Safety of intradermal Danish- vs. transcutaneous Tokyo-strain BCG vaccination in Korean children: A nationwide cohort study.

BACKGROUND

Whether differences in BCG strain and administration (intradermal Danish-1331 vs transcutaneous Tokyo-172) translate into meaningful variation in safety remains uncertain. South Korea's dual BCG policy enables a head-to-head comparison.

METHODS

We conducted a nationwide retrospective cohort of all children born 2014-2018 who received BCG in infancy (N = 1,217,695; Danish/intradermal n = 457,063; Tokyo/transcutaneous n = 760,632). National immunization, tuberculosis surveillance, claims registries were deterministically linked. Children were followed from vaccination until outcome or 31 December 2023. Incidence rates (IRs) were calculated, and Cox models adjusted for birth year, sex, socio-economic status, and region estimated adjusted hazard ratios (aHRs).

RESULTS

The study cohort comprised 1,217,690 infants born between 2014 and 2018 who received BCG vaccination in infancy. Of these, 457,063 (37.5 %) received intradermal BCG (Danish 1331 strain) and 760,632 (62.5 %) received transcutaneous BCG (Tokyo 172 strain). Lymphadenitis occurred more often after intradermal BCG (42,385/457,063; 9.27 %) than after transcutaneous BCG (67,466/760,632; 8.87 %) (p < 0.001), with an adjusted hazard ratio of 1.00 (95 % CI, 0.99-1.02). Osteitis was rare in both groups (142 vs. 209 cases; ∼0.03 % each; aHR = 1.01; 95 % CI, 0.80-1.26).

CONCLUSIONS

In this large real-world cohort, intradermal BCG (Danish-1331) and transcutaneous BCG (Tokyo-172) provided equivalent protection against pediatric tuberculosis. Intradermal vaccination was associated with only a very small increase in lymphadenitis, while serious adverse events were uncommon with both methods. These findings support programmatic flexibility in BCG vaccination policy: both methods are safe and effective, so selection can be guided by vaccine supply, delivery logistics, and local preferences rather than expected differences in outcomes.