Skip to main content

Table 3 Primary Outcomes: TB Incidence and Vaccine Efficacy under the Evaluated Prior and Current Vaccination Policies, by Country

From: A systematic review of BCG vaccination policies among high-risk groups in low TB-burden countries: implications for vaccination strategy in Canadian indigenous communities

Study Characteristics

Study Outcomes

Policy Recommendations

Ref, Year

CTR

ST

Study Pop

Current / prior BCG policy (cov) a

New policy (cov) a

Int description b

Cont description

N (int)

N (cont)

TB Incidence and Vaccine Efficacy

Risk groups for Targeted Vaccination

Observations and General Recommendations

TB Inc. (cont) n/N

TB Inc. (int) n/N

% Efficacy (95%CI)

[48], 1990

Can

CC

Alberta indigenous communities (Treaty Indians)

All Treaty Indians in Alberta are given BCG vaccination at birth since 1948

 

Cases - TB

No TB

160

314

  

57 (23.4–75.3)

Indigenous population

Vaccination in this population should continue until incidence rates fall (cut-off not specified) (as benefits in this high-incidence community outweigh risks and costs), however, vaccination at an older age (instead of at birth) warrants consideration given the finding of higher protection among those vaccinated after 6 months of age (63% efficacy after 6 months of age, 42% before 6 months). Diagnosis and treatment should be strengthened as control measures.

[11], 1994

Czh

CS

Non-BCG vaccinated children

Mass BCG vaccination in all newborns (discontinued in 3 regions of the country in 1986)

Vaccination only in high-risk infants

Non-vaccinated

NA

184,648

NA

LTBI: 283/184648

  

Unspecified high-risk infants

Risk of infection was low, so re-introduction of mass vaccination is not necessary. Overcrowded living conditions were associated with a higher risk of infection (also relevant to Canadian Indigenous context).

[10], 1993

Czh

QE

Czech children

Mass vaccination

Discontinuation of mass vaccination in one region in 1986

Region of ceased vaccination

Regions of continued mass vaccination

148,560

600,195

31/148560 (not vaccinated)

24/600195 (vaccinated)

80%

 

Although incidence increased in the region where mass vaccination was stopped, mass vaccination found to be obsolete due to the small number of children that were not vaccinated developing TB. The cessation of mass vaccination is additionally beneficial due to the restoration of the utility of the TST as a screening tool. Implications of Withdrawal: 25.6 cases could have been prevented through continuation of mass vaccination in the region where it was stopped.

[12], 2015

Fr

CS

French Children

Mass vaccination until 2007

Vaccination in high risk groups

Children in Ile-de-France (all vaccinated, as all considered high risk)

Children in other regions of France where mass vaccination was ceased

  

Regions where only at-risk children are vaccinated: 2000–2005:0.30, 2006–2011: 0.47

Ile-de-France (all considered at risk, all vaccinated): 2000–2005: 1.14, 2006–2011: 0.29 per million

  

There was no significant difference in incidence of TB meningitis before and after the changes in policy, suggesting that targeted screening is sufficient for TB control, however, given the possibility of incomplete coverage of high-risk children after suspension of mass vaccination, surveillance efforts should continue.

[49], 2011

Fr

QE

General French population

Universal BCG vaccination (47%)

Lower coverage due to de-commercialisation of vaccine

2008 coverage (lower)

2006 vaccination coverage (higher)

  

82 cases (in children under 3)

105 cases (in children under 3)

  

Implications of withdrawal: Increase in number of cases reported in 2008 (low vaccine coverage) compared to 2006, where vaccine coverage was higher - although it is unclear whether this change indicates an actual increase in transmission, or rather is the result of improved surveillance

[50], 1994

Fr

RC

General French population

Routine vaccination at birth (80% in children under 5 in 1990)

     

70 cases of TB meningitis in total population (1.2 per million inhabitants (95% CI: 0.9–1.5), 6 in under 5 age group

87.5% (30–98)

 

Routine BCG vaccination continues to play an important role in TB control in France, and should be maintained unless the risk of infection continuous to reduce.

