Risk Estimation Following Infarction Noninvasive Evaluation ICD efficacy (REFINE ICD) Trial.

Each year 40,000 Canadians have a cardiac arrest. Death ensues within minutes if the heart rhythm is not normalized by a shock to reset the heart. Life-saving shocks can be delivered externally or via an implantable defibrillator or ICD; a small device implanted under the skin similar to a pacemaker. An ICD requires surgery, specialized follow-up care and is expensive. Hence, it is only used in patients groups where it has been shown to save lives in a cost-effective manner. There are 1.5 million Canadians with heart disease. Identifying those at highest risk of a cardiac arrest is challenging. At present patients who need an ICD are identified by the presence of very poor heart pump function. Yet, 75% of people who suffer a cardiac arrest have only slightly impaired heart pump function. Finding people in this group that might benefit from an ICD is challenging. We found that the combination of slightly impaired heart pump function plus abnormal heart electrical signals increases the risk of a cardiac arrest by 6-fold. While promising, this information is not sufficient to change how we treat patients. That requires clear evidence that an ICD will save lives. REFINE ICD is a large international trial of 1,400 patients with slightly impaired heart pump function plus abnormal heart electrical signals identified on a 24 hour ECG (Holter). Participants are randomly assigned to standard care or standard care plus an ICD. The majority of funding for the study is being provided through unrestricted grants from Medtronic Inc., an ICD manufacturer, GE Healthcare, a Holter manufacturer, and the Alberta government. Funds from the CIHR are being requested to support study management (University of Calgary) and data coordination (Montreal Heart Institute). This study is testing whether we can save lives after a heart attack. If this strategy is shown to be cost-effective we expect that it will change how we treat patients after a heart attack.