The people I work with describe a pattern that keeps showing up with their doctor when they try to have a conversation with them about metabolic psychiatry. They go to their doctors, asking for support with ketogenic metabolic therapy, and the clinician replies with some version of “there is not enough evidence,” then treats the request as closed. That response is getting framed as evidence-based medicine. It is not. It is a shortcut that uses uncertainty as a conversation stopper, even though uncertainty was the exact problem the concept of evidence-based medicine was created to handle.

Evidence-based medicine was built to correct a specific failure mode that was seen in medicine at the time, and quite honestly still exists. Medical decisions were often being driven by seniority, anecdote, habit, and physiologic theories that sounded plausible but repeatedly failed when tested. The proposed correction was not “only follow randomized trials.” The correction was to use the best available evidence, judge its trustworthiness, interpret whether it applies to this patient, then to carefully integrate the evidence with clinical expertise and the patient’s values and preferences (Sackett et al., 1996).
PubMed

That last step is where many clinicians quietly abandon the method while still claiming the label. Evidence can estimate benefits and harms but is simply cannot decide which outcomes matter most to a particular person, how much risk they are willing to tolerate, or what tradeoffs are acceptable given their life constraints. When clinicians refuse to engage a patient’s preferences, they are not practicing evidence-based medicine. They are practicing clinician preference, then calling it science.

What the founders actually meant

The term evidence-based medicine is closely linked to McMaster University and is widely credited to Gordon Guyatt’s work in the early 1990s, with the phrase appearing prominently in the 1992 JAMA article by the Evidence-Based Medicine Working Group. But a guy named David Sackett is regarded by many as the father of evidence-based medicine, and his 1996 BMJ paper remains the clearest statement of what the approach was and was not.

Sackett had a core clarification that really matters because it directly contradicts how evidence-based medicine gets used today in real clinical settings. Evidence-based medicine was never meant to replace clinical expertise. It was meant to make expertise accountable to stronger forms of knowledge than personal conviction and tradition. Sackett’s strongest point that is relevant to what patients are experiencing in seeking out ketogenic therapies today, is that evidence alone does not dictate decisions. A clinician still has to interpret applicability, weigh harms, consider feasibility, and work with what the patient values most. Without that, the practice becomes rigid, and rigidity is not the same as rigor (Sackett et al., 1996). Although many doctors wanting to sound particularly smart without doing the cognitive work involved, would really like you think so.

This is also why a “guideline first” mentality is not equivalent to evidence-based medicine. A well-developed clinical practice guideline can be an efficient starting point, but evidence-based medicine still requires doctors to ask whether the guideline applies to the individual patient, in that context, and with the patients goals, constraints, and risk tolerances considered. When a clinician hides behind “the guideline” to avoid a values-based conversation, they are opting out of the method while keeping the branding. And I think for your own recovery, you become very aware to when that might be happening.

How “evidence-based” gets misused in real clinics

One common misuse is treating guidelines as the end of thinking rather than the beginning. Don’t be fooled. Guidelines are just summaries. You would see that if you ever had reason to pull one up and look. They are not a replacement for clinical reasoning about applicability. When evidence is limited, when a patient has characteristics under represented in trials, when comorbidity complicates tradeoffs, or when the outcome priorities differ from the outcomes emphasized in the literature, guideline obedience becomes a poor substitute for patient centered care. And I would go so far as to say an injustice. Evidence-based medicine does not demand automatic compliance with a document. It demands justified decisions and good judgement informed by evidence and grounded in what the patient values.

One of the things that bothers me most that I encounter on social media is any clinician misuses the intent of evidence-based medicine and misuses it by collapsing “no large randomized trials” into “no evidence.” Those are different statements. And I would expect a doctor who has been trained in the evaluation of research to be able to remember that evidence exists in layers.

Sometimes evidence is indirect, sometimes observational, sometimes mechanistic, sometimes early stage clinical. Evidence-based medicine does not pretend those layers are equal, but it also does not pretend that anything short of a large trial is meaningless. It asks for transparent appraisal and for clear communication of uncertainty. It even asks for shared decisions under uncertainty. That is the work of the doctor. Shutting down the conversation is not the work. It is a misuse of power in the doctor patient relationship. And it certainly does not deserve some kind of unspoken gold start for the delusion that they are practicing evidence-based medicine.

Another misuse of the concept of evidence-based medicine is confusing population averages with individual predictability. Randomized trials estimate average effects in the studied population but they simply do not tell you whether a particular person will be a strong responder, a partial responder, or a non responder. Evidence-based medicine never promised it would, so researchers need to knock off pretending that it should. That is why there is a long history of structured approaches aimed at learning what works for an individual, including N of 1 randomized trials designed to test treatments within a single patient using repeated, randomized comparisons (Guyatt et al., 1986; Lillie et al., 2011).

