Establishing a microbiological diagnosis of active TB is paramount for patient care and infection prevention at the healthcare and community levels. Multiple diagnostic tests are available for active and latent TB diagnosis, and a healthcare system’s choice of diagnostic tests and workflow
depends on patient volume and available resources. State TB Medical Consultants (every state has one, and they can typically be reached through the state health department) and Regional Centers for Disease Control and Prevention (CDC) TB Centers of Excellence (https://www.cdc.gov/tb/education/tb_coe/default.htm) can assist with obtaining diagnostic tests not routinely available.

Radiography At a minimum, standard anteroposterior chest radiographs should be obtained for all patients with suspected pulmonary and/or extrapulmonary TB. Classically, lower lobe infiltrates with pleural effusions are associated with primary infection, and upper lobe cavitary lesions are associated with reactivation disease. However, radiographic manifestations of TB are extremely variable, and pulmonary TB can be present in patients with normal chest radiographs. Atypical radiographic manifestations of TB are more common among immunosuppressed patients (eg, HIV).⁵ A parenchymal infiltrate with intrathoracic adenopathy should raise suspicion for pulmonary TB. Radiologists’ awareness of TB is important for infection prevention, as chest radiograph reports raising concern for TB have been associated with decreased time to obtaining appropriate diagnostic specimens.⁶

Microbiological Tests to Identify Active Tuberculosis Because TB may occur in almost any body site, a variety of specimens may be appropriate to collect, including sputum (expectorated or induced), bronchial washings or biopsy material, gastric aspirates, urine, cerebrospinal fluid, pleural fluid, purulent drainage, endometrial scrapings, bone marrow biopsy, or other biopsy or resected tissue. The 2016 U.S. National Guideline for Diagnosis of Tuberculosis lists essential laboratory tests for M tuberculosis diagnosis (Table 23-1). Since most institutions will not have all these tests on site, partnering with private and public health laboratories is necessary for effective usage of TB diagnostics. All specimens should be submitted at least for AFB smear and mycobacterial culture. All patients with culture-positive TB should have a first-line drug-susceptibility test.

Acid-Fast Bacilli Smear, Mycobacterial Culture, and Drug-Susceptibility Testing The detection of AFB in stained smears is the easiest and quickest procedure that can be performed, and it provides preliminary support for TB diagnosis. Also, the sputum smear aids in assessing the patient’s degree of infectiousness. The sensitivity and specificity of the AFB smears depend on the specimen source. Nontuberculous mycobacteria will also stain AFB positive. For example, the specificity of AFB smears performed on gastric lavage samples is compromised by the existence of commensal nontuberculous mycobacteria. Approximately 10 000 bacilli/mL are required for M tuberculosis detection in the AFB smear. Thus, AFB smear sensitivity is very low for specimens associated with disease sites with low bacillary loads (eg, CSF). The use of fluorescence microscopy allows the smears to be read much more rapidly than does standard microscopy and also increases sensitivity.⁷ Sputum specimen centrifugation and chemical processing to liquefy sputum are additional measures for increasing smear sensitivity.⁸ Overall, AFB smears are positive in 70% of pulmonary culture-confirmed TB cases. The average reported first specimen sensitivity is ˜54%. A second specimen increases sensitivity another approximately another 10%, and third specimen only increases sensitivity another 3%-5%. Thus, although collecting three sputum specimens is accepted practice in the United States, some experts suggest two sputum specimens suffice to aid in decision to discontinue airborne precautions.^9,10 Early-morning specimens have increased sensitivity, presumably due to pooling of respiratory secretions during sleep.

Many patients are unable to produce expectorated sputum, particularly when there is low burden of pulmonary disease. Obtaining an induced sputum is generally preferable prior to resorting to a bronchoscopy.¹¹ Sputum induction is an aerosol-generating procedure performed by administering nebulized hypertonic saline, and therefore adequate infection prevention measures should be observed (see Administrative and Environmental control sections). Blinded aspiration of bronchial secretions via endotracheal tube is an option for intubated patients. The number of required sputum to discontinue airborne precautions should remain the same, irrespective if spontaneous, induced, or obtained via blinded endotracheal suction. Since obtaining >1 bronchoscopy is generally impractical, only 1 respiratory sample will generally be available to aid in decision for discontinuation of airborne precautions among patients unable to produce expectorated and/or induced sputum. Many patients are able to produce expectorated sputa after a bronchoscopy, and
these samples may have higher yield compared to respiratory samples obtained during the bronchoscopy itself. Thus, the 2016 U.S. National Guidelines for Diagnosis of Tuberculosis recommend a postbronchoscopy sample for AFB smear and culture.¹¹ Although negative AFB smears are the most commonly used diagnostic test to aid in discontinuation of airborne precautions, it is important to recognize that smear-negative pulmonary TB cases are a source of a TB transmission.⁶⁵ Thus, it is important to have a plausible alternative diagnosis to TB prior to removing airborne precautions for patients with negative AFB smears.