[42], 1997

Ire

QE

Children (15 or younger) in Ireland

Dependent on area (see int/cont description) (38% in 1986 to 35% in 1991)

 

Areas with neonatal BCG vaccination

Areas without neonatal BCG vaccination

children < 15, 1986 and 91: 751813, 680,269

269,350, 282,489

1986: 14.1/100000, 1991: 13.4/100000

1986: 4.9/100000, 1991: 3.4/100000

In areas with vs. without neonatal BCG vaccination: IRR = 1.92 (95%

CI = 1.47–2.40)

p = 1.5 ×

10–5

in 1986; and IRR = 2.12 (95% CI 1.75–2.58),

p = 1.0 × 10–7 in 1991

Neonates

TB incidence in areas without a policy of neonatal BCG was significantly higher than areas where neonates were vaccinated.

As the number of vaccinations needed to prevent one case decreased over time (1986: 646, 1991: 551), this suggests that continuing the policy of neonatal BCG vaccination is beneficial. Also, based on the WHO’s recommendations that routine vaccination of newborns should not be continued until:

• The average incidence is < 5 cases per 100,000 per year in three successive years, −

• The annual risk of tuberculosis infection is ≤0.1%,

• A 10% decrease in tuberculosis annually for a period of 10 years is observed

• Or the incidence of childhood tuberculosis meningitis is < 1 case per 10 million per year

which was not yet the case in Ireland at the time of the study.

[51], 2009

LIC

M

General population in LIC

         

A common reason for the discontinuation of mass vaccination programs has been the trade-off between vaccinating and being able to effectively detect LTBI, with most low-incidence countries having suspended mass vaccination and focusing instead on the accurate identification and subsequent prophylaxis of LTBI, however, according to this model, low-incidence countries do not detect and treat LTBI at rates high enough to benefit from the cessation of mass vaccination programs. (Note that the model assumed that LTBI in previously vaccinated individuals is completely undetectable, regardless of time since vaccination, and thus represents a conservative estimate of the benefit of mass vaccination). Based on US life expectancy, incidence and cure rate data, mass vaccination begins to become beneficial at a vaccine efficacy of 50%. (Given that many of the studies included in the current review report higher efficacies in their populations, this modeling study suggests that mass vaccination continues to be beneficial in a low-incidence country context).

[52], 2008

LIIC

M

Children in LIIC

         

Universal vaccination was found to be beneficial in settings with a prevalence of approx. 30 sputum smear positive cases per 100,000. If prevalence is below 5 per 100,000, a universal vaccination strategy may not be beneficial due to high incidence of adverse events per case prevented. Note that the model assumes 100% BCG coverage and a vaccine efficacy of 80% against both meningitis and miliaryTB.

[34], 2009

Nor

QE

Native-born (non-immigrant) population < 30 years old

Vaccinates all 12–14 year olds.

NA

Vaccination policies in Sweden, Finland and Denmark

Norwegian vaccination policy

Py:

Nor: 16, 567, 951

Py:

Swe: 29, 851, 794

Den: 18, 042, 696

Fin: 18, 353, 224

Per100 000 py: (95%CI)

Nor:

0.45 (0.36–0.57)

Swe: 0.56 (0.48–0.65)

Den

1.41 (1.24–1.59)

Fin: 0.65 (0.54–0.78)

IRR:

Fin vs. Swe: 0.45 (95%CI 0.22–0.94)

61–64 (Norwegian 15–29-year-olds)

67–71 (after adjustment for coverage)

Average annual # of prevented cases 1.9–2.2.

Children with origins in high-incidence countries or travelling to high-incidence countries

Protective effect of BCG vaccination of newborns in Finland, and adolescent vaccination in Norway among persons at low risk of TB, however, a high number of vaccinations (21699–25,125) were needed to prevent one TB case among the low-risk population. The Norwegian Ministry of Health therefore decided in 2009 to discontinue vaccination among low-risk groups but continue it among high-risk groups.

[43], 2006

Swe

RC

Swedish population

Routine vaccination of newborns until 1975 (discontinued due to high incidence of bcg osteitis (more than 95% before 1975)

Targeted, in high-risk groups < 2% in 1976–1980, and 16% post 1980. (88% of targeted risk groups))

Born before 1974 (routine vaccination)

Born after 1974 (targeted vaccination)

  

For the birth cohort born in 1975 and followed up to 2004: 0.5/100,000py (227 cases) (6.4 per 100,000 in 2005)

85% in 1969–74

High risk infants include those with

- Family history of TB,

- Close contact of TB case- Origin from endemic countries (whether born in Sweden or not), − Planned travel to endemic countries

An increased incidence of TB in the mostly non-BCG vaccinated cohorts born after 1975 was observed vs those born during routine vaccination period (1969 to 1974).