What? You never heard of N=1 as something legit? Well it certainly is!
An N of 1 randomized trial is a clinical trial conducted within a single patient, where the patient serves as their own control. Instead of comparing two groups, the trial compares the patient to themselves across repeated treatment periods. The core goal is something relevant to the patient, not to estimate an average effect in a population. The goal is to determine whether one option worked better for this them, using planned measurement rather than impression and memory. It sort of looks like a traditional randomized controlled trial, but miniaturized to the individual. The patient goes through repeated pairs of treatment periods, usually one period on the active intervention and one period on placebo or an alternative. The order is randomized. Blinding can even be used when feasible. Outcomes are tracked each period using a defined target such as symptoms, function, or lab measures. The sequence continues until the pattern is persuasive enough to support a clinical decision for that person.

Why would something like N of 1 be considered evidence-based medicine at the individual level?

Because evidence-based medicine gets distorted when people treat population averages as if they settle individual care. They do not. As any unwell person who has already tried all the treatments supported by randomized controlled trials will tell you. N of 1 trials were created as an answer to that distortion. N of 1 experiments are mature and accept variability as a fact. They treat “try it and see” as a legitimate clinical question, then discipline it with randomization, repetition, and structured outcomes so the result is less vulnerable to expectation effects, natural symptom fluctuation, and selective recall. N of 1 trials tend to be most useful when the condition is chronic and relatively stable, the treatment has a reasonably quick onset and offset, and the outcome can be measured repeatedly without ambiguity.

Let’s look at the population asking about KMT as a treatment for mental illness. Do they fit as possible N of 1s?

• Chronic and relatively stable conditions
• Reasonably quick onset of KMT treatment over two to four weeks with four months sustained to see treatment effects
• Outcome can be measured without ambiguity using mood assessments, patient report and family member observations

So why isn’t your doctor offering you this method? And how come it is not more widely used?

I don’t know. People for whom the standard of treatment have failed tend to end up doing it on their own. The barrier too using it has rarely been scientific logic but doctor logistics. These little n of 1 trials take time to set up, time to monitor, and time to interpret. Many clinical environments ration clinician time so tightly that structured individual experimentation becomes hard to deliver, even when it would answer the most important question in the room, which is whether it worked for the person needing it.

Another barrier is that the medical culture often treats individualized trialing as less legitimate than population evidence, even though the whole purpose of individualized care is to make a decision for one person, not to publish a group average. N of 1 trials sit exactly at that tension point, and they expose how often “evidence-based” is used as a rhetorical posture rather than a method for disciplined learning or clinical practice.

Another misuse is the evidence pyramid image. Many clinicians (myself included) were trained on an oversimplified pyramid where “systematic review” sits at the top and everything below it is treated as increasingly irrelevant.

That image hides a basic categorical error in thinking. Systematic reviews and meta analyses are tools for consuming evidence, not study designs that inherently outrank the underlying primary studies in every circumstance. A systematic review can be excellent, or it can be misleading if the included studies are biased, heterogeneous, or poorly aligned with the patient in front of you. This critique has been made directly within the evidence-based medicine literature, including calls to reframe how evidence hierarchies are taught and used.

A public primer on systematic reviews

Another misuse is the double standard around mechanistic reasoning. In daily practice, clinicians often rely on physiology and plausibility when direct evidence is incomplete and this is normal and sometimes necessary. The problem is they are selective in their rigor. When a patient asks for something outside the clinician’s comfort zone, clinicians suddenly adopt an absolutist stance that only one kind of evidence counts, even while they routinely accept weaker forms of reasoning to justify their daily standard clinical choices. That is not methodological purity but cultural gatekeeping disguised as epistemology.

Why this matters for ketogenic metabolic therapy requests

Ketogenic metabolic therapy tends to trigger clinician discomfort for predictable reasons. It sits at the intersection of psychiatry, nutrition, metabolism, and behavior change all while the evidence base in psychiatric conditions is still developing. The intervention requires real implementation effort and can carry risks that demand appropriate screening, monitoring, and coordination. Those are all legitimate clinical issues. The failure happens when legitimate issues get turned into blanket refusals, and refusal is justified as evidence-based medicine.

Evidence-based medicine actually has a structured way to handle emerging interventions and uncertain evidence. It requires transparency about certainty, honest discussion of desirable and undesirable effects, and direct engagement with patient values and preferences. It also requires acknowledging feasibility, acceptability, cost, and equity issues that shape whether a plan can work in a real life patient context.

