Abstract
Tuberculosis control depends on successful case finding and treatment of individuals infected with Mycobacterium tuberculosis. Passive case finding is widely practised: the present study aims to ascertain the consensus and possible improvements in active case finding across Europe.
Recommendations from national guidelines were collected from 50 countries of the World Health Organization European region using a standard questionnaire.
Contacts are universally screened for active tuberculosis and latent tuberculosis infection (LTBI). Most countries (>70%) screen those with HIV infection, prisoners and in-patient contacts. Screening of immigrants is related to their contribution to national rates of tuberculosis. Only 25 (50%) out of 50 advise a request for symptoms in their guidelines. A total of 36 (72%) out of 50 countries recommend sputum examination for those with a persistent cough; 13 countries do not, even if the chest radiograph suggests tuberculosis. Nearly all countries (49 out of 50) use tuberculin skin testing (TST); 27 (54%) out of 50 countries also perform chest radiography irrespective of the TST result. Interpretation of the TST varies widely. All countries use 6–9 months of isoniazid for treatment of LTBI, with an estimated median (range) uptake of 55% (5–92.5%).
Symptoms and sputum examination could be used more widely when screening for active tuberculosis. Treatment of latent tuberculosis infection might be better focused by targeted use of interferon-γ release assays.
Despite investment towards tuberculosis (TB) control in Europe, the incidence of TB has increased between 2000 and 2005 in countries such as Norway, the UK and Ireland, in addition to Romania, Turkey, Macedonia and Bulgaria and all except Kyrgyzstan and Turkmenistan in the eastern part of the World Health Organization (WHO) European region 1. Passive case finding is standard and consistent with the WHO-recommended Stop TB Strategy 2. Active case finding includes all methods for the identification and then treatment of those with TB who have not reported to the healthcare system of their own accord. The aim is to reduce the transmission of TB by screening high-risk populations (i.e. those at an increased risk of exposure to TB infection, most notably contacts of infectious cases) and to detect and treat active disease earlier than would otherwise occur 3, 4. At the same time, latent tuberculosis infection (LTBI) may be detected and treated, and thereby prevent the later development of active TB.
In 2006, the Tuberculosis Network European Trials Group (TBNET) was established in order to conduct clinical trials and surveys on TB in Europe. The present survey was undertaken within the TBNET to determine whether national guidelines or policies within Europe recommend screening of the same high-risk groups, use the same methods and locations for screening and whether the nature and estimated uptake of preventive treatment is uniform. The present authors expected to see differences related to local epidemiology, such as screening for migrants, only in those countries where immigrants represent a significant proportion of cases of TB 5. Discrepancies between local epidemiology and national guidelines might encourage the redirection of resources to improve the treatment success rate of active TB, which is currently less than the World Health Assembly target of 85%. The current authors’ aim was to help clinicians and inform national committees creating guidelines of: 1) consensus, which would define the minimum standard of active case finding for TB across Europe; 2) variations, affected by local epidemiology; 3) guidelines of historical or sociological significance, that are no longer necessary; and 4) differences in the treatment of LTBI.
METHODS
The questionnaire
A questionnaire was developed consisting of three sections: who should be screened; what screening methods are used; and the nature of preventive treatment (see online supplementary material). The questions regarding a positive tuberculin skin test (TST) were based on the American Thoracic Society guidelines interpretation of the available evidence 6, whereby the clinical circumstances determine the cut-off value. Questions included bacillus Calmette–Guérin (BCG) vaccination policy as a check for replies to TST screening. Locations for screening were requested as an indication of the variety of screening programmes available. Questionnaires were in English and Russian; the Russian version was re-translated into English to confirm accuracy of translation.
Data collection
Representatives from each European country were selected on the basis of their official position in their own country's TB programme: all were participants in the TBNET, EuroTB (a WHO collaborating centre for the surveillance of TB in Europe) correspondents and/or members of the TB section of the European Respiratory Society (ERS). They were required to have sufficient command of English or Russian to understand the questionnaire, to have access to their national guidelines and to answer the questionnaire using these guidelines. The only personal opinion that was invited was their estimate of the uptake of LTBI treatment compared with the number eligible under their national guidelines. The questionnaire was returned to each respondent 1 yr after data collection; the respondent was then required to confirm that they had answered each question from their national guidelines and that no personal opinions were expressed, except for the estimate of uptake of LTBI treatment. Any corrections made at this stage were then incorporated into the final manuscript.
