NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Chronic Conditions (UK); Centre for Clinical Practice at NICE (UK). Tuberculosis: Clinical Diagnosis and Management of Tuberculosis, and Measures for Its Prevention and Control. London: National Institute for Health and Clinical Excellence (UK); 2011 Mar. (NICE Clinical Guidelines, No. 117.)

Cover of Tuberculosis

Tuberculosis: Clinical Diagnosis and Management of Tuberculosis, and Measures for Its Prevention and Control.

Show details

11BCG vaccination

Created: .

11.1. Overview

11.1.1. Overall introduction

Bacille Calmette-Guerin (BCG) was developed by Calmette and Guèrin, at the Pasteur Institute (Lille) using in vitro attenuation by repeated passage of an isolate of M. bovis from 1908 onwards; it was finally tested in humans in 1921. Since BCG has never been cloned and has been grown under different conditions and in different laboratories, genetic differences have developed between the various commercially used strains,{279} so called ‘antigenic drift’. Genome research has since shown that in the passaging of the organism, but before its distribution from the Pasteur Institute, a section of the genome, the RD1 region, was deleted. This deleted region common to all BCG strains contains antigens such as ESAT6, CFP10 and tb7.7 which are now used in interferon gamma based blood tests, and hence these blood tests are not affected by prior BCG vaccination (see section 5.1 for further details).

The efficacy of a vaccine is a measure of its activity on individuals given the vaccine and can be defined as the proportion of those vaccinated who gain protective immunity from the vaccination.{280} Huge variations in estimates of efficacy against pulmonary TB, ranging from 0% to >80%, have been shown for different BCG vaccines in various geographical settings.

While a number of explanations have been put forward for this, geographical latitude seems to have a particularly important effect, accounting for over 40% of the variability in efficacy.{281} Thus nearly zero efficacy against tuberculosis in India,{282} is contrasted with a 64% protective efficacy in people of Indian origin with the same vaccine in a higher, more temperate, latitude.{283} Though the effect of climate on environmental mycobacteria has been suggested as the cause of the latitude effect, this has not been proven.

A further conundrum in BCG efficacy is that even in parts of the world where there is little reported efficacy against tuberculosis, efficacies of 50–60% are reported against leprosy and Buruli ulcer, caused by other mycobacteria.{280} Yet another problem with interpreting the data is that although it was assumed that the tuberculin sensitivity induced by BCG vaccination correlated with protective efficacy, this is not so. In a large UK study there was no correlation between tuberculin sensitivity induced by BCG and protective efficacy; those individuals tuberculin negative after BCG vaccination derived just as much protection as those who became tuberculin positive.{284}

Many controlled trials have followed efficacy for 10–15 years and have shown some decline over time, but the total duration of any benefit was not known and could only be expressed as an efficacy lasting up to 15 years.{285} The only truly long-term follow-up of BCG vaccination, in a North American aboriginal population, reported in 2004, showed 50% protective efficacy lasting for at least 50 years.{286}

BCG is a live vaccine and as such is contraindicated{3} in a number of situations where the immune system may be compromised, particularly if the person is known or suspected to be HIV positive, because of the risk of generalised BCG infection. HIV testing, after appropriate counselling, is also an important consideration, but lies outside the scope of this guideline. Readers should be aware of the British HIV Association guidelines on TB/HIV co-infection{8} and those forthcoming on testing from the British Association for Sexual Health and HIV.

Current practice in vaccination is led by the advice of the Joint Committee on Vaccination and Immunisation, principally through the ‘Green Book’.{3},{21}

11.1.2. OVERALL RECOMMENDATIONS

R88.

When BCG is being recommended, the benefits and risks of vaccination and remaining unvaccinated should be discussed with the person (or, if a child, with the parents), so that they can make an informed decision. This discussion should be tailored to the person, be in an appropriate language, and take into account cultural sensitivities and stigma. D(GPP)

R89.

People identified for BCG vaccination through occupational health, contact tracing or new entrant screening who are also considered to be at increased risk of being HIV positive, should be offered HIV testing before BCG vaccination16. (See section 10.1 for details of further action in HIV-positive patients.) D(GPP)

11.2. For neonates

11.2.1. Clinical introduction

Neonatal BCG (up to age three months) is given in countries, or in subgroups defined by ethnicity and/or deprivation, with high rates of TB disease. Efficacy studies on neonatal BCG have used different end points which have contributed to some confusion about its efficacy in various settings. These have included the end points of pulmonary disease, death, TB meningitis, disseminated (miliary) disease, and laboratory-confirmed cases.

In England and Wales, which has had a selective neonatal BCG programme for over 20 years, assessments of coverage of appropriate infants have shown substantial variation in, and deficiencies in, both BCG policy and implementation.{287} These deficiencies and system problems were particularly in medium and low TB incidence districts which often had no system for identifying those neonates for whom BCG was recommended.

