In the New England Journal of Medicine, an article reported that a new shingles vaccine was 97.2 percent effective against the virus that causes shingles. I was stunned by that number.

In the NEJM study, the authors found that out of 15,216 subjects in the treated group, 18 developed shingles. That translates to a rate of 0.12 percent.

Out of 15,426 subjects in the placebo group, 470 developed shingles; a rate of 3 percent.

How do the authors’ report a 97 percent efficacy? They use the relative risk ratio, which is a statistical manipulation to exaggerate the benefits of a mediocre therapy.

It should not be used to make a clinical decision.

A better method would be to calculate the absolute risk difference, subtracting the illnesses in the treatment group (0.12 percent) from the illnesses in the placebo group (3 percent).

This gives an absolute risk difference of 2.88 percent.

The new shingles vaccine was 2.88 percent effective at preventing shingles. In other words, the vaccine failed more than 97 percent of the time.