Skip to main content

Table 2 Studies investigating the feasibility and performance of screening tools for paediatric LTBI (by study population)

From: Recommendations for the screening of paediatric latent tuberculosis infection in indigenous communities: a systematic review of screening strategies among high-risk groups in low-incidence countries

First Author
(Reference)
Year Study setting Study population Sample size Study objective Screening tool used Comparator tool Comparative screening results (SP, SE, discordance) Findings/Recommendations regarding screening tools
Indigenous Communities
 Alvarez [25] 2014 Nunavut, Canada A high-risk Indigenous community in Iqaluit, Nunavut 256 (with both TST and IGRA results) To evaluate the feasibility of the use of IGRAs for LTBI screening in the Nunavut Indigenous population IGRA TST 44/256 (17.2%) discordant results, most of which occurred in people with multiple BCG vaccinations or those who were vaccinated after infancy • 18% IGRAs positive, 32% TSTs positive.
• IGRAs are a valid screening tool for LTBI in Nunavut, as most of the community is BCG-vaccinated, making IGRAs more specific.
 Kwong [26] 2016 Sioux Lookout, Ontario, Canada Indigenous adolescents in northern Ontario (screened at age 14) 11 To evaluate the IGRA for LTBI screening in a Canadian Indigenous community IGRA TST 7/11 had a positive TST, of these 7, all had a negative IGRA and none developed symptoms of active TB disease. • Recommends use of IGRAsv due to high proportion of false-positive TST in BCG vaccinated adolescents.
 Reid [27] 2007 Saskatchewan, Canada Preschool children (0–4 yrs) living in Indigenous reserve communities in Saskatchewan 2953 (1086 BCG+, 1867 BCG-) To investigate the effect of BCG vaccination on TST results in Canadian Indigenous children TST None   • More positive TSTs among BCG+ children at 5 mm for 0–4 yrs., but no longer significant in 3–4 yrs. at > 10 mm
• Need to take BCG vaccination status and age into consideration when using TST to screen for LTBI in Indigenous communities
Immigrant or refugee children
 Howley [41] 2015 USA Children (2–14 yrs) immigrating to the US from Vietnam, the Philippines or Mexico 2520 To evaluate the QFT vs TST for LTBI screening in immigrant children QFT TST • kappa = 0.20.
QFT displayed stronger association with presence of risk factors for LTBI
• Recommends the use of QFT in immigrant children (2 yrs. and above)
 Losi [42] 2011 Modena, Italy Immigrant children and adolescents 621 To evaluate the QFT-GIT as a screening tool for LTBI in children immigrating to Italy IGRA (QFT-GIT) TST • 104/621 TST+
• 80/621 QFT+
• 50 TST+/QFT+,
• 30 TST inconcl./QFT+
4 active TB cases suspected, all QFT-GIT+/TST+
• QFTs useful for screening in low-burden settings, however, due to continued lack of research regarding sensitivity of IGRAs in younger children, TST positivity should not be disregarded
 Lucas [43] 2010 Perth, Australia African and Burmese refugee children resettling in Australia 524 To evaluate the QFT-GIT and T-SPOT.TB as screening tools for LTBI in refugee children resettling to Australia QFT-GIT and TB-SPOT.TB TST • QFT-GIT and TB-SPOT.TB showed high concordance (k = 0.78, p < 0.0001).
• Both had poor concordance with TST: 50%.
• High failure rates (14 and 15% respectively) of T-SPOT and QFT-GIT
• High proportion of inconclusive results for both IGRAs suggests TST remains useful alternative.
 Salinas [47] 2015 Spain Undocumented immigrant children (< 19 yrs) in Basque Country, Spain 845 To determine the prevalence of LTBI in undocumented immigrant teenagers using QFT QFT-GIT TST 63% overall, 57% positive and 96% negative concordance. • Screening high-risk subgroups of the population in low-incidence countries is recommended.
• The use of QFT as opposed to TST in this group reduced the provision of preventive treatment by 43%
Children considered at risk based on suspected TB exposure/contact with a TB case (among other risk factors, including immigration or adoption)