11.2.2. Current practice

The DH advises BCG vaccination for all neonates at higher risk of TB, with opportunistic vaccination of older children as necessary, according to criteria set out below in the recommendations.

The review of current services, conducted in the year prior to the introduction of neonatal vaccination and abolition of school-based vaccination, found that outside London, only two of 62 clinics (3%) (in the same HPU, an area of high notifications) reported universal neonatal BCG vaccination. In London, 12 of 31 clinics (39%) reported universal coverage. There was no consistency in the risk groups used for selected neonatal BCG. Many respondents did not name any explicit risk groups, but those who gave details mostly cited ethnicity, immigration and family history as the means for identifying neonates at higher risk.

11.2.3. Methodological introduction

Studies investigating the effectiveness of BCG vaccination administered in neonates and infants in preventing the development of TB infection or disease were sought. This was compared to unvaccinated groups in relevant populations. One meta-analysis, one cohort study and one case control study were found.

One meta-analysis conducted in the USA{288} included five RCTs and 11 case control studies in the analysis. The scope was international, but all RCTs were conducted in the northern hemisphere and were situated far from the equator relative to case controls, which were distributed across both temperate and equatorial regions. The analysis combined RCT and case control studies separately and did not use cross-design analysis since there were too few RCTs relative to case control studies. It was therefore appropriate to grade the evidence statements according to whether they were derived from the RCT (level 1) or case control results (level 2).

Factors for consideration raised by the meta-analysis included the following:

  • The duration of BCG vaccination protection administered in infancy was inadequately established despite information on this issue being available from six studies. This was due to the small numbers of TB cases when data was analysed separately by year of occurrence.
  • The impact of BCG strain on efficacy of immunisation was not associated with variation in the protection afforded by the vaccine in the studies reviewed.
  • Differences in the characteristics and methodological quality of individual studies were addressed by a sensitivity analysis, expressed as a study quality validity score.
  • Study quality validity scores accounted for 15.3% of the heterogeneity in the results of the nine case control studies, while RCTs were homogeneous.
  • Distance from the equator did not appear to be an important correlate of BCG efficacy reported by case control studies, while RCTs displayed homogeneity in terms of distance from the equator.

One cohort study conducted jointly in the Federal Republic of Germany (FRG) and the German Democratic Republic (GDR),{289} was published prior to the meta-analysis, but not cited in it. The study retrospectively focused on BCG vaccination administered to an entire population of neonates in the GDR over a three and a half year period compared to no vaccination in the FRG over the same time period to investigate the efficacy of the vaccine in preventing cases of TB meningitis.

A case control study conducted in Spain,{290} which was not cited in the meta-analysis was excluded due to methodological limitations presented in Appendix I.

11.2.4. Evidence statements

Evidence was found for the efficacy of BCG vaccination in infancy for preventing:

  • pulmonary TB disease
  • TB deaths
  • TB meningitis
  • laboratory-confirmed TB cases
  • disseminated TB.

Evidence for these five outcomes is presented in Table 34.

Table 34. Summary of evidence: neonatal BCG vaccination.

Table 34

Summary of evidence: neonatal BCG vaccination.

11.2.5. Health economics

The GDG considered the interactions between neonatal and school-age BCG vaccination programmes required population dynamic economic modelling, which is, at the time of writing, being commissioned by the DH. With this in mind, recommendations on neonatal BCG are presented purely on the basis of clinical evidence, pending the findings of the model.

11.2.6. From evidence to recommendations

Neonatal BCG is significantly better than no vaccine using the end points of pulmonary disease, death, meningitis, laboratory-confirmed TB and disseminated TB.

There is difficulty ensuring thorough vaccination coverage in primary care, where babies are not registered until the first appointment, compared to vaccination by midwives, for example, where coverage can be assured.

The GDG supported the explicit criteria set out by the WHO for discontinuing universal vaccination, but wished TB clinicians and service planners to be aware of possible future changes to the criteria in response to changing global epidemiology. The aim of this section is to guide clinicians in vaccinating those who are most at risk.

Given the conclusions of the health economics for school-based BCG vaccination in section 11.3, the recommendations seek to provide guidance for a neonatal BCG programme that will offer protection to all who are at risk. In a high-incidence area, this may be most easily provided by a universal programme.

The largest group of neonates who are at increased risk of TB are those whose families have immigrated from high-incidence countries. Neonates continue to be at risk even if their parents were also UK born because of continuing migration, home visits and exposure to increased levels of TB within communities. The recommendations therefore advise selection on the basis of a parent or a grandparent being born in a high-incidence country. GDG members were aware of selection being practised on the basis of skin colour or surname, and aimed to provide clear-cut recommendations to replace these practices.