In addition, the recommended age at vaccination was recommended to be increased from at birth to 6 months in 1994, to avoid vaccinating newborns who may later develop immune-deficiencies.

[13], 1991

UK

RC

Population of Oxfordshire, UK

Routine vaccination of 13 year olds in schools, until 1981

Cessation of routine vaccination

    

4 cases since 1981, of children who had not been routinely vaccinated at school. All four had other risk factors apart from being unvaccinated.

 

Neonates of African, Asian, Central and South

American and Middle Eastern origin, those with a family history of tuberculosis, contacts of TB cases, children newly immigrating from or travelling to endemic countries.

Given the continued decline in cases from 1973 to 89, (94 vs. 24 cases), does not recommend re-introduction of universal vaccination, but seeing as the 4 cases all had other risk factors, recommends improvement in identification and subsequent vaccination of at-risk groups

[53], 1991

UK

CC

Asian Children living in England

(51 and 64% in cases and controls, respectively)

 

Cases - TB

No TB

111

555

  

49 (14–62)

Children originating in endemic countries (or whose parents originate in these countries)

Continued vaccination of high-risk children, e.g. those originating in endemic countries. As the incidence in this group is decreasing, this policy may be reviewed in the future.

[54], 1989

UK

M

UK general population

Universal vaccination through school programs at age 14 (75–80%)

         

Gradual increase in vaccinations needed to prevent one case: 5800 in 1994, 9300 in 1999 (as risk of infection also decreased: 1:17000 to 1:26000).

Implications of withdrawal: If universal vaccination were stopped in 1991, case notifications would increase by 80, and 50 if stopped in 1996. (These new infections would mainly be in the 15–29 year old age group and would comprise existing as well as new sources of infection in the community)

[55], 2018

USA, Can, Grn

PC

Canadian, American and Greenland Indigenous populations

Mass vaccination of infants in some study years, but discontinued in others. (Greenland, Eeyou Istchee, Nunavik, and Nunavut: at least 80%; First Nations population of Alberta: 50–60%)

    

Decrease in annual TB incidence attributed to population-based BCG vaccination: −10% (95% CI −5 to −15) (adjusted for infant mortality and crowded housing)

Indigenous populations

Population-based vaccination was significantly associated with a decrease in TB incidence. This remained true when adjustment was made for improvements in overall health and socioeconomic status over the years of the study. This therefore suggests that population-level vaccination in high-risk groups (rather than vaccination based on individual risk factors) may be beneficial in the Indigenous community context.

Implications of withdrawal: Cessation of population-based vaccination in Indigenous communities may have led to exacerbation of TB incidence in these populations (e.g. Greenland)

[8], 2004

USA

QE

Native American (and Alaskan) Indigenous population

  

BCG vaccinated

Placebo

1483

1309

66/1309 (138 / 100,000 py)

36/1483 (66/100000 py)

52 (27–69)

Indigenous populations

Vaccine efficacy persisted for 50–60 years, suggesting long-term utility of vaccination

[30], 2001

USA

M

Homeless individuals

  

Vaccination

No vaccination

2 million

NA

   

Homeless population

Vaccination resulted in a 15.4% decline in TB cases among the chronically homeless and a 21.5% decline among the transiently homeless. Vaccination of 10% of the chronically homeless population led to a 10% decrease in TB cases and 2.4% decrease in TB-associated deaths over 10 years. Targeted BCG vaccination among the homeless HIV negative population is most beneficial if treatment for LTBI is available to those who have been vaccinated

  1. Can Canada, CC Case-Control, Cont Control, CS Cross-Sectional, CTR Country, Czh Czech Republic, Den Denmark, Fin Finland, Fr France, Grn Greenland, Inc Incidence, Int Intervention, Ire Ireland, IRR Incidence Rate Ratio, LI(I)C Low (and Intermediate) Incidence Countries, M Modelling Study, Nor Norway, PC Prospective Cohort, Pop Population, Py Person years, QE Quasi-Experimental, RC Retrospective Cohort, ST Study Type, Swe Sweden, UK United Kingdom, USA United States of America
  2. a At the time of the study
  3. b Or case group in case-control studies