Cost effectiveness of a targeted age-based West Nile virus vaccination program. Issue 23 (25th May 2017)

• Record Type: Journal Article
• Authors: Shankar, Manjunath B.; Staples, J. Erin; Meltzer, Martin I.; Fischer, Marc
• Is Part Of: Vaccine. Volume 35:Issue 23(2017)
• Page Start: 3143
• Page End: 3151
• Publication Date: 2017-05-25
• Subjects: West Nile virus -- Vaccines: cost-effectiveness -- Markov model -- Monte Carlo method -- Age-based targeting; Vaccines -- Periodicals; 615.372
• DOI: 10.1016/j.vaccine.2016.11.078
• Languages: English
• ISSNs: 0264-410X

Abstract

Highlights: We assessed the cost effectiveness of age based targeting of WNV Vaccination. We compared costs and health outcomes of vaccination versus no vaccination. Seven cohorts, spaced at 10 year intervals, were each followed until 90-years-old. Age-based targeting is more cost-effective than universal vaccination. Most important inputs are WNV Disease incidence, disability weights & vaccine cost. Abstract: Background: West Nile virus (WNV) is the leading cause of domestically-acquired arboviral disease in the United States. Several WNV vaccines are in various stages of development. We estimate the cost-effectiveness of WNV vaccination programs targeting groups at increased risk for severe WNV disease. Methods: We used a mathematical model to estimate costs and health outcomes of vaccination with WNV vaccine compared to no vaccination among seven cohorts, spaced at 10 year intervals from ages 10 to 70 years, each followed until 90-years-old. U.S. surveillance data were used to estimate WNV neuroinvasive disease incidence. Data for WNV seroprevalence, acute and long-term care costs of WNV disease patients, quality-adjusted life-years (QALYs), and vaccine characteristics were obtained from published reports. We assumed vaccine efficacy to either last lifelong or for 10 years with booster doses given every 10 years. Results: There was a statistically significant difference in cost-effectiveness ratios across cohorts in both models and all outcomes assessed (Kruskal-Wallis test p < 0.0001). The 60-year-cohort had a mean cost per neuroinvasive disease case prevented of $664, 000 and disability averted of $1, 421, 000 in lifelong model and $882, 000 and $1, 887, 000, respectively in 10-year immunity model; these costs were statistically significantly lower than costs for other cohorts (p < 0.0001). Vaccinating 70-year-olds had the lowest cost per death averted in both models at around $4.7 million (95%CI $2–$8 million). Cost per disease case averted was lowest among 40- and 50-year-old cohorts and cost per QALY saved lowest among 60-year cohorts in lifelong immunity model. The models were most sensitive to disease incidence, vaccine cost, and proportion of persons developing disease among infected. Conclusions: Age-based WNV vaccination program targeting those at higher risk for severe disease is more cost-effective than universal vaccination. Annual variation in WNV disease incidence, QALY weights, and vaccine costs impact the cost effectiveness ratios.