askthedoctor “Scientific” Tests T

Take a Second Look Before believing a claim based on a “scientific study,” think about how and why the study was done. If the study was done to help sell the product, be skeptical.

56 MILITARY OFFICER JANUARY 2015

Many products are advertised as “scientifi cally tested.” But as Rear Adm. Joyce Johnson, D.O., explains, not every study is objective, nor should you believe the results.

There are stringent regulatory re- quirements for the testing and approval of pharmaceuticals and some other products, but health claims for items such as athletic gear, exercise equipment, hair products, etcetera don’t require such controlled data. “Scientifi cally tested” implies some sort of study was performed, but it does not mean a health claim is true; it’s possible to design a study specifi cally to get the desired result. There are several ways to infl uence the

results of a study. Questions in a multiple- choice survey can be worded to skew an answer. A survey might require a person’s name and other identifying information (often to get a payment for completing the survey) and make respondents hesitate to provide negative answers. If people testing a product are briefed about its advantages before using it, the power of suggestion in- fl uences their responses. Some studies don’t have enough participants for statistical va- lidity, or the participants selected are those most likely to give the desired response — such as employees of the manufacturer. What factors make a study more valid?

First, the study population should be ap- propriate (e.g., to test whether a vaccine will work against a certain disease, the study population should be people who are susceptible and exposed to the dis- ease). Second, study participants should be selected randomly from that study pop- ulation. Third, the study design and how it is followed are critical, as they can add bias and make the results invalid.

A placebo-controlled study includes

at least two groups of participants, one receiving the test product and the other (the control group) receiving a placebo. The placebo should be similar to the test product, so study participants are unable to determine which they are receiving. A placebo could be a sugar pill, an injection of salt water, or some other harmless but fake treatment. Sometimes the “placebo” is the standard treatment for a disease. For example, it would be unethical to give a diabetic patient an injection of salt water while testing a new type of insulin. In- stead, the control group would receive the standard insulin treatment. A double-blind study is one in which neither the medical team (investigators) nor the patients (subjects) knows who is getting the test product or a placebo. An independent group collects the data and monitors results. This design eliminates much testing bias. Occasionally, a study population has extremely diff erent results from the control group. A study might end early if the study product is found to be dangerous or (rarely) if the study product is found to be signifi cantly better than the control product and it is determined to be unethical to deprive the control population and others of the study product.

MO

— Rear Adm. Joyce Johnson, USPHS (Ret), D.O., M.A., is a health care consultant in Chevy Chase, Md. Find more health and wellness resources at www.moaa.org/wellness. For sub- mission information, see page 6.

PHOTO: STEVE BARRETT