Discussion

Despite the high treatment initiation rate among TB contacts in this outbreak, only 16.9% eventually completed treatment. This poor outcome led to a cascade of infection and disease that lasted for 9 years.

The current incidence of TB among native Israelis is low (1.0 per 100 000 person-years) and most TB patients are foreign-born [Reference Bishara2]. Nevertheless, TB outbreaks in Israel have been reported in congregate settings, where crowded conditions facilitate the spread of TB [Reference Stein-Zamir3]. However, in such controlled environments, TB outbreaks are short-lived, because the implementation of TB control measures is simple; contacts are easily identified, examined and successfully treated.

We have previously reported high LTBI treatment completion rates (94%) among Ethiopian immigrants in Israel, with directly observed isoniazid treatment and free transportation provided [Reference Bishara4]. Those previous findings stand in total contrast to the low LTBI treatment adherence we encountered in this outbreak [Reference Stein-Zamir3, Reference Bishara4].

There were several barriers to health care, at both the patient and system levels, that may have caused such poor treatment adherence in this outbreak. At the patient level, attending clinic appointments posed challenges, such as the cost and inconvenience of commuting to and from the clinic. Commuting to the TB clinic from the provincial areas of the northern district was challenging for contacts without private transportation. Many of those who missed their follow-up appointment at the TB clinic, declined to return because of the inconvenience of public transportation to the TB clinic. Some individuals involved in this outbreak were unemployed or were employed at low wages. Poverty and financial hardships have been associated with poor treatment adherence [Reference Liu5]. For those who were employed, screening required a day off work to perform a TST and another day off (2 days later), to read the TST reaction. In addition, some were worried that they might lose their jobs for recurrent absence or for being stigmatised as a ‘suspected’ TB patient. This might explain why many contacts who initially presented for screening or LTBI treatment did not return for their scheduled follow-up appointment.

Furthermore, awareness of the possibility of TST cross-reaction with a previous BCG vaccination deterred many contacts from accepting TST ‘positive’ results as indicative of infection [Reference Liu5]. Patient 8 illustrates the pertinence of the obstacles mentioned above. She was a 25-years-old female and an Israeli of EI descent. She was examined in 2012 as a close contact of an active TB patient in this group and isoniazid treatment for LTBI was started. However, she stopped taking her medication after 2 months of treatment and did not report for her scheduled follow-up. She was examined in June 2015 because of a persistent cough and was diagnosed with smear-positive, culture-positive pulmonary TB. She had 102 contacts, mostly co-workers in a factory where she was employed. Of these contacts, 54.9% of those tested were TST positive. Considering the high rate of positive TST in this group, which might in some cases be related to their BCG vaccination and the uncertainties regarding the extent and intensity of exposure, we suggested performing interferon-gamma release assay (IGRA) test to resolve this issue. However, this did not work out and our treatment recommendation was based solely on the TST result. Eventually, these co-workers did not comply with the treatment recommendations. Overall, two of these co-workers contracted active TB.

All 10 male SGE contacts who developed TB in this outbreak abused alcohol or smoked illicit substances. Such high-risk behaviour has been reported among the youth of immigrant population [Reference Sharabany and Israeli6]. Unemployment and the diminished parental status that often accompanies immigration have also been associated with high-risk behaviour among youths of immigrant families. Alcohol abuse is associated with higher TB transmission and lower rates of sputum culture conversion [Reference Volkmann7]. Substance abuse is also related to an increased risk for TB and increased contagiousness [Reference Oeltmann8]. Finally, the stigma of TB is known to influence health-seeking behaviours and treatment adherence [Reference Datiko, Jerene and Suarez9]. TB primarily affects the poor and is associated with self-stigmatisation as well as a community-based stigmatizing response [Reference Datiko, Jerene and Suarez9]. Therefore, some patients were reluctant to reveal their contact details or provided incomplete information to hinder the contact tracing process. Patient 2 in this outbreak was a close contact of the index case. However, he was not reported as such and was not included in the index case's contacts investigation. Patient 2 was identified as part of this cluster only after he contracted active TB.

Apart from the barriers to treatment at the patient level described above, there were barriers at a system-level as well. National TB control programmes (NTP) are usually focused on finding and treating active TB. In Israel, the programme grants active TB patients (as well as recent EI's), monetary incentives during the treatment phase. These interventions likely promoted patient compliance [Reference Chemtob1, Reference Bishara4]. Monetary incentives (e.g. refund or cost paid for transportation) would sometimes be granted for close contacts on a case-by-case basis as well.

With hindsight, this outbreak could have been anticipated. The poor response to the call to be present for the screening and the low adherence to LTBI treatment should have alerted the health care professionals overseeing this outbreak to the risk of evolution of latent TB cases to active disease.

This ongoing TB transmission among second-generation immigrants raised our awareness regarding the need for a proper definition, identification and management of high-risk groups with obstacles to health care. This outbreak illustrates the increased risk for TB transmission among second-generation immigrants, not formally defined as a high-risk group and the need to recognise and act upon the specific obstacles to health care facing this community. Routine contact investigation practices in this outbreak fell short of stopping the cascade of TB transmission and disease, which lasted for 9 years.