Schmitt et al1 report that preexposure anthrax vaccination of US Postal Service (USPS) employees is more costly and results in more anthrax infections and deaths than postattack intervention. We believe that these conclusions are incorrect owing to a flawed study design and several questionable parameter estimates.
The most important flaw is the assumption that unvaccinated workers would be denied antibiotics after an attack, especially when 50% adherence to preattack vaccination is presumed. If instead these workers were modeled to receive antibiotics, quality-adjusted life years (QALYs) gained preventing anthrax cases would exceed QALYs lost to vaccine adverse events. Based on the presented incidence, duration and quality-of-life effects of vaccine adverse effects, 0.0000148 QALYs are lost per vaccination series. With a 50% adherence assumption, 2.6 QALYs would be lost because of adverse events. If one assumes that preexposure vaccination is 92.5% effective and considering the data in Table 3 of the article by Schmitt et al,1 a strategy providing antibiotics to exposed but unvaccinated workers would reduce anthrax cases and deaths by 39% (2.4 and 0.7, respectively), providing a gain in QALYs far outweighing QALYs lost to vaccine adverse events. These improved health outcomes can only be assigned to the addition of preattack vaccination to a postattack intervention using antibiotics. Adding postexposure vaccination for these workers would reduce cases and deaths and further improve QALYs gained.
Hopkins RJ, Waytes TA, Zink TK. Preattack Vaccination Against Anthrax May Be Cost-effective in Certain Populations. Arch Intern Med. 2008;168(1):114-115. doi:10.1001/archinternmed.2007.27