Analytical approach:
A decision analytic model was populated with data from the literature to compare the cost-effectiveness of the two vaccines. The time horizon was 180 days and the authors stated that a societal perspective was adopted.
Effectiveness data:
The effectiveness data were mainly derived from a published, randomised, double-blind, multinational study, which was a head-to-head comparison of the two vaccines in children aged under five years. Further data were derived from other published literature. The primary clinical parameters were the probability of uncomplicated and complicated influenza and vaccine-associated adverse events.
Monetary benefit and utility valuations:
The utility values for the different health states were obtained from published studies. The methods used to derive them were not reported.
Measure of benefit:
The authors used quality-adjusted life-years (QALYs) as the primary measure of benefit. The cases of influenza, acute otitis media, and lower respiratory infections, hospitalisations, emergency room visits, and out-patient physician visits avoided were secondary measures of benefit. The remaining life-years were appropriately discounted at an annual rate of 3%.
Cost data:
The economic analysis included the costs of vaccine acquisition, administration, and adverse events; hospitalisation and out-patient care due to influenza or vaccine-related adverse events; drugs due to influenza; and transportation. The indirect costs, such as caregiver absenteeism from work and usual activities due to vaccination for influenza, were also included. The resource use data were based on the principal study. The costs were either derived from official national sources or from the published literature. They were reported in US dollars ($) and appropriately adjusted and reported for the price year 2006. Some assumptions were required and were reported in detail.
Analysis of uncertainty:
The parameter uncertainty was investigated through one-way sensitivity analysis on all the model parameters. A scenario allowing secondary wild-type influenza virus transmission among household members was investigated. Probabilistic sensitivity analysis, using Monte Carlo simulation, was used to investigate the variation in clinical outcomes and resource utilisation on the results. The method used to derive the parameter distributions was only briefly reported.