All definitions used (except otherwise mentioned) were derived from WHO documents and ERS guidelines 2–4, 7. Epidemiological data for TB in Europe were taken from the EuroTB website 8. These included: the proportion of cases due to foreign-born transmission; the most recent cure/completion rate; and the proportion with concurrent HIV infection.
RESULTS
Respondents
Replies were obtained from 50 out of 53 countries of the WHO European region (all countries except Monaco, San Marino and Andorra; table 1⇓). Two countries have no national guidelines for TB and each gave an expert opinion on current practice (Greece) or guidelines to be issued in late 2008 (Austria). Five (10%) countries had two respondents: in the first survey, replies from four countries were identical, but one differed by 27 out of 65 items; differences were resolved by referral to a TB expert from that country. A total of 12 (24%) respondents provided a copy of national guidelines (eight with a translation of the relevant portions in English), which were used to check the accuracy of the replies; no discrepancies were identified. Eight countries provided modifications when prompted 1 yr later to reassess their responses against national guidelines and exclude personal opinion. In one, there were nine changes due to revised guidelines (Montenegro), another added missing data (Tajikistan) and the remaining six countries made one, two or three changes (two, three and one countries, respectively).
Contacts
Six (12%) out of 50 countries limit screening to contacts of sputum smear-positive pulmonary TB (table 1⇑). The remaining countries are equally divided between those who screen contacts of just pulmonary or all TB cases.
Immigrants
In total, 28 (56 %) countries screen immigrants. In 16 (32 %) out of 50 countries, foreign-born persons contribute >40% of all TB and only two in this category (Denmark and Austria) do not carry out screening of new immigrants. In countries with a moderate contribution (11–40%) of immigrants to TB (seven (14 %) out of 50), three have no screening programme, two carry out selective screening and two screen all new entrants. Six countries with ≤1% and four countries with <5% of TB due to foreign-born subjects have screening programmes for immigrants.
Other high-risk groups
Most countries screen for TB in those who are HIV positive (42 (84%) out of 50), prisoners (37 (74%) out of 50) and hospital contacts of in-patients (35 (70%) out of 50). In 32 (64%) out of 50 countries, information on HIV co-infection was available: two out of 16 countries with an estimated co-infection rate of >5% and two out of 13 countries with a rate of 1–5% have no policy for screening patients with HIV infection for TB. Screening of laboratory staff (33 (66%) out of 50 countries), the homeless (24 (48%) out of 50), teachers (19 (38%) out of 50) and the elderly in long-term care facilities (16 (32%) out of 50) occurs less often. Three replies mention screening for TB before the use of anti-tumour necrosis factor (TNF) treatments and in those whose immunity is depressed for any reason. Individual countries also screen intravenous drug users (Belgium, Macedonia, Portugal, Serbia), selected Roma people (Macedonia, Serbia), the mentally ill (Macedonia, Serbia), prostitutes (Austria) and food handlers (Moldova, Poland).
Process of screening
One-half of the countries specify questioning for symptoms of TB in those who are screened. Almost three-quarters (36 (72%) out of 50) ask for a sputum smear where someone reports cough lasting >3 weeks. Only one country screens all those at risk of TB with chest radiography alone, while the majority (49 out of 50) use TST and 27 (55%) of these obtain a chest radiograph irrespective of the TST result. A total of 12 (24%) countries out of 50 have access to mass radiography screening and seven (14%) out of 50 recommend its use in guidelines. Need for sputum examination is not stipulated by guidelines in 13 (26%) out of 50 countries when a chest radiograph suggests TB. Legal provisions for screening individuals for TB are available in 20 (40%) out of 50 countries, but comments as to how often this was used were not invited.
Many guidelines recommend screening for active TB in a larger group than that eligible for LTBI treatment. In 22 (44%) out of 50 countries, guidelines recommend TST in the same population that will receive LTBI treatment if LTBI is present (one for 0–5 yrs of age; 12 for age 0–16 yrs, five for those up to 35 yrs of age; and four for >35 yrs of age), using other tests to exclude active TB. The median estimated uptake of LTBI treatment in these countries is 55% (data available for 17 countries; range 10–90%), but lower (40%) in countries where TST is used indiscriminately (data available for 14 countries; range 5–92.5%).
