Cost-effectiveness of implantable cardioverter-defibrillators

Modelling A Markov model was constructed to simulate the clinical and economic outcomes associated with ICD and control therapy. The model served also to extrapolate short-term trial-based results to a lifetime horizon. The cycle length was one month. Patients who received an ICD were at risk of death from the implantation procedure. Patients who did not die from the procedure and patients assigned to control therapy entered the Markov tree and were followed until death. A patient could die from arrhythmia, non-arrhythmic cardiac causes, or noncardiac causes. If the patient survived, he or she remained well for the 1-month period. Patients who had an ICD could have a lead infection or failure that could cause them to discontinue treatment.