Testing Cold Treatments
New cold treatment ideas require testing in clinical trials to determine if they actually work. Testing is necessary because of four natural circumstances effecting the nature of colds:
- Colds are a self-limited illness from which virtually all people recover in a few days. (1) If colds were universally fatal, no formal testing of new treatment ideas would be necessary to determine if they work.
- The duration and severity of a cold varies from one person to another and also varies in the same individual over time. (2) If all colds were exactly alike in all people at all times, it would also not be necessary to perform clinical trials on new treatment ideas.
- Colds respond to placebo (sugar pills). (69) People given totally inactive substances such as small amounts of starch as a "treatment" will have significantly milder illnesses than those who are given no "treatment." Interesting examples are available of how curious subjects in clinical trials have sometimes been able to guess correctly about whether they were on the experimental drug or placebo and thus unblinded themselves. (70)
- The benefit of most effective treatments for most self-limited diseases are not so dramatic that it is always obvious that the treatment is effective in an individual case.
For these reasons, an individual taking a cold treatment based on the new idea may be mislead into believing the experimental treatment was effective because the cold being treated was naturally mild or because the person experienced a placebo reaction or both. Conversely, a person trying a new treatment idea might be misled into believing it did not work when actually it did provide some benefit, but the benefit was obscured by the patient having a particularly severe cold.
To accurately test the value of new cold treatment ideas requires conducting experiments in which these chance events and biases are adequately controlled. (71) Chance and bias can never be totally eliminated from a testing situation. The experimental design which most effectively reduces the influence of chance and bias is that of a properly "powered", "controlled", "randomized", "double blinded" clinical trial.
"Controlled" means that the treatment idea being tested is compared to either placebo (inactive substance, sugar pill) or to another treatment which has previously been shown to be effective. (71) Comparison to a placebo determines if the new treatment has any benefit at all. Comparison to an oldtreatment determines if the new treatment is better than the old one.
It is important that a clinical trial have enough subjects enrolled to give accurate results. This is called having adequate "statistical power". (71) It would not make sense to have only one or two people in each group of a clinical trial because people naturally vary in the length and severity of their colds. It would also not make sense to have 100,000 people in each group because a study of this size would be difficult and expensive to conduct. More importantly, an accurate answer to the question could be obtained with much fewer subjects. It would be a waste of resources. Mathematical calculations can be used to select the correct number of subjects ("sample size") for a particular clinical trial, depending on its purpose.
"Double blinded" ("masked") means that neither the investigator conducting the clinical trial nor the subjects in the trial know who is getting the experimental treatment and who is getting the placebo (or the older treatment). (71) Masking controls for the bias associated with the placebo effect and for any biases on the part of the investigator or subjects, who naturally would like to see a new idea for treatment succeed. Being biased does not mean that a person is dishonest. However, masking also helps control for any dishonesty on the part of the investigator, if such is present.
"Randomized" means that subjects are assigned to either the group receiving the experimental treatment or the control group (receiving placebo or comparator drug) in a random fashion, such as by flipping a coin. (71) Randomization also helps control for a bias such as might occur if most of the sicker cases were assigned to one group. If this happened, the test material given to the sicker group would not look as effective as that given to the group with milder illness.
What to Look for in Evaluating New Cold Treatment Ideas
- Do the investigators have a reputation for honesty and competency? This may be difficult or impossible to determine.
- Was the new treatment idea tested in a clinical trial?
- Were the results of the trial published in a "peer reviewed" medical journal? "Peer review" means that the paper was evaluated and approved by independent experts before it was accepted for publication.
- Was the clinical trial controlled? (71) Was it designed so that the new treatment was compared to placebo or an established form of treatment?
- Were "sample size" calculations done to determine that the study had adequate statistical power? Does the paper state what specific endpoints were being evaluated and why the number of subjects assigned to the treatment and control group was selected?
- Did the clinical trial have a double blinded design? Were both the investigators and the subjects unaware of what was being given to the subjects until after the study was completed?
- Was there proof that the new drug preparation (tablet, capsule, liquid, etc.) was indistinguishable from the placebo or comparator drug preparation? Was actual testing done to determine that people could not distinguish the new drug preparation from the placebo or comparator drug before the clinical trial was conducted? Were the preparations compared for such things as appearance, smell, taste, after-taste, and any other characteristics that might help subjects to guess whether they were receiving the experimental treatment or the control preparation?
- Were subjects randomly assigned to the treatment and the control group? This is usually done by a table of random numbers or a computer-generated list of random numbers.
- Have the investigators carefully looked for and reported the side effects of the experimental treatment?
If the evaluation of a new cold treatment idea meets these nine criteria, you can feel confident that it has been properly tested and that the results of the study are probably correct ("valid"). It is even better if other investigators had done clinical trials with the same experimental treatment and have obtained similar results. In general, the Food and Drug Administration requires three well done clinical trial before recognizing a new treatment as being effective.