Truths and Fallacies of Psychopharmacology

by | Jul 7, 2023 | Education

    Truths of Psychopharmacology

1. Your treatment is as good as your diagnosis.
2. Treat diseases, not symptoms.
3. All drugs are guilty until proven innocent.
4. All drugs are toxic; only the dosing and indication makes them therapeutic.
5. Always have an exit strategy.
6. Most psychiatric drugs are symptomatic, not disease-modifying. The clearest exception is lithium.
7. Older drugs are more effective than newer drugs.
8. Newer drugs are more tolerable than older drugs.
9. Treatment “resistance” usually reflects either misdiagnosis or an invalid diagnosis.
10. Course, not symptoms, reveals the diagnosis.

    Fallacies of Psychopharmacology

1. More is better.
2. What gets you well keeps you well.
3. Treat to remission.
4. Polypharmacy is good.
5. Make one change at a time.
6. 4-8 weeks is a sufficient trial of a drug.
7. Dose low, go slow.
8. Always taper a drug.
9. Treat the most severe symptom.
10. Always incorporate the patient’s preference.

One should never assume that what one is taught is true. C. S. Burwell, a Harvard professor of medicine, had it right; in paraphrase: What we are taught is about half true, and half false, and it’s important to know which half is which (1).

Over three decades of clinical experience and research in psychopharmacology, I have come to conclude that there are some basic concepts that need to be understood to practice well. Some of those concepts need to be understood as true, and some as false. In short, there are truths and fallacies in psychopharmacology, and it is important to know the difference.

I have discussed these principles in more detail elsewhere (2) but provide a summary here.

Truths of Psychopharmacology

1. Your treatment is as good as your diagnosis.

This axiom was told to the author by Dr. Edwin Cassem, former chairman of the department of psychiatry at Massachusetts General Hospital, and also a Jesuit priest. It reflects the central idea of the Hippocratic approach to medicine, which is that we should treat diseases not symptoms (3). So if the disease is unknown or not diagnosed, treatment of symptoms will only have partial benefit at best, often counteracted by many side effects. One of the problems in contemporary psychiatry is that it doesn’t pay much attention to diagnosis; many clinicians just treat symptoms. Another way it works is that diagnoses are made using the DSM system, which is mostly invalid, as explained elsewhere (4).

If the diagnosis is wrong – or worse – as is often the case with DSM ideology, if the diagnosis is invalid – the treatment will be ineffective.

2. Treat diseases not symptoms.

This basic Hippocratic principle was restated most clearly by the father of modern medicine, William Osler, at the turn of the 20th century. When a disease is treatable, then effective treatments will manage all symptoms, and even notable side effects will be outweighed by benefits for the overall disease. When a disease is unknown, untreatable, or invalidly defined, then treatment of its symptoms may occur to a partial or modest extent, but will never be complete. Further, since all drugs have side effects, as in the following principle below, those side effects often are outweigh partial or modest benefits. As an example, aspirin will never completely get rid of fever, but penicillin will. This is the case, even though penicillin has no direct anti-fever effect, and aspirin does. Similarly, in psychiatry, presumed antidepressants, which are proven mainly to improve current symptoms rather than to modify an underlying disease (see below), are less likely to completely rid patients of depressive symptoms, compared with a medication shown to modify an underlying disease that causes depressive episodes, such as lithium.

3. All drugs are guilty until proven innocent.

The 19th century physician Oliver Wendell Holmes Sr taught, again in the Hippocratic tradition, that all medications have side effects, and therefore they should not be used because side effects are always harmful.3 However, if medications are proven effective, and a disease is improved or cured, then those medications can be used since those benefits will outweigh side effects in most cases. This approach relates to the classic risk-benefit analysis of medical treatment.
Clinicians know that all treatment involves weighing of risks and benefits. But many clinicians start on the risk side: they first ask themselves which have the least side effects; often, the search for efficacy is quite limited or even ignored, and medications are simply used because they are apparently benign. It is important to always remember that no medication is truly benign. It is well established that, over time, apparently safe medications turn out to have been harmful.

Instead, start not with risks but with benefits. Focus first on what medications are proven effective. Then with that winnowed list, assess which ones are safest or have the fewest side effects. Start with efficacy, then assess side effects, not vice versa.

4. All drugs are toxic dosing and indication makes them therapeutic.

This truth, which is paraphrased from the teaching of the medieval founder of pharmacology, Paracelsus. All medications have side effects and harms, and thus in a sense they are all toxic. But side effects and toxicities vary with dose. Many times a medication can be quite harmful at high doses, and quite safe at low doses. So dosing is an important aspect of making judgments about the risks and harms of medications. Indication for treatment relates to efficacy and benefits of medications. If a medication is proven effective, as discussed above, for a certain diagnosis or disease or indication, then it can be used even if it has many side effects or toxicities.

5. Always have an exit strategy.

General Colin Powell once described, when thinking of the Vietnam war, that you should not start a war unless you know how you would plan to end it. You need to have an exit strategy from the beginning. The same holds for medications. Some might be continued indefinitely if one has clear evidence that benefits outweigh harms as described above, but this decision usually holds for drugs which treat diseases, not purely symptomatic drugs, as is the case with most psychiatric medications (see below).

