Assessing an outbreak of tuberculosis in an English college population

To the Editors:

In October 2008, a sixth form college student in England, UK, who had had a cough for 2 months was diagnosed with sputum smear-positive cavitatory pulmonary tuberculosis. An investigation led to the identification of 19 cases of active tuberculosis among 2,284 students. Here, we describe the outbreak and investigation into the factors associated with active tuberculosis.

The study population consisted of students enrolled at the college for daytime courses. All participants were aged >16 yrs. Members of staff who directly taught these students were also screened. Between October 2008 and December 2009, students, friends of the index case and staff at the college were interviewed. The interviews were used to collect demographic and clinical information, including symptoms suggestive of tuberculosis. At the same time, a blood sample was drawn from each participant for interferon-γ release assay (IGRA) testing [1, 2]. Following the screening of household contacts and friends of the index case, the students at the school were assessed in concentric circles of decreasing intensity of exposure in three groups (table 1) [3]. Those with a positive IGRA result from any screening round were recalled for chest radiography and clinical review at a respiratory medicine clinic. Preventative therapy was offered to all those with no evidence of active tuberculosis and a positive IGRA result who were <35 yrs of age, according to national guidelines [4]. Detailed results of the IGRA tests are described elsewhere [5].

Where relevant samples were available, following decontamination, direct microscopy and culture using a liquid culture system were performed and, where positive, subsequent testing for sensitivity to standard antituberculosis drugs, followed by 24-locus Mycobacterium interspersed repetitive unit–variable-number tandem repeats typing (MIRU-VNTR) [6]. In addition to a detailed clinical review, patients with suspected active tuberculosis had further tests, including computed tomography (CT) and bronchoscopy, where appropriate. Active tuberculosis was defined as any individual with culture-confirmed disease caused by Mycobacterium tuberculosis or a clinical presentation consistent with active tuberculosis where the clinician had decided to give a full 6 months of treatment.

Adjusted and unadjusted odd ratios were calculated using logistic regression analysis to investigate factors associated with active tuberculosis. All variables considered a priori to be risk factors for tuberculosis, including age, sex, place of birth, ethnicity, contact with the index case and lack of bacille Calmette–Guérin (BCG) vaccination, were modelled. We investigated interaction between risk factors for tuberculosis using the likelihood ratio test. Data were analysed using the statistical software STATA (version 11; StataCorp, College Station, TX, USA).

The median age of the students was 17.8 yrs (interquartile range (IQR) 17.3–18.5 yrs) and 1,055 (46.2%) were male. Among the students, 49.6% were BCG-vaccinated, 90.2% were of white ethnicity and 18.8% had recently travelled to a high-incidence country. The participants travelled to a number of countries, including China, Russia, India, Bangladesh and several countries in sub-Saharan Africa; 400 (17.5%) had a positive IGRA. The median age of the staff was 46.8 yrs (IQR 38.5–55.3 yrs). No active tuberculosis cases were diagnosed among staff members.

The index case attended the college for <3 weeks during September 2008. Excluding the index case, 19 cases of active tuberculosis were diagnosed and treated. Eight cases were culture confirmed as M. tuberculosis; seven cases were confirmed as having an indistinguishable 24-locus MIRU-VNTR genotype (42234 2742511334 422423255) as the index case; a further case had 23 identical digits with a missing 24^th digit. Of the eight culture-confirmed cases, two were sputum smear and culture positive; one case was sputum smear-negative but culture positive; four cases were positive on culturing washings obtained from bronchoalveolar lavage; and one case was positive on culture of pleural biopsy material. The two sputum smear-positive cases were not symptomatic while attending the college. The diagnosis of cases who were not microbiologically confirmed was largely based on radiological findings, including CT. In total, nine cases were asymptomatic.

During the first two rounds of screening, nine cases of active tuberculosis were identified (including three who were also friends or extended family). Two cases outside the groups screened then presented clinically, prompting a third round of screening that identified a further eight cases. The yield therefore was 3, 2 and 0.5% in groups 1, 2 and 3, respectively. table 1 shows the characteristics of all cases. All isolates were fully susceptible to first-line antituberculosis drugs. The median age of the active tuberculosis cases was 17.2 yrs (IQR 17–18.3 yrs) and 12 (63.2%) out of 19 were male. In the fully adjusted model (table 2), only age was associated with the risk of active tuberculosis, with younger students being at a higher risk of tuberculosis (OR 0.23, 95 CI 0.07–0.75).

This study found that the only independent risk factor for active tuberculosis was age. Although the incidence of tuberculosis has increased in the UK over the last two decades [7], the rate of active tuberculosis identified in this college (875 cases per 100,000 persons) is several times higher than the age-equivalent rate for England (14.6 cases per 100,000). A limitation of the study was the relatively small number of active tuberculosis cases, reducing our ability to investigate the risk factors for active disease. The time taken to screen the entire college possibly contributed to the emergence of secondary cases. While this may, in part, question the “stone in the pond” approach [8], we contend that this incident is unusual. A time-to-event analysis would have been more appropriate to account for right censoring. The lack of dates of contact with infectious cases precluded hazard analysis.

We conclude that this outbreak resulted in transmission within the college, as evidenced by indistinguishable DNA fingerprints of the M. tuberculosis strains. Public health services should ensure the prompt detection and management of such outbreaks.

• ©ERS 2011