INTRODUCTION — Important priorities for control and eventual elimination of tuberculosis (TB) include treatment of individuals with active TB and identification and treatment of individuals with latent TB infection (LTBI) [1-3].

In most individuals, Mycobacterium tuberculosis infection is contained initially by host defenses, the infection remains latent or is cleared, and the individual is asymptomatic and noninfectious [4]. However, latent infection has the potential to develop into symptomatic active disease at any time. Identification and treatment of LTBI can reduce the risk of development of disease by as much as 90 percent [5] and so has the potential to protect the health of the individuals as well as the public by reducing the number of potential future sources of infection [6,7].

There is no test that definitively establishes a diagnosis of LTBI. LTBI is a clinical diagnosis established by demonstrating an immune response to M. tuberculosis antigens and excluding active TB disease. Available tests to the immune response to TB antigens include the tuberculin skin test (TST) and interferon-gamma release assays (IGRAs). These measure immune sensitization (type IV or delayed-type hypersensitivity) to mycobacterial protein antigens that occurs following exposure to (and infection by) mycobacteria. There is no clear advantage of the IGRA or the TST for predicting future risk of active TB; the decision to select a test should be based on the setting, cost and availability.

The approach to diagnosis of LTBI in adults will be reviewed here; issues related to interpretation of the TST and IGRAs are discussed separately. (See "Use of the tuberculin skin test for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults" and "Use of interferon-gamma release assays for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults".)

Issues related to treatment of LTBI in adults are discussed separately. (See "Treatment of latent tuberculosis infection in HIV-uninfected nonpregnant adults".)

Issues related to LTBI in patients with HIV, patients receiving tumor necrosis factor-alpha inhibitors, and patients undergoing solid organ transplant are discussed separately. (See "Treatment of latent tuberculosis infection in nonpregnant adults with HIV infection" and "Tumor necrosis factor-alpha inhibitors and mycobacterial infections" and "Tuberculosis in solid organ transplant candidates and recipients".)

Issues related to LTBI in children and pregnant women are discussed separately. (See "Latent tuberculosis infection in children" and "Tuberculosis in pregnancy", section on 'Latent tuberculosis infection in pregnancy'.)

TERMINOLOGY — Tuberculosis terminology is inconsistent in the literature [8]. Relevant terms are defined in the table (table 1).

INDICATIONS FOR LTBI TESTING — The goal of testing for latent TB infection (LTBI) is to identify individuals who are at increased risk for the development of active TB disease and therefore would benefit from treatment of LTBI (table 2 and table 3). Only those who would benefit from treatment of LTBI should undergo testing, so a decision to test presupposes a decision to treat if the test is positive [1].

Testing for and treatment of LTBI should be pursued for individuals in two categories (table 2 and algorithm 1) [3,9-12]:

●Individuals at risk for new infection due to TB exposure

●Individuals with LTBI who are at increased risk of reactivation (ie, progression to active disease) due to underlying conditions

In general, repeat LTBI testing among individuals with prior negative test results should be performed only in the setting of new exposure since the last assessment (such as new contact with a case of untreated active respiratory TB, occupational exposure, or extended travel in regions where TB is endemic). (See "Use of the tuberculin skin test for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Repeat and serial testing' and "Use of interferon-gamma release assays for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Serial testing, conversion, and reversion'.)

Increased risk of new infection

Individuals with recent exposure (close and casual contacts) — Testing for latent TB should be performed in asymptomatic individuals for whom new infection is suspected (algorithm 1). High-priority groups for evaluation include close and casual contacts of patients with untreated active respiratory TB (table 2).

One study suggests that a definition of high-priority close contacts as persons sharing an enclosed space ≥4 hours per week can be used as a measure of risk [13]. This includes those living in the same household or frequent visitors to the house; it may also include contacts at work or school. Because the risk of developing active disease is very high in the first few years following exposure and new infection, all close contacts should be evaluated for LTBI regardless of age (table 2).

Most TB control programs evaluate close contacts first and expand the contact investigation to casual contacts if evidence of transmission is found. Casual contacts may be defined as individuals with less than four hours of contact per week. This may include health care workers (HCWs) and/or contacts at work or school. Evidence of transmission is defined by detection of active TB among contacts (secondary active TB) or observing prevalence of positive tuberculin skin tests (TSTs) or interferon-gamma release assays (IGRAs) that is greater than expected. In general, source TB patients with positive sputum smear are more contagious; casual contacts of smear-negative patients are usually not evaluated unless they are immunocompromised.

If the first test is negative, close contacts of patients with active pulmonary TB should undergo a second test eight to ten weeks later. Many TB programs test casual contacts only once at eight to ten weeks after last exposure. (See "Use of the tuberculin skin test for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Repeat and serial testing' and "Use of interferon-gamma release assays for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Serial testing, conversion, and reversion'.)

