Post-migration follow-up of migrants at risk of tuberculosis

Targeting screening and treatment for latent tuberculosis infection towards asylum seekers from high-incidence countries - a model-based cost-effectiveness analysis

Background

Enhancing tuberculosis (TB) prevention and care in a post-COVID-19-pandemic phase will be essential to ensure progress towards global TB elimination. In low-burden countries, asylum seekers constitute an important high-risk group. TB frequently arises post-immigration due to the reactivation of latent TB infection (LTBI). Upon-entry screening for LTBI and TB preventive treatment (TPT) are considered worthwhile if targeted to asylum seekers from high-incidence countries who usually present with higher rates of LTBI. However, there is insufficient knowledge about optimal incidence thresholds above which introduction could be cost-effective. We aimed to estimate, among asylum seekers in Germany, the health impact and costs of upon-entry LTBI screening/TPT introduced at different thresholds of country-of-origin TB incidence.

Methods

We sampled hypothetical cohorts of 30–45 thousand asylum seekers aged 15 to 34 years expected to arrive in Germany in 2022 from cohorts of first-time applicants observed in 2017–2019. We modelled LTBI prevalence as a function of country-of-origin TB incidence fitted to data from observational studies. We then used a probabilistic decision-analytic model to estimate health-system costs and quality-adjusted life years (QALYs) under interferon gamma release assay (IGRA)-based screening for LTBI and rifampicin-based TPT (daily, 4 months). Incremental cost-effectiveness ratios (ICERs) were calculated for scenarios of introducing LTBI screening/TPT at different incidence thresholds.

Results

We estimated that among 15- to 34-year-old asylum seekers arriving in Germany in 2022, 17.5% (95% uncertainty interval: 14.2–21.6%) will be latently infected. Introducing LTBI screening/TPT above 250 per 100,000 country-of-origin TB incidence would gain 7.3 (2.7–14.8) QALYs at a cost of €51,000 (€18,000–€114,100) per QALY. Lowering the threshold to ≥200 would cost an incremental €53,300 (€19,100–€122,500) per additional QALY gained relative to the ≥250 threshold scenario; ICERs for the ≥150 and ≥ 100 thresholds were €55,900 (€20,200–€128,200) and €62,000 (€23,200–€142,000), respectively, using the next higher threshold as a reference, and considerably higher at thresholds below 100.

Conclusions

LTBI screening and TPT among 15- to 34-year-old asylum seekers arriving in Germany could produce health benefits at reasonable additional cost (with respect to international benchmarks) if introduced at incidence thresholds ≥100. Empirical trials are needed to investigate the feasibility and effectiveness of this approach.

Using country of origin to inform targeted tuberculosis screening in asylum seekers: a modelling study of screening data in a German federal state, 2002-2015

BACKGROUND

Screening programmes for tuberculosis (TB) among immigrants rarely consider the heterogeneity of risk related to migrants' country of origin. We assess the performance of a large screening programme in asylum seekers by analysing (i) the difference in yield and numbers needed to screen (NNS) by country and WHO-reported TB burden, (ii) the possible impact of screening thresholds on sensitivity, and (iii) the value of WHO-estimated TB burden to improve the prediction accuracy of screening yield.

METHODS

We combined individual data of 119,037 asylum seekers screened for TB in Germany (2002-2015) with TB estimates of the World Health Organization (WHO) (1990-2014) for their 81 countries of origin. Adjusted rate ratios (aRR) and 95% credible intervals (CrI) of the observed yield of screening were calculated in Bayesian Poisson regression models by categories of WHO-estimated TB incidence. We assessed changes in sensitivity depending on screening thresholds, used WHO TB estimates as prior information to predict TB in asylum seekers, and modelled country-specific probabilities of numbers needed to screen (NNS) conditional on different screening thresholds.

RESULTS

The overall yield was 82 per 100,000 and the annual yield ranged from 44.1 to 279.7 per 100,000. Country-specific yields ranged from 10 (95%- CrI: 1-47) to 683 (95%-CrI: 306-1336) per 100,000 in Iraqi and Somali asylum seekers, respectively. The observed yield was higher in asylum seekers from countries with a WHO-estimated TB incidence > 50 relative to those from countries ≤50 per 100,000 (aRR: 4.17, 95%-CrI: 2.86-6.59). Introducing a threshold in the range of a WHO-estimated TB incidence of 50 and 100 per 100,000 resulted in the lowest "loss" in sensitivity. WHO's TB prevalence estimates improved prediction accuracy for eight of the 11 countries, and allowed modelling country-specific probabilities of NNS.

CONCLUSIONS

WHO's TB data can inform the estimation of screening yield and thus be used to improve screening efficiency in asylum seekers. This may help to develop more targeted screening strategies by reducing uncertainty in estimates of expected country-specific yield, and identify thresholds with lowest loss in sensitivity. Further modelling studies are needed which combine clinical, diagnostic and country-specific parameters.

A cluster of multidrug-resistant Mycobacterium tuberculosis among patients arriving in Europe from the Horn of Africa: a molecular epidemiological study

Background

The risk of tuberculosis outbreaks among people fleeing hardship for refuge in Europe is heightened. We describe the cross-border European response to an outbreak of multidrug-resistant tuberculosis among patients from the Horn of Africa and Sudan.

Methods

On April 29 and May 30, 2016, the Swiss and German National Mycobacterial Reference Laboratories independently triggered an outbreak investigation after four patients were diagnosed with multidrug-resistant tuberculosis. In this molecular epidemiological study, we prospectively defined outbreak cases with 24-locus mycobacterial interspersed repetitive unit-variable number tandem repeat (MIRU-VNTR) profiles; phenotypic resistance to isoniazid, rifampicin, ethambutol, pyrazinamide, and capreomycin; and corresponding drug resistance mutations. We whole-genome sequenced all Mycobacterium tuberculosis isolates and clustered them using a threshold of five single nucleotide polymorphisms (SNPs). We collated epidemiological data from host countries from the European Centre for Disease Prevention and Control.

Findings

Between Feb 12, 2016, and April 19, 2017, 29 patients were diagnosed with multidrug-resistant tuberculosis in seven European countries. All originated from the Horn of Africa or Sudan, with all isolates two SNPs or fewer apart. 22 (76%) patients reported their travel routes, with clear spatiotemporal overlap between routes. We identified a further 29 MIRU-VNTR-linked cases from the Horn of Africa that predated the outbreak, but all were more than five SNPs from the outbreak. However all 58 isolates shared a capreomycin resistance-associated tlyA mutation.

Interpretation

Our data suggest that source cases are linked to an M tuberculosis clone circulating in northern Somalia or Djibouti and that transmission probably occurred en route before arrival in Europe. We hypothesise that the shared mutation of tlyA is a drug resistance mutation and phylogenetic marker, the first of its kind in M tuberculosis sensu stricto.

Funding

The Swiss Federal Office of Public Health, the University of Zurich, the Wellcome Trust, National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC), the Medical Research Council, BELTA-TBnet, the European Union, the German Center for Infection Research, and Leibniz Science Campus Evolutionary Medicine of the Lung (EvoLUNG).