In accordance with the Green Book,{3} tuberculin skin testing is not routinely recommended prior to BCG vaccination for children under six years of age.

11.2.7. RECOMMENDATIONS

R90.

Neonatal BCG vaccination for any baby at increased risk of TB should be discussed with the parents or legal guardian. D(GPP)

R91.

Primary care organisations with a high incidence of TB17 should consider vaccinating all neonates soon after birth. D(GPP)

R92.

In areas with a low incidence of TB17, primary care organisations should offer BCG vaccination to selected neonates who: D(GPP)

  • were born in an area with a high incidence of TB17, or
  • have one or more parents or grandparents who were born in a high-incidence country,18 or
  • have a family history of TB in the past five years.
R93.

Mantoux testing should not be done routinely before BCG vaccination in children younger than six years. D(GPP)

Cross-referring:

11.3. For infants and older children

11.3.1. Clinical introduction

Following clinical trials in the early 1950s, BCG vaccination was introduced for previously unvaccinated adolescents aged 10–14.{284} Age 10–14 was selected for vaccination in 1953 because at that time, in what was nearly entirely a white UK-born population, TB was most common in those aged 15–29 (with a second peak in older people). This cohort, now aged over 70, have the highest TB rates among white UK-born people (see Appendix G). The rationale therefore was to give vaccination at this age to try to prevent acquisition of pulmonary disease before this peak, and it became known as the ‘Schools BCG Programme’. During the writing of this guideline, the DH abolished the programme, replacing it with neonatal vaccination based on the criteria given above.

Tuberculosis rates fell through the 1950s and early 1960s by almost 10% per annum, and continued to fall at a lower rate until 1987 (approximately), since when there has been an increase. However, over this time, both the proportion of cases and rates of disease in the white UK-born ethnic group have continued to fall. The proportion of cases in this ethnic group was 85% in 1985, 43% in 1993, 37% in 1998, and is now under 30%.{140} Rates of TB in white UK-born children aged 10–14 years, the cohort of previously unvaccinated children to whom the schools programme applies, are between one and two cases per 100,000 for both sexes (see Appendix G).

International criteria for discontinuation of unselective BCG vaccination

The International Union against Tuberculosis and Lung Disease published their criteria for discontinuation of BCG programmes in countries of low prevalence in 1993.{291} This set out general considerations and criteria. The general criteria to be met in a country before stopping or modifying BCG programmes were:

  • there is a well functioning TB control programme
  • there has been a reliable monitoring system over the previous five years or more enabling the estimation of the annual incidence of TB by age and risk groups, with particular emphasis on TB meningitis and sputum smear-positive pulmonary TB
  • due consideration has been given to the possibility of an increase in the incidence of TB resulting from HIV infection.

The criteria for discontinuing a BCG vaccination programme in a country with a low prevalence of TB were:

  • the average annual notification rate of sputum smear-positive pulmonary TB should be five cases/100,000 population or less during the previous three years, or
  • the average annual notification rate of TB meningitis in children under age five years of age should be less than 1 case per 10 million general population over the previous five years, or
  • the average annual risk of TB infection should be 0.1% or less.

Additional considerations were also suggested.

Cost: with it being advisable, but not essential, to calculate the number of cases which would be prevented by continuing BCG vaccination, so that the saving can be expressed in terms of preventing human suffering and also in saving of cost of treatment.

Adverse reactions to BCG: documentation of the rate of adverse reactions to BCG vaccination in a country are helpful. A low incidence rate of active tuberculosis, coupled with a high rate of adverse reaction tends to reinforce a decision to stop or modify the BCG vaccination programme. The reported rates of serious adverse reactions varies from country to country, with vaccination technique used, the preparation of BCG vaccination used, and doctors' awareness of reactions being factors influencing the reported rates.

Risk groups: in the event of discontinuation of the BCG vaccination programme for the general population, it may be advisable to continue vaccination in certain well-defined population groups with a known high notification rate of active tuberculosis.

11.3.2. Current practice

The Department of Health no longer recommends BCG vaccination for school children between ages 10–14 years.

11.3.3. Methodological introduction

The focus was on studies investigating the effectiveness of BCG vaccination administered in a school-aged population in preventing TB infection or disease. One RCT and two cohort studies were found that addressed the topic.

One RCT conducted in the UK{285} reported on the protective efficacy of BCG vaccination against tuberculosis (TB) disease in vaccinated and unvaccinated groups of school-aged subjects in England over a 20-year follow-up period. Two cohort studies, both conducted in the UK,{292},{293} retrospectively identified notified cases of TB disease who had been eligible for BCG vaccination within the schools vaccination scheme when aged 13.{292},{293} These studies estimate the protective efficacy of the BCG vaccine in this general population and in the white ethnic group. Sutherland and Springett{292},{293} estimate the numbers of additional TB notifications that would be expected among young white adults annually, if the schools BCG scheme were to be discontinued at specific dates. Both cohort studies incorporated data from the RCT cited above.

