Recap of Today's ISNR Discussion Group - Insurance Companies Still Consider NFB Experimental
Susan Alvarez
👉 Yes—most major U.S. health insurers still classify neurofeedback as “experimental” or “investigational” in 2026, although there are some exceptions and gradual changes.
- Policies from large insurers (e.g., Blue Cross, Cigna, Aetna in some cases) commonly state that neurofeedback/EEG biofeedback is “experimental, investigational, or unproven.” [scienceinsights.org]
- Example: A Blue Cross medical policy explicitly states neurofeedback is**“experimental/investigational” due to insufficient evidence of improved outcomes.** [bcbsm.com]
- Because of that classification, claims are often denied by default.
- Medicare generally does NOT cover neurofeedback for mental health or cognitive conditions, classifying it as investigational and not “reasonable and necessary.” [biologyinsights.com]
- It typically only covers traditional biofeedback (not brain-based) for specific physical uses like muscle rehabilitation. [legalclarity.org]
- ✅ Covered more often:
- Traditional biofeedback (muscle/physiological)
- ❌ Often denied:
- Neurofeedback (EEG-based brain training), especially for
- ADHD
- anxiety
- depression
- PTSD
- Overall:
- “Most major insurers still classify neurofeedback as experimental or investigational.” [neurolaunch.com]
- Some insurers approve it case-by-case (especially with strong documentation)
- Coverage is more likely if:
- Other treatments have failed
- It’s combined with psychotherapy
- It’s billed under broader biofeedback or mental health codes
- Certain plans (or states/employers) do allow reimbursement, but not as a standard benefit [legalclarity.org]
“Experimental” doesn’t always mean “never covered”—it usually means “not routinely covered.”
- Evidence is considered inconsistent, evolving, or not definitive
- There’s no broad clinical consensus on protocols or outcomes
- It’s not yet widely accepted as a standard first-line treatment
- ✅ Yes — The health insurance industry largely still considers neurofeedback experimental/investigational in 2026
- ❗ This leads to frequent denial of coverage
- 🔄 But:
- Coverage is increasing slowly
- Some plans approve it under specific conditions
-
Appeals and documentation can sometimes succeed
Best regards,
Susan Alvarez
Executive Director, ISNR
International Society for
Neuroregulation and Research
2146 Roswell Road
Suite 108, PMB 736
Marietta, GA 30062
Office: (703) 848-1994
off...@isnr.org
Follow ISNR on Linked In:
https://www.linkedin.com/in/isnr-net-neurofeedback-and-research-90a12981
Follow ISNR on Facebook:
https://www.facebook.com/ISNRorg/?fref=ts
Learn about the latest research in ISNR's peer-reviewed, Open Access journal, NeuroRegulation: www.neuroregulation.org
Mark Trullinger
--
You received this message because you are subscribed to the Google Groups "ISNR_Members_Forum" group.
To unsubscribe from this group and stop receiving emails from it, send an email to isnr_members_fo...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/isnr_members_forum/BYAPR07MB5368F385493A6AF687632B67A80F2%40BYAPR07MB5368.namprd07.prod.outlook.com.
sarahprinsloophd
We now have randomized controlled trials, placebo‑controlled trials, dosage‑finding studies, and mechanistic trials for neurofeedback. In other words, the field has produced all of the evidence, and in some cases more, that pharmaceutical and device manufacturers must generate to achieve formal approval. Yet the pathway for neurofeedback remains unclear and disproportionately difficult.
I’ve read about the extensive work being done around CPT codes, and it’s impressive. But it also raises the question: what else can be done, and why does the bar seem so much higher for neurofeedback than for other neuromodulation approaches?
Take TMS as an example. It is currently one U.S.-based RCT, approximately 129 participants, away from being endorsed as a treatment for pain. Only a small number of European trials have demonstrated TMS efficacy for pain, yet the FDA’s requirement is simply that a single RCT be conducted domestically. That’s a remarkably streamlined path compared to what is being asked of neurofeedback.
So why the apparent leniency for TMS and the barriers for NFB? Is it because TMS already holds FDA clearance for depression, and pain is considered an extension into a new indication rather than a new technology? Or is it because, outside of research settings, TMS must be administered under the supervision of an MD, which may make regulators more comfortable with perceived oversight and risk management?
These discrepancies raise important questions about how neuromodulation technologies are evaluated, who gets to define “acceptable evidence,” and what structural factors influence regulatory momentum.
BTW, the n=129 was recommended by the FDA...
David Cantor
To view this discussion visit https://groups.google.com/d/msgid/isnr_members_forum/2f284800-39a1-4e24-8d6b-5550580146e3n%40googlegroups.com.
 |
| ||||||||||||
“Success is not the key to happiness. Happiness is the key to success. If you love what you are doing, you will be successful.”
