Diagnosis of tuberculosis infection in children with a novel skin test and the traditional tuberculin skin test: An observational study

Abstract

Background: A novel skin test-called Diaskintest (DT)-containing specific M. tuberculosis antigens is in clinical use in the Russian Federation (RF). This test may improve diagnosis of tuberculosis (TB) infection. The use and performance of the DT was described and compared to the tuberculin skin test (TST).

Methods: Data on children <18 years referred to a TB reference centre (Jan/2018- Dec/2019) with ≥1 DT and TST result available were analysed. An immune correlate of TB infection was defined as a positive TST (≥10 mm induration) or a positive DT (any induration).

Results: Of 2710 included cases, the median age was 9.0 (IQR 5.7-13.1) years and 97.5% were BCG immunised. Overall, 1976 (79.9%) were TB uninfected, 724 (26.7%) had an immune correlate of TB infection and 10 (0.4%) TB disease. Reasons for referral were: positive or increasing skin test results in routine screening (992, 36.6%), screening before admission to a health care institution (501, 18.5%) and TB contact (457, 16.9%). DT was positive in 11.7% (308/2625) and TST in 63.1% (467/740) (Kappa 0.08, 95% CI:0.013-0.14). A positive DT was associated with older age (OR 1.16 (95% CI: 1.13-1.19) per year). Among TB contacts DT positivity was associated with contagiousness: highest proportion of positivity of 12.0% was observed when the index case was smear positive.

Conclusion: In a setting with universal BCG vaccination and regular screening with TST, DT was used to rule out TB infection as TST was commonly positive. We found an association of DT positivity and contagiousness of the index case in children contacts. These observations may suggest improved specificity and sensitivity of DT compared to TST.

Fig 1. The Flow diagram shows children referred to the TB dispensary.

Eligible for the final analysis were 2710 children and reasons for exclusion and numbers are displayed. Children were divided into TB disease, immune correlate of TB infection and TB non-infected. As immune correlate of TB infection, we defined for the purpose of this study the presence of at least one TST with induration of ≥10 mm and/or DT with induration of any size. Since this definition differs from the one used for TB infection and indications for TB treatment in clinical practice in the Russian Federation only a proportion of the children was started on treatment.

Fig 2. Venn diagram using exact Euler calculation showing positive test results for TST, DT and initiation of treatment in children with immune correlate of TB infection.

TST positivity was defined as with induration ≥10 mm and DT positivity as any induration. The area of the ellipses is proportional to the number of cases. As immune correlate of TB infection, we defined for the purpose of this study the presence of at least one TST with induration of ≥10 mm and/or DT with induration of any size. Since this definition differs from the one used for TB infection and indications for TB treatment in clinical practice in the Russian Federation only a proportion of the children was started on treatment.

Fig 3. Final diagnosis of children referred to the TB dispensary by reason for referral.

Among children referred for an abnormal routine annual screening and following a contact with a TB case the proportion of immune correlate of TB infection was highest with 62.3% and 16.2%, respectively (see also S1 Table).

Fig 4. DT and TST positivity in children who had contact to an TB index case, stratified by the smear results of the contact.

Children exposed to a smear positive index case were more likely to have a positive DT and TST (in red) compared to children exposed to smear negative or unconfirmed contacts.