Tuberculin skin testing and QuantiFERON™-TB Gold Plus positivity among household contacts in Vietnam

SETTING: TB infection (TBI) is diagnosed using the technique-dependent tuberculin skin test (TST) or costly, more accurate interferon-gamma release assays. The TST (⩾10 mm) threshold was indicated by previous research among household contacts in Vietnam, but routine implementation with a different tuberculin reagent showed unexpectedly low TST positivity. OBJECTIVE: TST (⩾5 mm and ⩾10 mm) results were compared to QuantiFERON™-TB Gold Plus (QFT) results in household contacts during community campaigns in 2020 and 2021. DESIGN: This was a cross-sectional multi-center implementation study. RESULTS: Among 1,330 household contacts in 2020, we found a TBI prevalence of 38.6% (QFT), similar to TST ⩾5 mm (37.4%) and higher than TST ⩾10 mm (13.1%). QFT+/TST+ was higher for TST ⩾5 mm (20.7%) than TST ⩾10 mm (9.4%). QFT was not discordant with TST ⩾5 mm (McNemar’s test = 0.6, P = 0.5) but was discordant with TST ⩾10 mm (McNemar’s test = 263.9, P < 0.01). Older age and Southern region increased odds for positive TST ⩾5 mm and QFT with weaker associations for TST ⩾10 mm. Agreement and discordance were similar in 2021 for 1,158 household contacts. CONCLUSION: Tuberculin reagents affect TST positivity rates. High TB burden countries should monitor reliability of TBI diagnosis, including tuberculin potency, cold chain, and TST technique to optimize eligibility for TB preventive treatment.

QFT+/TST+ was higher for TST ⩾5 mm (20.7%) than TST ⩾10 mm (9.4%).QFT was not discordant with TST ⩾5 mm (McNemar's test = 0.6, P = 0.5) but was discordant with TST ⩾10 mm (McNemar's test = 263.9,P < 0.01).Older age and Southern region increased odds for positive TST ⩾5 mm and QFT with weaker associations for TST ⩾10 mm.Agreement and discordance were similar in 2021 for 1,158 household contacts.CONCLUSION: Tuberculin reagents affect TST positivity rates.High TB burden countries should monitor reliability of TBI diagnosis, including tuberculin potency, cold chain, and TST technique to optimize eligibility for TB preventive treatment.
A pproximately one-fourth of the world's popula- tion has TB infection (TBI), 1 of whom an estimated 5-10% are at risk for progression to TB disease. 2 The highest risk for progression is within the first 2 years of infection, especially among children younger than 5 years old and those with immune suppression, including people living with HIV (PLWH), silicosis and kidney disease. 3As a reservoir of TB disease, TBI is a major barrier to ending TB. 4,5 In high TB burden countries, TB preventive treatment (TPT) was previously prioritized in PLWH and child contacts (less than 5 years) of bacteriologically confirmed pulmonary TB disease.In these groups, TPT required TB disease exclusion but not TBI diagnosis. 6The WHO's updated consolidated guidelines for latent TBI 2 recommends that TPT may be given to HIV-negative household contacts of all ages without TB disease.High TB burden countries must determine whom to treat: excluding TB disease alone expands eligibility for TPT, while diagnosing TBI is complicated by suboptimal testing methods.The tuberculin skin test (TST) remains the TBI diagnostic standard of choice in high TB burden settings, 5 where it is less specific due to cross-reactivity with antigens from recent bacilli Calmette-Guérin (BCG) vaccination 7 and environmental non-tuberculous mycobacteria.TST administration and interpretation are operator-dependent, 8 and results may vary according to the purified protein derivative (PPD) tuberculin used. 9Global shortage of tuberculin and growing numbers of locally manufactured tuberculin reagents are a challenge to standardization and quality assurance. 10Interferon-gamma release assays (IGRAs) are more specific than TST and better at predicting progression to TB disease, 11,12 but are expensive and require laboratory capacity.There is no gold standard to diagnose TBI since methods only detect immune response to TB antigens. 2,13To determine clinical sensitivity, TBI diagnostic tests are conducted among individuals with bacteriologically confirmed TB disease and compared against WHOrecommended IGRA reference tests. 13,14ietnam is eleventh among 30 countries with the highest TB burden; 15 an estimated 169,000 individuals developed TB disease in 2021. 16Ending TB requires reducing TB incidence by 90%, and modelling shows that both TB disease and TBI interventions are required to achieve this goal.The Vietnam National TB Programme (NTP) is implementing active case-finding (ACF) community campaigns to detect TB disease and link individuals to treatment.These campaigns initially prioritized TB disease and, since 2020, integrated TBI for household contacts of all ages.
Both TST and IGRA (QuantiFERON ® -TB Gold Plus [QFT]; Qiagen, Hilden, Germany) are approved in Vietnam to diagnose TBI, but TST is preferred for routine implementation.The TST ⩾10 mm threshold was based on Vietnam's first national tuberculin survey in child contacts (6-14 years) of bacteriologically confirmed index pulmonary TB, which was nested within the 2007 TB prevalence survey; 17 TBI prevalence (TST ⩾ 10 mm) was 16.7% using tuberculin RT23/Tween 80 (Statens Serum Institute, Copenhagen, Denmark).Another study in Vietnam showed 25.8% TST ⩾10 mm positivity among contacts of new TB patients and 40.8% among contacts of multidrug-resistant TB (MDR-TB) patients. 18In 2020, the Vietnam NTP implemented community campaigns using a different tuberculin reagent (BulBio, BB-NCIPD Ltd, Sofia, Bulgaria) than the prevalence survey.TBI prevalence (TST ⩾10 mm) among household contacts was lower than expected, prompting QFT testing in a subset of household contacts to evaluate agreement with TST and determine the best TST threshold (⩾5 mm or ⩾10 mm) for TBI.TST and QFT were administered in a different subset of household contacts in 2021.

