Vaccine 30 (2012) 2795–2800 Contents lists available at SciVerse ScienceDirect Vaccine jou rn al h om epa ge: www.elsevier.com/locate/vaccine Estimating the age-specific duration of herpes zoster vaccine protection: A matter of model choice? Joke Bilcke a,∗ , Benson Ogunjimi a , Frank Hulstaert b , Pierre Van Damme c , Niel Hens a,d , Philippe Beutels a a Center for Health Economic Research and Modeling Infectious Diseases (CHERMID), Vaccine and Infectious Disease Institute (Vaxinfectio), University of Antwerp, Universiteitsplein 1, 2610 Wilrijk, Belgium b Belgian Health Care Knowledge Center (KCE), Administrative Centre Botanique, Doorbuilding (10th floor), Boulevard du Jardin Botanique 55, 1000 Brussels, Belgium c Center for the Evaluation of Vaccination (CEV), Vaccine and Infectious Disease Institute (Vaxinfectio), University of Antwerp, Universiteitsplein 1, 2610 Wilrijk, Belgium d Interuniversity Institute for Biostatistics and Statistical Bioinformatics (I-BioStat), Hasselt University, Agoralaan, 3590 Diepenbeek, Belgium a r t i c l e i n f o Article history: Received 14 June 2011 Received in revised form 16 September 2011 Accepted 20 September 2011 Available online 1 October 2011 Keywords: Zoster Shingles Waning Model uncertainty a b s t r a c t Introduction: The estimation of herpes zoster (HZ) vaccine efficacy by time since vaccination and age at vaccination is crucial to assess the effectiveness and cost-effectiveness of HZ vaccination. Published estimates for the duration of protection from the vaccine diverge substantially, although based on data from the same trial for a follow-up period of 5 years. Different models were used to obtain these estimates, but it is unclear which of these models is most appropriate (if any). Only one study estimated vaccine efficacy by age at vaccination and time since vaccination combined. Recently, data became available from the same trial for a follow-up period of 7 years. Aim and methods: We aim to elaborate on estimating HZ vaccine efficacy (1) by estimating it as a function of time since vaccination and age at vaccination, (2) by comparing the fits of a range of models, and (3) by fitting these models on data for a follow-up period of 5 and 7 years. Results: Although the models’ fit to data are very comparable, they differ substantially in how they estimate vaccine efficacy to change as a function of time since vaccination and age at vaccination. Discussion: An accurate estimation of HZ vaccine efficacy by time since vaccination and age at vaccination is hampered by the lack of insight in the biological processes underlying HZ vaccine protection, and by the fact that such data are currently not available in sufficient detail. Uncertainty about the choice of model to estimate this important parameter should be acknowledged in cost-effectiveness analyses. © 2011 Elsevier Ltd. All rights reserved. 1. Introduction Herpes zoster (HZ), also known as shingles or zona, is a viral dis- ease characterized by a painful dermatomal skin rash. It is caused by the varicella-zoster virus (VZV), which initially causes chicken- pox (varicella), but after recovery remains dormant within the body and may reactivate in later life causing HZ [1–3]. HZ occurs more commonly with increasing age and immunosuppression [4]. The duration and severity of HZ-related pain increases with age [5,6]. Recently, a vaccine has been shown to be efficacious in preventing Abbreviations: HZ, herpes zoster; VZV, varicella-zoster virus; AICc, corrected Akaike Information Criterion; SPS, Shingles Prevention Study; FDA, Food and Drug Administration. ∗ Corresponding author. Tel.: +32 3 265 28 95; fax: +32 3 265 27 52. E-mail addresses: joke.bilcke@ua.ac.be (J. Bilcke), benson.ogunjimi@ua.ac.be (B. Ogunjimi), Frank.Hulstaert@kce.fgov.be (F. Hulstaert), pierre.vandamme@ua.ac.be (P. Van Damme), niel.hens@uhasselt.be, niel.hens@ua.ac.be (N. Hens), philippe.beutels@ua.ac.be (P. Beutels). HZ (Zostavax ® ) [7]. Many countries are considering implementa- tion of HZ vaccination programs and therefore the effectiveness and cost-effectiveness of such programs is evaluated [8–17]. All of the ten currently published cost-effectiveness analyses of HZ vaccination have shown that the assumed duration of protection of the HZ vaccine determines to a large extent the cost-effectiveness of its implementation [8–17]. Most studies also found that the effec- tiveness and cost-effectiveness of HZ vaccination depends on the assumed age at which people are vaccinated [9–12,15,16]. Hence, the estimation of HZ vaccine efficacy by time since vaccination and age at vaccination is crucial to assess the cost-effectiveness of HZ vaccination. Indeed, four of the cost-effectiveness studies published esti- mates for the duration of protection from the HZ vaccine against the occurrence of HZ [9,13–15] (with two of them [9,13] referring to the same estimates). Surprisingly, these estimates range from 7.5 years [15] and 10 years [14], up to lifelong protection [9,13], although based on data from the same ‘Shingles Prevention Study’ (SPS) trial [7,18]. This trial followed subjects for up to 5 years after 0264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2011.09.079