Sometimes the patient and the clinician are optimizing different outcomes. The patient may be asking for KMT support because they care most about sleep, cognition, energy, daily functioning, or reducing medication burden. The clinician may be prioritizing symptom scale change, relapse prevention, short-term safety, or guideline consistency. Evidence-based medicine requires naming that mismatch, because the “best” choice depends on which outcomes are being weighted. If the clinician does not surface this in the discussion, a values based decision gets presented as a neutral evidence-based decision, and the patient’s preferences are silently excluded.

This is also where clinician humility becomes part of evidence-based practice. Some clinicians simply are not trained to support metabolic interventions. That does not justify dismissal. It just defines a boundary of competence. A clinician can say, directly and professionally, that they are not the right person to manage the metabolic and medication monitoring aspects of KMT, then offer collaboration, referral pathways, or coordination so the patient is not abandoned. Evidence-based medicine was never meant to protect clinician comfort. It was meant to protect patients from unsupported certainty.

What an evidence-based KMT conversation sounds like

A clinician practicing evidence-based medicine can start by treating the patient’s request as legitimate, even when their awareness or understanding of the evidence-base is limited. The clinician can say that evidence is emerging, that uncertainty remains about who benefits and how reliably, and that safety considerations require appropriate oversight. Then the clinician can ask what the patient is trying to change, which outcomes matter most, what tradeoffs they are willing to accept, and what constraints they face in daily life. If they are truly practicing evidence-based medicine, this is not extra work on their part. This is the clinical decision process, because patient preferences determine how benefits and harms should be weighed.

From there, the clinician can move into shared decision making in a disciplined way and present reasonable options, including standard care options and adjunctive options, describe what is known and unknown, and support the patient in deliberating based on what matters most to them. Shared decision making models were developed for exactly this purpose, to make preference sensitive decisions transparent and clinically usable.

Instead of treating KMT as something the clinician must endorse or reject in principle, the clinician can treat it as a time limited, monitored experiment designed to answer one practical question, whether it helps this patient on outcomes that matter to them without causing unacceptable harm. That approach can shift the conversation away from ideological debates and toward measurable goals, clear safety parameters, and agreed upon criteria for continuing, modifying, or stopping based on what actually happens. This is where N of 1 logic becomes clinically powerful. N of 1 trials are designed to determine the optimal intervention for an individual patient using objective, data driven criteria, recognizing that average effects do not determine individual response.

Sure, in real world practice, a clinician may not run a formal randomized N of 1 protocol, but they can still adopt the core discipline and help patients define outcomes, monitoring, time windows, stopping rules, and decide in advance how the patient and clinician will interpret the results. THAT, my friends, is evidence-based practice applied to the individual patient rather than outsourced to population averages. And none of that requires overstating what KMT can do. It requires doing the actual job of evidence-based medicine.

Appraise evidence honestly.

Communicate uncertainty clearly.

Integrate clinical expertise.

Respect patient preferences.

Build a safe plan.

Learn from the patient’s response.

The hard truth

When clinicians claim evidence-based medicine while refusing to engage patient preferences, they are not being rigorous. They are being paternalistic with better branding and creating a disrespectful power differential with their patient. Sackett warned against practice being tyrannized by evidence in ways that are inapplicable or inappropriate for the individual patient, and he framed clinical expertise as the safeguard that keeps evidence from becoming dogma.

The broader critique within medicine that evidence-based medicine is sometimes misapplied, sometimes industrialized, and sometimes used to serve interests other than patient care. These critiques are not fringe. They include major arguments that the movement needs renewal and refocusing, and that evidence can be distorted by conflicts of interest and system incentives. And I would say you should not put your recovery on hold waiting for the evidence-based treatment movement to achieve such a refocused revival of purpose and momentum. If the method has been replaced by institutional habit, your job is to find a way to recover in spite of it.

Ketogenic metabolic therapy does not need to be treated as a universal answer to be treated as a legitimate patient preference. A patient asking for KMT support is asking for informed partnership under uncertainty. Evidence-based medicine, properly practiced, requires that partnership.

Refusing it while claiming evidence-based purity is not science. It is a refusal to do the work. And we all need to call it out when we see it happen. But more importantly, for you, we need to help you flow around it towards the wellness that you want and deserve.

The purpose of this blog post is not to make you mad at your doctor. It is to help you see the limitations of your doctor, so you can stop hoping and pretending you will get the support you want and need. The purpose of this blog post is to help you let go of and accept the limitations being practiced by that medical provider, so you can be free to keep looking for a true evidence-based practitioner. If you have to, be patient and keep hope. The number of metabolic psychiatrists and other truly evidence-based clinicians are continuing to grow.

