RetourNumerous studies have identified risk factors for TB disease in people with positive or negative tests for TB infection; these risk factors may be epidemiological (e.g. contacts), demographical (e.g. age) or clinical (e.g. HIV infection). These studies have recently been summarized, and pooled estimates of absolute and relative risks provided, in an aggregate data meta-analysis (6), and two large-scale individual patient data meta-analyses (4, 5). The risk of TB disease varies widely among those with different epidemiological or clinical risk factors. In large-scale cohort studies of untreated people with positive TB infection tests but no other identifiable risk factors, the annual incidence of TB disease ranged from 0.3 to 1.0 per 100 000 (6, 16, 17). By contrast, the cumulative risk of TB disease in those aged 2–5 years ranged from 50 to 250 per 100 000 in close contacts, particularly young children, or in people living with HIV (People with HIV) (i.e. more than 100 times higher).
Other reviews have summarized risks relative to incidence in general population samples (i.e. populations with positive TB infection tests but no other risk factors). The magnitude of the difference in risk varies; for example, it is 25 times greater in People with HIV without treatment and in recipients of dialysis, solid organ transplants or tumour necrosis factor (TNF)-alpha inhibitors, whereas it is two to five times greater in people with diabetes mellitus or alcohol use disorders, or who are underweight living in insanitary conditions and of low socioeconomic status (18). This underscores the recommendations to consider TPT only for those at high risk, as listed in Box 2.1.