The risk of disease in the first two years following infection is age dependent. The decline of risk with increasing age reflects greater innate and acquired immunity [14]:

●Infants (ages ≤1 year): 50 percent

●Children (ages 1 to 2 years): 12 to 25 percent

●Children (ages 2 to 5 years): 5 percent

●Children (ages 5 to 10 years): 2 percent

●Age >10 years: 1 to 2 percent

Health care workers — In regions with low TB incidence rate, HCWs in facilities with low TST conversion rate need not undergo routine serial TB screening. In such settings, HCWs should undergo initial TB screening with individual risk assessment and symptom evaluation. For individuals without documented prior TB disease or LTBI, baseline TB testing with an IGRA or a TST should be performed. Thereafter, in the absence of a known exposure or ongoing transmission, no routine serial TB testing at any interval after baseline is warranted [15]. TST interpretation thresholds are summarized in the table (table 4). (See "Use of the tuberculin skin test for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Serial testing' and "Use of interferon-gamma release assays for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Serial testing, conversion, and reversion'.)

Serial TB screening may be reasonable for HCWs at increased risk for occupational exposure to TB (such as pulmonologists or respiratory therapists) or for HCWs in certain settings (such as emergency departments). Such policies should be individualized based on factors such as the regional and institutional incidence of pulmonary TB, whether delays in initiating airborne isolation have occurred, or whether prior serial screening has demonstrated ongoing transmission; such determinations may be made in consultation with local or state health departments [15]. The approach to serial testing is described below. (See 'Approach to serial testing' below.)

In the setting of a recognized exposure, HCWs with a baseline negative LTBI test and no prior TB disease or LTBI should undergo testing (IGRA or TST) when the exposure is identified; if IGRA is newly positive, or TST is newly positive at ≥5 mm, this should be interpreted as a evidence of new TB infection. If the test is negative, another test should be performed 8 to 10 weeks after the last exposure [15]. In this case, we define TST conversion if the induration of the repeat TST measures ≥10 mm and has increased by ≥6 mm over the previous test (ie, the initial negative TST) [16]. Alternatively, one test may be performed 8 to 10 weeks following the end of exposure.

For HCWs with untreated new LTBI, treatment is encouraged unless medically contraindicated [15]. HCWs with new LTBI who do not complete treatment should be monitored annually with symptom evaluation to detect early evidence of TB disease and to re-evaluate the risks and benefits of LTBI treatment [15]. In addition, these HCWs should be educated about the signs and symptoms of TB disease that should prompt immediate evaluation between screenings.

The above approach reflects guidelines issued by the United States Centers for Disease Control and Prevention and National Tuberculosis Controllers Association in 2019 [15], which serves as an update to the prior 2005 guidelines (table 5) [16]. This approach is supported by a number of factors. TB incidence rates in the United States declined by 42 percent between 2005 and 2017 (from 4.8 to 2.8 per 100,000 population) [17]. In addition, the 2019 guideline statement refers to an unpublished systematic review of eight studies including more than 63,000 HCWs; there was no disease occurrence among these individuals [15]. Furthermore, the use of TSTs and IGRAs for serial testing of HCWs has been associated with a number of limitations [18].

The role of IGRA for serial testing in HCWs is uncertain, in large part due to high rates of false conversions in most studies in HCWs undergoing serial testing with IGRA. This issue is discussed further separately. (See "Use of interferon-gamma release assays for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Reproducibility'.)

Homeless shelters and correctional facilities — Annual screening is warranted for residents and employees of homeless shelters and correctional facilities, given the risk of TB exposure in congregate settings [19]. (See "Epidemiology of tuberculosis", section on 'Community settings'.)

The approach to annual screening is discussed below. (See 'Approach to serial testing' below.)

Increased risk of reactivation

General principles — Identifying individuals who warrant LTBI testing and treatment depends on the risk for reactivation due to underlying conditions (relative to healthy individuals) (table 2 and table 6 and algorithm 1). In general, individuals with LTBI who are otherwise healthy have an annual risk of at most 0.1 percent (1 per 1000) of developing active TB disease [20]; one review noted a much lower rate of 0.03 percent (0.3 per 1000) among otherwise healthy individuals with a positive TST (table 7 and table 8) [21].

The relative likelihood of a true-positive versus false-positive test result should be considered carefully. A web-based algorithm can be used to estimate the likelihood that an individual with a positive TST or IGRA has a true-positive test, and the cumulative lifetime risk of developing active TB in that individual, taking into consideration various conditions that increase risk of reactivation; it also factors in the age-related risks of isoniazid [22]. This algorithm has been validated in a cohort of including more than 37,000 foreign-born individuals in Canada; however, risks may be overestimated in the highest risk groups [23].