All specimens should be inoculated in both solid and liquid culture media. Cultures are more sensitive compared to AFB smears and require 10-100 bacilli/mL to become positive. Liquid medium has higher sensitivity and lower time to positivity compared to solid medium. However, contamination rates are higher in liquid medium compared to solid medium.¹¹ Obtaining drug-susceptibility testing is paramount for patient and contacts care. Traditionally, M tuberculosis susceptibility testing is performed by culture-based phenotypic methods. Genotypic methods for predicting M tuberculosis susceptibility are a rapidly evolving field,^12,13 and we advise expert consultation to assist with test characteristics and limitations, particularly when phenotypic and genotypic testing yields discordant results. Most laboratories perform phenotypic susceptibility tests to first-line drugs rifampin, isoniazid, and ethambutol. Given technical challenges, not all laboratories perform pyrazinamide susceptibility testing. Second-line drug-susceptibility testing (eg, fluoroquinolones) is usually performed in specialized laboratories upon request. The CDC offers a molecular detection of drug resistance test for first- and second-line drugs upon request (https://www.cdc.gov/tb/topic/laboratory/default.htm). Although culture-positive samples are preferred, nuclear acid amplification-positive samples are considered.

Establishing TB diagnosis in pediatric patients is especially challenging, given the paucibacillary nature of the disease and inability to participate in sputum-collecting procedures. Fasting early-morning gastric aspirates can be collected in an effort to obtain samples for cultures when sputum cannot be obtained (a protocol for this procedure is available at https://www.currytbcenter.ucsf.edu/products/pediatric-tuberculosis-guide-gastric-aspirate-procedure).¹¹ Additionally, the American Academy of Pediatrics Red Book offers further recommendations for pediatric TB diagnosis, including indications for lumbar puncture in children <2 years of age.

Nucleic Acid Amplification Tests The 2016 U.S. National Guidelines for Diagnosis of Tuberculosis recommends obtaining a nuclear acid amplification test (NAAT) for all patients suspected of having pulmonary TB.¹¹ Two NAATs are endorsed by these guidelines, the Hologic Amplified Mycobacterium Tuberculosis Direct (MTD) test (San Diego, CA) and the Cepheid Xpert MTB/Rif test (Sunnyvale, CA). The main advantages of NAATs are quick turnaround time compared to conventional culture and ability to differentiate nontuberculous mycobacteria from M tuberculosis in AFB smear-positive samples. The above-mentioned NAATs have high specificity for M tuberculosis (85%-98%).^14,15

Sensitivity is also high when performed in AFB smearpositive samples (96%), but decreased in smear-negative samples (66%). Thus, the combination of a positive AFB smear and positive NAAT is highly suggestive of TB, whereas the combination of a negative AFB smear and negative NAAT does not exclude TB. The combination of a positive AFB smear with negative NAAT is highly suggestive of the presence of nontuberculous mycobacteria, but the presence of polymerase chain reaction inhibitors should be considered. Both the above-mentioned NAATs have positive controls that allow for assessment of inhibitors. Many laboratories offer in-house NAATs that should be interpreted with caution given substantial heterogeneity in performance, particularly if no validation procedures have been performed.¹⁶ The MTD and Gene Xpert MTB/Rif retain specificity when performed on extrapulmonary specimens, but sensitivity is decreased.

In addition to detecting M tuberculosis DNA, the Gene Xpert MTB/Rif detects rifampin resistance-associated mutations in the rpoB gene. Rifampin resistance is a surrogate for multidrug-resistant TB (MDR-TB) (resistance to at least rifampin and isoniazid). Although this test has good specificity, the positive predictive value is low in populations with very low prevalence of rifampin-resistant TB.

