Economic evaluations of varicella vaccination programmes: a review of the literature.

Chickenpox infections are generally mild but due to their very high incidence among healthy children they give rise to considerable morbidity and occasional mortality. With the development of a varicella vaccine in the early 1970s and its progressive licensing in many countries, interest in the efficiency of varicella immunisation programmes grew. The objective of this review was to discuss the methodological aspects and results of published economic evaluations of varicella vaccination. From this, we attempted to make recommendations. A computerised search was carried out; 17 full economic evaluations of varicella vaccination were retrieved. The review identified the methodological divergences and similarities between the articles in four areas: study design, epidemiological data, economic data and model characteristics. We assessed to what extent the applied methods conform to general guidelines for the economic evaluation of healthcare interventions and compared the studies' results. The desirability of a universal vaccination programme depends on whose perspective is taken. Despite variability in data and model assumptions, the studies suggest that universal vaccination of infants is attractive to society because large savings occur from averted unproductive days for parents. For the healthcare payer, universal vaccination of infants does not generate savings. Vaccination of susceptible adolescents has been proposed by some authors as a viable alternative; the attractiveness of this is highly dependent on the negative predictive value of anamnestic screening. Targeted vaccination of healthcare workers and immunocompromised individuals appears relatively cost effective. Findings for other target groups are either contradictory or provide insufficient evidence for any unequivocal recommendations to be made. High sensitivity to vaccine price was reported in most studies. This review highlights that some aspects of these studies need to be further improved before final recommendations can be made. First, more transparency, completeness and compliance to general methodological guidelines are required. Second, because of the increasing severity of varicella with age, it is preferable and in some cases essential to use dynamic models for the assessment of universal vaccination strategies. Third, most studies focused on the strategy of vaccinating children only while their results depended heavily on disputable assumptions (regarding vaccine effectiveness and impact on herpes zoster). Since violation of these assumptions could have important adverse public health effects, we suggest pre-adolescent vaccination as a more secure alternative. This option deserves more attention in future analyses.