Sociodemographic predictors of variation in coverage of the national shingles vaccination programme in England, 2014/15.

INTRODUCTION: In September 2013, England introduced a shingles vaccination programme to reduce incidence and severity of shingles in the elderly. This study aims to assess variation in vaccine coverage with regards to selected sociodemographic factors to inform activities for improving equity of the programme.

METHODS: Eligible 70year-olds were identified from a national vaccine coverage dataset in 2014/15 that includes 95% of GPs in England. NHS England Local Team (LT) and index of multiple deprivation (IMD) scores were assigned to patients based on GP-postcode. Vaccine coverage (%) with 95% confidence intervals (CIs), were calculated overall and by LT, ethnicity and IMD, using binomial regression.

RESULTS: Of 502,058 eligible adults, 178,808 (35.6%) had ethnicity recorded. Crude vaccine coverage was 59.5% (95%CI: 59.3-59.7). Coverage was lowest in London (49.6% coverage, 95%CI: 49.0-50.2), and compared to this coverage was significantly higher in all other LTs (+6.3 to +10.4, p<0.001) after adjusting for ethnicity and IMD. Coverage decreased with increasing deprivation and was 8.2% lower in the most deprived (95%CI: 7.3-9.1) compared with the least deprived IMD quintile (64.1% coverage, 95%CI: 63.6-64.6), after adjustment for ethnicity and LT. Compared with White-British (60.7% coverage, 95%CI: 60.5-61.0), other ethnic groups had between 4.0% (Indian) and 21.8% (Mixed: White and Black African) lower coverage. After adjusting for IMD and LT, significantly lower coverage by ethnicity persisted in all groups, except in Mixed: Other, Indian and Bangladeshi compared with White-British.

CONCLUSIONS: After taking geography and deprivation into account, shingles vaccine coverage varied by ethnicity. White-British, Indian and Bangladeshi groups had highest coverage; Mixed: White and Black African, and Black-other ethnicities had the lowest. Patients' ethnicity and IMD are predictors of coverage which contribute to, but do not wholly account for, geographical variation coverage. Interventions to address service-related, sociodemographic and ethnic inequalities in shingles vaccine coverage are required.