For now, take the label off the clinician misusing it and put it back on the method. Evidence-based medicine means shared decisions under uncertainty, not being shut down. If your provider cannot do that work with you, keep looking until you find someone who will. You do not need to abandon evidence-based medicine to pursue KMT support. You need a clinician who practices it as Sackett defined it, with evidence, expertise, and your values driving the plan. If your current provider refuses that partnership, it is not your job to stay stuck. Your recovery does not need to wait for a system wide culture shift. It needs informed collaboration, careful monitoring, and respect for what you are trying to change.

Find the clinician who can meet you there.

Brain Fog Recovery Program

References

Davidson, K. W., Silverstein, M., Cheung, K., Paluch, R. A., & Epstein, L. H. (2021). Personalized (N-of-1) Trials: A Primer. JAMA Pediatrics, 175(4), 404–409. https://doi.org/10.1001/jamapediatrics.2020.5801

Duan, N., Kravitz, R. L., & Schmid, C. H. (2013). Single-patient (n-of-1) trials: A pragmatic clinical decision methodology for patient-centered comparative effectiveness research. Journal of Clinical Epidemiology, 66(8, Supplement), S21–S28. https://doi.org/10.1016/j.jclinepi.2013.04.006

GRADE handbook. (n.d.). Retrieved December 21, 2025, from https://gdt.gradepro.org/app/handbook/handbook.html

Guyatt, G., Cairns, J., Churchill, D., Cook, D., Haynes, B., Hirsh, J., Irvine, J., Levine, M., Levine, M., Nishikawa, J., Sackett, D., Brill-Edwards, P., Gerstein, H., Gibson, J., Jaeschke, R., Kerigan, A., Neville, A., Panju, A., Detsky, A., … Tugwell, P. (1992). Evidence-Based Medicine: A New Approach to Teaching the Practice of Medicine. JAMA, 268(17), 2420–2425. https://doi.org/10.1001/jama.1992.03490170092032

Guyatt, G., Sackett, D., Taylor, D. W., Ghong, J., Roberts, R., & Pugsley, S. (1986). Determining Optimal Therapy—Randomized Trials in Individual Patients. New England Journal of Medicine, 314(14), 889–892. https://doi.org/10.1056/NEJM198604033141406

Lillie, E. O., Patay, B., Diamant, J., Issell, B., Topol, E. J., & Schork, N. J. (2011a). The n-of-1 clinical trial: The ultimate strategy for individualizing medicine? Personalized Medicine, 8(2), 161–173. https://doi.org/10.2217/pme.11.7

Lillie, E. O., Patay, B., Diamant, J., Issell, B., Topol, E. J., & Schork, N. J. (2011b). The n-of-1 clinical trial: The ultimate strategy for individualizing medicine? Personalized Medicine, 8(2), 161–173. https://doi.org/10.2217/pme.11.7

Lillie, E. O., Patay, B., Diamant, J., Issell, B., Topol, E. J., & Schork, N. J. (2011c). The n-of-1 clinical trial: The ultimate strategy for individualizing medicine? Personalized Medicine, 8(2), 161–173. https://doi.org/10.2217/pme.11.7

Sackett, D. L., Rosenberg, W. M., Gray, J. A., Haynes, R. B., & Richardson, W. S. (1996a). Evidence based medicine: What it is and what it isn’t. BMJ (Clinical Research Ed.), 312(7023), 71–72. https://doi.org/10.1136/bmj.312.7023.71

Sackett, D. L., Rosenberg, W. M., Gray, J. A., Haynes, R. B., & Richardson, W. S. (1996b). Evidence based medicine: What it is and what it isn’t. BMJ (Clinical Research Ed.), 312(7023), 71–72. https://doi.org/10.1136/bmj.312.7023.71

Shamseer, L., Sampson, M., Bukutu, C., Schmid, C. H., Nikles, J., Tate, R., Johnston, B. C., Zucker, D., Shadish, W. R., Kravitz, R., Guyatt, G., Altman, D. G., Moher, D., Vohra, S., & CENT group. (2016). CONSORT extension for reporting N-of-1 trials (CENT) 2015: Explanation and elaboration. Journal of Clinical Epidemiology, 76, 18–46. https://doi.org/10.1016/j.jclinepi.2015.05.018

The Truth About Evidence-Based Medicine with Dr. Gordon Guyatt. (n.d.). Metabolic Mind. Retrieved December 21, 2025, from https://www.metabolicmind.org/resources/news-views/podcasts/metabolic-mind-podcast/the-truth-about-evidence-based-medicine-with-dr-gordon-guyatt/

Vohra, S., Shamseer, L., Sampson, M., Bukutu, C., Schmid, C. H., Tate, R., Nikles, J., Zucker, D. R., Kravitz, R., Guyatt, G., Altman, D. G., & Moher, D. (2015). CONSORT extension for reporting N-of-1 trials (CENT) 2015 Statement. BMJ, 350, h1738. https://doi.org/10.1136/bmj.h1738