the bmj | 28 March 2015 31 What is the next investigation? If clinical evidence of active tuberculosis is excluded by the methods described above, the next step is to assess the patient for evidence of latent tuberculosis infection. This can be identifed by: •  The tuberculin skin test—intradermal injection of purifed protein derivative and measurement of skin erythema and induration, which corresponds to a delayed type hypersensitivity reaction in patients with previous exposure to tuberculosis (including latent tuberculosis). Previous vaccination with BCG will also cause a skin reaction but of lower magnitude than afer exposure to tuberculosis •  Interferon γ release assays (two commercially available: QuantiFERON-TB Gold In-Tube assay (QFT) and T-SPOT.TB), which measure interferon γ release by T cells in blood samples mixed with M tuberculosis antigens. 4 QFT measures the interferon γ concentration (IU/mL) using ELISA (enzyme linked immunosorbent assay) technology, whereas T-SPOT. TB reports the number of T cells producing interferon γ (spot forming cells). Table 1 summarises the benefts and limitations of the tuberculin skin test and interferon γ release assays. A major problem with identifying latent tuberculosis in patients due to start biological drugs is that the tuber- culin skin test and, to a lesser extent, interferon γ release assays are thought to be less reliable in immunosuppressed patients. 10  11 A recent meta-analysis of the performance of inter- feron γ release assays versus tuberculin skin tests patients with rheumatic disease before starting biological drugs reported pooled concordances between the QFT assay and tuberculin skin test of 72% (65% to 78%) and between T-SPOT.TB and tuberculin skin test of 75% (67% to 83%). It also showed that, compared with the tuberculin skin test, interferon γ release assays have a lower rate of false negative and false positive results in patients treated with corticosteroids and those with a history of BCG vaccina- tion. 5 It is not currently possible to determine the true sen- sitivity or specifcity of the tuberculin skin test or inter- feron γ release assays because a gold standard test for presence of latent tuberculosis infection is lacking. However, a closer correlation has been found between interferon γ release assays and risk factors for latent tuber- culosis infection, including personal history, tuberculosis contact history, and a suggestive chest radiograph. 12  13 Some studies have suggested using only an inter- feron γ release assay when screening for latent tubercu- losis because, in the context of immunosuppression, it is more specifc than the tuberculin skin test and may therefore reduce the proportion of patients needing anti- biotic chemoprophylaxis. 9  14 However, observational studies suggest that, given the discordance between EDUCATION PRACTICE 1 Chest and Allergy Department, St Mary’s Hospital, Imperial College NHS Trust, London W2 1NY, UK 2 Imperial College London, St Mary’s Campus, London, UK Correspondence to: O M Kon onn.kon@imperial.nhs.uk Cite this as: BMJ 2015;350:h1060 doi: 10.1136/bmj.h1060 This series of occasional articles provides an update on the best use of key diagnostic tests in the initial investigation of common or important clinical presentations. The series advisers are Steve Atkin, professor of medicine, Weill Cornell Medical College Qatar; and Eric Kilpatrick, honorary professor, department of clinical biochemistry, Hull Royal Infirmary, Hull York Medical School. To suggest a topic for this series, please email us at practice@bmj. com. A 27 year old white woman born in the United Kingdom presented with diarrhoea and rectal bleeding. A diag- nosis of Crohn’s disease was made after urgent flexi- ble sigmoidoscopy and she was treated with high dose intravenous corticosteroid. She relapsed on conversion to oral drugs and was start on infiximab, a tumour necrosis factor α (TNF-α) inhibitor. Introduction Biological drugs that target components of the immune system, such as TNF-α antagonists, interleukin 1 receptor antagonists, and interleukin 6 receptor antagonists, are licensed for the treatment of immune mediated infamma- tory conditions such as rheumatoid arthritis and Crohn’s dis- ease. These agents suppress the host immune response, so are associated with increased risk of opportunistic infections including viral infections (odds ratio 1.91, 95% confdence interval 1.02 to 3.58), such as herpes simplex virus and varicella zoster virus, and mycobacterial infections (3.73, 1.72 to 8.13), although risk varies with diferent agents. 1 In about 95% of people infected with Mycobacterium tuberculosis the infection is controlled by the immune sys- tem. This results in a state of “latent tuberculosis infec- tion,” with no features of active disease. 2 However 5% of those with latent tuberculosis later develop reactivation of tuberculosis and active disease. 2 TNF-α antagonists can cause such reactivation of latent infection and active dis- ease. 3 People with reactivation have a signifcantly higher risk of disseminated tuberculosis, and deaths have been recorded. 3 Crucially, screening for latent tuberculosis before starting biological drugs and chemoprophylaxis reduced the risk of reactivation by 78%, from a tuberculo- sis incidence of 522 (95% confdence interval 369 to 738) to 117 (29 to 470) per 100 000 patient years. 2 Initial clinical assessment to exclude active tuberculosis Before starting biological drugs, undertake a full clinical his- tory and examination. This should focus on symptoms of active tuberculosis infection including cough, fever, weight loss, and night sweats; history of exposure to tuberculosis infected contacts; and history of tuberculosis. Perform chest radiography in all patients to look for signs of previous or active tuberculosis. Refer patients with symptoms of active tuberculosis, a history of tuberculosis, or an abnormal chest radiograph to a specialist for further assessment. RATIONAL TESTING Screening tests for tuberculosis before starting biological therapy Richard J Hewitt, 1 Marie Francis, 1 Aran Singanayagam, 1 Onn Min Kon 1 2 THE BOTTOM LINE • Biological drugs are associated with an increased risk of progression from latent to active tuberculosis • Before starting treatment, exclude active tuberculosis by asking about symptoms (such as cough, fever, weight loss, and night sweats) and possible exposure to or history of tuberculosis, and with a chest radiograph • Check for latent tuberculosis with the tuberculin skin test or an interferon γ release assay; a combination of both tests may be the most sensitive approach thebmj.com Previous articles in this series Ж Investigating young adults with chronic diarrhoea in primary care (BMJ 2015;350:h573) Ж Investigating sepsis with biomarkers (BMJ 2015;350:h254) Ж Investigating asymptomatic invisible haematuria (BMJ 2014;349:g6768) Ж Neutropenia in primary care (BMJ 2014;349:g5340)