False-positive tests occur more commonly with TST but also can occur with IGRA. True-positive tests are more likely in individuals with relevant epidemiologic exposure (such as close contact with a case of contagious TB or residence in a high-incidence country). True-positive tests are very uncommon in individuals born in the United States (or other low-incidence countries) after 1960 who have no other risk of exposure.

In general, individuals at increased risk of reactivation require a single test for diagnosis of LTBI; if the result is negative, no further testing is needed. Given that the likelihood of latent TB infection is very low in most individuals born in the United States since 1960, the likelihood of a false-positive test result must be considered carefully. Dual testing (with IGRA as well as TST) can be useful in certain circumstances. (See 'Role of dual testing' below.)

Balancing risk of disease and risk of treatment — The approach to balancing the risk of disease and the risk of treatment depends in part on the treatment regimen utilized. (See "Treatment of latent tuberculosis infection in HIV-uninfected nonpregnant adults" and "Treatment of latent tuberculosis infection in nonpregnant adults with HIV infection".)

The relationship between the risk of adverse events and increasing age is well documented for isoniazid; the risk of serious hepatotoxicity in individuals >65 years, 50 to 65 years, and 35 to 50 years is >5 percent, 2 to 5 percent, and <2 percent, respectively [2,24,25]. The risk of hospitalization for liver toxicity associated with isoniazid therapy is also related to older age, and presence of comorbid conditions [26]. In individuals with conditions conferring moderate or slightly increased risk who would be treated with isoniazid, we carefully assess the balance of risks and harms in individuals >50 years of age (especially individuals >65 years of age). Guidelines in the United States [1,2] and Canada [27] do not set any upper age limit for testing for and treatment of LTBI; guidelines in the United Kingdom set an upper age limit of 65 years [28].

For other LTBI regimens, such as daily rifampin for four months (4R) or weekly isoniazid-rifapentine for three months (3HP), the relationship between the risk of adverse events and increasing age is less clear. In secondary analysis of two trials in adults, the risk of hepatotoxicity (grade 3 to 4) likely attributable to 4R was significantly lower than the risk attributable to daily isoniazid for nine months, and was not age related [29]. These findings suggest that 4R is likely to be safe for older patients; therefore, if 4R is available, older age is not a contraindication for testing. However, use of concomitant medications may be more likely among older patients, increasing risk of drug-drug interactions. In a randomized trial evaluating 3HP for treatment of LTBI, serious adverse reactions including hypersensitivity were significantly more common with older age [30]. (See "Treatment of latent tuberculosis infection in HIV-uninfected nonpregnant adults".)

●High risk − We consider individuals with high risk for reactivation of latent TB to be those with the risk of reactivation is at least six times higher than the risk in healthy individuals. These include individuals with major immunocompromising conditions (eg, lymphoma, leukemia, head and neck cancer, chemotherapy, solid organ transplant, HIV infection, tumor necrosis factor [TNF]-alpha inhibitors), and individuals with chest radiograph demonstrating fibronodular changes typical of healed TB (so-called "inactive TB") [31]. All individuals in these categories should have a single test to evaluate for LTBI.

●Moderate risk − We consider individuals with moderate risk for reactivation of latent TB to be those with risk of reactivation that is three to six times higher than the risk in healthy individuals. These include individuals with diabetes mellitus (regardless of insulin dependence) or those on corticosteroid therapy. For individuals with moderate risk for reactivation, latent TB testing should be restricted to those from groups with increased prevalence of LTBI (these include homeless individuals, injection drug users, contacts of active TB cases, and foreign-born individuals who immigrated as adults from countries with TB incidence >100/100,000 (table 3)).

●Slightly increased risk − We consider individuals with slightly increased risk for reactivation of latent TB to be those with risk of reactivation that is 1.5 to 3 times higher than the risk in healthy individuals. These include individuals who are underweight, smoke cigarettes, or have small granulomas on chest radiograph. For individuals with slightly increased risk for reactivation, latent TB testing should be individualized; we favor testing for such individuals from groups with increased prevalence of LTBI (these include homeless individuals, injection drug users, contacts of active TB cases, and foreign-born individuals who immigrated as adults from countries with TB incidence >100/100,000 (table 3)).

Risk groups — Patients at increased risk of reactivation include individuals with HIV infection or malignancy. In such groups, exclusion of active TB is especially important. (See 'Excluding active TB' below.)

HIV infection — All newly diagnosed HIV-infected patients should be screened for latent TB with TST or IGRA (algorithm 1) [32,33]. Subsequently, for those who tested negative, annual screening is warranted when there are ongoing risk factors for TB (eg, incarceration, living in communal settings, active drug use). (See "Primary care of adults with HIV", section on 'Tuberculosis'.)

In addition, for HIV-infected patients who initially tested negative in the setting of a CD4 cell count <200 cells/microL, repeat testing should be performed once the CD4 cell count has increased above this threshold, because of the possibility of false-negative results in the setting of immunosuppression [32,34-36]. Testing should not be repeated in patients who have previously had a positive test.