11.3.4. Evidence statements

Efficacy of BCG vaccination for preventing TB disease

One RCT{285} and one cohort study{292} found that BCG given in school-aged children led to a reduction in the annual incidence of TB disease in vaccinated compared to unvaccinated individuals. Evidence is presented in Table 35.

Table 35. Summary of evidence: vaccinated and unvaccinated children of school-going age.

Table 35

Summary of evidence: vaccinated and unvaccinated children of school-going age.

BCG vaccination in school-aged children and longitudinal trends in TB prevention

Evidence was found on BCG vaccination use in school-aged children in England and Wales and the following longitudinal trends:

  • decrease in the efficacy of BCG and the incidence of TB notifications
  • the estimated risk of notified TB in the white ethnic population eligible for the school's BCG vaccination scheme
  • TB notifications prevented by BCG vaccination in the white school-aged population
  • TB notifications as a consequence of discontinuing the BCG schools vaccination scheme for the white ethnic population
  • the estimated risk of notified TB in the white ethnic group if the school's BCG vaccination scheme were discontinued.

The evidence is presented in Table 36.

Table 36. Summary of evidence: vaccination and longitudinal trends in TB among children of school-going age.

Table 36

Summary of evidence: vaccination and longitudinal trends in TB among children of school-going age.

11.3.5. Health economics

A decision analytic model was used to estimate the cost-effectiveness of the current school BCG programme. The model distinguished between a ‘high-risk’ group of children who should have already been offered BCG before the school programme (through neonatal or new entrant schemes) and a ‘low-risk’ group, which is the remainder of the 10–14-year-old cohort. The school BCG programme is potentially beneficial for low-risk children and as a catch-up for previously unvaccinated high-risk children. The model relies on the assumption that there is negligible transmission between the high-risk and low-risk groups.{294}

The model is a simple decision tree that estimates the number of primary cases for a cohort of 10–14-year-olds, the consequent number of secondary cases in the population, and the associated costs and health outcomes, with and without a school BCG programme. The effectiveness of school BCG for the low-risk group and the number of secondary cases per primary case were taken from Saeed et al (2002),{295} updating the work of Sutherland and Springett in 1989.{293} The benefits for unvaccinated high-risk children were then estimated. It is important to note that this method can only give approximate results for an infectious disease such as TB. A population dynamic model would be expected to provide more reliable results.

Whenever possible, the input parameters and assumptions for the model were based on best available empirical evidence. However, we could not find evidence to inform all of the important parameters. In such cases, estimates are based on judgement by the guideline economist and the GDG. There is some uncertainty over the results of the model due to uncertainty over some of the input parameters for the analysis. In particular, the results are sensitive to the proportion of 10–14-year-olds in ‘high-risk’ groups, the estimated QALY loss due to TB, and the estimated cost of treating a case of TB.

Cost-effectiveness of school BCG for the low-risk group

The economic model suggests that the schools programme is not cost-effective for the low-risk group alone – with 0% in the high-risk group, the incremental cost per QALY gained (incremental cost-effectiveness ratio, ICER) is over £150,000 if we assume 15-year protection from BCG, and over £750,000 if we assume only 10-year protection. School BCG appears to be cost-effective for the ‘low-risk’ population only if their 10–15-year risk is very high: approximately 0.13–0.15%. This compares with current estimates of 0.03% (age 15–24) or 0.05% (age 15–29) (see Table 37).

Table 37. Cost-effectivensss of school BCG for low-risk group only by baseline risk of TB.

Table 37

Cost-effectivensss of school BCG for low-risk group only by baseline risk of TB.

Cost-effectiveness of school BCG as a catch-up for unvaccinated high-risk children

Based on the assumptions that 64% of high-risk children have been previously vaccinated, that they have a relative risk of 40 (compared with the low-risk group), and that BCG offers protection for 10 years, the schools programme appears to be cost-effective for areas with around 25–30% or more children in the high-risk group. If we assume 15-year BCG protection, school BCG appears cost-effective with around 10–15% or more in the high-risk group (see Table 38).

Table 38. Cost-effectiveness of school BCG by percentage of cohort in high-risk group.

Table 38

Cost-effectiveness of school BCG by percentage of cohort in high-risk group.

These results are sensitive to the estimated mean cost of treatment and QALY loss per case of TB age 15–24/29.

11.3.6. From evidence to recommendations

The GDG noted that the schools BCG programme was for those at low risk of TB and previously unvaccinated, whilst those at higher risk of TB (see section 10.2) receive BCG vaccination either at birth or on entry to the UK.

