Scientific research into alternative treatments and new drugs are commonly “doubled up” by the combination of looking at two alternatives simultaneously, and by “blinding” both patients and providers as to which patients are getting which of the alternatives.

This serves to minimize bias due to the “placebo effect”. If patients and providers knew which patients were getting the real drug, both might be moved to perceive effects that are not there, or that are due to the “mind-body connection”, rather than the effects of the drug, per se. With “controls” getting placebos, the effect of “pure belief” in the efficacy of the drug would be uniform for all cases, and can be subtracted from the total effect of treatment to calculate the impact of the drug alone.

Comparisons of specific therapies that do not involve drugs are normally between the “usual treatment” and the specific therapy under study. These are more difficult to “double blind”, since the provider of the treatment will certainly know whether the patient is getting the usual treatment or the studied alternative. The patient may not know what the usual treatment is, of course, so that “placebo bias” may not be present, or a “sham” version of the therapy may be used as an actual placebo. It is often difficult to “fool” the patient, however, unless the “sham treatment” is perceived as the real thing.

But there are new reasons for adding a third option – “no treatment” to the usual vs. new therapy alternatives or drugs are being studied. For one thing, only the addition of “no treatment” will enable what is the equivalent of a placebo to treatment trials, since usual treatment involves a treatment rather than a placebo. And when comparing drugs, only the no-drug option will be totally devoid of the placebo effect, and provide the basis for learning whether either treatment is as cost-effective as “watchful waiting” for example. This should only be done when ethically acceptable, of course, and when patients are informed.

Moreover, since treatments often have negative side effects, only the no treatment option should be devoid of these, and permit a full and balanced comparison of the logical options. While placebo drugs should have no side effects, usual care may have some, and the net positive vs. negative effects of treatments should be compared to no treatment, where feasible and acceptable, to enable greater learning about what is the best option.

When no treatment is used, there should be no placebo effect for patients who get none. However, the total effects of both real and sham treatment should be included for comparison with no treatment, since the placebo effect is a real benefit to the patient, even if not due to a particular treatment, but to the patients’ belief in its efficacy. Subtracting the placebo effect would eliminate what may be a major portion of the actual benefits patients gain from being treated compared to not.

As we enter the era of “value-based medicine”, and “value-based” purchasing and payment thereof, it makes sense to look at the no treatment option in most cases where it is ethically permissible. Certainly consumers should know how proposed treatments differ in outcomes, including side effects, from both usual care and any new alternative, and so should providers. It is already far too common for providers to use or refer patients to the alternatives with which they are most familiar and comfortable, and perhaps the ones that generate volume and revenue for their practices, as well.

By including a third option of no treatment, and perhaps comparisons to complementary and alternative medicine (CAM) options as well, both consumers and providers, and hopefully payers as well, will have a more complete picture of the full range of options available. Since there are often multiple CAM alternatives as well as multiple options in traditional medicine, all options will probably not be included in any given trial, but multiple trials can cover all reasonable options, while comparing all to each other and the no treatment alternative to all.

For similar reasons, “triple blinding” such studies also makes sense. There has been too much evidence already that the sponsorship of a given study tends to bias the results reported by those who analyze the data. Only if those who perform the analysis are blind as to both which patients got the no treatment option, the usual care, or the alternative being studied, whether a drug or a particular therapy is involved, — and ideally as to who is sponsoring the study — will such bias likely be eliminated.

Bias may be as likely to arise from unconscious optimism or leanings on the part of analysts as from a deliberate attempt to “fudge the data”, but whether conscious or not, analyst bias should be eliminated as well as patient or provider bias. By combining triple options with triple blinding, though this may require higher expenditures for the studies affected, the reliability and validity of results would at least be improved or protected, and we would learn that much more from each study reported.

This might also promote far greater reporting of results that do not promote the interests of the study sponsors. If those who perform the study are separated from sponsors entirely, not only will bias potential be reduced, but we will all gain from the reporting of far more study results than seems to be the case today under current clinical trial methods, where sponsors control which and whether results are reported.

There is a fourth approach that would also add greatly to the public good in regard to clinical trials, of course. That would be the willingness of the media to avoid writing and publishing sensationalist and premature stories about the latest “wonder drug” or treatment. As was pointed out in two books, the first for medicine [SK Sarnoff Sanctified Snake Oil: The Effects of Junk Science on Public Policy Praeger 2001] and the second for CAM [RB Bausell Snake Oil Science: The Truth About Complementary and Alternative Medicine Oxford 2007], the “science” on drugs and treatments is often anything but.

The media tend to publish whatever medical care reports will interest the public, while either not understanding or not caring enough about the validity and reliability of the science behind it. But at least one journalist has recently taken a vow to report anything that is not based on: “… large, randomized, placebo-controlled, and double-blind clinical trials published in high-quality, peer-reviewed medical journals.” [J. Adler “A Big Dose of Skepticism” Newsweek Dec 1, 2007]

He reported being “shamed into” such a vow by reading the Bausell book that described the kind of statistical analysis that journalists should demand to see before writing their reports of findings. When compared to placebos, far too many treatments fail to offer any added benefit, and publishing stories about them, while promoting placebo effects, enables too many people to either make or pay money for useless treatments and nostrums. While this could go too far if popular media only published studies that meet scientific standards and journalists understand the science behind them, given the average statistical acumen of journalists, it could prevent a lot of “snake oil” from doing harm to and wasting money of consumers.