Study population and setting

From March 2016 until September 2016, five (20% of a total of 25) Public Health Services (PHS) in urban (n = 2) and rural (n = 3) parts of the Netherlands replaced mandatory CXR entry-screening with screening for both TB disease and LTBI using a health assessment and tuberculin skin testing (TST) with confirmatory Interferon Gamma Release Assay (IGRA) or IGRA alone. Clients with a positive IGRA test result or clients with symptoms suggestive for TB disease, relevant comorbidities or using immunosuppressive medication were additionally screened with CXR and consulted by a TB physician.

We included immigrants, not applying for asylum, from non-EU countries with a TB incidence of >50/100,000 population, of all ages, with an intended stay in the Netherlands of at least six months, and no history of TB. Immigrants who did not meet inclusion criteria were screened according to current policy: LTBI screening for those age <18 years and CXR for those age ≥18 years. A detailed description of current TB screening policies in the Netherlands can be found elsewhere [4, 5, 11].

LTBI screening and treatment process

We developed a study protocol on which we trained PHS staff. Clients received an information brochure about the purpose of the LTBI screening either by mail or at registration at the PHS. Medical Technical Assistants (MTA), i.e., healthcare assistants trained to perform TB screening activities (symptom screening, tuberculin skin test and chest X-ray), BCG vaccination, to give information on TB screening procedures, to document screening results and to perform specific administrative tasks, handled registration of the client and gave further explanation about the screening procedures. Consequently, clients completed a health questionnaire which was evaluated by MTAs to identify clients with: 1) a history of TB disease or LTBI treatment, 2) symptoms suggestive for TB disease, 3) BCG vaccination or recent (other) vaccinations, and 4) relevant comorbidities or immunosuppressive medication use, which may affect the validity of test results. (S1 Fig)

Two PHSs screened directly with IGRA and three PHSs used TST followed by confirmatory IGRA. Children <12 years with normal immunity were screened with initial TST followed by confirmatory IGRA. PHS used PPD Tuberculin mammalian (BulBio, Sofia, Bulgaria, 0.1 ml intracutaneous) for TSTs. MTAs interpreted TST results after 48–72 hours by measuring the size of induration (mm): indurations between 3 and 10mm were double read by a second MTA. According to the national policy for LTBI testing, TST indurations ≥5mm were confirmed with IGRA: using QuantiFERON-TB Gold Plus (QFT-Plus; Qiagen, Germantown, MD) according to manufacturer’s instructions. [12] The LTBI diagnosis was considered confirmed in those with a QFT-plus test result ≥0.35 IU/ml, after exclusion of TB disease through physical examination, CXR and -if indicated- bacteriological investigation.

TB physicians initiated and monitored LTBI treatment of clients following the national guideline for LTBI treatment. After confirmation of the LTBI diagnosis, TB physicians ruled out contra-indications for LTBI treatment and transaminase levels were assessed in clients with a history of liver disease, alcohol abuse, HIV infection, age >35 years, pregnancy and during the first three months of post-partum. If eligible for LTBI treatment, TB physicians educated the client on the purpose and possible side-effects of LTBI treatment and offered each client LTBI treatment with 3 months rifampicin /isoniazid combination therapy. In accordance with national guidelines, clients with contra-indication or clients who were not motivated to initiate LTBI treatment were offered bi-annual follow-up screening for a period of two years. [13]

TB physicians consulted clients on LTBI treatment every month for adverse events (including hepatotoxicity tests) and treatment adherence. A CXR was repeated after one month and at the end of the treatment to rule out the development of thoracic TB disease. During regular contact moments, TB nurses interviewed, educated and supported the client during the treatment in a demand-driven, tailored way. [13]

Data collection

Data analyses

We double entered data from questionnaires in MS-Access (Microsoft Corp, Seattle WA, USA) and checked for inconsistencies against the raw data. After merging the datasets, we checked inconsistencies between the three datasets, for which we approached the PHS for validation and substitution of missing data. We calculated proportions for clients’ characteristics, and the cascade of screening and care: LTBI test results, LTBI treatment initiation and completion, including reasons for not initiating or completing the LTBI treatment.

