No doubt they are one of the miracle tools of modern medicine: A metal case barely the size of a match box supervising the rhythm of the heart by tiny electrodes. In case of life threatening cardiac arrhythmias they trigger electroshocks which might save the life of the patient.
Something being theoretically a convincing concept was tested thoroughly in clinical studies: Today ICD devices are well established in primary prevention of ventricular fibrillation. From study to study the criteria for an implantation have been expanded further. The more than 6 years old MADIT II-study still serves as a rule: A relative decrease of 31 percent of the total mortality and an absolute risk reduction of 5.6 percent were the results in patients after a myocardial infarction with an ejection traction of a maximum of 30 percent after an observation time of about 20 months. Accordingly 18 ICD save one life in this collective.
Many ICD candidates are actually too sick
So far, so undisputed. Admittedly there are discussions time and again about the very broad MADIT-II criteria. Not just a few plead for a broadening of the QRS complex in the ECG as an additional criterion. Then the "number needed to treat" would turn out considerably more favourable, thus shown in subgroup analyses of the MADIT-II study. But on the whole the MADIT-II criteria are widely accepted. Nonetheless, cardiologists of the St. Louis University spoke up in the June edition of the American Heart Journal with an extremely critical article. The internist Professor Paul Hauptmann and his team of physicians took a close look at clinical reality of implantation of ICD devices and the devices for cardiac resynchronisation ("biventricular pacemakers") mainly combined with the ICD.
Their conclusion: Way too often, ICD devices and biventricular pacemakers are implanted in patients who, due to their bad general condition, do not live long enough to profit from those devices to an extent worth mentioning. "Originally those implants were not planned for very sick patients hospitalized with a heavy cardiac insufficiency – not to mention that they were tested for those patients in any studies", explains Hauptmann. According to him exactly this is happening in clinical reality: Patients with a low life expectancy receive an ICD or a resynchronization device because they formally meet the criterion of a low cardiac pump function after myocardial infarction.
Inotropic drugs as a predictor for an early death?
Methodically, the US physicians proceeded as follows: They used a national hospital database as a basis for their interpretation. In 240 hospitals they identified about 28,000 patients with an ICD or a biventricular pacemaker due to chronic cardiac insufficiency in the years 2004 and 2005. Then they checked how many of those patients already died in hospital and what costs incurred. The cost effectiveness of implants in patients dying early did not exactly come as a surprise, which was not half as effective as for those dying at a later point of time.
What the physicians actually were interested in were the predictive factors for an early death to enable them in the future to easier identify those patients really profiting from an implant because they will live long enough.
Here the use of positive inotropic substances has a high predictive value: Every fourth in-patient who received positive inotropic substances, died there at the hospital. According to the opinion of the authors, a therapy with inotropic drugs therefore should prompt a reconsideration of an ICD indication. "It is clear that patients needing or most likely requiring an inotropic therapy have a rather low chance profiting from a cardiac implant. The guide lines of ICD implantation emphasize as well that the life expectancy should at least be one more year", says Hauptmann. The expert therefore pleads for listing criteria speaking for a lower life expectancy in the ICD guide lines.