Given fast turnaround time, there is interest in NAAT usage in lieu of AFB sputum smear to aid decision to maintain airborne precautions among hospitalized patients suspected of pulmonary TB. The Gene Xpert MTB/Rif is FDA-approved for this indication based on a study indicating a negative predictive value of 99.7% for smear- and culture-positive pulmonary TB.^17,18 Two serial Gene Xpert MTB/Rif tests increased negative predictive value to 100%. As expected, the sensitivity was decreased for smear-negative culture-positive pulmonary TB. Implementation of an infection control algorithm using the Gene Xpert MTB/Rif can decrease median patient time on airborne isolation, time to hospital discharge, and hospital cost.¹⁹

At the time this chapter was written, several non-FDA-approved NAATs were available, including the Gene Xpert Ultra (more sensitive compared Gene Xpert MTB/Rif) and the MTBDRline probe assays (Hain Lifescience, Nehren, Germany).¹³ Additionally, a new Gene Xpert cartridge for second-line drug resistance testing is under development.²⁰

Genotyping Genotyping of M tuberculosis is used to determine the clonality of M tuberculosis strains and is essential for studying TB epidemiology and investigating outbreaks, including nosocomial transmission.²¹ Additionally, this technology has been instrumental in the identification of several pseudo-outbreaks caused by laboratory culture cross-contamination. The CDC National TB Genotyping and Surveillance Network aims to genotype at least one isolate from all cases of culture-confirmed TB in the United States. Currently, a cultured specimen is required for genotyping, another reason to pursue culture confirmation of all TB cases. Accomplishing this goal requires private laboratories to send all cultures positive for M tuberculosis complex to the state public health laboratories. Each state laboratory submits isolates to central genotyping labs funded by the CDC. Several genotyping techniques are available; the CDC implemented whole genome sequencing technology in 2018.²²

Adenosine Deaminase and Free Interferon Gamma Establishing the diagnosis of extrapulmonary TB can be difficult. To this end, the ancillary assays adenosine deaminase and free interferon gamma are endorsed by the 2016 U.S. National Guidelines for Diagnosis of Tuberculosis for selected extrapulmonary sources.¹¹

When considering the diagnosis and treatment of latent TB infection, it is always necessary to carefully exclude active disease. TSTs and IGRAs do not distinguish active from latent TB, and it is prudent to obtain a chest radiograph (in addition to history and physical) before a positive test is ascribed to latent infection (even when a person is asymptomatic). When interpreting LTBI diagnostics, it is important to understand there is no gold standard for LTBI diagnosis. Thus, test sensitivity is generally derived from testing subjects with known current or past active TB disease, and test specificity is generally derived from testing subjects unlikely to have TB infection. The TST and IGRA are the two diagnostic modalities to investigate M tuberculosis exposure in current use.¹¹ Repeat testing for people with prior positive test due to latent or active TB infection is not recommended as a negative test does not indicate M tuberculosis clearance and a persistently positive test does not necessarily indicate re-exposure. Importantly, both active and latent TB can cause a positive TST/IGRA, and, conversely, both active and latent TB can be associated with negative TST/IGRA results. Both tests measure delayed-type hypersensitivity reaction to M tuberculosis antigens, which develops within 2-10 weeks of exposure. Thus, if a TST/IGRA is negative soon after an identified exposure, a repeat test should be obtained 8-10 weeks after the exposure has ended.²³ The CDC guidelines were updated in 2019 to recommend against routine serial TB screening of HCP. However, for any identified exposure to active pulmonary TB, postexposure testing of HCP is recommended, and results should be compared with the baseline testing (which should be done at the time of hire).²³ The same recommendations apply in selected cases of exposure to extrapulmonary TB (eg, operating room personnel performing incision and drainage of M tuberculosis abscesses without adequate use of respiratory protection).

Tuberculin Skin Test The tuberculin skin test (TST) remains the most commonly used method to investigate evidence of prior exposure to M tuberculosis.^11,24 The test is performed by intradermal administration of 0.1 mL purified protein derivative (PPD) tuberculin into the skin of the volar surface of the forearm (Mantoux technique). Tests should be read by a trained health professional between 48 and 72 hours after injection. The basis of reading is the size of induration, which should be measured transversely to the long axis of the forearm and recorded in millimeters. The CDC website describes this process (https://www.cdc.gov/tb/topic/testing/tbtesttypes.htm). The induration results from a delayed-type hypersensitive reaction to the PPD antigens. False-negative test results can occur due to impaired cellular immunity (eg, HIV infection), and falsepositive results can occur due to environmental mycobacteria exposure and BCG vaccination. Causes of false-negative and false-positive reactions are shown in Table 23-2.

