Introduction NICE guidelines recommend a Mantoux threshold of 15 mm induration to trigger investigation for tuberculosis (TB), with IGRA tests used as second line in selected groups. Little data are available about the role of the two tests in the diagnosis of active TB. Emerging evidence suggests IGRA tests may be more sensitive in identifying TB infection in children.

Aim To explore the relationship between Mantoux and IGRA in children attending a paediatric TB clinic in Manchester.

Hypothesis A Mantoux threshold of ≥15 mm induration underestimates TB infection in children.

Method All children aged 0–17 years referred to the TB service at Central Manchester Foundation Trust between Jan 2009 and May 2011 were included. Initial screening included symptom review and Mantoux testing. Based on induration at 48–72 h, Mantoux tests were defined as negative <6 mm or positive ≥6 mm for those with no prior BCG, and negative <10 mm, borderline 10 to ≥15 mm or positive ≥15 mm for those with prior BCG. Children with borderline or positive Mantoux test results, or in whom there was clinical concern, were referred for consultant assessment and/or IGRA (Quantiferon Gold). Data were collected retrospectively from case notes.

Results 976 children were referred. 756 completed initial assessment (388 (51%) male, mean age 6.2±4.6 years, range 0.16–16.36 years). BCG history was known in 754 (99.7%; 516 BCG). 403 patients were discharged without intervention, 63 were offered BCG vaccination, two were referred elsewhere and 288 were referred for consultant assessment. Of these 288, 108 were notified with TB, 46 received chemoprophylaxis, 117 received no treatment, 5 received BCG and 12 failed to attend. 252 children had paired Mantoux and IGRA. Of these, 18/44 (41%) of those with a borderline Mantoux had a positive IGRA. 126/252 had TB infection (91 active and 35 latent TB)—see Abstract P175 table 1. A Mantoux threshold of ≥15 mm identified 77 (61%) children with TB infection, IGRA identified 92 (73%) and the two tests combined identified 100 (79%) children.

Conclusion Using a Mantoux threshold of ≥15 mm induration significantly underestimates the number of children with TB infection compared with using Mantoux and IGRA together.