Introduction The risk of active tuberculosis (TB) disease is estimated to be increased 40-fold in people with HIV (PHIV). Effective antiretroviral treatment (ART) may reduce this significantly. UK national guidance recommends using blood interferon gamma release assay (IGRA) +/- tuberculin skin testing (TST) for latent TB (LTBI) diagnosis but there are little supporting health economic data. We sought to evaluate the cost-effectiveness of different testing strategies using data from a prospective contemporary cohort.
Methods Subjects receiving ambulatory HIV care were recruited by stratified selection within our HIV centre. TST, IGRA (TSpot. TB), frontal chest radiograph (CXR) and single induced sputum for mycobacterial culture were performed. The yield was used to model a screening programme that utilised current UK HIV demographics (Public Health England). Costs were based on the BNF, local costs or published literature (TST £16, IGRA £60, CXR £50, induced sputum £42, treatment for latent and active TB £786 and £7619 respectively). Uptake and LTBI treatment efficacy were both estimated at 65%. We assumed a lifetime reactivation rate with TST+/IGRA+ of 10% and TST+/IGRA- of 2%; and that all those with evidence of LTBI would be given treatment.
Results Over 13 months, 211 people were recruited. 26% were female and 26% black African. LTBI rates amongst subjects from sub-Saharan Africa, medium and low TB incidence countries were 8/55 (15%), 2/31 (6%) and 4/125 (3%) respectively. One patient had a persistently indeterminate IGRA. Subclinical TB disease was diagnosed in two (1%) subjects.
Using these data to model TB testing nationally, the British HIV Association (BHIVA) testing algorithm was the most cost-effective with an incremental cost effectiveness ratio (ICER) of £21,475. The NICE algorithm both cost more and prevented fewer cases of active TB (Table 1). More comprehensive strategies were associated with increasing cost.
Conclusion Testing only those at highest risk of progression to active TB disease in an HIV population with high ART use was cost-effective, whilst most strategies testing all comers and for active TB cost considerably more than the £20–30,000/QALY gained threshold used in the UK.