Background : Elimination of tuberculosis (TB) requires the management of latent tuberculosis infection (LTBI) in key populations. Thus, people with LTBI need to be diagnosed and treated with appropriate regimens. It is unknown how different screening and treatment strategies of key populations affect TB transmission and progress toward elimination. Objective : The objective of this project was to estimate the potential of various LTBI screening and treatment strategies in reducing transmission and to assess their contribution in moving towards elimination of TB in Europe. Methods : A novel, comprehensive, deterministic TB transmission model for four European Union countries (the Netherlands, the Czech Republic, Portugal and Spain) was developed. The selected countries were used as examples for other low-incidence countries/settings in order to pilot-test the model. This model accounts for transmission within and between the general population and different key population groups (i.e. migrants from TB high-endemic countries, homeless people, people who inject drugs, and prisoners). The TB natural history is characterised by two stages of LTBI (recent and remote), asymptomatic TB, and active pulmonary TB. Infected people can progress and regress between the different stages, depending on transition rates. The different stages correspond to different chances after a positive chest X-ray (CXR), tuberculin skin test (TST), interferon gamma release assays (IGRA), and culture. Transmission is governed by infection rates within and between population groups, as well as an external force-ofinfection due to travel to endemic countries. The quantified model was used for four representative countries (the Netherlands, the Czech Republic, Portugal, and Spain) as a pilot to predict the impact of different strategies to screen for LTBI and provide preventive treatment to those testing positive. The model assumes equilibrium situations, both regarding the dynamics of population groups and transmission within and between these groups. The 4% annual decreasing trend in notified TB incidence observed in the Netherlands was used to assess the longterm impact of LTBI screening strategies. Results : Outcomes are expressed as trends in pulmonary TB incidence and LTBI prevalence for the total population as well as for different population groups. Screening and treatment for LTBI in people who inject drugs (PWID) and homeless people usually results in the steepest decrease in pulmonary TB incidence, followed by screening prisoners or migrants from high-endemic countries at entry. With a combination of screening and treatment for LTBI in all three key population groups, the incidence of pulmonary tuberculosis (PTB) could be reduced by 15 to 45%, over a period of 20 years, depending on the country. In all countries analysed and for all strategies, the impact is slightly better for screening with TST than with IGRA, due to a slightly higher sensitivity of the test. These findings build upon very optimistic assumptions, including maximum coverage, no imported infections due to travel and migration, and applying an additional 4% annual decrease. According to the model, the elimination threshold of a TB incidence of <1/1 000 000 will only be achieved about 50 years after implementation of LTBI screening of at-risk groups for all countries studied. The predicted number of (averted) pulmonary TB cases does not change significantly when using reasonable alternative parameter values in the sensitivity analysis. Conclusions : This is the first TB transmission model that includes various interacting risk and age groups in low-incidence settings. The findings obtained from the model indicate that the World Health Organization's TB elimination target cannot be achieved by 2050 using LTBI screening as the sole control strategy for the disease. This is largely due to the remaining presence of LTBI in the population, including the general population and migrants already residing in the country of interest. Neither the general population nor migrants are currently considered for LTBI screening since their infection may have been acquired a long time ago and has a low risk of progressing to active disease. LTBI screening shows more potential for people who inject drugs/homeless people and prisoners than for new migrants from high-endemic countries. A high coverage of screening and completion of treatment are important to further increase effectiveness of LTBI control. Better diagnostic tests and shorter LTBI treatment would be welcomed as well, but to ensure that the reservoir of LTBI eventually disappears from the population, dedicated screening and more time is needed.