Whilst BCG in school-age children has a protective efficacy of 75–80% lasting 10–15 years, the incidence of active TB in those at low risk is now in the order of 1 case per 100,000, with a continuing downward trend.

England and Wales meet the accepted international criteria for the cessation of universal BCG vaccination in a low-prevalence country,{291} and have done so at least since 2000.

Economic modelling shows that the schools programme is not cost effective, and extremely expensive with an incremental cost-effectiveness ratio between £696,000 and £767,000 for low-risk individuals.

The schools programme becomes cost-effective only if 15% or more of the children included are at higher risk and previously unvaccinated.

For these reasons, it was felt that routine BCG vaccination of children aged 10 to 15 in schools should not continue. Those children at risk will either have been vaccinated neonatally (see section 11.2) or on entry to the UK (see section 11.4). Where universal childhood screening and vaccination is thought appropriate for an area because of very high local incidence, then this would be better achieved by a local universal neonatal BCG policy.

11.3.7. RECOMMENDATIONS

R94.

Routine BCG vaccination is not recommended for children aged 10–14.

  • Healthcare professionals should opportunistically identify unvaccinated children older than four weeks and younger than 16 years at increased risk of TB (see section 10.2) who would have qualified for neonatal BCG and provide Mantoux testing and BCG (if Mantoux negative). C
  • This opportunistic vaccination should be in line with the Chief Medical Officer's advice on vaccinating this age group following the end of the school-based programme20 D(GPP)
R95.

Mantoux testing should not be done routinely before BCG vaccination in children younger than six years unless they have a history of residence or prolonged stay (more than one month) in a country with a high incidence of TB21. D(GPP)

11.4. For new entrants from high-incidence countries

11.4.1. Clinical introduction

The incidence of tuberculosis in new entrants from countries of high incidence (40/100,000 per year or greater) is high, peaking 2–3 years after first entry, and falling significantly after 10 years, but remaining well above general UK population rates (see Appendix G). Up to 30% of such recent arrivals from the Indian subcontinent are tuberculin negative.{296},{297} Since they will be living in communities with a rate of TB some 25 times that of the white UK-born community, they may benefit from BCG vaccination to reduce the risk of acquiring TB disease. Such a BCG policy would however have to take into account the possibility of false negative Mantoux test from HIV co-infection.

11.4.2. Current practice

In the Department of Health's Immunisation against infectious diseases (the Green Book) 1996,{3} the following recommendation is made for new entrants from countries with a high prevalence of tuberculosis, their children and infants wherever born.

‘New entrants to the UK, including students, from countries with a high prevalence of tuberculosis, and all refugees and asylum seekers, should be tuberculin tested as part of the initial screening procedure unless there is definite evidence of a BCG scar. Those with positive reactions should be referred for investigation as they may require chemoprophylaxis or treatment. BCG immunisation should be offered immediately to those who are tuberculin negative.’

Under section 32.4.1d of the same document HIV-positive individuals are listed as one of the contraindicated groups to whom BCG vaccine should not be given with the following comment:

‘BCG is absolutely contraindicated in symptomatic HIV positive individuals. In countries such as the UK where the risk of tuberculosis is low, it is recommended that BCG is withheld from all subjects known or suspected to be HIV positive, including infants born to HIV positive mothers. There is no need to screen mothers for HIV before giving BCG as part of a selective neonatal immunisation programme (see 32.3.2(e))’.

The newly updated chapter of the draft 2006 Green Book{21} states:

‘BCG immunisation should be offered to… previously unvaccinated, tuberculin-negative new entrants under 16 years of age who were born in or who have lived for a prolonged period (at least three months) in a country with an annual TB incidence of 40/100,000 or greater.’

Readers should also be aware of the recommendations made for neonates (see section 11.2).

11.4.3. Methodological introduction

Studies investigating the effectiveness of BCG vaccination in new entrants from high-risk countries in preventing TB infection or disease were targeted. No systematic reviews, randomised controlled trials, cohort or case control studies were found that directly addressed the area.

One meta-analysis conducted in the USA{298} demonstrated that BCG vaccine had protective efficacy across a wide range of study conditions, BCG strains, populations, age ranges and vaccine preparation methods. BCG efficacy in new entrants from countries with a high TB incidence was not addressed.

Since the meta-analysis did not use cross-design analysis, it was appropriate to grade evidence statements according to whether they were derived from the RCT (level 1), clinically controlled trial (level 2) or case control study (level 2) results.