For HIV-infected patients in high TB prevalence settings (particularly those with CD4 cell count <200, which confers greater risk for development of TB disease), LTBI treatment may be reasonable regardless of test results, once active TB has been excluded [2,32,37]. (See 'Excluding active TB' below and "Treatment of latent tuberculosis infection in nonpregnant adults with HIV infection".)

For HIV-infected patients in high TB prevalence settings where LTBI testing is not available, all patients should receive treatment for LTBI regardless of CD4 cell count [38,39]. LTBI therapy administered in the absence of LTBI testing has been associated with a 40 to 50 percent reduction in active TB disease among persons living in areas with very high TB incidence [40]. (See "Treatment of latent tuberculosis infection in nonpregnant adults with HIV infection".)

Malignancy — Routine LTBI testing is warranted for patients with hematologic malignancies, head and neck cancer, and lung cancer, given substantially increased risk for reactivation in these groups. In patients with solid tumors, a risk-stratified approach is reasonable, since the risk for reactivation appears to be lower. Among patients at low risk for hepatotoxicity (based on age, other comorbid illnesses, and chemotherapy regimen), we perform LTBI testing for individuals with an expected five-year survival >25 percent. Among patients at increased risk for hepatotoxicity, we perform LTBI testing for individuals with an expected five-year survival >50 percent.

These issues have been evaluated in systematic reviews [31,41]. One review included 13 studies with more than 920,000 patients; it reported incidence rate ratio (IRR) for active TB of 3.5 (95% CI 1.5-7.6) for adults with hematologic malignancies and IRR of 2.3 (95% CI 2.0-2.6) for adults with solid tumors [41]. Another review included 23 studies with more than 300,000 patients; in the six studies in the United States published after 1980, the IRR was 26 for hematologic malignancies, 16 for head and neck tumors, 9 for lung cancers, and 4 for breast and other solid tumors [31].

Individuals from high-incidence settings — The prevalence of LTBI is increased among individuals from high-incidence countries (table 3) [42]. Testing and treatment of individuals from high-incidence countries may be beneficial, particularly if they also have risk factors for reactivation of disease (eg, fibronodular scarring on chest radiograph, immunosuppression from diabetes, HIV infection, chronic kidney disease, or other causes) (table 2).

Testing and treatment of individuals from high-incidence countries who do not have additional risk factors for reactivation is more controversial. In the United States and United Kingdom, guidelines favor LTBI testing for such individuals [11]; in Canada, guidelines favor LTBI testing only for those with additional risk factors for progression to active disease [37]. (See 'Guidelines vary by country' below.)

The increased incidence of TB among individuals from high-incidence countries is related principally to the increased prevalence of LTBI in this group, not to an increased risk of LTBI reactivation [43]. One modeling study concluded that treatment of LTBI in individuals from high-incidence countries who had resided in the United States for more than five years was moderately cost-effective, but the individual benefit was very small [42].

CHOOSING BETWEEN TST AND IGRA

Overview

General principles — There are two major types of tests for identification of TB infection: the tuberculin skin test (TST) and the interferon-gamma release assay (IGRA) blood test (table 4 and table 9 and algorithm 1) [44,45]. Both test types evaluate cell-mediated immunity. There is no clear advantage of the IGRA or the TST to predict future risk of active TB. The decision to select a test should be based on the setting, cost and availability. (See 'Predictive value of IGRA versus TST' below.)

Guidelines for use of TST versus IGRA vary by country. (See 'Guidelines vary by country' below.)

Guidelines for the diagnosis of active and latent TB were published in 2017 by the American Thoracic Society, United States Centers for Disease Control and Prevention (CDC), and Infectious Diseases Society of America [1]. The guidelines state the following:

●For individuals with low-to-intermediate risk of progression to active disease (table 6), the IGRA is preferred over TST for diagnosis of LTBI. IGRA is especially useful for patients who are unlikely to return to have the TST read and for patients with a history of Bacille Calmette-Guérin (BCG) vaccination (administered at birth in most TB-endemic countries) [1]. The TST is an acceptable alternative to IGRA, especially in situations where IGRA is not available or is too costly (even though it is less specific than the IGRA).

●For individuals with high risk of progression to active disease (table 6), either the IGRA or TST may be used; in such cases, the higher false-positive rate of the TST is acceptable. A dual testing strategy (perform one test and, if negative, perform the other) may be used, in which a positive result from either test would be considered positive.

