The rate of diagnosis of active TB in HCWs has consistently been reported to be higher than that for the general population in a number of studies conducted in countries with low and high TB prevalence. For example, in England and Wales, HCWs are 2.4 times more likely to contract active TB than other people, after taking into account socioeconomic status, ethnic group, age, and sex7. Recently, Chu et al.8 conducted a nationwide, population-based cohort study based on the Taiwan National Health Insurance Database for the years 2000–2010. They found that the incidence of active TB was higher for HCWs than for matched subjects, particularly for pulmonary TB compared with extrapulmonary TB. A study in South Korea investigated the prevalence of TB among HCWs employed at a university hospital and found that the prevalence ratio for nurses working in a TB-related department (medical intensive care unit, pulmonary medicine ward, or emergency department) was higher than for the general population, suggesting that some of the HCWs developed active TB disease via a hospital infection9.

In addition to active TB, it is also well known that HCWs have a substantially increased risk of occupationally acquired TB infection. Menzies et al.10 showed that the median prevalence of latent TB infection (LTBI) in HCWs was 63% in low- and middle-income countries and 24% in high-income countries, suggesting the risk of LTBI is high for HCWs worldwide. Recently, Casas et al.11 reported that the cumulative incidence of tuberculin skin test (TST) conversion was higher in HCWs who worked in high-risk areas in a tertiary referral hospital in Spain. In South Korea, a prospective study evaluated the annual incidence of LTBI among newly employed nurses at a tertiary care university hospital, and concluded that the annual risk of LTBI for these nurses was at least 3%12.

According to the CDC guidelines12, a TB infection control program should be based on the following three levels of hierarchy: (1) administrative control, which reduces the risk of exposure; (2) environmental control, which prevents the spread of the disease and reduces the concentration of droplet nuclei; and (3) respiratory protection control, which further reduces the risk of exposure in specific areas and circumstances. Of these, the first and most important level is the use of administrative measures to reduce the risk of exposure to people who might have TB disease. According to the CDC guidelines, this level comprises the implementation of a range of activities12; these are summarized in Table 1.

The administrative control measures according to the WHO policy3, which covers TB infection control in congregate settings (i.e., institutions or environments where people live in close proximity to each other) and households as well as health care facilities, are rather different from those recommended by the CDC. Along with environmental control and personal protective equipment, the WHO's policy for administrative control in health care facilities includes the following3.

Screening of HCWs for LTBI is another key component of the administrative control of hospital TB infection control programs. According to the facility risk classification recommended by the CDC (Table 2), the majority of general hospitals and university hospitals in South Korea correspond to the medium risk level for TB. According to the CDC guidelines1, in a setting classified as medium risk, all HCWs should receive a baseline TB screening test when they first start employment with the institution using a two-step TST to test for infection with M. tuberculosis. After this baseline testing, HCWs should receive annual TB screening using a TST (serial TST) (Table 2). Any HCWs with a baseline positive or newly positive TST result, or with the documentation of previous treatment for an LTBI or TB disease, should undergo chest radiography once to exclude active TB disease.

A TST conversion is defined as a TST induration that has increased in size by ≥6 mm and has a total size ≥10 mm. If such a TST conversion is observed in a HCW with a documented negative (<10-mm) baseline two-step TST result, this indicates a recent TB infection. Recent TST converters have a higher risk of TB development (by as much as 15-fold) than subjects with no risk factors, and approximately 5% of people newly infected with M. tuberculosiswill develop the disease in the 1–2 years after infection1314. Therefore, among people infected with M. tuberculosis, a TST converter is generally classified as high risk for the development of active TB15. Accordingly, the Korean TB guidelines revised in 2014 have classified anyone with a TST conversion within the previous 2 years as a high-risk group16. HCWs whose TST result has recently converted to positive should therefore be regarded as candidates for preventive chemotherapy.