Factors for consideration raised by the meta-analysis included:

  • differences in the characteristics and methodological quality of individual studies were addressed by a sensitivity analysis, expressed as a study quality validity score
  • among 13 prospective trials, study validity explained 30% of the between-study variance in the trials, and geographical latitude accounted for 41% of the variance
  • among the 10 case-control studies, data validity score was the only variable to explain a substantial amount (36%) of the heterogeneity
  • different strains of BCG were not associated with more or less favourable results in the 13 trials, as differing BCG strains administered in the same populations provided similar levels of protection.

One non-analytic study from the UK{299} was excluded due to methodological limitations presented in Appendix I.

11.4.4. Evidence statements

Evidence was found for the efficacy of BCG vaccination in preventing:

  • pulmonary TB disease
  • TB deaths
  • TB meningitis
  • disseminated TB.

Evidence for these four outcomes is presented in Table 39.

Table 39. Summary of evidence: BCG vaccination for new entrants.

Table 39

Summary of evidence: BCG vaccination for new entrants.

11.4.5. From evidence to recommendations

The GDG noted that there was little data in this field. The high rates of tuberculosis in recently arrived new immigrants from high incidence countries was also noted from epidemiological data over the last 25 years.

Although there is no direct evidence in this group in the UK, the meta-analysis cited above was regarded as applicable.

Analysis of the evidence on BCG efficacy has shown no evidence for persons aged over 35. The GDG felt that for this pragmatic reason, BCG vaccination should be limited to those under 36, unless they have occupational risk factors.

11.4.6. RECOMMENDATIONS

Readers should also be aware of the recommendations under new entrant screening (section 12.8). This process should include Mantoux tests on appropriate new entrants and risk assessment for HIV prior to vaccination.

R96.

BCG vaccination should be offered to Mantoux-negative new entrants22 who:

  • are from high-incidence countries, and
  • are previously unvaccinated (that is, without adequate documentation or a characteristic scar), and B
  • are aged:
    • younger than 16 years, D(GPP) or

    16 to 35 years23 from sub-Saharan Africa or a country with a TB incidence of 500 per 100,000.

11.5. For healthcare workers

11.5.1. Clinical introduction

Although earlier studies had not shown an association, in the 1990s healthcare workers were shown to have twice the expected incidence of TB, allowing for age, sex and ethnic factors.{300} Because of the risk of exposure, it became standard practice to recommend BCG vaccination to people commencing healthcare work who would have contact with patients or clinical material, if they had not had prior BCG vaccination, and were Mantoux test negative.

11.5.2. Current practice

In Immunisation against infectious disease (the Green Book),{3} the Department of Health recommended BCG vaccination for all those at higher risk of tuberculosis. Under section 32.3.2a this included:

Health service staff who may have contact with infectious patients or their specimens. These comprise doctors, nurses, physiotherapists, radiographers, occupational therapists, technical staff in microbiology and pathology departments including attendants in autopsy rooms, students in all these disciplines, and any others considered to be at high risk. It is particularly important to test and immunise staff working within maternity and paediatric departments, and departments in which patients are likely to be immunocompromised, eg transplant, oncology and HIV units.’

The newly updated chapter of the draft 2006 ‘Green book’{21} states:

‘People in the following occupational groups are more likely than the general population to come into contact with someone with TB:

  • healthcare workers who will have contact with patients or clinical materials
  • laboratory staff who will have contact with patients, clinical materials or derived isolates…’

11.5.3. Methodological introduction

Studies investigating the efficacy of BCG vaccination in health care workers for preventing the development of TB infection or disease in comparison to unvaccinated healthcare workers were targeted. One systematic review was found that addressed the topic.

One systematic review conducted in the USA{301} included two randomised controlled trials, two prospective cohort studies, one historically controlled study, one retrospective cohort study and six non-analytic studies. Information on the study methods and results was reported for only four of the six non-analytic studies. The scope was international, but all 12 studies were conducted in the northern hemisphere, 10 in temperate zones situated far from the equator, the eleventh in California, and for the twelfth, the specific setting was unknown.

The systematic review was methodologically sound, and hence it could technically be given a grading of 1+. However, the review did not conduct a meta-analysis due to the heterogeneity of study designs and methodological limitations in each of the studies. The methodological limitations of individual studies contained within the review meant that there was insufficient robust data from which to derive evidence statements for this area. The review authors noted that despite methodological limitations, all six controlled studies reported a protective effect for BCG vaccination.

11.5.4. From evidence to recommendations

Whilst the systematic review was sound, all of the studies had multiple methodological flaws. There was however a consistent trend to benefit in the six controlled studies. Also, given the weight of evidence for the efficacy of BCG in other settings, it seemed unlikely that BCG would not be effective in this population. The GDG also noted that potential TB exposure continues throughout a career in individuals with patient or clinical material contact, and is not age limited.

There is not sufficient age-specific evidence to make recommendations on BCG vaccination for people over 35 but vaccination is recommended for healthcare workers of all ages because of the increased risk to them – and consequently the patients they care for – if they remain unvaccinated.