TST interpretation is risk-stratified based on probability of M tuberculosis exposure and individual risk factors for developing active TB. A reaction ≥5 mm is considered positive in persons with HIV infection or severe immunosuppression, persons with close contacts of infectious TB cases, or persons with abnormal chest radiographs consistent with TB.¹¹ A reaction ≥10 mm is classified as positive in persons who do not meet the above criteria but who have other risk factors for TB, such as HCPs and people living with diabetes (reference¹¹ for complete list of criteria). Finally, a reaction ≥15 mm is considered positive when M tuberculosis exposure is unlikely and/or risk for developing active TB is low. TST interpretation among BCG vaccine recipients should consider age at time of vaccination, number of BCG doses, and time between vaccination and testing. The background TST reactivity among those receiving a single BCG dose at birth and tested ≥10 years later is low.²⁵

Repeated tuberculin skin testing does not induce falsepositive reactions. However, the test can restore reactivity among those previously exposed to M tuberculosis (and other mycobacteria that cross-react with PPD). This socalled booster effect can be seen on a second test done as soon as a week after the initial stimulating test, and the booster effect can persist for a year and perhaps longer. The booster effect can lead to apparent TST conversion.
To avoid this problem, a two-step testing protocol is recommended upon HCP hire (preplacement) and when serial testing is recommended. In two-step testing, an initial TST is performed. If the reaction to the first test is negative, a second test should be given 1-3 weeks later. If the reaction to the second of the initial two tests reaches the appropriate cut point for a positive result in the patient, this probably represents a boosted reaction. On the basis of this second test result, the person should be classified as being previously infected and managed accordingly. An increase in TST reaction size ≥10 mm within 2 years is considered a “conversion” and warrants evaluation for active TB and consideration for LTBI treatment.²⁶

Several PPD preparations are available globally. Aplisol (Parkedale Pharmaceuticals, Rochester, MI) and Tubersol (Pasteur Mérieux Connaught USA, Swiftwater, PA) are the current products licensed by the U.S. Food and Drug Administration.²⁷ Aplisol and Tubersol reactivities are similar enough to replace each other, particularly at times of product shortage.^27,28

Interferon-Gamma Release Assays Interferon-gamma release assays (IGRA)s are blood-based in vitro tests to investigate M tuberculosis exposure.¹¹ There are two commercially available FDA-approved tests, the QuantiFERON-TB Gold Plus (QFT Gold Plus, Cellestis Limited, Carnegie, Victoria, Australia) and the T.SPOT TB Test (T-SPOT, Oxford Immunotec Ltd., Abingdon, United Kingdom). Both tests use mixtures of synthetic peptides simulating two proteins present in M tuberculosis: early secretory antigenic target-6 (ESAT-6) and culture filtrate protein-10 (CFP-10). ESAT-6 and CFP-10 are secreted by all M tuberculosis and pathogenic M bovis strains, but are absent from BCG vaccine strains and from commonly encountered nontuberculous mycobacteria except Mycobacterium kansasii, Mycobacterium szulgai, and Mycobacterium marinum. Both tests include negative (nil) and positive (mitogen) controls. The QuantiFERON test quantifies soluble interferon gamma released by antigen stimulation using an enzyme-linked immunosorbent assay (ELISA), whereas the T.SPOT is based on the number of T cells releasing interferon gamma upon antigen stimulation using an enzyme-linked immunospot assay (ELISPOT).

These tests are interpreted by calculating the difference between interferon-gamma release induced by mycobacterial antigens and by negative control. An “indeterminate” result is reported when there is high interferon-gamma release by the negative control and/or insufficient interfer-on-gamma release by the positive control. Indeterminate results could be due to technical issues (eg, inappropriate incubation time) or immunosuppression. It is reasonable to repeat the test as recurring low mitogen response could indicate inability to generate an immune response (“anergy”). The T.SPOT has an additional “borderline” result category that should be interpreted in conjunction with the prerest probability of M tuberculosis exposure. As opposed to the TST, IGRA interpretation is not risk-stratified. However, it is well described that positive-to-negative reversions are more common at the lower end of positivity.²⁹ It has been recognized that TST can boost IGRAs, but to date, this phenomenon has been reported only among subjects with pre-TST placement positive IGRAs. Boosting can occur starting 3 days after TST placement and may persist for months.³⁰

