Treatment of latent tuberculosis infections in the Darwin region

As it is estimated that one-third of the world population have a latent tuberculosis infection (LTBI), treatment to prevent active tuberculosis is an essential component of the World Health Organization “End TB Strategy”.1

To inform and evaluate practice, we undertook a cohort study of people in the Darwin region (estimated population, 140 000; 25% Indigenous Australians, 25% overseas-born residents) diagnosed with LTBI according to Northern Territory guidelines2 during June 2013 – July 2014. Diagnosis was based on a positive Mantoux test result2 and the absence of radiological and clinical evidence for active tuberculosis. Demographic and treatment acceptance and compliance data were collected from the sole treatment centre serving the region. The recommended therapy was 9 months’ treatment with isoniazid, or 4 months’ treatment with rifampicin if isoniazid was contraindicated.2 Treatment adherence was assessed monthly on the basis of clinic attendance, self-reported adherence, and collection of the medication.

During the study period, 573 people were diagnosed with LTBI, of whom 422 (74%) were overseas-born, 81 (14%) were Indigenous Australians, and 70 (12%) were non-Indigenous Australians. The age range was 0–77 years (median, 32 years); 61% were male. The proportions of people diagnosed with LTBI who were offered, accepted and completed treatment are shown in the Box. Uncertainty on the part of the physician about an individual’s ability to complete treatment was the most common reason for not offering treatment, including to members of transient populations, such as those with short prison sentences and immigration detainees. Most patients who did not complete therapy had been lost to follow-up, either moving interstate (31%) or defaulting without a reason being recorded (25%). Outcome data for people moving interstate were not collected, as there are no mechanisms for routinely sharing such data between Australian states, and forwarding addresses were unavailable. The 55% completion rate therefore probably underestimates the proportion of those who completed treatment.

Parents and guardians of all 28 children under 6 years of age accepted treatment for their children, 12 of whom (43%) completed treatment, including six of 11 who were contacts of people with active tuberculosis, four of 16 immigration detainees, and three of seven refugees. Five children did not, however, commence treatment (three immigration detainees, two refugees), and 11 moved interstate without completing treatment. Nine of those moving interstate were immigration detainees, highlighting the transiency of this population and the uncertainty of outcomes arising from a lack of feedback between states about LTBI treatment compliance.

Indigenous Australians were significantly more likely to accept treatment than overseas-born people (odds ratio [OR], 4.46; 95% CI, 1.55–12.8) or non-Indigenous Australians (OR, 7.69; 95% CI, 2.33–25.4). Overseas-born patients were less likely to complete treatment than Indigenous (OR, 1.34; 95% CI, 0.66–2.72) or non-Indigenous Australians (OR, 1.38; 95% CI, 0.56–3.41). This finding, however, was not statistically significant, and potentially confounded by the fact that all 36 patients who moved interstate were overseas-born, so that completion for this population was possibly higher.

Similar to the findings of other Australian studies, 45% of patients who accepted treatment did not complete it, representing missed opportunities for preventing disease.3,4 Uncertainty about treatment adherence by overseas-born people moving interstate indicates that national data sharing and collaboration between tuberculosis services should be improved. LTBI treatment could then be evaluated according to WHO recommendations, and targeted measures to improve treatment outcomes for this high-risk population implemented.1,3 Further, the reasons for not completing treatment were often unknown; communicating with non-adherent patients would identify problems and enable targeted interventions for improving compliance.

Encouragingly, we found high uptake of treatment by Indigenous Australians, which may help reduce the disproportionately high incidence of active tuberculosis in this population, compared with non-Indigenous Australians.