- Albert Schweitzer
"Success has nothing to do with what you gain in life or accomplish for yourself. It's what you do for others. "
= Danny Thomas
lrob...@gmail.com
Dave, Sarah, Mark, and others:
My primary concern and reason for posting this is: as people turn more and more to AI… this may discourage the general public from considering neurofeedback because they are not thinking at the level we as professionals do… they simply read: “Policies from large insurers (e.g., Blue Cross, Cigna, Aetna in some cases) commonly state that neurofeedback/EEG biofeedback is “experimental, investigational, or unproven.”
We need to begin to address this issue for the general public.
Rob
To view this discussion visit https://groups.google.com/d/msgid/isnr_members_forum/CAA6yvVYsHrb5nB6yZmQdA2V1uzdrrUY2KL93%3Dsv9Npgr8ifu2w%40mail.gmail.com.
David Cantor
To view this discussion visit https://groups.google.com/d/msgid/isnr_members_forum/002401dceaad%2425559600%247000c200%24%40gmail.com.
Mark Trullinger
Rob and all,
Thank you for sending this and for keeping this conversation moving. I think Rob’s concern is important because the general public is not reading these issues the way we do as professionals. They are seeing a simple message: large insurers still call neurofeedback experimental, investigational, or unproven. For many people, especially as they increasingly turn to AI for quick answers, that may be enough to discourage them from considering neurofeedback at all. That is a problem we need to address directly.
In my dissertation, I looked closely at why neurofeedback has not progressed along a typical adoption trajectory for a medical device innovation in mental healthcare. That context matters. Medical device innovations in mental healthcare generally move much more slowly than medications, with an average path closer to 40 years compared to roughly 10 years for a pharmaceutical treatment. When I first studied this, I framed neurofeedback’s under-adoption in comparison to TMS, which appeared to be following a more typical trajectory and perhaps even moving slightly ahead of schedule. It is hard to believe I was collecting those data almost 10 years ago, around the same time I became involved with the CPT project and insurance reimbursement.
At the same time, I do not think the comparison to TMS is always completely fair. There are important differences in how the two fields developed, how they are regulated, and how they are used. Neurofeedback has been around for many years and has had Category I CPT codes since 1978, under the original biofeedback codes that were intended to broadly cover medicine. However, our field also has a complicated history. We have sometimes over-promised what neurofeedback can do, rapidly adopted new protocols and new indications before developing strong internal consensus (within our field before even addressing outside of our field with mainstream medicine), allowed use across a very wide range of professionals and settings, and blurred the boundary between clinical treatment and non-clinical peak performance work. There has also been a longstanding countercultural element in parts of the field that did not want to be part of mainstream medicine and, at times, actively pushed back against it.
I am not saying this describes the whole field, and I am not raising these points simply as criticisms. Some of these qualities are also part of what made neurofeedback innovative, flexible, and clinically meaningful. But all of them slow the path toward mainstream medical acceptance. They make it easier for insurers, regulators, and medical policy committees to say the field lacks standardization, lacks oversight, or lacks consensus, even when the evidence base itself is much stronger than those labels suggest.
This is where the FDA issue becomes especially important. Relatively little formal FDA work has been completed in neurofeedback, and I do not think that is because the bar is impossibly high. I think it is partly because we have not consistently prioritized fitting into the structures mainstream medicine actually uses. Too often, we have wanted to define the rules ourselves rather than ask what the FDA, CPT, insurers, and medical policy committees are actually requiring.
Sarah’s point about TMS is a good example. TMS reportedly had an FDA-recommended pathway involving one U.S.-based randomized controlled trial of approximately 129 participants for pain. That is not an impossible number. More importantly, neurofeedback already has an example within our own broader space. How many people in our field are aware that there is already FDA clearance for a neurofeedback-based device to assist in the treatment of PTSD? That clearance was based on just over 80 participants, included a 3-month follow-up, required a defined amount of improvement in at least 50% of participants, and did not require a control group. This device came from and mostly exists outside of the main frameworks of our field as we currently exist. That should be concerning. Moreover, some biofeedback devices with specific indications have also received clearance by relying on existing research, including research involving different devices. So, while a primary indication would likely require more, the starting point is not nearly as unrealistic as many in our field assume.
