Table 1.
Summary of epidemiological studies investigating BCG NSEs.
| Country (study period) | Sample size | Subject follow-up period | % Reduction in all-cause mortality | Risk of bias | References |
|---|---|---|---|---|---|
| Canada (1933–1945) |
609 | 6–14 years | 6% (−32; 33) | Moderate | (6) |
| USA (1935) |
3008 | 2 years | 9% (−99; 59) | Moderate | (7) |
| USA (1941–1960) |
451 | Up to 13 years | 58% (−35; 87) | Moderate | (8) |
| Benin (1983–1987) |
294 | 4-36 months | 32% (−23; 62) | High | (9) |
| Guinea-Bissau (1984–1996) |
1657 + 695 + 4418 |
6-8 months | 37% (−33; 70) 95% (54; 99) 44% (16; 63) |
High | (10, 11) |
| India (1987–1989) |
3072 | 12 months | 40% (−97; 82) | High | (12) |
| Papua New Guinea (1989–1994) |
3937 | 1-6 months | 83% (66; 91) | High | (13) |
| Malawi (1995–1997) |
751 | 8 months | 55% (−23; 84) | High | (14) |
| Senegal (1996–1999) |
4421 | 2 years | 2% (−90; 50) | High | (15, 16) |
| India (1998–2002) |
10274 | 6 months | 56% (34; 71) | High | (17) |
| Guinea-Bissau (early: 2002–2004) |
105 | 1 month | 72% (−37; 94) | Low | (18, 19) |
| (main: 2004–2008) | 2343 | 1 month | 45% (11; 66) |
The studies systematically reviewed by the SAGE were determined to be inconclusive, as they did not link the reduction in all-cause mortality with mechanisms of NSEs. Studies assessed as having a very high risk of bias are not included. The percentage of reduction in all-cause mortality consists in a 95% Confidence Interval (CI) further specified by (1-risk ratio). The classification of the risk of bias is as reviewed by the SAGE Working Group.