An alternative, much less expensive approach to assess the protective effect of BCG is the use of a retrospective study design.  The best known example is the case-control study.  Here one starts with the outcome and looks at the exposure.

Cases of tuberculosis (the outcome) are assembled and a group that is comparable in every respect but the outcome (they must be free of disease) is chosen as a control group.  The prevalence of BCG vaccination is then assessed and compared in the two groups.  Rates cannot be calculated as this type of study is not population-based.  The measure here is the odds ratio.  It approximates the relative risk for rare diseases.  The protective effectiveness is similarly calculated as 1 minus the odds ratio, usually expressed as a percentage.

Case-control studies are more likely to be biased than prospective studies as the selection of appropriate controls is often hampered with difficulties that cannot always be easily overcome.

To top