That is why I think the FDA issue is not simply a research problem. It is a prioritization problem. It is not that we have no evidence. As Sarah stated well, the field has randomized controlled trials, placebo-controlled trials, dosage-finding studies, mechanistic studies, and decades of clinical work. The problem is that the evidence has not always been organized, framed, and operationalized in the way mainstream medicine expects. We have not created a clear blueprint that says: here is the indication, here is the protocol, here is the patient population, here is the outcome measure, here is the dosage, here is the follow-up period, and here is the path to regulatory and reimbursement acceptance.
Rob’s prior terminology crosswalk is a perfect example of this same problem. We use terms like home training, remote training, and telehealth in ways that do not map neatly onto the language used by the FDA, CPT, and mainstream medicine. We could choose to adopt common medical terminology, but instead we often keep translating between our language and theirs. That may seem small, but it affects research design, advertising, documentation, regulatory strategy, and reimbursement. If we cannot even agree to use the language of medicine when speaking to medicine, then we should not be surprised when medicine struggles to understand what we are doing.
The CPT project is a major step in the right direction. While I cannot share confidential details, I do want to say clearly that people with access to that information can verify that, over the last 10 years, there were multiple points when we faced a very real risk of losing our current Category I CPT codes. That would not have been a small setback. It could have moved the field significantly backward and made recovery very difficult if not impossible. We survived that process, and now we have an opportunity to use our presence in the CPT world to move toward medical policy development, reimbursement progress, and broader acceptance by mainstream medicine.
But we need to decide what we actually want as a field. Do we want mainstream acceptance? If so, are we willing to do what that requires? Are we willing to invest in FDA strategy, CPT maintenance, medical policy work, research organization, and terminology alignment? Are we willing to accept that initial coverage may be limited to a few indications and may require more established protocols before broader flexibility is possible? Are we willing to take an honest look at ourselves, including the criticisms that may be uncomfortable but sometimes valid?
Based on my research, my work with insurance companies and CPT, and conversations within medicine more broadly, I think several steps would move us forward.
First, FDA strategy needs to become a real priority. FDA clearance gives us a stronger position for marketing, reimbursement, policy advocacy, and public credibility. The path is not as impossible as we sometimes make it sound. The PTSD neurofeedback device clearance and existing biofeedback device pathways suggest there are realistic starting points if manufacturers and professional organizations choose to prioritize them.
Second, we need to continue investing in CPT work. We now have a presence in that world, not only to protect what we have, but to help shape the field in accepted medical terminology. The last 10 years came with real costs, although we worked hard to keep those costs much lower than what other organizations have had to spend. But if we disappear from that process again, history will likely repeat itself.
Third, we need a better research translation strategy. The issue is not simply that we need more research or a different kind of research. The issue is how the research is framed. We need ISNR, perhaps in coordination with FNNR, to develop a regularly updated research review that places neurofeedback in context with other accepted interventions, identifies where the evidence is strongest, and provides a practical blueprint for moving the field forward. This could function as a living white paper, updated every few years, and could lead to periodic published summaries.
Fourth, we need to keep pushing insurance conversations forward. Medical policies are not the same as coverage policies, and we should not accept blanket denials as if they are simply the natural order of things. Some Medicaid plans, VA systems, employer-funded plans, and private plans do cover neurofeedback under certain conditions. That inconsistency matters. If some systems can determine coverage using the same evidence, then blanket denial by others deserves to be challenged. Mental health parity also matters here, especially when Category I CPT codes are denied across the board as experimental, investigational, or not medically necessary for any diagnosed condition.
I do think we are closer than we have been in a long time. We may be on the edge of meaningful reimbursement progress, possibly even broader coverage in some areas. But that progress will likely come first for a limited number of conditions and may initially require more established protocols. The question is whether we can accept that as a starting point while continuing to make room for the broader clinical flexibility that has always been part of this field.
I appreciate everyone engaging in this conversation, and I hope we continue it. More than that, I hope we act on it. If mainstream acceptance is truly what we want, then we need to fund it, organize around it, speak the language required for it, go through the steps that are commonly accepted like FDA, and build the regulatory and research pathways that make it possible with clear intent and motivation.
Talk is important, desires matter, but at the end of the day intentional targeted action and perseverance until it is complete is what ultimately reveals our priorities.
So, what do we want?
Best,
Mark Trullinger, PhD
To view this discussion visit https://groups.google.com/d/msgid/isnr_members_forum/002401dceaad%2425559600%247000c200%24%40gmail.com.
Robert Thatcher
John Mekrut
John Mekrut
Brandi Johnston
On May 23, 2026, at 12:49 PM, 'Robert Thatcher' via ISNR_Members_Forum <isnr_memb...@googlegroups.com> wrote:
To view this discussion visit https://groups.google.com/d/msgid/isnr_members_forum/1258000307.3234866.1779562123330%40mail.yahoo.com.
