Analytical approach:
A deterministic multicycle discrete-time Markov model was used synthesise the evidence. Two hypothetical cohorts were analysed: all-age and 39 years or younger. The time horizon was the lifetime of the patient receiving the blood transfusion. The authors reported that a societal perspective was adopted.
Effectiveness data:
Clinical and effectiveness data were derived from the authors’ organisation (Blood Systems, which represents 8% of the USA blood supply) and previously published studies. The authors searched PubMed for relevant studies. The main measure of effectiveness was effectiveness adjustment factors (defined as the percentage of confirmed cases each strategy would interdict out of the total reported). This measure of effectiveness was derived from data from Blood Systems.
Monetary benefit and utility valuations:
Health state utilities were identified from previously published studies.
Measure of benefit:
Quality-adjusted life-years (QALYs) gained. These were discounted at an annual rate of 3%.
Cost data:
Direct costs were: antibody test (including labour), confirmatory tests, annual clinical workup (including visit, X-ray, ecochardiogram and laboratory tests), drug treatment and follow-up and treatment of cardiomyopathy, heart failure and surgery. Costs of screening were derived from data supplied by Blood Systems. Other costs were derived from previously published studies and the Arkansas 2007 Healthcare Cost and Utilization Project. Indirect costs were costs of early mortality due to heart failure or Chagas disease. These estimates were estimated using the friction-period method and assumed a six-month replacement period at average USA household income.
Analysis of uncertainty:
The authors reported a series of one-way, two-way and probabilistic sensitivity analyses. For the probabilistic sensitivity analysis, probability distributions were fitted alongside each model variable and 10,000 Monte Carlo simulations were performed. Results were presented using a cost-effectiveness acceptability curve.