A meta-analysis published in 2008 found that TST, prior versions of the QuantiFERON test (QFT Gold and QFT Gold In-Tube) and T.SPOT have similar specificity, while IGRAs (in particular T.SPOT) have higher sensitivity compared to TST.³¹ However, TST specificity was decreased among those with history of BCG vaccination. Prior QuantiFERON versions had one M tuberculosis antigen tube, whereas the currently available QuantiFERON test (QFT Gold Plus) has two M tuberculosis antigen tubes.³² The additional tube contains peptides designed to stimulate CD8+ T cells and is purported to have increased sensitivity for active TB disease, following recent M tuberculosis exposure, and among immunosuppressed patients, including people living with HIV. There are limited data on QFT Gold Plus among lowrisk populations, such as U.S.-based HCP.³³

Choice of Screening Test for M. tuberculosis infection or Latent TB infection (LTBI) The 2016 U.S. National Guidelines for Diagnosis of Tuberculosis offer recommendations for choice between TST and IGRAs according to the age, likelihood of M tuberculosis infection, and risk for progression to active TB¹¹ (Table 23-3).

The advantages of the IGRA tests are that they require only one patient visit and retain specificity among BCG-vaccinated subjects. IGRA’s disadvantages include need for phlebotomy and to process blood timely (see package insert for updated recommendations). There are limited data on IGRA performance in pediatric populations. Some evidence suggests decreased sensitivity compared to TSTs and increased risk for invalid test results among children <5 years old.^34,35 IGRAs were initially heralded as test with less variability compared to TSTs. However, there are various sources of IGRA variability and discordant test results are not uncommon, especially when initial result is close to cutoff value.³² This represents a disadvantage for IGRAs, particularly when serial testing is indicated. Discordant TST and IGRA results are not uncommon, including among HCP, and there is no consensus on how to interpret discordant results.^11,29 Current U.S. guidelines recommend dual testing in selected situations.¹¹

LTBI testing is relevant to TB infection prevention chiefly for the screening of HCP. As discussed above, HCP should have LTBI testing at baseline and after any identified exposure or high-risk activity.²³ If TST is chosen for baseline testing, a two-step testing is recommended to increase the sensitivity for detecting true LTBI. Positive results of baseline testing for LTBI among HCPs should be interpreted with consideration of possibility of prior exposure to active TB. For many HCPs (U.S. born and born in the 1980s or later), the pretest probability for having LTBI is so low that a falsepositive test result is more likely than a true-positive result. For this reason, baseline testing in such HCP has been recommended to use a 15-mm cutoff for PPD positivity, and, in institutions where IGRA tests are used, an initial positive IGRA should be followed by a second confirmatory test (either a second IGRA or a TST).²³ Testing HCP after an identified or suspected exposure to active TB should employ the same test modality that was used for baseline testing.

A review of healthcare-associated TB outbreaks in the United States reveals factors that have contributed to TB transmission and informed current policies and procedures.⁴² Before effective chemotherapy was available,
almost 10% of people admitted to a New York state sanatorium between 1920 and 1953 were physicians. However, TB was not recognized as an occupational hazard to HCP until well into the 20th century. In fact, early 20th century authoritative figures described concerns for healthcare-associated TB transmission as “fairy tales” and “phthisiophobia.”^43,44 To some extent, the epidemiology of healthcare-associated TB in the United States mirrors the TB epidemiology in the general population.

With the decline in TB incidence among young U.S. adults in the 1920s, the baseline prevalence of TST reactors among new nursing and medical students also decreased. This allowed for the recognition of increased TST conversions among these students. For example, the TST conversion rate was 100% among nursing students enrolled in a Philadelphia-based teaching hospital between 1935 and 39.⁴⁵ Among University of Minnesota medical school graduates between 1912 and 1932, 7% developed radiographic lesions consistent with active TB. During this era, infection control procedures were not implemented due to the belief that curbing infections was impossible.⁴⁴ In the early 1930s, the University of Minnesota implemented serial TST for nonreactors at medical school enrollment, pre-employment TST and chest x-ray for HCP, and universal screening of all inpatient admissions with a TST and chest x-ray, regardless of presenting symptoms.⁴⁶ Additionally, emphasis was placed on teaching medical students how to recognize active TB and implementation of “contagious disease techniques.” TST conversion rates decreased from 68% in the 1936 medical school class cohort to 16% in the 1951 cohort. Similarly, a dramatic decrease in TB among Los Angeles HCP was reported after implementation of routine hospital admission chest x-rays for all patients.⁴⁷ There was a decrease in the importance of TB infection control measures following to development of effect TB chemotherapy and the decrease in TB incidence in the United States. However, during the mid-1980s and early 1990s TB resurgence, multiple U.S. healthcare-associated TB outbreaks occurred and were closely linked to the HIV/AIDS epidemic and also and associated with the emergence of MDR-TB (in New York City).³⁶ These outbreaks renewed attention to TB infection control culminating in the publication of the 2005 CDC guidelines on the prevention of nosocomial transmission of TB.⁴⁸ Overall, these outbreaks show the tremendous rapidity and extent of spread that can occur when persons who have undiagnosed TB are brought together with highly vulnerable, immunosuppressed persons in an enclosed and relatively densely populated environment in the absence of adequate infection control precautions.