When a TST is used for baseline testing, two-step testing is recommended for HCWs whose initial TST results are negative when a booster effect is considered1718. The concept and the interpretation of the two-step TST are as follows. (1) If the first-step TST result is negative, the second-step TST should be administered 1–3 weeks after the first TST was read. (2) If either the baseline first-step TST result is positive, or the first-step result is negative but the second step is positive, the HCW can be diagnosed as having a LTBI after active TB has been excluded. (3) If the first and second-step TST results are both negative, the person is classified as not infected with M. tuberculosis1. A positive result of the second step of a baseline two-step TST is thought to be due to a boosting effect that may result from a remote infection of M. tuberculosis. The CDC guidelines recommend that all HCWs receive a baseline two-step TST when first employed (Figure 1)1. In South Korea, most centers have not adopted this approach because of executive difficulties. However, according to one report, the proportion of HCWs who convert to positive in the second step of the two-step TST is not rare in South Korea; among 556 newly employed HCWs at a university hospital, 14.2% experienced a boosted reaction on the second TST19.

Although there is considerable experience of using the TST over several decades, this test has a number of drawbacks. In particular, the TST requires two or more patient visits to conduct the test because the results are only available 48 to 72 hours later. In addition, a bacillus Calmette–Guerin (BCG) vaccination can cause a false-positive result. Third, although rare, the injections may result in an adverse skin reaction. Finally, the TST result needs to be interpreted by a well-trained person.

Compared to the TST, the interferon-gamma release assay (IGRA), including the QuantiFERON-TB Gold In-Tube assay (QFT-GIT) and T-SPOT.TB test, has several distinct advantages. It requires only one patient visit, and the results are unaffected by HCW perception or bias and are available in 24 hours. In addition, a BCG vaccination will not cause a false-positive result. In this context, serial testing with IGRA is attractive because the approach avoids subjective measurement, which can be repeated without sensitization and boosting in subsequent tests, and eliminates the need for multiple visits for the interpretation of the results of the two-step test20.

Since the CDC published guidelines in 2005 indicating that the IGRA may be used in all circumstances in which a TST is recommended, including for the serial screening of HCWs1, there have been several reports of its use. Slater et al.21 reported data regarding the reproducibility of serial QFT-GIT in a large cohort of HCWs in a single North American health care institution. Among 9,153 HCWs who underwent two or more QFT-GIT tests, 8,227 showed an initially negative result, of whom 361 (4.4%) showed a second QFT-GIT result conversion over 2 years. A total of 261 of these HCWs (72.3%) with conversions underwent repeat short-term testing after the first positive result, with 169 (64.8%) reverting. Interestingly, these authors found that the QFT-GIT cut-off would need to be increased by more than 10-fold to yield an equivalent historical TST conversion rate. These results suggest that the manufacturer's definition of QFT-GIT conversion results is an inflated conversion rate, incompatible with the low-risk setting in North America.

Another important study involving HCWs at four health care institutions in the United States also indicated that IGRA is associated with a significantly higher conversion rate than the TST, and that many IGRA conversions appear to be false-positives22. In South Korea, Park et al.23 evaluated the conversion and reversion rates of IGRA in prospectively enrolled HCWs who were in contact with patients with active pulmonary TB. Of the 48 HCWs who received monthly QFT-GIT, 25 (52%) showed inconsistent results in the serial testing, confirming that fluctuations in QFT-GIT results were also common in HCWs in South Korea. The poor reproducibility and high conversion rate of QFT-GIT appear to stem from variability inherent within the test itself rather than being due to host or pathogen factors24. This finding has particular relevance for HCWs who make up the majority of those tested repeatedly. In this context, it is notable that Canadian guidelines published in 2013 discourage the use of serial IGRA tests for HCWs for this reason4. Further studies are required concerning the availability and the optimal cut-off level for QFT-GIT as a serial test for detecting LTBI in HCWs in South Korea.

In contrast to the conclusions of these reports, a recent study has suggested that T-SPOT.TB tests that adopt a borderline zone are reliable for HCW serial screening without the possibility of poor reproducibility or an inflated conversion rate25. However, given the limited number of studies of T-SPOT. TB tests with such a borderline zone for the evaluation of LTBI in HCWs, further studies will be needed to confirm this finding.