Paediatric tuberculosis during universal and selective Bacillus Calmette-Guérin vaccination policy: a nationwide population-based retrospective study, Finland, 1995-2015

Retrospective cohort study exploring the impact of universal Tuberculosis (TB) vaccination cessation on the epidemiology of paediatric TB in Ireland, 2011-2021

BACKGROUND

Since 2010, Ireland's Tuberculosis (TB) crude incidence rate (CIR) remains below 10 per 100,000 population defining it as a low TB incidence country. Ireland maintained a universal BCG vaccination programme until its discontinuation in 2015 due to lack of vaccine supply. This study explores the impact of discontinuing a national universal BCG vaccination programme on the epidemiology of paediatric TB cases.

METHODS

We retrospectively analysed TB notifications aged 0-6 years old reported to the Irish National TB Surveillance System between 2011 and 2021. Key epidemiological characteristics and temporal trends in TB age specific incidence rates (ASIRs) were compared between 0 and 6 year old cases born during a period of universal BCG vaccination (2007-2015) and 0-6 year old cases born after BCG vaccination ceased (2015-2021).

RESULTS

No significant temporal trend was detected in the overall 0-6 year old ASIR by notification year during 2011-2021 (IRR:0.95; 0.86-1.1). However, the temporal trend for cases born during universal vaccination showed a significant decline (0.74; CIR: 0.62-0.89) while cases born after BCG vaccination ceased had a non-significant increase (1.2; CIR: 0.73-1.86). A significantly declining temporal trend was detected among cases born in Ireland during universal vaccination (IRR:0.73; 0.62-0.86), but no significant trend was detected in the cases born outside Ireland during universal vaccination (IRR:0.83; 0.53-1.31). No significant trend was detected in cases born after vaccination ceased in either cases born in Ireland (IRR:1.0; 0.60-1.65) or those born outside Ireland (IRR:0.64; 0.29-1.40).

CONCLUSIONS

Universal BCG cessation has not yet directly impacted on TB cases among 0-6 year olds in Ireland. However, interruption of the previously declining temporal trend in this cohort during universal vaccination may be an early warning of a future increase. Paediatric TB cases remain an important cohort for timely surveillance to monitor trends in this primarily unvaccinated cohort to evaluate the long-term effects.

Lower incidence of hospital-treated infections in infants under 3 months of age vaccinated with BCG

BACKGROUND

Live vaccines potentially have non-specific effects that protect against other infections than those the vaccines are targeted against. The national vaccination program (NVP) in Finland was changed on September 1st, 2006: before BCG vaccine was given to all newborn babies and afterwards to babies in risk groups only. We used this natural experiment to study the non-specific effects of BCG in the frame of NVP using before-after design.

METHODS

We compared the incidence of several outcomes obtained from Finnish health registers between children born between July 1st, 2004, and June 30th, 2006 (BCG-eligible) and an age- and season-matched reference cohort born between July 1st, 2007, and June 30th, 2009 (BCG-non-eligible) using Poisson regression. These cohorts were restricted to full-term children whose parents were born in Finland. Follow-up began at birth and lasted 3 months, which is the scheduled age for DTaP-IPV-Hib vaccination, and from 4 months until first birthday. The outcomes included all infections, pneumonia and injuries as a negative control outcome.

RESULTS

The incidence rate ratio (IRR) of the BCG-eligible cohort (N = 93,658) compared to BCG-non-eligible cohort (N = 94,712) for hospital-diagnosed infections was 0.89 (95 %Cl 0.86-0.93) for the 3-month follow-up. The decrease was mainly caused by respiratory infections. In 4-12 months follow-up the BCG-eligible had slightly more infections than BCG-non-eligible children (IRR 1.03, 1.01-1.06).

CONCLUSIONS

BCG vaccination was associated with a lower incidence of all hospital-diagnosed infections during the first three months of life. The difference cannot be attributed to lung tuberculosis, since only few paediatric cases occurred in Finland during 2000s. The disappearance of non-specific effect after administration of an inactivated vaccine is compatible with previous studies.