Interpretation of the TST varies widely across the European region (table 2⇓). In 28 (56%) out of 50 countries, interpretation of the TST is not modified according to the presence or absence of a BCG scar; two comment that BCG vaccination is compulsory and five have no BCG programme. In the latter five countries, immigration contributes significantly to the incidence of TB (median 44% (range 9–60%)). Most countries set a positive TST for those with concurrent HIV infection (70%) and BCG-negative contacts (17 out of 17 where mentioned) at 5 mm. Induration of 10 mm was the commonest cut-off value for healthy subjects at risk of TB (25 (54%) out of 47), immigrants (20 (41%) out of 49) and BCG-positive contacts (nine (53%) out of 17 where specified).
All countries offer LTBI treatment consisting of either 6 or 9 months of isoniazid. Alternative regimens are offered in 13 (26%) out of 50 countries in specified circumstances: 11 offer 3 months of rifampicin and isoniazid, eight offer 4 months of rifampicin and four offer 2 months of rifampicin and pyrazinamide. One country (Portugal) permits the use of 2 months of rifampicin, isoniazid and pyrazinamide.
LTBI treatment is required in 26 (52%) out of 50 countries, while radiographic follow-up is recorded as an option in 27 (54%) out of 50 countries. Most countries will recommend LTBI treatment for HIV-positive contacts of patients with sputum smear-positive pulmonary TB (40 (80%) out of 50) and for babies born to mothers with pulmonary TB (42 (84%) out of 50; one only if the mother is not on treatment at the time of delivery).
DISCUSSION
The present study is the first survey of national guidelines comparing screening for active and latent TB across the WHO European region. There is a general consensus that advocates screening of TB contacts and offers treatment for individuals with LTBI. Discrepancies between a low detection rate and excessive screening suggest that some programmes could be improved.
Priorities in contact tracing
The two main reasons for contact tracing are identification of active cases and those infected who are at risk of developing active TB later in life. Most patients with active TB have symptoms, such as cough, fever, night sweats and weight loss 9, yet only half of European countries recommend asking for symptoms of active TB in those who are screened. Indeed, a productive cough for >3 weeks can be used as a simple tool to refine the investigation of those with suspected contagious TB 10, 11. The evidence supports both screening of contacts of only pulmonary TB (especially sputum smear-positive TB, which is the most contagious form) 12, 13 and of nonpulmonary TB, the latter on the grounds that they represent an especially high-risk group 13–17. Indeed, DNA fingerprinting has shown that strains of Mycobacterium tuberculosis found in contacts may differ from, rather than be the same as, the strain obtained from the index case 18, 19.
The recognition of LTBI before developing TB is uncertain. A positive TST is most frequent in contacts of those with smear-positive TB 12, 20. Even when the index has nonpulmonary TB, a positive TST is still more common than in any of the other high-risk groups screened 13. Therefore, the current authors advocate that after effective treatment of active TB is established, contacts of all TB patients are screened before any other high-risk group.
Other high-risk groups
Active screening is expensive, and indiscriminate screening is inefficient 5. Several countries had high TB incidence and poor treatment success rates despite extensive screening; higher efficacy and lower cost programmes seem feasible 4, 21. However, mathematical models suggest that active case finding can have a significant impact on the incidence of TB 22, especially in those co-infected with HIV 23. There is a general consensus supporting the screening of patients with HIV infection for TB 1, 7, 13, 24. Treatment of LTBI reduces the risk of active tuberculosis in HIV-positive individuals with positive TSTs, in countries with high 25 and low 26 incidences of TB, although the absolute number of prevented cases in low incidence countries is small. This policy is not yet implemented widely, especially in countries from the Eastern European region, where the TB/HIV co-infection rate is increasing. The newer T-cell interferon-γ release assays (TIGRAs) may be more helpful than the TST in detecting LTBI in individuals infected with HIV 27.