We assessed factors associated with LTBI diagnosis and LTBI treatment initiation using full case analyses: cases with missing values were excluded from the analyses (n = 33 and n = 7 respectively). As probabilities for outcome occurrence (18% and 52% respectively) were not rare, univariable Poisson regression models with robust variance estimators were used to calculate risk ratios. [15] For LTBI diagnosis, we also developed a multivariable Poisson regression model using backward elimination of the initial model with variables yielding a p-value <0.1 in univariable analysis, guided by changes in regression coefficients and changes in the fit, as indicated by -2loglikelihood. We checked for interaction between independent variables before fitting multivariable Poisson regression models. Due to small numbers, we combined categories of variables in regression analyses, we did not perform multivariable analysis for LTBI treatment initiation and we did not calculate risk ratios for LTBI treatment completion. We recoded and analyzed data using SPSS Statistics 25.0 (IBM Analytics, Chicago, IL, USA).

We verbally transcribed all audiotaped interviews. After familiarization with the data, IS developed a coding scheme to guide the coding of all transcripts. We refined the coding scheme along the coding process. In regular meetings, IS and JS discussed coding and interpretation of the data. [16] We used MAXQDA (Version 11, VERBI GmbH, Berlin, Germany) to assist in analyses of qualitative data.

Characteristics of clients screened for LTBI

Of 566 clients screened for LTBI, the majority were female, aged between 18–34 years, had higher levels of education, intended to stay more than one year in the Netherlands, and originated from countries with a TB incidence >200/100,000 population. At the time of the screening, 157 (28%) clients did not have health insurance. (Table 1)

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptive statistics of the study population.

https://doi.org/10.1371/journal.pone.0219252.t001

LTBI screening and treatment results

Of 317 clients tested with TST, 227 (72%) had a positive TST, of whom 50 (22%) had a positive IGRA. Of 249 clients screened directly with IGRA, 51 (20%) tested positive. In total 101 (18%) clients had a positive IGRA, of whom 94 (17%) were diagnosed with LTBI, three clients (0.5%) were diagnosed with extrapulmonary TB disease, three clients (0.5%) had (treated) TB disease in the past, and one (0.2%) client had fibrotic lesions of presumed tuberculous origin. (Fig 1)

The risk of being diagnosed with LTBI was higher for men (1.5 times; 95% CI 1.0–2.2) than for women; higher for clients aged 25–34 years (2.5 times; 95%CI: 1.3–4.8) or 35 years and older (3.5 times; 95%CI: 1.8–6.8) than for clients aged 0-24-year; higher for clients from countries with a TB incidence >100/100,000 population than for those from countries with lower TB incidence; and higher for clients with lower levels of education (1.6 times; 95% CI 0.9–2.6) than for clients with higher levels of education. (Table 2)

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Results of Poisson regression model with robust variance estimators.

https://doi.org/10.1371/journal.pone.0219252.t002

Of 94 clients diagnosed with LTBI, 49 (52%) initiated and 34 (69%) completed LTBI treatment. Treatment initiation proportions differed considerably between PHSs (29–86%) and declined with increasing age, from 7% among those 0–17 years to 27% among those 45 years and older. (Table 1) Clients were more inclined to initiate LTBI treatment when their intended stay in the Netherlands was ≥5 years, and when having lower levels of education. (Table 2) TB physicians reported contra-indications as a reason for eight (18%) of 45 clients not initiating LTBI treatment. Other common reasons reported for not initiating treatment were: no perceived advantages of LTBI treatment by client and return of the client to country of origin in the foreseeable future. The most common reason for discontinuation of treatment (31% of clients) was side-effects. Challenges encountered by TB physicians during client’s LTBI treatment are summarized in Table 3. TB physicians reported that they did not perceive language as a barrier when consulting most (81%) of their clients. In case of language barriers, usually a family member was used as an interpreter during consultations. Most (76%) clients who initiated LTBI treatment received support by the TB nurse. (Table 3)