The National Tuberculosis Controllers Association (NTCA) issued guidance for latent TB screening, diagnosis, and treatment in 2021 [46]. In this guidance, IGRAs are preferred for most non-United States-born patients who received (or may have received) BCG vaccination; for other individuals, either a TST or IGRA may be used, depending on availability and cost. The NTCA advises against routine use of dual testing (with both TST and IGRA), but that this approach may be considered for patients at risk for a poor immune response to these tests, for patients at risk of severe forms of TB disease, and for patients in whom TB infection is strongly suspected. The NTCA also suggests that if patients at low risk for TB infection required testing, either an IGRA or TST should be used; if the result is positive, a second test (with the same or a different method) should be performed to confirm the test result and enhance specificity.

Updated guidelines for the diagnosis and treatment of latent TB were published by the World Health Organization in 2020 [3]. These guidelines state that either TST or IGRA may be used to test for LTBI, and neither test is preferable over the other for predicting progression to active TB disease. Performing TSTs may require fewer resources than IGRAs and may be more familiar to practitioners in resource-constrained settings; however, global TST shortages have reduced use of TST in scaling up programmatic management of LTBI.

Predictive value of IGRA versus TST — There is no clear advantage of the IGRA or the TST for predicting future risk of active TB; the decision to select a test should be based on the setting, cost and availability. In very high-risk populations, there may be an advantage of performing both TST and IGRA, to enhance sensitivity. In general, among individuals with a given risk factor, those with positive IGRAs have higher rates of active TB than those with positive TST, likely reflecting greater specificity of IGRA (table 7).

The key value of a test for LTBI is the accuracy of the test in predicting future TB disease, since this risk determines the need for LTBI treatment. There have been many longitudinal studies evaluating the risk of active TB in untreated individuals with a positive TST and/or IGRA. In many of these studies, participants underwent more than one test at the outset, allowing comparison of the predictive accuracy of different tests. Several systematic reviews of these studies have been performed.

Reviews have focused on studies performing TST and at least one IGRA in specific populations that are frequently candidates for LTBI testing: children, immunocompromised individuals, and migrants from high- to low-TB incidence countries. In one review including 17 studies, children with positive IGRA had higher risk of TB in two studies but not three others, there was no difference between TST and IGRA in pooled risk of TB among immunocompromised individuals, and three studies including migrants demonstrated inconsistent results [47]. Subsequently, the same authors found that sensitivity to predict future incident TB was greatest with the TST if a cut-point of 5 mm was used (in migrants and children), while the QFT had highest sensitivity in immunocompromised individuals. In all three populations, the highest specificity was observed using a TST of 15 mm as criteria for a positive test [48].

Findings of earlier reviews differed from one another; one concluded there were no significant differences between TST or IGRA [49], and another concluded that IGRA had significantly higher positive predictive values [50]. These disparate findings reflected differences in inclusion criteria, quality assessments, and meta-analysis methods. However, in both reviews the pooled risk of active TB among individuals with a positive IGRA was 2.1 percent; emphasizing that the great majority of individuals with a positive IGRA do not develop active TB. Therefore, the decision to treat also must be based on clinical and epidemiologic assessment of risk for progression to active TB, not just the test result.

Role of dual testing

Use of IGRA to confirm TST — Use of IGRA as a confirmatory test for TST has been shown to be effective in contact tracing. In this approach, individuals with positive TST undergo subsequent IGRA testing; LTBI treatment is offered only to those with positive IGRA (algorithm 1). In one study including more than 1800 contacts in school-based outbreaks in Korea, one-third of TST-positive schoolchildren were also IGRA positive; this strategy identified all TST-positive schoolchildren who later developed active TB [51]. In another study including more than 870 household contacts in Spain, this sequential testing strategy was found to be noninferior to tuberculin testing alone for TB prevention [52]. These studies suggest that this strategy may be reasonably used to identify those at highest risk of developing active TB.

If an IGRA test is needed to confirm a TST result, it has been suggested that blood be drawn for IGRA within three days of TST placement (ie, at the time of the TST reading), due to a potential concern that prior TST may boost subsequent IGRA results. However, in a systematic review including five studies and more than 200 individuals persons, no significant boosting effect from prior TST on subsequent IGRA results was observed [53]. In theory, TST boosts IGRA results only in individuals with prior exposure to M. tuberculosis antigens; it is well established that the TST itself does not induce a cell-mediated immune response to M. tuberculosis. However, the studies that described boosting of IGRA by TST were not designed to distinguish whether this occurred only in those with true LTBI. Regardless of remaining uncertainties regarding potential boosting, a positive IGRA result obtained after prior TST should be interpreted as a positive test for LTBI.

Use of IGRA and TST to assess TB risk — Use of both IGRA and TST can be useful in certain circumstances [21]. If a first test is positive, but the risk of LTBI treatment is high or the pretest likelihood of LTBI is very low, a positive second test (with an assay of the alternative type, eg, using TST for the second test if IGRA was used for the first test, or vice versa) increases the risk of active disease. Similarly, if a first test is negative, but the likelihood of LTBI is high, or, if infected, the risk of disease is high (eg, in the setting of immunosuppression), then results of a second test (with an assay of the alternative type) may be useful; in such a case, a discordant result on the second test implies a higher risk of disease than concordant negative tests. The risk of TB disease is highest with concordant positive results, lowest with concordant negative results, and intermediate with discordant results [21].