Study population
This was a cross-sectional, multi-center study implemented under routine program conditions during community campaigns screening for TB disease and TBI (2020 and 2021).Campaign sites were conveniently selected in nine districts of five provinces that reported approximately 20% of TB notifications in Vietnam and represented the Northern, Central and Southern regions, which vary in terms of TB prevalence. 19,20District campaign sites in each province differed by year, with the strategy to cover all districts over the implementation period.Campaign size was estimated from districts' index patients with pulmonary TB disease and their household contacts.The QFT subset was conveniently sampled among household contacts in Southern and Northern/Central regions; sample size was determined by capacity and budget for QFT testing.
Index patients were adults diagnosed with pulmonary drug-susceptible TB disease (clinically or bacteriologically confirmed) and notified in NTP registers, within 2 years of each campaign.Home visits identified household contacts, defined as those who 1) slept in the same house with a pulmonary TB patient for at least one night/week for 3 months before diagnosis, or 2) stayed in the same house with a pulmonary TB patient for at least 1 h/day over 5 consecutive days/week for 3 months before TB disease diagnosis.Campaign exclusion criteria were COVID-19 positivity or no verbal consent.ACF campaigns screened other vulnerable populations (elderly, smokers, diabetics) for TB disease, but TBI testing prioritized household contacts.

Testing for TBI diagnosis
District NTP staff administered TST for household contacts excluding those younger than 5 years or pregnant.Using the Mantoux method, 0.1 mL of 5 tuberculin units of BulBio tuberculin was injected intradermally into the forearm.Transverse diameter induration at the injection site was measured using pen and palpation after 48-72 h.QFT was conducted by or under the supervision of Tekmax Co, Hanoi, Vietnam), the main vendor in Vietnam performing IGRA.Venous blood (4 mL) was collected for QFT analysis immediately before TST injection and processed according to the manufacturer's instructions, with ⩾0.35 international units (IU)/mL as the positivity cut-off.TB disease was excluded among household contacts using chest X-ray and sputum Xpert ® MTB/RIF or Xpert ® Ultra (Cepheid, Sunnyvale, California, USA) testing.

