Recurrence of tuberculosis at a Sydney chest clinic between 1994 and 2006: reactivation or reinfection?

Objective: To estimate the incidence of recurrence of culture-positive tuberculosis (TB) and the relative contributions of reinfection and reactivation (based on DNA fingerprinting). Design, setting and participants: Retrospective analysis of all culture-positive TB notifications between 1994 and 2006 from Liverpool Chest Clinic in the south-west of Sydney. Patients with more than one notification of culture-positive TB during this period were identified. Genotyping of Mycobacterium tuberculosis was used to determine whether recurrence was due to reinfection or reactivation. Main outcome measures: Estimation of the incidence of recurrence of culture-positive TB (cases per 100000 person-years of follow-up), and the proportions of reinfection and reactivation. Results: Three cases of recurrent culture-positive disease were identified (incidence of recurrence: 57.7 per 100000 person-years of follow-up). All three patients were treated with directly observed therapy. Two of these patients had evidence of reinfection with different strains; both were natives of a country with a high incidence of TB and had returned to that country after the initial episode. The other patient had evidence of reactivation of the initial strain, indicating secondary failure of treatment. This patient had poor adherence to treatment. Conclusions: Our observations suggest there is a very low rate of reactivation of tuberculosis. The low incidence of recurrence due to reinfection reflects the low incidence of tuberculosis in Australia. When reinfection does occur, this probably has been sustained during residence in a country with a high incidence of tuberculosis.