COVID-19 vaccination and LTBI testing — There is no biologic reason to suspect that inactivated vaccines affect TST or IGRA results; however, unless testing for LTBI is considered urgent, the NTCA 2021 guidance recommends deferral of testing until four weeks after administration of the last coronavirus disease 2019 (COVID-19) vaccine dose. Further guidance may be found online [54].

History of NTM infection — For individuals with known prior history of nontuberculous mycobacterial (NTM) infection, the IGRA is preferred over TST since NTM infection may be associated with a false-positive TST result. Exceptions include Mycobacterium kansasii, Mycobacterium szulgai, and Mycobacterium marinum, which affect both the TST and IGRA.

Low risk settings — In low-risk situations where LTBI testing is not medically indicated (eg, individuals at low risk for progression to active disease if latent infection is present) but required by law or credentialing bodies, an initial positive test result (IGRA or TST) warrants confirmatory testing (TST if the first test was an IGRA or vice versa); the person should be considered to have LTBI only if both tests are positive [1].

Test interpretation — Issues related to interpretation of the TST and IGRAs are discussed separately. (See "Use of the tuberculin skin test for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults" and "Use of interferon-gamma release assays for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults".)

Guidelines vary by country — The approach to LTBI testing varies by country. Some guidelines favor use of the TST over IGRA, some favor use of either assay interchangeably, and some favor a two-step approach (TST followed by IGRA) in certain circumstances (table 9). Many guidelines favor more than one approach, depending on the risk group tested.

United States — United States guidelines are discussed above [1,2,46]. (See 'Choosing between TST and IGRA' above.)

Canada — Guidelines on IGRAs were published in the seventh edition of the Canadian TB Standards in 2013 [55]. According to the guidelines, both TST and IGRA are acceptable alternatives for LTBI diagnosis. Either test can be used for LTBI screening in any of the situations in which testing is indicated, with preferences and exceptions noted below.

●Circumstances in which neither TST nor IGRAs should be used include:

•Testing individuals who have a low risk of infection and a low risk of progression to active TB disease if they are infected. However, low-risk individuals are commonly tested before exposure, when repeat testing is likely. In this situation, TST is recommended (discussed under the third heading below); if the TST is positive, then an IGRA may be useful to confirm a positive TST result to enhance specificity.

•Diagnosis of active TB

•Routine or mass screening for LTBI of immigrants

•Monitoring treatment response to antituberculous therapy

●Individuals for whom IGRAs are preferred for testing (but TST is acceptable) include:

•Individuals who received BCG as a vaccine after one year of age and/or have received BCG vaccination more than once

•Individuals from groups that have historically poor rates of return for TST reading (such as homeless individuals or illicit drug users)

●The TST is recommended for testing (and an IGRA is NOT acceptable) for circumstances in which repeat testing is planned to assess for risk of new infection (ie, conversions); examples include repeat in a contact investigation or serial testing of individuals with potential for ongoing exposure (such as health care workers [HCWs], corrections staff, or prison inmates).

Routine dual testing with both TST and IGRA is not warranted. However, there are circumstances in which results from both tests (sequentially, in any order) may be helpful to enhance overall sensitivity:

●Risk of infection, progression to disease, and a poor outcome are high (table 6).

●An initial IGRA result is indeterminate, borderline, or invalid and a reason for testing persists.

United Kingdom — The United Kingdom National Institute for Health and Care Excellence guidelines were updated in 2016 [28]; they were previously published in 2006 and 2011 [56,57].

General recommendations include the following:

●LTBI testing should be offered to close contacts of patients with pulmonary or laryngeal TB, individuals who are immunocompromised and at high risk of TB, and new entrants from high-incidence countries presenting for health care.

●The upper age limit for offering to test and treat latent infection is 65 years.

●In any patient, TST is considered positive if skin induration is ≥5 mm, regardless of BCG history.

●Patients should be assessed for active TB if any test for latent infection is positive; if active TB is excluded, treatment for latent TB infection should be offered.

Specific recommendations include:

●TST is recommended for adult contacts (aged 18 to 65 years) of people with pulmonary or laryngeal TB. If the TST is positive (induration ≥5 mm, regardless of BCG history) treatment for latent TB infection should be initiated once active TB has been ruled out.

●Both TST and IGRA are recommended for immunocompromised adults (although intended to increase sensitivity in this high-risk group, there is no specific evidence to support this practice); LTBI treatment is warranted if either test is positive. Immunocompromised children should be referred to a specialist.

