1. HPA East of England Regional Epidemiology Unit, Cambridge, England
2. Department of Primary Care and Public Health, Cardiff University, Wales
3. Bedfordshire and Hertfordshire Health Protection Unit, Bedfordshire, England
Several tuberculosis outbreaks in schools have been reported in the last decade, usually following transmission from an infectious staff member. The index case is rarely a child, although outbreaks of this nature have been documented [1,2,3,4,5].
In March 2007, a case of smear negative, culture positive tuberculosis in a nine-year-old was reported to the Bedfordshire and Hertfordshire Health Protection Unit in the United Kingdom. This preliminary report describes an ongoing outbreak investigation in the school attended by the child.
The school is situated in a town with an ethnically diverse urban area and a population of around 200,000 inhabitants. The incidence of tuberculosis cases is 45 per 100,000, meaning that the town meets the World Health Organization's definition of a high incidence area . The school involved is attended by 190 children, over half of whom are from non-white or mixed ethnic backgrounds.
The first case (presumed index case) was a nine-year-old boy who had been symptomatic with cough and fatigue since January 2007. He initially received a diagnosis of asthma but after having had haemoptysis he was investigated for tuberculosis. Sputum obtained was smear negative but Mycobacterium tuberculosis was isolated from sputum culture at the end of March 2007. The strain was found to be fully sensitive to anti-tuberculosis drugs and he was started on chemotherapy.
A total of 11 household and close contacts of the child were identified and screened in accordance with national guidelines for the United Kingdom . Five of these contacts were the immediate family of the first case and lived in the same house. The other six were relatives that lived together in another house. His parents and siblings had strongly positive Mantoux tests, as did all individuals in his relative’s house. They were all commenced on chemoprophylaxis. Additionally, an eight-year-old cousin showed changes in the chest X-ray and was started on chemotherapy with isoniazid, rifampicin and pyrazinamide.
The DNA fingerprinting from the first case was identical to the tuberculosis strain found in an adult close contact from 2001. The child had not been screened at the time and had not had further contact with this case in the previous two years.
As the child was symptomatic while at school for about three months and all the initial close contacts were strongly Mantoux positive, a decision was made to screen the 36 children who were in the same year as him and 12 staff in close contact with his class. This initial screening revealed that a large proportion of children (29/36) had strongly positive Mantoux tests. Six of these children also had chest X-ray changes consistent with active tuberculosis and were commenced on chemotherapy. Following this, in view of the high number of children requiring chemoprophylaxis or chemotherapy, a decision was made to screen all the children at the school. The school caters for children between the ages of eight to 12 years. There are two classes for each year group.
Following national advice, this was carried out by gamma-interferon T spot (T SPOT.TB Oxford Immunotec) blood test to expedite the process before the half-term holiday break and in view of the large numbers of children involved. All children in the school who were not on chemoprophylaxis or chemotherapy for tuberculosis (ie all pupils except those in the class of the first case) were invited for screening by T spot blood testing and chest X-ray examination in a mobile chest X-ray van on 22 and 23 May. An additional nine children who had left the school since September 2006 were also invited to be screened.
Just over 100 adults who had had more than eight contact hours with the school since September 2006 were invited for a chest X-ray to exclude them as the source of infection.
Preliminary epidemiological findings
There were a total of 189 children eligible for screening in the school, 64 of whom had latent infection and required chemoprophylaxis. To date, 17 children are receiving chemotherapy due to chest X-ray changes compatible with active tuberculosis, excluding the first case (17/189=9%). Seven of these are from the same class as the first case. Three children are still awaiting repeat T spot and three were not screened as they were absent from school on the screening days – they are being followed up.
During the clinical follow-up from the screening, another child in the school from a different class than the first case showed a positive T spot and chest X-ray changes. Gastric washings revealed acid fast bacilli indicating active respiratory tuberculosis. This child had had a productive cough and had been generally unwell for five months before diagnosis. Investigation of this child’s close contacts is currently ongoing. Samples from this child have been sent for culture and DNA fingerprinting to investigate whether or not this case is microbiologically linked to the first case.
The initial epidemiological investigation showed that the attack rates for both active and latent infection were highest in the year group of the first case (Year 5). The highest rate was in the first case’s class (Class 5, n=26 excluding the first case) with 85% (22/26) of children requiring chemoprophylaxis or chemotherapy. The overall attack rate in the school was 43% (81/189), with 64 requiring chemoprophylaxis only and 17 requiring chemotherapy.
The class teacher of the first case had a positive T spot but had a negative chest X-ray and was not symptomatic. She was therefore not the source of the outbreak. Ninety-two adults have had a negative chest X-ray to date and the others are being followed up for chest X-ray.
Attack rates were similar across all ethnic groups, suggesting that the network of transmission was predominantly within a classroom setting.
Interpretation of data collected so far
These initial findings are unprecedented. Although school outbreaks have been documented, the source of infection is usually an adult and no adult source has yet been identified here. Additionally, the presumed index case (first case) was only nine years old and was sputum-smear negative. National Institute for Health and Clinical Excellence guidelines  do not currently recommend screening the class of a pupil diagnosed with tuberculosis unless they have sputum smear-positive tuberculosis. However, in this case, because the child was symptomatic in school for three months and a large number of household and close contacts were strongly Mantoux positive, the decision was made to screen the class.
There is published evidence to support the increased relative risk of acquisition of latent or active tuberculosis with use of poorly ventilated or small classrooms and gymnasiums within school settings [5,8]. Visits to the school have revealed that it is relatively poorly ventilated, with teachers predominantly keeping doors closed in the winter months as the classrooms only have a blower heater. There is also a high level of mixing between different classes, especially within those in the same year group, which may partially explain the high attack rates in Classes 5 and 6.
The first case and the case of active tuberculosis picked up following the screening had both been unwell over a period of months. There should be a high index of suspicion of tuberculosis in patients presenting with persistent cough or other tuberculosis-related symptoms, especially in high incidence areas. Further epidemiological investigations are ongoing.