Tuberculosis in the United Kingdom: epidemiology, immigration and control

Tuberculosis (TB) continues to be a public health concern in high-income countries, such as the UK, with disease notifications primarily, and disproportionately, occurring in foreign-born immigrants who have a higher TB incidence than the local-born population. Underlying this epidemiology is the synergy of migration from high TB burden regions and the reactivation of imported latent TB infection (LTBI) in the initial years after migration which has refocused attention on immigrant screening for TB. In addition, amongst foreign-born individuals, TB disease primarily manifests as an extrapulmonary phenotype although the relative importance of demographic, host, mycobacterial lineage and environmental factors remains unclear. International, and national (UK-specific), surveys of the current systems in place for immigrant screening were undertaken. Using empirical UK immigrant screening data obtained from a multi-centre assessment of single-step IGRA testing and a prospective, community-based evaluation of different screening modalities, estimates for LTBI prevalence, yields for screening and the relative performance of different screening methods were computed. Decision-analysis models were constructed, with the data obtained from these empirical studies, to assess the cost-effectiveness of screening for LTBI at different incidence thresholds with different screening tools with and without port-of-arrival chest radiography to enable specific policy recommendations to be made. To assess demographic, mycobacterial lineage and environmental factors associated with specific clinical phenotypes, data from three independent Northern hemisphere TB databases (Birmingham, UK; US CDC and London, UK) were analysed. High-income countries prioritised screening for active TB but LTBI screening, paradoxically, was limited and those countries that did screen for LTBI varied significantly in whom they targeted and how they screened. Similar patterns were seen in the UK with heterogeneity particularly evident in the incidence threshold at which screening was instigated and the screening tools used. Moreover, screening for LTBI was inversely related to TB burden in the responding area. Data from the empirical studies of immigrant screening revealed that LTBI prevalence varied from 17-30% (depending on screening tool) and that prevalence was independently associated with increasing TB incidence in country of origin. The proportion of latently infected individuals identified (with LTBI) decreased as the TB incidence screening threshold increased. Health economic analyses revealed that the most cost-effective screening model would be to dispense with port-of-arrival chest radiographic screening and screen for LTBI at an intermediate screening threshold (250 or 150/100000) using single-step IGRA testing (with the QuantiFERON Gold in-tube). Data on the association of demographic and mycobacterial lineage factors revealed that, on multivariate analysis, extrapulmonary disease was most strongly associated (and more strongly than lineage) with non-White ethnicity. Mycobacterial lineages were geo-ethnically restricted and certain lineages (Euro-American and East African Indian) were preferentially associated with pulmonary disease whilst others, such as the East Asian lineage, were associated with extrapulmonary disease. In a separate assessment of demographic and environmental factors, local-born Whites had a significantly lower proportion of extrapulmonary disease compared to local-born Non-Whites; both groups had a significantly lower proportion of extrapulmonary disease as compared to foreign-born Non-Whites. In foreign-born Non-Whites, the proportion of cases that were extrapulmonary significantly increased as time elapsed between migration and the development of active TB before stabilising. In the US dataset, states with higher levels of sunshine had significantly lower levels of extrapulmonary TB. In the London cohort, individuals with purely extrapulmonary tuberculosis (14.1 nmol/L) had significantly lower mean vitamin D levels compared to individuals with pulmonary tuberculosis only (25.8 nmol/L)(p=0.0003). Severe vitamin D deficiency (