●TST is recommended for new entrants from high-incidence countries who present to health care services; an IGRA can be offered if TST not available.

●IGRA alone may be used for outbreak situations and other circumstances in which follow-up for TST interpretation may be difficult.

Europe — The European Centre for Disease Prevention and Control guidelines on IGRAs were published in March 2011 [58]. The main recommendations are:

●IGRAs should not replace the existing standard diagnostic methods for the diagnosis of active TB, and a negative IGRA result does not exclude active TB disease.

●For LTBI diagnosis, IGRAs may be used as part of the overall risk assessment to identify individuals for preventive treatment (eg, immunocompromised persons, children, close contacts, and recently exposed individuals).

●In high TB incidence countries, there is no added value in using IGRAs to diagnose LTBI, as the focus of prevention and control is to identify and treat active cases. In low TB incidence countries, given the evidence available, IGRAs could be used in contact-tracing algorithms applying the two-step approach (following TST, in TST-positive subjects).

●In immunocompromised individuals, as it is essential to maximize sensitivity, the simultaneous use of TST and IGRAs could be beneficial in identifying LTBI. However, in immune-compromised individuals, IGRAs should not be used to exclude LTBI and/or active TB.

Other countries — World Health Organization guidelines published in 2020 recommend that either TST or IGRA may be used but noted that IGRAs are more costly and more technically complex to perform than the TST [3]. Hence, the decision to use TST or IGRA should be based on operational considerations including the availability of trained personnel, laboratory facilities, and adequate resources to pay for the tests.

Approach to serial testing — Serial testing (eg, annual screening) is warranted for residents and employees of homeless shelters and correctional facilities, given the risk of TB exposure in these settings. In addition, serial testing may be reasonable in some health care facilities for HCWs at increased risk for occupational exposure to TB (such as pulmonologists or respiratory therapists) or for HCWs in certain settings (such as emergency departments) [15]. (See 'Homeless shelters and correctional facilities' above and 'Health care workers' above.)

Approximately half of the lifetime risk of reactivation occurs in the first one to two years following infection, so it is important to target those who convert their TST or IGRA for treatment. (See "Treatment of latent tuberculosis infection in HIV-uninfected nonpregnant adults".)

United States guidelines recommend use of either TST or IGRA for serial testing screening (algorithm 2) [1]. Conversion of a TST or IGRA from negative to positive indicates new infection. Given the difficulties of interpretation of serial IGRA testing, we favor use of TST for LTBI screening in individuals who are tested on a regular basis.

If the TST is used, a two-step TST protocol should be followed (ie, if the first TST is negative, then a second TST is performed within one to four weeks). For the TST, conversion is defined as a new positive test (with an increase of ≥10 mm) in the size of the TST reaction compared with the previous two years. (See "Use of the tuberculin skin test for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Serial testing'.)

For the IGRA, conversion is defined as a change from a negative to a positive result as determined by the cut point of the assay used; however, the use of a simple IGRA conversion cut point may be problematic due to inherent variability of these tests [18,59]. (See "Use of interferon-gamma release assays for diagnosis of latent tuberculosis infection (tuberculosis screening) in adults", section on 'Serial testing, conversion, and reversion'.)

EXCLUDING ACTIVE TB — All individuals with a positive test for TB infection (positive tuberculin skin test or interferon-gamma release assay result) warrant evaluation to exclude active TB prior to initiation of treatment for LTBI. Such evaluation is important to minimize the risk of drug resistance associated with inadvertent monotherapy of active TB.

The evaluation includes clinical history, physical examination, and chest radiograph. Active TB may be asymptomatic in patients with HIV infection [60,61].

Patients with relevant clinical manifestations (cough >2 weeks' duration, fevers, night sweats, weight loss) and/or abnormal chest radiograph should submit three sputum specimens (obtained via cough or induction at least eight hours apart and including at least one early-morning specimen) for acid-fast bacilli smear, mycobacterial culture, and nucleic acid amplification testing [62]. Symptoms related to extrapulmonary sites should also be evaluated as appropriate. (See "Diagnosis of pulmonary tuberculosis in adults".)

For specimens collected in a health care setting, standard respiratory precautions should be used (negative pressure room and N95 mask for health care workers). For inpatients (who should be in a respiratory isolation room), sputum collection should be performed within the room. For outpatients, sputum should be collected with care to minimize exposure to others. (See "Tuberculosis transmission and control in health care settings".)

The chest radiograph is normal in 85 percent of patients with latent TB [63]. A chest radiograph is considered abnormal if it demonstrates parenchymal abnormalities, particularly opacification of the upper lobe or superior segment of the lower lobe. Radiographs demonstrating stable upper lobe fibro-nodular disease or calcified granulomas are considered to demonstrate evidence of previous active TB and indicate the patient is at increased risk of reactivation (table 6). In a systematic review including 25 studies of patients with LTBI, 15 percent (95% CI 12-18 percent) had findings on chest radiograph suggestive of prior TB infection (such as calcified nodules, noncalcified nodules, pleural thickening, and fibrotic scarring) [63]. These lesions may be difficult to distinguish from lesions of active TB; therefore, in the setting of any radiographic abnormality, we have a low threshold for obtaining sputum mycobacterial cultures.