Rob Longo
John:
You wrote: “This is another opportunity to teach AI search engines what the current truth is. If someone better informed than me can write a piece on the current state of affairs (Mark?), we can place this in the ISNR education documents which will inform our Members, our own FAQ section (and our new ISNR GPT) and AI at large.”
I so appreciate your enthusiasm and ideas. What I am hoping you and others here following this thread will understand is the following:
Teaching an AI search engine to search for the truth on any topic related to QEEG, NFB, photo-biomodulation, neuroregulation, etc. etc., is not that simple. I have a close friend who is a major AI player and developer and has worked with google. (and keep in mind that students can now major and get degrees in AI use and development). And I mentioned in a previous post on the listserv… Most people barely know how to do a basic search using Google and other search engines. Using Chat GTP, CoPilot, etc., is no different. A simple inquiry can generate misinformation (PLEASE SEE THE ARTICLE LISTED AT THE BOTTOM OF THIS POST)
He listed for me the following brief bullet statements regarding AI and how it works (this is a simplified version):
- Most people use social media and AI is full of misinformation from Social Media. AI is listening to everything we do on line. Thus, if we are pulling in bad info, AI is learning that on our behalf.
- One must develop tools and generate rules and code for AI to follow. This requires time and a thoughtful process.
- The tools (when properly created) help build a unified code that AI uses to generate correct information and rule out bad info.
- This is a process that takes time.
- Once you generate AI information one needs to reverse the process to see if everything submitted was used and transposed accurately. So, it must be reverse checked much like translating a document from English to another language like Spanish. The Spanish version needs to be translated back into English to make sure it is accurate. The same needs to be done with AI. Are any references generated quality references from peer reviewed journals, or just info on line and from news casts?
- A single AI article on the ISNR Web Cite is not going to be helpful at all.
- ISNR would need to have a qualified professional generate a customized AI tool for ISNR members to use.
In summery: A good AI tool would certainly help increase the growth of ISNR, but that AI tool needs to be developed for ISNR by someone skilled in AI development.
Of course this is a different item than ISNR generating “our own FAQ section”. And, to create “our new ISNR GPT) and AI at large” will entail what I have mentioned above.
I hope this helps.
AI-generated fake citations are flooding scientific literature across publications, scientists warn
Rob Longo
To view this discussion visit https://groups.google.com/d/msgid/isnr_members_forum/8167e7f4-b943-4b62-b334-aabe0bc4ff7bn%40googlegroups.com.
Siegfried Othmer
David Cantor writes:
“The same argument can be made for EMDR. While the body of evidence is relatively limited, because it is coded under psychotherapy, it is generally approved. In many cases, insurance companies lump together any procedure using "sensors" as not being psychotherapy, making the codes arguable.”
It helps that there is an identification of EMDR with a particular discipline, which is not the case for NF. My thoughts go in multiple directions. Firstly, EMDR is based on a physiological mechanism. That means the boundary between the realm of the mind and that of physiology remains breached within the field of psychology. That de facto boundary was effectively established when psychology walked away from Biofeedback, which it totally owned at one time, and also distanced itself from NF. Since instrumental Biofeedback had already eclipsed Autogenics and Progressive Relaxation, it also forgot about those.
Given what we now know about the centrality of early trauma among the most challenging of cases in mental health, combined with our understanding that the trauma response system is largely physiologically rooted, the field of psychology is curiously distancing itself from its frontier as well as from its history.
This brings us back to EMDR. It stumbles just when it encounters those more intractable cases of early trauma, steeping the client further into disregulation. It falters just where it is most needed. Health professions do not consider the costs they impose on those with whom they fail. They are left worse off than before, both psychologically and physically.
As Bessel van der Kolk frames it, “Any therapy that does not place regulation front and center is likely to miss the mark.” In our emerging reality, NF is the indispensable remedy for deep trauma. And in my parochial perspective, the remedy lies in combining Endogenous Neuromodulation in the Infra-Low Frequency regime for restoration of self-regulatory integrity with traditional Alpha-Theta training for psychological resolution and the encounter with the core self. Deep trauma does not have to be remembered in order for the trauma response system to be deconstructed. It rests on physiological memory. In both domains, ILF and A/T, the process is inner-directed. In our attempt to alter our nature, we must act in accord with our nature.
Regulation of health practices does not lead; it follows. The reality on the ground has to change first, just as it did in the case of EMDR. Self-interest in augmentation of their skill set should suffice to induce psychologists to incorporate self-regulation technologies into their practices, independently of the issue of reimbursement.
Siegfried Othmer, Ph.D.
Chief Scientist, The EEG Institute
Los Angeles