Measuring outcomes in tuberculosis care

Tuberculosis services in the UK are guided by the British Thoracic Society's code of practice. The recently introduced Enhanced Local Surveillance of Tuberculosis system has greatly improved the timely collection of epidemiological information. The next step should be to monitor the outcome of treatment of all patients. The Birmingham Chest Clinic TB Aftercare Office receives notifications of tuberculosis from the city of Birmingham (population ∼ 1 million with about 300 cases each year). One year after each notification is received, a questionnaire with the following choices is sent to the consultant in charge of the case. Satisfactory completion of treatment _ (or satisfactory progress on continuing treatment) |_| Transferred to another hospital or clinic |_| Wrong diagnosis |_| Died but not of TB |_| Died of TB |_| Defaulted from attendance |_| Persistent problems - still culture positive |_| Persistent problems - not culture positive |_| These categories have been chosen because they require a minimum of recall by the clinician. The first four indicate satisfactory performance of TB management services and the second four don't. In Birmingham in 1998 there were 254 notifications of tuberculosis. A year later we sent out questionnaires for each of them. We received 174 replies. Ninety per cent of the responses indicated satisfactory outcomes according to these definitions and 10 per cent were unsatisfactory. One patient was still culture positive after a year. He had multi-drug resistance and later died. Measuring outcomes in tuberculosis is practical and valuable. It should be introduced nationally.