 Bergamini [56] 2009 Italy At-risk children (contacts of TB cases and recent immigrants)
(0–19 yrs)
496 To investigate the effect of age on IGRA effectiveness IGRAs (QFT-GIT, QFT-G, T-SPOT.TB) None • TST: uncertainty of accuracy in BCG-vaccinated children. QFTs: more indeterminate results in those < 4 yrs. compared to older children
• T-SPOT: lower proportion of indeterminate results compared to QFTs
Need to take into account age and BCG vaccination status when using the TST for LTBI screening
 Connell [38] 2006 Melbourne, Australia High-risk children (suspected contact with an active case, recent arrival from high-incidence country, clinical suspicion) 106 To evaluate the effectiveness of LTBI screening with IFN-y vs. TST IFN-γ assay TST • IFN-/TST+ in 70% of TST+ cases
• Kappa = 0.3
TST recommended for screening among high-risk children, as TST positivity was a better predictor of the possibility of actual LTBI (those with household contacts with TB were more likely to be positive by TST than by IFN-γ assay, although this does not necessarily translate into better predictiveness of the TST given that not all household contacts will necessarily develop LTBI).
 Grare [39] 2010 Nancy, France Children considered at risk for TB due to clinical suspicion, an adult case contact or recent immigration from an endemic region
(< 18 yrs)
51 (44 with test results) To evaluate the effectiveness of QFT-GIT vs. the TST for the identification of paediatric LTBI QFT-GIT TST 84% agreement between the two tests • TST remains a useful predictor of LTBI.
• High levels of inconclusive IGRA results suggest more studies are needed to determine its increased effectiveness vs. TST
 Sali [45] 2015 Italy Children (0–14 yrs) with suspected active TB, exposure to an adult case, or healthy adopted children 621 To evaluate the QFT-GIT for paediatric active TB diagnosis and LTBI screening QFT-GIT None   • Usefulness of IGRAs in younger children: 0–12 month age group more likely to have indeterminate QFT results (p = 0.001).
• Children < 8 months: impaired ability to respond to mitogen compared to those > 8 months (but not statistically significant)
 Salinas [46] 2011 Spain Children < 17 yrs. in Basque Country, Spain, that had previous TB contacts 160 To compare the QuantiFERON-TB gold in-tube test to the TST for the detection of LTBI in children with TB contacts QFT-GIT TST 95–96% concordance (100% in non-vaccinated children and children < 5 yrs) • QFT-GT reduced preventive treatments by 28–34% and is therefore recommended in low-incidence countries
Other Paediatric Populations
 Grinsdale [40] 2016 San Francisco, USA Children < 15 yrs. screened for TB at 20 community clinics in San Francisco 1092 To assess the concordance of QFTs and the TST in a low-burden setting QFT-GIT TST • 79% discordance (TST+/IGRA-) in BCG-vaccinated foreign-born children vs. 37% discordance in non-vaccinated US-born children.
• Children > 5 yrs. also significantly more likely to have discordant results
• QFT vs. TST discordance was high, however, QFT has high NPV, as no TST+/QFT- children developed active TB disease in the 5 years of follow-up.
• QFT may be a better predictor of risk of progression to active disease in low-burden setting
 Mekaini [44] 2014 Abu Dhabi, UAE Children (1–19 yrs) attending health centres in Abu Dhabi for routine care 699 (669 gave blood sampling consent) To evaluate the QFT-GIT for paediatric LTBI screening QFT-GIT Risk factor questionnaire • QFT positivity was low (4/669, 0.6%), however it identified two LTBI cases that would have been suspected negative based on the risk factor questionnaire alone QFT is recommended, depending on the estimated prevalence of TB in the population
 Rose [29] 2014 Toronto, Canada Paediatric HIV patients (< 19 yrs) 81 To evaluate the QFT-GIT for LTBI screening in HIV-positive children QFT-GIT TST 96% TST-/QFT- concordance, but TST+/QFT+ concordance was low. Use of QFT in HIV-positive children is valid, although low correlation with risk factor assessment (5 mm cut off used for the TST)