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Evaluation of latent tuberculosis treatment by the physician.

https://doi.org/10.1371/journal.pone.0219252.t003

Enhancers and barriers for LTBI screening and treatment

Most interviewed PHS staff endorsed the importance of a future LTBI screening and treatment program among immigrants to reach a 25% decline in TB incidence in the Netherlands. However, PHS staff did experience some barriers for execution of the program.

Providing information about screening procedures is a legal obligation of PHS. Despite the available information brochures about the LTBI screening, some clients were still uninformed during the LTBI screening because a few PHSs (n = 2) did not send the information brochure prior to the LTBI screening as screening appointments were made by phone. Other clients had been misinformed about screening procedures through information channels (colleagues, companies, relatives) and therefore wrongly believed that the screening comprised a CXR. PHS staff said that clients who were not correctly informed about screening procedures were often more agitated because the LTBI screening was more time-intensive than the expected CXR. Another misunderstanding among some clients was that screening outcomes would affect their residence permit.

The health questionnaire is considered an essential part of the LTBI screening. However, we noted frequently that reported comorbidity and medication were inaccurate and was not well evaluated by the MTAs. Furthermore, and despite training, the importance of the questions was not sufficiently clear to all MTAs, which negatively influenced the quality of the evaluation of the answers and initiation of the required follow-up steps.

MTA PHS1: “When you evaluate the health questionnaire and a client asks you why certain information is required, you notice that you need more in-depth understanding of underlying reasons for asking those questions.”

We found suboptimal LTBI treatment initiation (52%). Some TB physicians were under the impression that they could influence the clients’ choice on initiation of LTBI treatment. This is reflected by the broad treatment initiation range (29–85%): the PHS with the “intention to screen is intention to treat” attitude achieved initiation proportion of 86%. TB physicians of the three PHSs with the lowest proportions (29–50%) were more skeptical about offering LTBI treatment to immigrants with a short-intended duration of stay in the Netherlands. This is reflected by the higher risk ratio for those with intended stay of ≥5 years (RR: 1.3 (95%CI: 1.1–1.4)). TB physicians argued that for immigrants with a short-intended stay and high risk for re-infection in the home country, the individual and public health benefits would not outweigh the risks of LTBI treatment.

TB physician PHS4: “I am not in favor of initiating LTBI treatment when a client will return to Sub Saharan Africa because it is impossible for this client to live in Sub Saharan Africa without getting re-infected with TB. (….) However, for clients from northern Africa LTBI treatment might be favorable.”

TB physician PHS1: “For someone who will not even be here for one year the chance for that client to develop TB disease and therefore endanger our public health is low. Therefore, I would like to discuss duration of stay before we implement LTBI screening and treatment as national policy.”

The guidelines states that, following the recommendations of the manufacturer, LTBI diagnosis is confirmed in clients with an IGRA result ≥0.35 in whom TB disease or history of TB disease is excluded. However, some TB physicians question this recommended cut-of-value, particularly among clients with a low reported risk of exposure. For this reason, six clients at one PHS were not offered LTBI treatment.

TB Physician PHS2: “It [LTBI diagnosis] depends on exposure, recent or old infection, logistical factors, a lot of factors. So, IGRA results between 0.35 and 1.00 -the so-called grey area- can go either way. You can repeat the test, or you can just say: no, it is not an LTBI if anamnesis factors are not in favor. So, basing LTBI diagnosis solely on an IGRA result of 0.35 is not enough.”

Quantitative results showed that clients with lower levels of education were more inclined (RR: 1.3 (95%CI: 1.1–1.4)) to initiate LTBI treatment than those with higher levels of education. TB physicians perceived clients with difficulties to understand the concept of LTBI and who confused their diagnosis with TB disease, and clients with fear for developing TB disease were more inclined to initiate LTBI treatment.