Ethics review
Ethics approval was not required since community campaigns were routinely implemented to diagnose and treat TB disease and TBI; QFT was conducted as quality assurance to evaluate TST.Participants gave verbal consent.

Statistical analysis plan
Comparative analyses were used to evaluate the positivity rate, agreement, and discordance of TST (⩾5 mm and ⩾10 mm) and QFT.Raw agreement was calculated as a proportion of the total TST ⩾5 mm/⩾10 mm and QFT results for which there was positive agreement (QFT+/TST+), negative agreement (QFT-/TST-), and discordance (QFT+/TST-and QFT-/TST+).Agreement between TST ⩾5 mm/⩾10 mm and QFT was measured using Cohen's κ.Discordance was measured using McNemar's test to determine if TST ⩾ 5 mm and TST ⩾ 10 mm had the same marginal probabilities as QFT.χ 2 tests of independence for TST ⩾5 mm/⩾10 mm and QFT positivity by age group, sex and region were conducted.Univariable and multivariable logistic regression modelling of TST ⩾5 mm/⩾10 mm and QFT positivity included all available variables (age group, sex, region and symptoms) as predictors.Logistic regression reference groups were female sex; district with the highest participants each year; and age group with lowest odds for test positivity from univariable analysis (10 to less than 20).All analyses were conducted in Stata v17 (Stata, College Station, TX, USA).Data from 2020 and 2021 were analyzed separately due to different conditions in 2021, when campaigns with full-body personal protective equipment for health staff immediately followed COVID-19 vaccinations.
From July to October 2021, movement in Vietnam restricted due to COVID-19, preventing community campaigns that resumed after COVID-19 vaccinations.From October to December 2021, TST and QFT were performed in 1,158 household contacts from five districts in the Northern, Central and Southern Regions (Figure 2, Tables 1 and 2); the mean age was 40.3 years (SD 21.2); 63.1% were female, and 70.4% Southern.TBI prevalence was lower overall in 2021 than in 2020 but similar for QFT (24.1%) and TST ⩾ 5 mm (26.3%) compared to TST ⩾ 10 mm (11.1%) (Table 3).Agreement and discordance between QFT and TST ⩾ 5 mm/TST ⩾ 10 mm were similar to 2020 (Table 4).
Associations for age group and region with test positivity were weaker overall in 2021 (Supplementary Table S1): older age had higher odds for positive QFT and TST ⩾ 5 mm and the South had higher odds for positive QFT (aOR 1.5, 95% CI 1.1-2.1).Males had higher odds for positive TST ⩾ 5 mm (aOR 1.6, 95% CI 1.2-2.1) in 2021, which was not found in 2020.Age, sex and region were not associated with positive TST ⩾ 10 mm in 2021.S2 shows results from pairwise comparisons of age-specific positivity rates for QFT, TST ⩾5 mm and TST ⩾10 mm.QFT = QuantiFERON-TB Gold Plus; TST = tuberculin skin testing; + = positive; -= negative.

