g. non-cardiac comorbidities, the average lifespan-gain may not be large. Conversely, lower risk patients may have a low rate of non-arrhythmic death and therefore much longer exposure to the risk of arrhythmia, albeit it at a lower annual rate, and so the extent of survival benefit may not be obvious during the relatively short time-horizon of most trials. It is therefore important to consider lifespan-gain from ICD implantation across the spectrum of risk and for longer duration of follow-up than assessed through trials.
Trials have focussed on recruiting patients with dominant cardiac disease and excluded patients with severe co-morbidities. However, in clinical practice, patients with advanced heart failure often suffer from multiple co-morbidities and are at increased risk of non-sudden cardiac death (pump failure) which may discourage ICD implantation.
Lifespan-gain from ICD implantation was assessed using data from the landmark trials. As these necessarily had relatively short durations of follow-up, we modelled survival in the post-trial period using a Gompertz–Makeham model. We assessed (i) the effect of baseline mortality (ii) the interplay of sudden and non-sudden cardiac death and (iii) competing risk from non-cardiac causes on the lifespan-gain from ICD implantation.
Abstract and Introduction
Abstract
Background In at-risk patients with left ventricular dysfunction, implantable cardioverter defibrillators (ICDs) prolong life. Implantable cardioverter defibrillators are increasingly implanted for primary prevention and therefore into lower risk patients. Trial data have demonstrated the benefit of these devices but does not provide an estimate of potential lifespan-gain over longer time periods, e.g. a patient's lifespan.
Methods Using data from landmark ICD trials, lifespan-gain was plotted against baseline annual mortality in the individual trials. Lifespan-gain was then extrapolated to a time-horizon of >20 years while adjusting for increasing 'competing' risk from ageing and non-sudden cardiac death (pump failure).
Results At 3 years, directly observed lifespan-gain was strongly dependent on baseline event rate (r = 0.94, P < 0.001). However, projecting beyond the duration of the trial, lifespan-gain increases rapidly and non-linearly with time. At 3 years, it averages 1.7 months, but by 10 years up to 9-fold more. Lifespan-gain over time horizons >20 years were greatest in lower risk patients (~5 life-years for 5% baseline mortality, ~2 life-years for 15% baseline mortality). Increased competing risks significantly reduce lifespan-gain from ICD implantation.
Conclusion While high-risk patients may show the greatest short-term gain, the dramatic growth of lifespan-gain over time means that it is the lower risk patients, e.g. primary prevention ICD implantation, who gain the most life-years over their lifetime. Benefit is underestimated when only trial data are assessed as trials can only maintain randomization over limited periods. Lifespan-gain may be further increased through advances in ICD device programming.
Introduction
Implantable cardioverter defibrillators (ICDs) reduce the risk of sudden arrhythmic death in clinical trials of patients with chronic heart failure who have, or have not already experienced a life-threatening arrhythmia. Despite this, many patients who meet the guidelines for implantation do not receive an ICD, and conversely ICD implantation often falls outside evidence-based recommendations. Early studies of ICDs focused on patients at high risk of sudden cardiac death with a high event rate, so that relatively short follow-up could demonstrate a significant mortality effect. However, >70% of ICDs are now implanted for primary prevention and these patients are often at the lower end of the risk spectrum.
High-risk patients obtain the clearest early benefit, but as many patients with heart failure die for reasons other than arrhythmia, e.g. non-cardiac comorbidities, the average lifespan-gain may not be large. Conversely, lower risk patients may have a low rate of non-arrhythmic death and therefore much longer exposure to the risk of arrhythmia, albeit it at a lower annual rate, and so the extent of survival benefit may not be obvious during the relatively short time-horizon of most trials. It is therefore important to consider lifespan-gain from ICD implantation across the spectrum of risk and for longer duration of follow-up than assessed through trials.
Trials have focussed on recruiting patients with dominant cardiac disease and excluded patients with severe co-morbidities. However, in clinical practice, patients with advanced heart failure often suffer from multiple co-morbidities and are at increased risk of non-sudden cardiac death (pump failure) which may discourage ICD implantation.
Lifespan-gain from ICD implantation was assessed using data from the landmark trials. As these necessarily had relatively short durations of follow-up, we modelled survival in the post-trial period using a Gompertz–Makeham model. We assessed (i) the effect of baseline mortality (ii) the interplay of sudden and non-sudden cardiac death and (iii) competing risk from non-cardiac causes on the lifespan-gain from ICD implantation.
Source...