Strengthening of infection control procedures and a general decrease in TB cases in the United States have dramatically decreased the burden of TB among U.S. HCPs.^49,50 Though the frequency of nosocomial transmission and size of healthcare-associated outbreaks in the United States have markedly decreased in the 21st century, they continue to be reported^51,52,53 and, as such, demonstrate the ongoing need for vigilance in the implementation of effective infection control. Below, we review factors that influence the epidemiology of healthcare-associated TB and illustrate those with published healthcare-associated outbreaks.

The factors that influence the epidemiology of healthcareassociated TB are the joint probabilities that exposure to M tuberculosis will occur, exposure will result in infection, and infection will lead to active TB (Fig. 23-1).⁵⁴ Events or conditions that alter any of these probabilities (the probability of exposure, infection, or progression to active TB) may result in changes in the epidemiology of TB in a healthcare facility.

Factors That Affect the Likelihood of Exposure In a healthcare facility, the likelihood of exposure to M tuberculosis may be affected by factors such as the prevalence of infectious TB among patients and HCP; the degree of crowding in the facility (eg, shared rooms); the effectiveness of the facility’s TB infection control program in rapidly identifying, isolating, and treating persons with infectious TB; and the effectiveness of engineering controls, such as directional airflow and booths for cough-inducing procedures, in preventing the spread of contaminated air throughout the facility.

The single most important factor increasing the likelihood of exposure to M tuberculosis in healthcare and community settings has been failure to promptly identify patients or HCPs with infectious forms of TB.⁵⁵ It is common for a person with TB to have multiple healthcare encounters prior to diagnosis.⁵⁶ Almost 16% of persons diagnosed with TB in California between 2005 and 2011 had a hospitalization or emergency department visit with a respiratory complaint prior to TB diagnosis. Thus, it is not surprising that patients and HCPs with unsuspected TB are a common source of reported healthcare-associated exposure and outbreaks in the United States.^{53,57,58,59,60} Immunocompromised hosts at risk for multiple concomitant infections presenting with atypical forms of TB can be particularly challenging for infection control practice. Delayed TB diagnosis in an HIV-positive patient with concomitant Pneumocystis
jiroveci pneumonia and disseminated Mycobacterium avium was associated with a 19% TST conversion rate among exposed HCPs.⁶¹ Some of these lapses in hospital infection prevention were highly consequential to TB control efforts in the United States. For example, most MDR-TB cases in New York City between 1990 and 1993 were linked to nosocomial transmission.⁵⁷

Exposure to patients undergoing aerosol-generating procedures has been extensively documented as a risk factor for TB exposure, including exposure to patients receiving aerosolized pentamidine (at one time an extremely common treatment for people living with HIV) and among HCPs performing bronchoscopy.^59,62,63 HCPs regularly performing aerosol-generating procedures may have increased risk of TB exposure. Pulmonary fellows had higher rates of TST conversion compared to infectious diseases fellows in U.S.-based training programs in the early 1980s.⁶⁴ Although sputum AFB smear positivity is highly associated with TB infectiousness, a significant proportion of culture-positive TB cases in a low-burden setting may be smear-negative.⁶⁵ Healthcare-associated TB exposure from smear-negative patients has been documented, particularly after aerosol-generating procedures.⁶² Healthcareassociated TB exposures have also occurred following surgical manipulation of extrapulmonary sites. Following incision and drainage of a tuberculous soft tissue abscess, 80% of surgical suite HCP had a TST conversion.⁶⁶ High TST conversion rates were also reported among HCPs performing bedside dressing changes and debridement for a patient with cutaneous TB without wearing appropriate respiratory tract protection.⁶⁷