In resource-limited settings (where chest radiography and laboratory sputum evaluation may be less readily available), HIV-infected patients who are not on antiretroviral therapy (ART) and have none of the four symptoms may proceed with initiation of LTBI treatment [38,64]. This approach is based on a meta-analysis including more than 9000 HIV infected patients who were not on ART (among whom the prevalence of active TB among HIV-infected patients was 5 percent) in which absence of these clinical manifestations had a negative predictive value of 98 percent [60]. However, in HIV-uninfected children ≥5 years, HIV-uninfected adults, and HIV-infected individuals who are on ART, the sensitivity of symptom screen alone is less than 50 percent, and chest radiography is warranted [3].

For people living with HIV infection (PLHIV) and CD4 <200 (and especially if CD4 <50) in high TB incidence settings, a urine lipoarabinomannan (LAM) test may be a useful adjunctive tool; however, the implications and optimal management of PLHIV with positive urinary LAM test warrants further study. In a retrospective study including 563 PLHIV with CD4 <50 (including 28 with a positive urine LAM test and 535 with a negative urine LAM test), the rate of active TB was 18 versus 3 percent, respectively [65]. (See "Diagnosis of pulmonary tuberculosis in adults", section on 'Patients with HIV infection'.)

Patients with LTBI warrant HIV testing. (See "Screening and diagnostic testing for HIV infection".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Diagnosis and treatment of tuberculosis".)

SUMMARY AND RECOMMENDATIONS

●The goal of testing for latent tuberculosis infection (LTBI) is to identify individuals who are at increased risk for the development of active tuberculosis (TB) disease and therefore would benefit from treatment of LTBI. Only those who would benefit from treatment of LTBI should undergo testing, so a decision to test presupposes a decision to treat if the test is positive. (See 'Indications for LTBI testing' above.)

●Testing for and treatment of LTBI should be pursued for individuals in two categories (table 2 and table 3 and table 6 and algorithm 1 and algorithm 2):

•Individuals at risk for new infection

•Individuals at risk for progression from latent to active TB because of underlying conditions

●Individuals at risk for new infection include contacts of patients with untreated active respiratory TB. Health care workers (HCWs) at risk of exposure in regions with high TB incidence rate should undergo pre-employment baseline testing prior to exposure, followed by subsequent annual testing (algorithm 2). Those who live or work in homeless shelters or correctional facilities should also undergo annual testing. (See 'Increased risk of new infection' above.)

●In regions with low TB incidence rate, HCWs in facilities with low tuberculin skin test (TST) conversion rate need not undergo routine serial TB testing, in the absence of a known exposure or ongoing transmission. Routine serial TB testing may be reasonable for HCWs at increased risk for occupational exposure to TB (such as pulmonologists or respiratory therapists) or for HCWs in certain settings (such as emergency departments); such institutional policies should be individualized. (See 'Health care workers' above.)

●We favor use of the TST for serial testing, given the difficulties of interpretation of serial interferon-gamma release assay (IGRA) testing. If the TST is used, a two-step protocol should be followed (ie, if the first TST is negative, then a second TST is performed within one to four weeks). (See 'Approach to serial testing' above.)

●In individuals with increased risk of reactivation, the approach to testing should be individualized depending on the underlying condition, as summarized in the tables (table 2 and table 6) (see 'Increased risk of reactivation' above and 'Choosing between TST and IGRA' above):

•For diagnosis of LTBI in adults with low-to-intermediate risk of progression to active disease, IGRA is preferred over TST, particularly for patients who are unlikely to return to have the TST read and for patients with a history of Bacille Calmette-Guérin vaccination.

•For diagnosis of LTBI in adults with high risk of progression to active disease, either IGRA or TST is acceptable. A dual testing strategy may be used, in which a positive result from either test is considered positive.

●The approach to balancing the risk of disease and the risk of treatment depends in part on the treatment regimen utilized. The relationship between the risk of adverse events and increasing age is well documented for isoniazid. In individuals with conditions conferring moderate or slightly increased risk who would be treated with isoniazid, we carefully assess the balance of risks and harms in individuals >50 years of age (especially individuals >65 years of age). For other LTBI regimens, such as isoniazid-rifapentine for three months or rifampin for four months, the relationship between the risk of adverse events and increasing age is less clear. Until further data on the age-related risks of adverse events with these regimens are available, it is reasonable to adhere to the same cautions as for isoniazid. (See 'Balancing risk of disease and risk of treatment' above.)