Is An Herbal Supplement Good for Long-Term Management of Asthma?


You are counseling an adolescent patient about the long-term management of her asthma while she is away at college. She brought an article she found on the Internet that promotes an herbal supplement purported to be a "safe, natural" treatment for asthma. You recall a recent study in the medical literature showing inhaled steroids to be statistically superior to the advertised herbal supplement at preventing asthma exacerbations at the p < 0.05 level.
You explain to her that this means which of the following?

A. The inhaled steroids are 5% better than herbal treatment

B. A critical threshold for medical significance has been reached

C. Patients will not benefit from the herbal treatment 5% of the time

D. The odds are less than 1 in 20 that the differences observed were only a chance variation

E. It would be unethical to use herbal treatment'

Answer / Explanation:

The answer is d, The odds are less than 1 in 20 that the differences observed were only a chance variation.

The probability given is an estimation of the odds that the observed differences could have occurred by chance alone. More precisely, assuming one therapy was no different than the other (relative risk = 1.0, termed the null hypothesis), the p-value is the probability of obtaining an association as strong or stronger than the one observed. Typically, the p-value most often selected in the medical literature to test for "statistical significance" is .05; the smaller the p-value, the "more significant" the result. The interpretation of these results depends on an assessment of factors, such as the study design, the size of the sample, the type of controls used, the severity of the disease, the side effects, and the importance of the treatment. The tendency for negative results to remain unpublished should also be kept in mind.