DISCUSSION
Vietnam's End TB Strategy prioritizes scaling up TPT among all household contacts, but lower than expected TST positivity rates in routine implementation limited contacts' eligibility for TPT.QFT comparison showed higher positive agreement and lower discordance with TST ⩾ 5 mm than ⩾10 mm thresholds.Lower tuberculin potency was identified as the likely explanation for these programmatic results.Patterns of regional TBI prevalence in our study are consistent with TB disease prevalence in Vietnam, which is highest in the Southern and lower in the Northern and Central Regions. 19,20Our results show that among household contacts, living in the Southern Region and older age are stronger predictors of TBI than sex.TBI prevalence normally increases with age for adult household contacts and general populations in high TB burden settings due to increased risk for TB exposure over time, while TBI prevalence declines in the >70 years from waning immunity.In our population, TST ⩾ 5 mm and QFT positivity increased, as expected with age, consistent with studies in India 21 and China, 22 while TST ⩾ 10 mm positivity did not vary with age.Lower potency tuberculin could result in a delayed hypersensitivity response robust enough for induration ⩾5 mm, but not ⩾10 mm.
Being male did not increase the odds for TBI positivity, except for TST ⩾5 mm in 2021.This suggests that sex is not as strong a risk factor for TBI as it is for TB disease, for which the male:female ratio is approximately 4:1 in Vietnam.A previous study in Vietnam's Southern province of Ca Mau showed that more than one third of the general adult population had TBI, which was higher in males, but the sex difference between males and females was even greater for TB disease notification. 23TBI risk may be similar for male or female household contacts, while vulnerability for TB disease is significantly higher for males due to factors that promote TB progression or transmission, such as smoking, alcohol use, or exposure in work and social settings.Studies in other countries have shown no or minimal difference for TBI prevalence by sex among groups with higher TB disease prevalence in males. 21,24ST and IGRA (QFT-Plus or QuantiFERON-TB ® Gold In-Tube [QFT-GIT]) positivity rates vary in high-burden countries. 21,22,25One study in India 26 showed a TBI prevalence of 71% among household contacts tested with both QFT-GIT and TST (⩾5 mm and ⩾10 mm).Neither QFT-GIT nor TST at either threshold predicted incident TB disease after 24 months.Various tuberculin reagents manufactured in India have yielded different TBI prevalence rates among household contacts. 21,27A study of TB outbreaks in Chinese schools 28 calibrated the TST threshold against QFT-GIT, which was lower for contacts in the same classroom as the index (TST ⩾ 9 mm) than contacts in other classrooms (TST ⩾ 10.5 mm); China's guideline recommends TST ⩾15 mm. 28n vitro and animal studies 9 or direct in vivo comparison 29 with tuberculin standards evaluate potency for tuberculin reagents.Variability in the molecular composition of reagents limits root cause identification when potency differs.In Vietnam,  * QFT-indeterminate results were excluded from all analyses.QFT = QuantiFERON-TB Gold Plus; IU = international units; IQR = interquartile range.
BulBio tuberculin's potency was identified as the likely explanation for lower TST positivity in routine implementation, after evaluating cold chain and addressing TST administration skills.
In the Netherlands, BulBio tuberculin produced larger indurations than RT23 and Tubersol (5TE), 30 suggesting variable BulBio potency.For over 50 years, tuberculin potency was standardized using one product (PPD-Standard), but multiple standards are now used globally to evaluate tuberculin potency. 10Changes in tuberculin reagent can affect TST results. 31To mitigate challenges for TPT eligibility, high-burden countries may choose to give TPT to household contacts of all ages after excluding TB disease and assessing clinical risk if TBI diagnostic testing is unavailable or unreliable. 2,32Countries using TST to determine eligibility for TPT should monitor tuberculin reagent, potency, and cold chain, as well as TST quality assurance.
Despite training and ongoing monitoring, inter-and intra-operator variability for TST administration remains a limitation of our results.Our study findings on TST ⩾ 10 mm positivity differed from those using BulBio tuberculin in a Vietnamese study, 18 which was conducted in one province vs. our programmatic implementation in five provinces.We did not compare TST induration for BulBio head-to-head with tuberculin standards (e.g., RT23).Campaigns involved routine programming without controlling for clinical or demographic variables, with no data collected on BCG vaccination or scar.We evaluated TBI testing in household contacts, in whom clinical sensitivity of TST and the QFT reference cannot be calculated, as would be possible with confirmed TB disease.Our TST/QFT comparative analyses were also subject to potential QFT variability in field applications.Finally, TBI prevalence overall was lower in 2021 for both QFT and TST than in 2020.Explanations include the effect of the 2021 COVID-19 lockdowns on campaigns, the complexity of COVID-safe campaigns impacting testing quality by both methods, and recent COVID-19 vaccination.Inactive vaccines should not impact the immune response for TST or QFT, but data on this are limited.
New TBI diagnostic methods are needed for the programmatic expansion of TPT. 33TB antigen-based skin testing is approved by  the WHO to diagnose TBI, 34 with similar specificity to IGRA, similar sensitivity to TST and IGRA, and one (5 mm) threshold. 35Near point-of-care QIAreach QuantiFERON ® -TB (Qiagen; Carnegie, VIC, Australia) is a less expensive, simpler alternative to QFT. 14 Programmatic testing for TBI in high-burden countries requires accurate, quality-assured methods that can be decentralized and are affordable at scale.High TB burden countries expanding TPT to all household contacts should monitor TBI diagnostic performance, including quality assurance for tuberculin and TST.TPT strategies should align with availability and reliability of testing to exclude TB disease and diagnose TBI, guided by local epidemiology for TB disease and TBI.Public Health Action (PHA) welcomes the submission of articles on all aspects of operational research, including quality improvements, costbenefit analysis, ethics, equity, access to services and capacity building, with a focus on relevant areas of public health (e.g.infection control, nutrition, TB, HIV, vaccines, smoking, COVID-19, microbial resistance, outbreaks etc).
This is an Open Access article distributed under the terms of the Creative Commons Attribution License CC-BY 4.0 published by The Union (www.theunion.org).Contact: pha@theunion.orgInformation on PHA: http://www.theunion.org/what-we-do/journals/pha

