ISNR Conference Review, Part II

by Siegfried Othmer | September 22nd, 2005

Frank Deits and his wife Mary happened to be driving through the Denver area at the time of the conference, so they came to the exhibit hall and made their rounds. Frank has a new pacemaker installed, and thereby hangs a tale. Fortunately I have not yet had reason to become acquainted with the jargon of the field of cardiology, so my description must remain superficial. Frank had reason to believe that the demand-driven pacemaker was at times putting his heart into tachycardia, so he wired up some monitors and watched the heart signal along with the pacemaker signal. He convinced himself that there was a problem, so he took all his data to the electrical engineer associated with the project. “There would have been no point in taking it to my cardiologist,” said Frank.

Together they sat down and monitored his heart function at different drive levels of the pacemaker. The drive voltage had been set at five volts. (It is sometimes set as large as nine.) On experimenting it was found that a mere 1.5 volts was quite adequate to do the job. Throw in fifty percent on top for margin, and all is well.

This illustrates what is surely a quite general trend within medicine, and that is the increasing technical “intensity” of medical procedures. No ordinary mortal could hope to stay on top of it all. What will it come to mean to be a medical expert in such an environment? The doc on the front lines will come to serve a kind of integrating role, weighing different pieces of information to form a coherent understanding of the issues. The doc himself may be an expert on very few of the underlying procedures.

So how likely is it that the MD will in time feel the need to develop a personal competence in the field of neurofeedback? Outside of a psychiatry specialty, that is probably not going to happen except in isolated cases. That means there will be a need for a large group of clinicians trained in neurofeedback to serve in medical clinics and other such settings. These people will have no need for the qualification of licensure for independent practice.

I had a long conversation with Daniel Hoffman, MD, who feels that the key to the psychiatry community in general may lie in a two-step process, the first of which would be their discovery that QEEG data can readily be used to determine response to medication. This is through the work of Suffin and Emory, who have chosen not to publish their voluminous research on this topic over the years for the sake of a proprietary scheme that is now commercially available. So, the essential discovery has been made, but the psychiatry community has yet to discover that.

One of the most striking applications of the technique is in confirming when someone has been inappropriately medicated, and no one sees a way forward. One such case was a woman who had gradually accumulated prescriptions to the point where she was taking 22 different medications. Her problems, meanwhile, were getting worse. For purposes of the test, she had to be taken off those medications, and the test confirmed that she was not a candidate. The medications had in fact been compounding the problems. Another such case was a teenager on Adderall whose behavior was still unmanageable, with episodic explosive behavior, etc. The test showed that he was inappropriate for Adderall. When the mom came in for the debriefing, she already had good news. Her son’s behavior had turned around when the Adderall was interrupted for the test, and she never gave him any more. He was fine.

Similar results are being obtained at the Amen Clinic with their categorizations by means of SPECT scans. But of course the QEEG data will be seen as having more direct relevance to neurofeedback.

Once psychiatrists are oriented toward the QEEG at least to the point of recognizing its value, then the step toward the acceptance of neurofeedback will be much smaller. This recognition of the value of QEEG in pharmacology can happen well before a significant fraction of psychiatrists actually adopts the technique.

Svetlana Serova, of the University of North Texas, presented a Student Paper on the “Effects of Mood States on Objective Measures of Attention.” This nicely documents something we have observed consistently over the years. In fact, a number of years ago David Kaiser compiled TOVA data relating to primary diagnoses of depression, anxiety, and sleep disregulation and compared them to our data on ADHD. The aggregated TOVA profiles for these different clinical categories were indistinguishable. Ironically, the attempt by the neurofeedback deniers of the AABT (Lohr & Co) to indict the TOVA by saying that it does not offer good reliability in the diagnosis of ADHD is off the mark.

The CPTs cannot offer differential sensitivity simply because the ways in which attention can fail is largely common to ADHD, depression, and anxiety, etc. That is not an intrinsic deficit of CPTs, just a case of expecting too much from such a test. In consequence, we find the CPTs useful across the board in measuring progress in training. This also makes sense because the CPTs take us much closer to the actual target of training than the diagnostic categories. Any neurofeedback that results in improvement of CPT scores is likely to also impinge favorably upon symptoms of anxiety, depression, and disregulation of sleep. In all of these cases, the target of training is disregulation in the realm of cerebral timing and frequency. The result is an enhanced capacity in the control and maintenance of states.

Sue heard Leslie Sherlin say something to the effect that “If we believe that the problems we are concerned with are reflected in EEG anomalies, and if we believe that targeting those anomalies remediates the conditions….” Let’s hold it right there. The problem is that often the most significant assumptions in our work are the unexamined ones. The above is represented as being essentially beyond question. And everything that follows from those assumptions is then also sold as “science.” One unexamined assumption begets another. I recall the honest appraisal by David Joffe at a prior ISNR meeting at the very same hotel in which he allowed that “we knew as far back as 1995 that our assumptions were not being confirmed in the real world.”

The argument of a correlation between EEG data and clinical efficacy of targeted neurofeedback is strongest when we look at coherence anomalies. Here the rapidity of response is such that one gets the very strong impression that training a specific anomaly gets a rather specific result. But even here counter-examples exist. The inter-hemispheric training we are now doing is coming very close to being a kind of universal protocol, applicable to anyone who walks in. And the rate of clinical improvement is impressive. We expect to see discernible consequences of the neurofeedback training on a timescale of minutes, permitting refinement of the training parameters so that within a session or two or three everyone has their training optimized. This training is best regarded as reinforcement of desynchronization of the EEG, under conditions where this appears to be most needed. So a coherence model applies here as well.

Ironically, then, the movement from amplitude and power-based training in the QEEG formalism, along with traditional up-training in the mechanisms perspective, has given rise to another comparable pairing in the coherence training regime which again complements the specific, targeted, QEEG-based training with a generic, dynamic desynchronization approach at standard sites. Which is better? Matters are unlikely to divide that way. For some conditions such as TBI and specific learning disabilities, the targeted training has an advantage. For general conditions of disregulation the generic training no doubt holds an edge.

Another article of faith is illustrated by the following quotation from Vince Monastra from one of his recent publications: “Although significant clinical improvement was reported in approximately 75% of the patients in each of the published research studies, additional randomized, controlled group studies are needed in order to provide a better estimate of the percentage of patients with ADHD who will demonstrate such gains in clinical practice.”

Oh, really? Under normal circumstances, the findings in research leave unanswered the question of how things will fare in the actual clinical world, where patients don’t come in with one diagnosis or one symptom, aren’t paid to show up, etc. In our field, matters have gone entirely the other way. For some time now, clinical findings have been ahead of the research in terms of both protocols and efficacy. It is amazing to see the degree to which this reality is being ignored by those who are discomfited by it.

The myth is maintained by the presumption that the best protocols are those that are the most extensively reported in the literature. But that is a self-reinforcing reality that started out with a very stacked deck. How dare they call this science? If we were getting anything like 75% success right now, our clinicians would either be quitting in despair or they would be ushered out the door. In fact, I believe that we have “zero-defect” neurofeedback within our sights. That is to say, we are getting close to the point at which every person doing neurofeedback for the usual clinical conditions will be better off for the experience. In their own or their family’s judgment, the exercise will have been worthwhile. One clinician who years ago tracked outcome against expectations going in found that the goals of training were met or exceeded 98% of the time. That is not difficult to do. Significantly, those expectations were formulated after initial discussions with the clinician about what might be possible with neurofeedback. So an alternative perspective on the 98% figure is that the clinician cannot have been over-promising.

Leave a Reply