Use of electronic health records to investigate vaccination inequalities in older individuals in England

Abstract

The aim of this thesis was to measure inequalities in the burden of vaccine-preventable disease and vaccine uptake amongst older individuals in England, using primary care electronic health records (EHR) linked to hospitalisation and deprivation data. Social factors previously associated with vaccination amongst older individuals in Europe were first determined by conducting a comprehensive systematic review. Methods were developed to identify and investigate the recording of these factors in the linked UK EHR data. These methods were then applied in two cohort studies, focussing on herpes zoster (a common debilitating condition in older populations), to identify the social determinants of (a) zoster incidence in the decade before zoster vaccine introduction, 2003-2013 and (b) uptake of zoster vaccination in 2013-15 (the first two years after vaccine introduction). The methodological study showed that, among 591,037 individuals aged ≥65 years, completeness of recording of individual social factors varied from 1.6-80%. The ethnic distribution, and prevalence of deprivation, living alone, living as a couple and care home residence, were all comparable with data from the 2011 English Census. In the first cohort study of 862,470 older individuals, those at higher risk of zoster in the prevaccination era included females, those in care homes, those of White ethnicity and nonimmigrants, with increased zoster incidence in these groups ranging from 10-100%. Known clinical risk factors for zoster (co-morbidities and immunosuppressive treatment) explained little of these increased risks. In the second cohort study of 35,333 individuals, social factors associated with lower uptake of zoster vaccination included: care home residence (adjusted odds ratio (aOR):0.64 (95% confidence interval: 0.57-0.73)), living alone (aOR:0.85 (0.81- 0.90)), and being of non-White ethnicity (for example: Black ethnicity versus White ethnicity: aOR:0.61 (0.49-0.75)). Uptake decreased by increasing deprivation: aOR (most deprived areas versus most affluent): 0.69 (0.64-0.75). Lower uptake was also seen amongst females in the older catch-up group. The findings from this thesis should help inform specific interventions to mitigate zoster vaccine inequalities, including amongst doubly disadvantaged groups (with higher zoster burden and lower vaccine uptake) such as care home residents. The methods developed 5 can also be used to examine other health inequalities in older UK populations. Future linkages to other data sources, such as the Census, would further enhance the availability of information for studies of the social determinants of health.