11.5.5. RECOMMENDATIONS

R97.

BCG vaccination should be offered to healthcare workers, irrespective of age24 who: D(GPP)

  • are previously unvaccinated (that is, without adequate documentation or a characteristic scar), and
  • will have contact with patients or clinical materials, and
  • are Mantoux (or interferon gamma) negative.

Cross-referring:

11.6. BCG vaccination for contacts of people with active tuberculosis

11.6.1. Clinical introduction

Contacts of cases of pulmonary tuberculosis are at risk of contracting TB. This is particularly the case with household or close contacts of sputum smear-positive disease, where up to 10% become infected (see section 12.2). It may take several weeks to develop an immune response to infection, as judged by a positive tuberculin skin test. A second Mantoux test has to be performed in those whose initial test is negative, six weeks after the initial negative one and a decision made with the second result.{6} Those with serial negative skin tests are deemed not to have been infected, but BCG vaccination up to and including the age of 35 years is recommended. The index case should be rendered non-infectious within a few weeks by anti-tuberculosis drug treatment, but tuberculin-negative contacts remain at risk if there are secondary cases.

11.6.2. Current practice

The Department of Health's Immunisation against infectious disease (the Green Book) 1996{3} recommended BCG vaccination for all those at higher risk of tuberculosis.{3} Under section 32.2d this included:

Contacts of cases known to be suffering from active pulmonary tuberculosis. Contacts of a sputum smear positive index case may have a negative tuberculin skin test when first seen but be in the early stages of infection before tuberculin sensitivity has developed. A further skin test should be performed six weeks later and immunisation only carried out if this second test is negative. (If the second skin test is positive, the patient has converted and must be referred for consideration of chemoprophylaxis). However, if for some reason a further test is impossible, vaccine may be given after the first test. Newly born babies should be given prophylactic isoniazid chemotherapy and tuberculin tested after three to six months. If the skin test is positive, chemoprophylaxis is continued; if negative, BCG vaccine is given provided the infant is no longer in contact with infectious tuberculosis. Newly born contacts of other cases should be immunised immediately.’

The newly updated chapter of the draft 2006 Green Book{21} states:

‘BCG immunisation should be offered to… previously unvaccinated tuberculin-negative contacts of cases of respiratory TB (following recommended contact management advice – currently Joint Tuberculosis Committee of the British Thoracic Society 2000 [{6}] and National Institute for Health and Clinical Excellence 2006 [this document]…’

11.6.3. Methodological introduction

The focus was on studies investigating the efficacy of BCG vaccination in contacts of those with diagnosed active tuberculosis disease in comparison to unvaccinated contacts from the same population. One cohort study and five non-analytic studies were identified. All studies addressed BCG vaccination of contacts prior to their exposure to the index case.

One prospective cohort study conducted in South Korea{302} over a period of approximately two and a half years reported on the protective efficacy of BCG vaccination against TB disease in child contacts. Four studies{278},{303},{304},{305} reported contact tracing results that included stratification of contacts by BCG vaccination status. BCG vaccination status was not the primary variable used to generate group allocation or to stratify the analysis of the results, and for this reason the studies were classified as non-analytic. One study was conducted in the UK (England, Wales and Scotland) and two studies in Scotland. A fourth study conducted in Brazil dealt with contacts of index cases diagnosed with MDR TB. Although the latitude effect could have influenced the study findings, the study was included since it focused on BCG vaccination in a contact population at risk of acquiring MDR TB disease. MDR TB is not addressed in the three UK-based studies.

A fifth non-analytic study was excluded due to methodological limitations, which are presented in the appendix I.

11.6.4. Evidence statements

Evidence on the efficacy of BCG vaccination in preventing TB disease was found for contacts:

  • of index cases
  • of index cases diagnosed with MDR TB
  • belonging to different ethnic groups

The evidence is presented in Table 40.

Table 40. Summary of evidence: BCG vaccination for contacts of people with TB.

Table 40

Summary of evidence: BCG vaccination for contacts of people with TB.

11.6.5. From evidence to recommendations

The appraised evidence shows some protective efficacy for BCG vaccination given before contact with tuberculosis, but none of the studies addressed the efficacy of BCG administered to tuberculin-negative contacts after exposure to TB. However, such individuals may be at increased risk from secondary TB cases if not vaccinated. As for new entrants, the potential benefit of BCG vaccination is reduced with age, and there is no reason to change the upper age limit of 35 years, which is currently widely used.

RECOMMENDATION

R98.

BCG vaccination should be offered to Mantoux-negative contacts of people with respiratory TB (see section 12.2 for details of contact tracing) if they are previously unvaccinated (that is, without adequate documentation or a characteristic scar) and are: D(GPP)

  • aged 35 or younger
  • aged 36 and older and a healthcare or laboratory worker who has contact with patients or clinical materials (see section 11.5).

