A Markov model was developed and populated with efficacy and cost data from one centre. The model had an initial rapid phase, with a one-month cycle length, for two months, followed by a long-term phase, with a one-year cycle length, for over 30 years. The health states included ESHF, transplant, convalescence, and death. The non-parametric transition probabilities were based on patient-level data until the last point that these data were available, which was six years for the pre-ECMO group and two years for the post-ECMO group, and then they were extrapolated for another 30 years. The authors reported that the analysis was carried out from a health service perspective.
The authors stated that the effectiveness of ECMO bridging was based on a comparison of a cohort of children with ESHF treated in the era before ECMO (between 1994 and January 2001; n=34), who were offered conventional intensive care, and a cohort of children with ESHF treated after ECMO was introduced (after January 2001 until 2004; n=41). The study included children aged between 1 and 18 years with severe ESHF due to dilated cardiomyopathy, who were referred from throughout the United Kingdom to a tertiary paediatric cardiac transplant centre, between 1994 and 2004. Severe ESHF was defined as poor cardiac function, with a shortening fraction of less than 20%, New York Heart Association function class IV, and receipt of intravenous inotropic agents. Children were included even if they became unsuitable for transplant due to an evolving contraindication or recovery. They were excluded if they had less severe heart failure, if they had ESHF due to congenital heart disease (including postoperative patients), if they were younger than one year old, or if they had a known contraindication for transplant before admission. Their demographics (age, weight, highest inotrope score, mechanical ventilation, cardiac arrhythmias, and cardiac arrest) were compared, and the authors concluded that there was no evidence that the case-mix changed between the two eras. The mean follow-up was 4.39 years, with an interquartile range of 1.83 to 5.74 years.
Monetary benefit and utility valuations:
An expert panel consisting of two child psychologists and one specialist doctor estimated the utility values based on the Health Utilities Index 2.
Measure of benefit:
The primary measures of benefit were the quality-adjusted life-year (QALY) and the life-year gained (LYG). The discount rate was 3.5%.
The cost categories included ECMO support, ECMO assessment, assessment only, ICU stay, ward stay, and transport. The costs were based on a review of patients treated between 2004 and 2006. They were reported in UK pounds sterling (£) and discounted at a rate of 3.5%.
Analysis of uncertainty:
One-way sensitivity analyses were performed on QALY weights, costs, and the long-term survival rate for transplant recipients.