Ethical Issues in Research Design

In the study of the effect of letrozole on the recurrence of breast cancer, the placebo-controlled study was aborted at the point when clear statistical evidence had been accumulated demonstrating efficacy of the medication. In the words of Andrew von Eschenbach, Director of the National Cancer Institute, "because of the observation of benefit that was occurring with the drug, the study was stopped prematurely, and the results were made available." More specifically, he said, "as soon as there was statistically significant evidence of benefit, the study was stopped, so that the benefit could be made available to all the women in the study..." (Lehrer News Hour, October 9, 2003)

This is a remarkable statement in some ways. It recognizes the ethical morass we are in with placebo-controlled studies where significant mortality is involved. Twice as many women were seeing recurrences of cancer in the placebo group, and once firm knowledge was gained that a better approach has passed the requisite scrutiny, one could not ethically stand by and allow the placebo group to incur additional risk.

How might this apply to Neurofeedback? We know that significant risk attaches to brain surgeries for seizure disorder. One might therefore call for a placebo-controlled study of Neurofeedback, to establish a lower-risk, non-invasive alternative. But that does not pass ethical muster because a viable treatment, namely surgery, already exists. To hold off on surgeries would be to expose patients to additional risks from the seizures.

Perhaps instead one ought to do a comparison study of Neurofeedback and surgeries. By the above standard, once a statistical superiority of Neurofeedback has been established, the study would have to be stopped, and everyone would have to be given Neurofeedback prior to authorizing surgery.

Following that logic, the study could be stopped before it is even organized. Because it is already true that Neurofeedback has less risk, since there is essentially no risk of loss of function, and no risk from general anesthesia, and essentially no risk of seizures actually getting worse with the procedure. And the potential benefit is huge, with at least 80% of medically refractory epileptics responding significantly to Neurofeedback (greater than 30% reduction in seizure incidence.)

In one of the early NIH-funded controlled studies of Neurofeedback for seizures, subjects were drawn from a wait list for brain surgery, according to the Principal Investigator Barry Sterman (of the Sepulveda VA and the UCLA School of Medicine). When it was realized that some subjects no longer needed surgery after they experienced the benefit of Neurofeedback, the referrals to the research program were terminated. The study showed 60% average improvement in seizure incidence among 24 subjects, with a number of them becoming seizure-free.

A second ethical quagmire concerns Parkinsonism. The benefits of Neurofeedback are generally comparable to those of Deep Brain Stimulation. Again, a study comparing the two would for ethical reasons have to be stopped before it ever got started, because of the demonstrably lower risk that attaches to Neurofeedback by virtue of being non-invasive.

A third ethical conundrum is presented by Electro-Convulsive Shock Therapy, which entails the risk of general anesthesia, and is accompanied by significant side effects. Again, one might be tempted to propose a study comparing Neurofeedback to ECT, but the study would have to be stopped before it started because it is already apparent that Neurofeedback would involve considerably less risk to the person.

A fourth issue concerns hormone replacement therapy in women. The hazards of HRT have become so clear that women have been advised to consider abandoning the therapy, or at least to reduce the dosage. Neurofeedback is a non-invasive alternative for the management of post-menopausal symptoms, one that is not associated with any of the risks that attach to HRT. Again, if a comparison study were to be proposed, it would have to be stopped before it started, because the risk side of the ledger is so clearly favorable to Neurofeedback.

A fifth issue is the medications given to children for ADD. It is unfortunately not generally known that there have been deaths of children due to Ritalin and to Clonidine. This gave rise to great consternation because there seemed to be no way to predict these deaths. Children were dying who were appropriately medicated, and there was no known contra-indication. If there were any doubt about the cause of death in one Ritalin case, it was resolved on autopsy when the presiding doc announced incredulously that they boy had a "cocaine heart," a heart such as he had only seen in advanced cocaine addicts. Yet it was certain that this child could not have had such a drug history. He had been on Ritalin for only a few months. For this child, Ritalin had acted on the heart like cocaine. Some four children died within a short space of time.

When these news first hit, we were visited by Professor Dennis Cantwell of the UCLA Child Psychiatry Department. He was interested in investigating non-drug alternatives to Ritalin, and he had heard about our work with neurofeedback. He said, "ADHD is a terrible thing, but you don't die from it, and you should not be at risk of dying from the remedy." Dennis Cantwell died shortly thereafter, and our nascent collaboration never bore fruit.

These children's deaths have disappeared from the news, and there is no very visible effort to respond to this particular risk in our society. Ritalin is deemed to be a "good" drug, so adverse effects, including deaths, are effectively swept from view. There is no point in dwelling on the down-sides, particularly since these are fortunately very rare. In this case, studies comparing Ritalin and Neurofeedback have been done. And Neurofeedback matches Ritalin in all respects. Additionally, it has been shown in a large-scale Canadian study that more than 80% of children no longer benefited from Ritalin after they did Neurofeedback.

And another recent study in which Neurofeedback was combined with Ritalin, it was found that when the Ritalin was stopped after a year, only those who had received the Neurofeedback maintained their good function. If we now look at the risk side of the ledger, knowing that children occasionally die from Ritalin and Clonidine should be enough to shift the balance in favor of Neurofeedback.

The FDA was pressured to remove the drug Rezulin from the market a few years ago because of drug-induced liver failure. The agency was very reluctant to do so for a long time. To deal with the mounting evidence of risk, they kept recommending more frequent medical examinations. But those examinations were never able to identify the people at risk of dropping dead with sudden liver failure. The FDA was ultimately embarrassed by the number of deaths into removing the drug from the market. Their reluctance to do so was driven by the fact that they saw no good alternatives to the drug. But again there is Neurofeedback. Type II diabetics can be significantly helped with Neurofeedback in terms of their quality of life. It can be helpful with peripheral neuropathy pain, for example. At this point, it is not even necessary to argue that Neurofeedback is as worthy of consideration as Rezulin, because the latter is no longer an option. One could not even do a comparison study at this point.

We have argued in the above that an assessment of the issue of risk in comparison trials would strongly favor Neurofeedback. On the benefit side of the ledger, which we have hardly discussed, it also needs to be mentioned that Neurofeedback yields benefits that are not even offered by the medications. Thus in the case of ADHD, Neurofeedback helps with oppositionality and Conduct Disorder, for which no medication has been shown effective. In cases of seizure disorder Neurofeedback offers more than just seizure control. It is likely to improve cognitive function if that is impacted, and the quality of sleep, and of emotional regulation. In cases of depression that are slated for shock therapy, Neurofeedback offers a more comprehensive remedy involving improved sleep, mood regulation, energy level, and general brain stability. In the case of Parkinson's, the level of function of almost any individual can be enhanced with Neurofeedback, irrespective of medication status. It is therefore not even an issue of Neurofeedback versus medication. One needs to do both. In the case of diabetes, benefit for peripheral neuropathy pain has already been mentioned. There is no known medication that offers comparable benefit.

Neurofeedback appeals to what works in the brain. It reinforces and strengthens the brain's own capacities to function. It should be our first resort, not our last. It often complements medical management. And, above all, it involves negligible risk. That all by itself should lower the bar for consideration by both patients and their caregivers.

Siegfried Othmer