Assessment and management of latent tuberculosis infection in a refugee population in the Northern Territory

Around two billion individuals worldwide have latent tuberculosis infection (LTBI)1 — that is, asymptomatic infection with Mycobacterium tuberculosis. Diagnosis of this condition is an important intervention in tuberculosis control, as treatment markedly reduces the potential for dormant infection to progress to active disease.2 Isoniazid for 6 to 12 months is 60%–90% effective in preventing this and is generally well tolerated.3 However, side effects occur, leading to debate over its appropriateness and the optimal duration of therapy.4 Moreover, rates of LTBI treatment completion are low, and targeting groups who are most at risk of reactivation, such as refugees, is recommended.5

In the Northern Territory, refugee health screening on arrival is centralised, as all newly arriving refugees enter through the capital city, Darwin. We aimed to assess LTBI prevalence, and treatment acceptance and completion among refugees arriving in the NT.

Nine months after the end of the study, further data were sought from other sources, including CDC-NT’s paper-based charts and electronic records, and records from the Royal Darwin Hospital, including all available hospital files and electronic radiology and pathology test results.

Where data were incomplete and a patient’s current state of residence was unknown to us, NT refugee support organisations assisted with providing forwarding locations. All Australian jurisdictional tuberculosis control units were then contacted and asked to provide follow-up information where available, allowing completion of outcome data for 10 individuals.

From 1 February 2006 to 31 January 2009, the CDC-NT refugee clinic saw 471 refugees, all holding Humanitarian Program visas. Their demographic characteristics and aspects of screening on arrival are shown in Box 1 and Box 2. Three of six WHO regions were represented, and Burma, the Democratic Republic of the Congo and Liberia were the most common countries of birth. Most refugees were born in the WHO region of Africa; 81.1% were born in continental Africa.

Outcomes of assessment and treatment for all individuals are shown in Box 3. One refugee required treatment for clinically active, smear-negative, culture-negative pulmonary tuberculosis. A further five gave a history of previously treated active tuberculosis, leaving 465 for TST administration. Seven of these people did not return for TST reading, leaving 458 of those eligible for testing (98.5%) adequately screened for LTBI.

No severe adverse reactions, hospitalisations or deaths occurred during the study period. The 52 refugees who discontinued treatment (including 11 who were lost to follow-up) did so after a median of 55 days (IQR, 20–88 days). Reasons for discontinuation were medication side effects (30), patient choice (8) and not restarting after intercurrent illness or pregnancy (3). Reported side effects were gastrointestinal (nausea, vomiting, abdominal pain, anorexia) (12), rash (7), lethargy (6), depression (3), raised liver enzymes (3), headache (1), visual symptoms (1) and joint pain (1). One of the three patients who stopped treatment due to raised liver enzymes was positive for hepatitis B surface antigen (hepatitis B e antigen negative), while none of the 12 who stopped treatment due to gastrointestinal side effects was. The mean age of those ceasing treatment due to medication-related side effects was significantly greater than for those completing treatment (27.2 years v 19.8 years; t test, P = 0.009).

Overall, 28 refugees were lost to follow-up, including 21 with LTBI, 11 of whom had commenced treatment.

Our study shows that LTBI treatment completion was achieved in about half of those who accepted treatment. Rates of acceptance and treatment completion were higher among younger refugees. Importantly, the group with the lowest rate of treatment acceptance — refugees from the Eastern Mediterranean region — also had the highest prevalence of LTBI. The most common reasons for discontinuation of treatment were medication side effects, most frequently gastrointestinal. All reported side effects were recognised adverse effects of isoniazid.

We were able to obtain complete data for a large proportion of participants owing to the centralised screening processes in the NT and the support of refugee organisations and interstate tuberculosis clinics. However, significant attrition occurred from loss to follow-up at various stages of the assessment and treatment process, contributing to the finding that only 28% of refugees diagnosed with LTBI completed therapy.

The prevalence of active tuberculosis among refugees in this study was relatively low,7 at 2.1 per 1000 (one refugee in 471), probably reflecting effective pre-departure screening and timely assessment on arrival, and implying little immediate risk of infection to the general community. However, in the developed world, most cases of active tuberculosis result from reactivation of LTBI, and LTBI is increasingly recognised as an important reservoir for tuberculosis globally.8 In Australia, most LTBI diagnoses are made in foreign-born individuals,9 and refugees are at particularly high risk, as their exposure often occurs shortly before arrival in Australia.10 For these reasons, most developed countries target new immigrants and refugees for tuberculosis screening on entry, although approaches vary considerably.11,12

Our finding of LTBI in 31.9% of refugees is comparable to other Australian data.13,14 It is likely that the association between increasing age and LTBI reflects cumulative exposure in highly endemic regions. We were unable to assess the impact of HIV infection and BCG vaccination status, which are potentially important associations. LTBI prevalence rates in countries of birth are unknown, but regional prevalences of active tuberculosis in 2007 were: Africa, 475 cases per 100 000; Eastern Mediterranean, 139; and South-East Asia, 280.15 While we found a higher prevalence of LTBI in refugees from the Eastern Mediterranean region, most of these individuals were from Somalia and Sudan (reflecting recent political unrest), which had prevalences of active tuberculosis of 494 and 206 per 100 000, respectively, in 2009.16,17

In 13 of the 15 refugees in our study who were not offered LTBI treatment, a reasonable contraindication could be substantiated — most often age, pregnancy, breastfeeding or comorbidities. The refusal rate for LTBI treatment in our study (23.1%) is comparable to previous studies, including the large North American Tuberculosis Epidemiologic Studies Consortium (TBESC) study (17.1% overall, 23.4% of foreign-born individuals).18

The efficacy of prophylactic isoniazid increases with treatment adherence.3 The rate of failure to complete treatment in our study (55.9%) was similar to the rate in the TBESC study (52.5%), regardless of overseas-born status.18 One study in a group that included a high proportion of overseas-born individuals found an overall completion rate of 38.6%;19 while others20,21 included high proportions of immigrants but differed in their assessment of the effect of being foreign-born. A systematic review of studies in the United States and Canada concluded that individual trials were inconsistent, but overall results for treatment completion were suboptimal and few factors consistently predicted adherence.22 A study of refugees in Canada23 reported completion of 6 months of therapy in 69% (24/35), although few studies have specifically examined refugee groups.

Qualitative studies of tuberculosis treatment have demonstrated personal, structural, social and health-service factors to be important.24 However, few data are available for LTBI treatment. High rates of completion for LTBI treatment have been attained through cultural approaches, such as linking case managers to refugees of the same cultural background.25

Our findings show that timely screening and treatment of LTBI in refugee groups is feasible. Measures directed at maintaining contact with refugees after arrival, provision of culturally and linguistically appropriate support, better understanding of treatment barriers, and care for individuals with medication-related side effects are all likely to be effective strategies for improving treatment completion.

3 Summary of outcomes and numbers of newly arrived refugees included at each stage of assessment and treatment for latent tuberculosis infection (LTBI)

TST = tuberculin skin test.