TB Physician PHS3: “I believe that people from certain countries do not understand the difference between TB disease and LTBI and therefore want treatment. I can explain the difference a hundred times, but they just do not see it. (…) They say: ‘Oh, yes I have TB therefore I need treatment’.”

PHS staff said that some higher educated clients could not believe they had been diagnosed with LTBI. These clients often perceived TB as a disease of the poor or thought TB disease did not occur in their home country anymore. PHS staff said those clients were less likely to initiate treatment.

TB nurse PHS4: “They say: ‘There is no TB in our country!’. That is when I showed the TB incidence list of the World Health Organization and told them: ‘Sorry, but the World Health Organization disagrees with you’.”

MTA PHS5: “I can remember a client from India, who had a positive IGRA test result. The client said: ‘How is this even possible, this only occurs among the poor’.”

Furthermore, PHS staff perceived unfamiliarity with LTBI treatment as a barrier for LTBI treatment initiation. Clients who did understand prevention and the benefits of LTBI treatment were more inclined to initiate LTBI treatment.

MTA PHS4: “I remember this client in whom you [TB nurses] had invested a lot of time. She came to me and said: “I finally understand: he [the TB bacterium] is sleeping. So, he is present in the body but not active. So, you do not want to awake him. That is why you need to take medication”.

Some PHS staff said that some client’s lack of trust in the Dutch TB control system complicated treatment initiation. They experienced that some clients, especially from India, China and Indonesia, consulted their physician in their home country about their LTBI diagnosis and treatment options. PHS staff perceived that those physicians were unfamiliar with purposes of LTBI screening and treatment activities in low TB incidence countries and therefore advised their clients against LTBI treatment.

TB physician PHS4: “What I noticed, you do not win much trust from some clients. They always want to consult a physician in their own country. Especially clients from India. (….) they [physicians in home country] will tell what will eventually happen.’

TB nurses said that the frequency and intensity of treatment support generally depends on level of education: clients with lower levels of education need more intensive treatment support. A barrier that some TB nurses experienced was that clients were hard to reach during office hours for phone appointments, which are used to monitor treatment progress.

TB nurse PHS1: “Some clients were sometimes hard to reach by phone. We wanted to call to ask how they were doing. But because of work they did not answer our calls: they would call back at 11 at night. We need to draw a line somewhere…”

PHS staff experienced the LTBI screening and treatment program as more time-intensive than CXR screening and some therefore questioned the cost-effectiveness. Some other PHS staff, however, noted that with LTBI screening as policy, cost and time would be saved from cancelled CXR follow-up screening among those with a negative IGRA test result. PHS staff noted some conditions, related to testing and treatment costs, if LTBI screening and treatment would become national policy in the future. Some PHS staff said that LTBI screening with only IGRA would be a more efficient pathway: the client only must visit the PHS once instead of twice, and the PHS would save time in extra consultations because of the high rate of positive TST. However, PHSs consider IGRAs to be expensive at this moment (prices in the Netherlands range from 50–104 euros for a single IGRA test compared to 27 euros for a single TST test). They expected with an increased demand for IGRAs, laboratories should be able to offer them cheaper. Additionally, TB physicians reported that the costs of LTBI treatment -which we provided for free during this study- would be a major barrier for clients to initiate treatment (roughly 180 euros for LTBI medication, excluding hepatotoxicity tests and CXRs).

TB physician PHS1: “I am in favor of screening and treating clients at entry for LTBI. I think it is a key strategy towards elimination of TB in the Netherlands. (…) But, the strategy is not only of individual interest but also of public interest. Therefore, medication must be offered free of charge. That is, I think, the only way the strategy will be successful.”

TB physician PHS2: “A few clients had to -accidentally- pay for their medication at the pharmacy. They were shocked by the price. So, the financial compensation, I think, plays a key role in the success of the program.”