Laurent Coudeville, PhD
Sanofi Pasteur
69367 Cedex 07, Lyon, France
Annelies Van Rie, MD, PhD
Department of Epidemiology
University of North Carolina
Chapel Hill, NC 27599
To the Editor.
In their recent article, Lee et al1 addressed the question of vaccinating adolescents and adults against pertussis. Their analysis favors vaccination of adolescents but not of adults, which is in accord with previous analyses of adolescent vaccination in the United States2,3 and Canada4 but at odds with a recent modeling exercise.3 This exercise, while acknowledging uncertainties in data used, reported break-even costs for adult vaccination in line with those for adolescents. We feel that the conclusions regarding adult vaccination by Lee et al are overstated.
We disagree with the authors' justifications for not including herd immunity in the base case. A major reason to vaccinate adults is precisely the possibility of herd immunity.5 Narrowing the assessment to direct benefits only results in an incomplete economic evaluation that fails to account for major potential heath benefits attributable to the strategy.6 This is not mitigated by the 1-way sensitivity analysis of the impact of routine adult vaccination on infant transmission. This aspect should not be relegated to a single point estimate (it is properly the base case), and sensitivity analyses should be conducted around it.
A second issue relates to the use of Massachusetts surveillance data, which provide a much higher pertussis incidence in adolescents (155 of 100 000) than adults (11 of 100 000), guaranteeing the economic results in favor of adolescent vaccination. Data from other states with enhanced surveillance7 and serological studies8 indicate much less of a difference between adolescents and adults. Although incidence was included in the sensitivity analyses, results for adult vaccination are not reported and the ratio of adult to adolescent incidence was not considered.
Another concern is that the utilities assigned for disease are very different for adolescents and adults. This bias (likely resulting from the use of parent proxy responders for adolescents) sways the results in favor of adolescent vaccination and was not tested in sensitivity analyses. Moreover, it is hard to believe that the average person would knowingly take a 5% to 8% risk of death to avoid a sore upper arm after vaccination. These values, which reflect problems with the utility methodology, are responsible for the "dominated" result of most of the strategies and further contribute to the very poor showing of the adult strategy.
Finally, and perhaps most importantly, performing comparison between adolescent and adult strategies as if they were mutually exclusive does not properly address the public health viewpoint. The issue is not whether to fund either adolescent or adult vaccination (you can do both); rather, the interest is in determining whether the benefits of these new immunization strategies warrant any additional investment required. To address this question, the analysis must consider the new strategies with respect to the current one, not to each other.
We believe the merits of targeted or routine adult immunization from a public health perspective remain unclear and demand more extensive and careful analysis.
REFERENCES
Related articles in Pediatrics:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||