FIGURE 1
FIGURE 1 Age-specific prevalence for QFT, TST ⩾5 mm and TST ⩾10 mm positivity among household contacts in 2020 and 2021.Upper: Positivity rates for QFT (solid line), TST ⩾5 mm (dotted line) and TST ⩾10 mm (dashed line) by age groups in 2020 (left) and 2021 (right).Middle: Rates of concordant positivity for QFT and TST ⩾5 mm (open triangle); TST ⩾5 mm-only positivity (open square); and QFT-only positivity (open circle) by age groups in 2020 and 2021.All TST ⩾5 mm curves show an increase in TST positivity with increasing age.Lower: Rates for concordant positivity of QFT and TST ⩾10 mm (closed triangle); TST ⩾10 mm-only positivity (closed square); and QFT-only positivity (closed circle) by age groups.TST ⩾10 mm-only positivity (QFT-/TST+, closed squares) is low and does not vary with age, in contrast to TST ⩾5 mm.Corresponding with Middle and Lower panels, Supplementary TableS2shows results from pairwise comparisons of age-specific positivity rates for QFT, TST ⩾5 mm and TST ⩾10 mm.QFT = QuantiFERON-TB Gold Plus; TST = tuberculin skin testing; + = positive; -= negative.

FIGURE 2
FIGURE 2Participants in TB ACF campaigns evaluated with QFT and TST in 2020 and 2021.A subset of household contacts from ACF campaigns was evaluated for TBI with both TST and QFT.ACF participants were screened for TBI and TB disease; those with Xpert-confirmed pulmonary TB disease or with QFT-indeterminate results were excluded from analyses.TBI = TB infection; + = positive; TST = tuberculin skin testing; QFT = QuantiFERON-TB Gold Plus; ACF = active case-finding.

TABLE 1
Baseline characteristics of household contacts evaluated with QFT and TST in 2020 and 2021 * Some data are missing for age and sex.†QFT-indeterminate results were excluded from all analyses.QFT = QuantiFERON-TB Gold Plus; IU = international units; TST = tuberculin skin testing.

TABLE 3
Positive prevalence for QFT, TST ⩾5 mm and TST ⩾10 mm among household contacts by region, sex and age groups in 2020