A survey of several European countries has confirmed that the prevalence of TB in prisons is nearly 100-fold greater than in the general population 28, especially in those subjects with HIV infection 29. The risk of TB in i.v. drug users, prostitutes and those with mental illness is also associated with concurrent HIV infection 30–32. TB is spread by aerosols and not by food; the inclusion of food handlers in TB guidelines is therefore a historical anomaly 33.
The problem of the effectiveness of active case finding has particularly exercised national experts in regard to immigrant screening. While immigration can account for a significant proportion of a country's TB notifications, the number of immigrants can be so large that the likelihood of finding a case of TB for an individual clinic or entry point to the country can be extremely small 34. However, the yield may vary depending on the subgroup of immigrants screened. For example, in Belgium, the incidence of TB was significantly higher in asylum seekers than in other immigrants 35. In Israel, Ethiopians and those from the former Soviet Union have more intensive screening 36. Many countries have adopted a selective screening policy based on incidence of TB in the country of origin (table 1⇑). Most immigrants are healthy when they arrive and, therefore, screening by chest radiography is of little benefit. National surveys have shown that most immigrants present with active TB >5 yrs after arrival 37, 38. Programmes with continued screening of immigrants, even on a voluntary basis, can be effective 39, but most transmission is still from individuals who do not attend screening 40. Many have argued that the detection of LTBI can make screening for TB among immigrants cost effective, but only if treatment of LTBI is completed and set-up and administrative costs are excluded 34. Where multidrug-resistant and extensively drug-resistant TB is likely 41, 42, screening can be cost effective in treating active disease and permitting follow-up to recognise the development of drug-resistant TB early 43. Selection of immigrants most likely to have TB (e.g. in deprived city areas) 44, or to adhere to treatment of LTBI 45 may be more effective than general screening.
LTBI
The average number of secondary TB cases acquired from a primary case is affected most by those who develop TB after a latent period 46. Treatment of LTBI holds the greatest benefit in controlling TB after treatment of active, infectious cases. However, few contacts with positive TSTs are offered, and even fewer adhere to, treatment of LTBI 47.
TST is widely undertaken, but surveys to define the best cut-off point for discrimination between those most likely to develop disease and those in whom LTBI is unlikely are rare in Europe 48, with the best such survey being from Canada 49. There is a general consensus that a TST induration diameter of >5 mm should be the cut-off value for those without BCG vaccination or with HIV co-infection (table 2⇑). However, in HIV infection anergy may be an all-or-none phenomenon, so that the size of induration is not helpful 50. Most countries considered TST >10 mm positive in those with a BCG vaccination, but a meta-analysis has suggested that 15 mm might be a better cut-off 51.
TIGRA and LTBI
Large longitudinal studies of TIGRAs have yet to define their role in assessing who will benefit from treatment of LTBI. However, contacts with a positive TIGRA may have a higher risk of developing TB than contacts with a positive TST 52. As merely 1.7% of tuberculin-positive contacts develop active TB 53, the UK National Institute for Health and Clinical Excellence guidelines limit those offered treatment of LTBI by requiring a TST of ≥15 mm and a positive TIGRA 13. Transient TIGRA responses have been observed in TST-negative contacts, suggesting the possibility of resolving acute infection 54. The use of a TIGRA should be prioritised to those most likely to benefit from treatment of LTBI (contacts aged <16 yrs, HIV co-infected individuals and those receiving anti-TNF treatment) 55.
Treatment of LTBI
Treatment of individuals with TST conversion is cost-effective and is important to decrease the global incidence of TB 55. However, completion rates are unsatisfactory in the countries that were surveyed and the relative benefit of treatment of LTBI is very low (to prevent one case of active TB, 72 subjects (95% confidence interval 50–143) need to be treated 53, although this is comparable to the benefit of secondary prevention in myocardial infarction 56). Adherence is often as low as 11–30% 34, 38, 47. While the current authors cannot argue against the individual benefit of completing a course of treatment of LTBI, the public health benefit of the strategy relies critically on large acceptance and completion rates. Clearly, if treatment of LTBI is not going to be undertaken (e.g. where the risks from drug-induced hepatitis outweigh the benefits or indeed the individual does not agree to take any treatment after the risks have been explained clearly to them, unless follow-up is important as in the instance of contact with drug-resistant TB), then TST or TIGRA need not be performed and active TB should merely be excluded (e.g. by symptoms and sputum smear) at the first visit and by appropriate follow-up.
