Background: Herpes zoster is common and has serious consequences, notably post-herpetic neuralgia (PHN). Vaccine efficacy against incident zoster and PHN has been demonstrated in clinical trials, but effectiveness has not been studied in unselected general populations unrestricted by region, full health insurance coverage, or immune status. Our objective was to assess zoster vaccine effectiveness (VE) against incident zoster and PHN in a general population-based setting.
Methods and findings: A cohort study of 766,330 fully eligible individuals aged ≥ 65 years was undertaken in a 5% random sample of Medicare who received and did not receive zoster vaccination between 1st January 2007 and 31st December 2009. Incidence rates and hazard ratios for zoster and PHN were determined in vaccinated and unvaccinated individuals. Analyses were adjusted for age, gender, race, low income, immunosuppression, and important comorbidities associated with zoster, and then stratified by immunosuppression status. Adjusted hazard ratios were estimated using time-updated Cox proportional hazards models. Vaccine uptake was low (3.9%) particularly among black people (0.3%) and those with evidence of low income (0.6%). 13,112 US Medicare beneficiaries developed incident zoster; the overall zoster incidence rate was 10.0 (9.8-10.2) per 1,000 person-years in the unvaccinated group and 5.4 (95% CI 4.6-6.4) per 1,000 person-years in vaccinees, giving an adjusted VE against incident zoster of 0.48 (95% CI 0.39-0.56). In immunosuppressed individuals, VE against zoster was 0.37 (95% CI 0.06-0.58). VE against PHN was 0.59 (95% CI 0.21-0.79).
Conclusions: Vaccine uptake was low with variation in specific patient groups. In a general population cohort of older individuals, zoster vaccination was associated with reduction in incident zoster, including among those with immunosuppression. Importantly, this study demonstrates that zoster vaccination is associated with a reduction in PHN. Please see later in the article for the Editors' Summary.