Cross-referring:

11.7. Other groups

The Department of Health currently recommends BCG vaccination for a range of other people who may be at risk from TB.{21} This guideline concentrated on the groups given individually above but for completeness this section addresses the other groups at risk, who stand to benefit from BCG vaccination. For veterinary surgeons, abattoir workers and other people working with animals, there are a number of possible sources of infection, but no standard occupational health screening. Workplace screening is likely to be provided by private sector firms, and is therefore outside the remit of NICE. However, a number of regulations apply:

  • the Reporting of Injuries, Diseases and Dangerous Occurrences Regulations 1995, which require employers to notify the Health and Safety Executive
  • the Management of Health and Safety at Work Regulations 1999, which require general standards of risk assessment
  • the Control of Substances Hazardous to Health Regulations 2002, which require employers to assess infection risk and prevent or control exposure.

11.7.1. RECOMMENDATION

R99.

BCG vaccination should be offered to previously unvaccinated, Mantoux-negative people aged35 or younger in the following groups at increased risk of exposure to TB, in accordance with the ‘Green Book’:{21}D(GPP)

  • veterinary and other staff such as abattoir workers who handle animal species known to be susceptible to TB, such as simians
  • prison staff working directly with prisoners
  • staff of care homes for elderly people
  • staff of hostels for homeless people and facilities accommodating refugees and asylum seekers
  • people going to live or work with local people for more than 1 month in a high-incidence country.26

See 11.5 for advice on healthcare workers.

See the British HIV Association guideline for details of further action in HIV-positive patients. Available from www​.bhiva.org.

As defined by the HPA; go to www​.hpa.org.uk and search for ‘tuberculosis rate bands’.

Go to www​.hpa.org.uk and search for ‘WHO country data TB’.

Available from www​.dh.gov.uk

More than 40 cases per 100,000 per year, as listed by the Health Protection Agency (go to www​.hpa.org.uk and search for ‘TB WHO country data’).

People who have recently arrived in or returned to the UK from high-incidence countries.

The draft 2006 Green Book recommends BCG for new entrants only up to the age of 16. However in this guideline BCG is recommended for those up to 35 years who come from the countries with the very highest rates of TB because there is some evidence of cost-effectiveness.

As outlined in the Green Book, there is not sufficient age-specific evidence to make recommendations on BCG vaccination for people older than 35 (see full guideline for details). However, in this guideline BCG vaccination is recommended for healthcare workers of all ages because of the increased risk to them – and consequently the patients they care for – if they remain unvaccinated.

Go to www​.hpa.org.uk and search for ‘WHO country data TB’.

Footnotes

16

See the British HIV Association guideline for details of further action in HIV-positive patients. Available from www​.bhiva.org.

17

As defined by the HPA; go to www​.hpa.org.uk and search for ‘tuberculosis rate bands’.

18

Go to www​.hpa.org.uk and search for ‘WHO country data TB’.

19

Some of these would be secondary additional notifications outside the age group 15–29 years of age.

20

Available from www​.dh.gov.uk

21

More than 40 cases per 100,000 per year, as listed by the Health Protection Agency (go to www​.hpa.org.uk and search for ‘TB WHO country data’).

22

People who have recently arrived in or returned to the UK from high-incidence countries.

23

The draft 2006 Green Book recommends BCG for new entrants only up to the age of 16. However in this guideline BCG is recommended for those up to 35 years who come from the countries with the very highest rates of TB because there is some evidence of cost-effectiveness.

24

As outlined in the Green Book, there is not sufficient age-specific evidence to make recommendations on BCG vaccination for people older than 35 (see full guideline for details). However, in this guideline BCG vaccination is recommended for healthcare workers of all ages because of the increased risk to them – and consequently the patients they care for – if they remain unvaccinated.

25

Using Cox's regression test, ethnicity was no longer associated with incidence of TB disease.

26

Go to www​.hpa.org.uk and search for ‘WHO country data TB’.

Copyright © 2006, Royal College of Physicians of London.

For 2006 original guideline text, no part of the content may be reproduced in any form (including photocopying or storing it in any medium by electronic means and whether or not transiently or incidentally to some other issue of this publication) without the written permission of the Royal College of Physicians of London. Applications for the Royal College of Physicians of London's written permission to reproduce any part of this publication should be addressed to the publisher.

For 2011 updated text, the material may be freely reproduced for educational and not-for-profit purposes. No reproduction by or for commercial organisations, or for commercial purposes, is allowed without the express written permission of NICE.

Bookshelf ID: NBK97836
PubReader format: click here to try

Views

  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (4.2M)
  • Disable Glossary Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...