Assessment of the survival benefits of listing advanced heart failure for heart transplantation through a monte carlo microsimulation model

SANTIAGO G. MIRIUKA; ROSANA POGGIO; CHRISTIAN CAROLI; MARIANO GIORGI; HERNÁN COHEN ARAZI

New Orleans

Congreso; 60th Annual Scientific Sessions of the American College of Cardiology; 2011

Objectives: To simulate a clinical trial where ambulatory patients with advanced heart failure (aHF) are randomized to heart transplant listing (HTL) or medical treatment (MT). Methods: A model was built using a Monte Carlo microsimulation model. Probabilities of events were obtained from databases, including ISHLT (period 2001-2006) and UNOS for heart transplant events and ISHLT and INTERMACS for assist ventricular devices. Heart failure mortality was taken from a recent publication about a cohort with aHF (J Am Coll Cardiol 2009;53:334). In the model, Markov nodes included all mayor possibilities, including admission, improvement, VAD implantation, heart transplantation (elective, urgent, from VAD and retransplantation). None of the probabilities were assumed. The model also included crossover from MT to HTL. Markov cycles were monthly, and total follow up was 60 months. Internal validity was assessed by tracking mortality in each group. The model was built in the Treeage 2009 software. Survival curves were analyzed by Cox regression model using STATA 11.0. Results: Survival benefit was found for HTL group. After 60 months, 32.2% of the patients in the HTL group died, compared to 42.3% in the MT group (Risk Difference: 10.1% (95%CI 8.77-11.43)). The survival curve displayed two periods: First, up to 24 month approximately, there was a non-significant difference between groups (Hazard Ratio (HR): 1.02 (95%CI: 0.96-1.08): p=0.46). After this time point, curves significantly diverge and a significant difference was reached by the end of the follow-up (HR: 0.47 (95%CI 0.53-0.5); p