Conclusion
The minimum standard for active case finding is screening all contacts of patients with sputum smear-positive tuberculosis. In addition, screening contacts of those with tuberculosis but without a positive sputum smear and individuals with HIV infection is broadly supported. In order to improve the detection of active tuberculosis, symptoms should be sought and sputum sent for analysis from all those with an abnormal chest radiography suggestive of tuberculosis. Many guidelines could be improved by using T-cell interferon-γ release assays to confirm latent tuberculosis infection and limit testing to those who would agree to be treated. Selective screening, especially of immigrants, should be guided by local epidemiology. While a 6–9 month regimen of isoniazid monotherapy is widely recommended for the treatment of latent tuberculosis infection in Europe, Mycobacterium tuberculosis resistance to isoniazid is increasing and alternative well-tolerated short-term regimens need to be explored.
This article has supplementary material accessible from www.erj.ersjournals.com
Earn CME accreditation by answering questions about this article. You will find these at the back of the printed copy of this issue or online at http://www.erj.ersjournals.com/current.shtml
Support statement
This work was supported by the NHS Culyer allocation, UK.
Statement of interest
None declared.
Acknowledgments
The TBNET contributors are as follows. Official national representative for EuroTB contributors: H. Hafizi (Tirana, Albania); N. Mezhlumyan (Yerevon, Armenia); S. Mamedova (Baku, Azerbaijan); M. Wanlin (Brussels, Belgium); H. Zutic (Sarajevo, Bosnia and Herzegovina); D. Stefanova (Sofia, Bulgaria); A. Simunovic (Zagreb, Croatia); C. Hadjianastassiou (Nicosia, Cyprus); D. Falzon (Saint-Maurice, France); A. Salakaia (Tbilisi, Georgia); S. Georgia (Athens, Greece); T. Blöndal (Rekjavik, Iceland); D. Chemtob (Jerusalem, Israel); B.K. Khasanovna (Almaty, Kazakhstan); A. Elmira (Bishkek, Kyrgystan); J. Leimans (Riga, Latvia); E. Davidaviciene (Vilnius, Lithuania); A. Vidoevska (Skopje, Macedonia); A.P. Asciak (Qormi, Malta); O. Bojovic (Niksic, Montenegro); B.A. Winje (Oslo, Norway); M. Korzeniewska (Warsaw, Poland); I. Solovic (Vysne Hagy, Slovakia); D. Erzen (Golnik, Slovenia); S. Saidaliev (Dushanbe, Tajikistan); D.M. Khaldurdyevna (Ashgabat, Turkmenistan); O. Sakalska (Kiev, Ukraine); and U.D. Djumabaevna (Tashkent, Uzbekistan).
Other TBNETcontributors: B. Schmidgruber (Vienna, Austria); O.M. Kalechitz (Minsk, Belarus); G. Vankersshaever (Brussels, Belgium); S. Kos (Mirosov, Czech Republic); P. Ravn (Herlev, Denmark); K. Kliiman (Tartu, Estonia); R. Ruohonen (Helsinki, Finland); H-P. Mallet (Paris, France); O. Bock-Hensley (Heidelberg, Germany); G. Loytzved (Würzburg, Germany); A. Somoskövi and I. Horvath (Budapest, Hungary); J. Keane (Dublin, Ireland); L.R. Codecasa (Milan, Italy); F. Chaussade (Luxembourg, Luxembourg); V. Soltan (Chisinau, Republic of Moldova); S. Arend (Leiden, the Netherlands); M. Villar (Amadora, Portugal); I. Husar and P Stoicescu (Bucharest, Romania); L. Rybka (Moscow, Russia); D. Pesut (Belgrade, Serbia); J.M. Garcia (Asturias, Spain); J. Bruchfeld and I. Julander (Stockholm, Sweden); J-P. Janssens (Geneva, Switzerland); and O. Ergonul (Istanbul, Turkey).
The authors thank R. Centis, Tradate, Italy, for her kind support and advice.
- Received January 24, 2008.
- Accepted May 23, 2008.
- © ERS Journals Ltd