P166 Diagnosing pulmonary tuberculosis: how useful is the chest x-ray report?

Introduction and Objectives The diagnosis of pulmonary tuberculosis (TB) is frequently delayed, resulting in increased morbidity and mortality. One factor may be a delay in referral for chest X-ray (CXR). The NICE quality statement of January 2017 1 suggests that all patients with imaging suggestive of active pulmonary tuberculosis are assessed by the next working day. We wanted to understand how many patients with confirmed TB have an initial CXR which correctly reported the diagnosis. Methods We performed a retrospective analysis of the patient databases of two London hospitals of patients with confirmed pulmonary TB. We examined the report of the CXR prior to diagnosis in each case for mention of tuberculosis or referral to respiratory medicine. Results 247 adult patients were treated for pulmonary TB between January 2014 and May 2017. Of these, 231 (94%) had a CXR reported at our hospitals prior to diagnosis. 206 (89%) patients had a CXR which was reported as abnormal and 106 (52%) had a report which suggested TB as the diagnosis. In 29 (12%), the report did not mention TB, but did suggest referral to a respiratory physician, meaning that overall, 65% of cases would have been assessed. 198 patients (86%) had culture-positive disease, and of these, 91 (46%) had a CXR which suggested the diagnosis. In 123 cases (50%), the sputum was smear-positive for AFB and of these, 63 (51%) had a CXR reported as TB. Of those patients whose CXR was reported as TB, 92 (87%) had a positive culture, and 63 (51%) were smear-positive. Conclusions In our study, the initial CXR was abnormal in most cases of pulmonary tuberculosis, but in only 52% of these was TB suggested as the diagnosis. Thus almost half of diagnoses were missed on initial CXR, including those with smear-positive disease, and an automatic referral for assessment would not have been triggered. We think that these data highlight the importance of considering radiologist training as part of TB control efforts. Reference National Institute for Health and Care Excellence. Tuberculosis 2017;QS141. Available at: nice.org.uk/guidance/qs141