Hence, when patients get better with medications for current symptoms, clinicians and patients should not been conclude that those medications should be continued indefinitely.

6. Most psychiatric drugs are symptomatic, not disease modifying. Lithium is an exception.

This is a controversial concept, and it cannot be fully explained or defended here. Readers should consult a full paper length elsewhere for complete analysis.5 Here only conclusions will be given without full explanatory support. The claim here is that most psychotropic drugs are symptomatic in their effects, and do not alter or improve the underlying disease process that might be causing those symptoms, which is reflected clinically in improving the long-term course of illness. Evidence exists to support this view with antipsychotics (which do not improve the course of chronic psychosis in schizophrenia), antidepressants (where long-term prevention of mood episodes can be questions based on potentially invalid maintenance clinical trial designs as explained more fully elsewhere), benzodiazepines (where benefits for anxiety are shown short-term but not proven in proven with long-term randomized trials), and amphetamines (which have immediate benefits for attention, but long-term attention deficit resolution in the majority of subjects has not been shown in randomized trials). Evidence for the above statements with citation is provided in a separate full paper discussion.5 An exception to this overall rule is that mood stabilizers like lithium do alter the course of bipolar illness and unipolar depression, and prevent mood episodes, sometimes completely (6).

7. Older drugs are more effective than newer drugs.

The most effective mood treatments are ECT, discovered in the 1930s, the monoamine oxidase inhibitor class and tricyclic antidepressants and lithium, from the 1950s. The most effective treatment for schizophrenia is clozapine, from the 1970s. In short, since the 1970s, new psychotropic medications are not more effective than prior medications.

8. Newer drugs are more tolerable older drugs.

The flipside of the prior truth is that the newer generation of medications in the past few decades, like SRIs and newer antipsychotics, has been developed to have fewer side effects than older medications, but they are not more effective than older medications. In many cases, they are less effective than older medications, but still effective enough to be used clinically, and with their lower side effect burden, they are more feasible to use.

9. Treatment “resistance” usually reflects either misdiagnosis or an invalid diagnosis

The current attention to treatment resistance in psychiatry is a reflection of acceptance of the current diagnostic system as valid, while looking for better treatments for those diagnoses. If the current DSM diagnostic system is only partially valid,4 then insufficient benefit with medications could reflect the fact that the clinical clinical phenotypes being treated are not real enough to produce notable benefits with treatments that affect biological mechanisms. Even if clinicians do apply diagnostic criteria correctly, poor treatment results would follow.

10. Course, not symptoms, reveals the diagnosis.

This is an axiom of psychiatric diagnosis based on the tradition of Emil Kraepelin. The DSM system takes an opposite approach, using symptoms to make diagnoses in almost all cases, without reference to course of illness. Since many different illnesses can have many different symptoms, often overlapping, this symptomatic approach leads to diagnostic complexity that is both unnecessary and harmful. Many illnesses reveal themselves over time by their course, even if they overlap in symptoms to a great extent. This classic insight led Kraepelin to differentiate schizophrenia (dementia praecox) from manic-depressive illness. Pay attention to the course of illness: it is equivalent to pathology for psychiatry. It tells you which symptoms reflect which diseases, as opposed to assuming that each symptom is its own disease.

Fallacies of Psychopharmacology

1. More is better.

The first fallacy of psychopharmacology has to do with a belief that is logical, but false. Examples abound of useful and effective medications which are harmful at too high of a dose. Aspirin at very low doses prevents heart attacks and strokes, but at average to high doses has no further preventive benefit, and instead causes gastrointestinal bleeding. In low to average doses, lithium is very effective for manic episodes, but at higher doses, it is toxic and can cause kidney failure. The disadvantage of high doses is not always related to toxicity. Some medications are less effective at higher doses. For instance, aripiprazole is more effective for treatment of the acute depressive episode at less than 10 mg per day compared to more than that dose (7).

It is not a rule in a general sense that increasing doses routinely will lead to more benefit. In fact, it is a much more general rule that after a certain dose range, increasing doses will lead to a plateau of effects, with usually more side effects, toxicity, or harm.

2. What gets you well keeps you well.

This is the happily ever after fallacy. The idea is that if medications are helpful acutely, then they should be continued and the patient will continue to do well, happily ever after. This belief is consistent with common sense, but unfortunately common sense is not consistent always with scientific truth. There are many examples of drugs which work acutely not preventively, and vice versa. For example, sumatriptan is effective acutely for migraine, but does not prevent it. In contrast, propanolol is effective for prevention of migraine, but does not treat it acutely. Penicillin and other antibiotics are generally effective acutely for various infections, but do not prevent infections if continued long-term. Steroids are generally effective acutely for autoimmune disease exacerbations, such as lupus, but continuation of steroids do not prevent future exacerbations effectively. In short, there is no scientific or biological reason to believe that what gets you well keeps you well.

In psychiatry, this is one of the most common and harmful fallacies. SRIs are given acutely for depressive and anxiety symptoms, with some benefit, and then continued forever, for years and decades, with no exit plan, and with the eventual occurrence of very severe withdrawal syndrome if any attempt is made to come off them.

3. Treat to remission.

This fallacy has become the mantram of many psychopharmacology researchers. The view is that if patients are not improving sufficiently, it is simply because clinicians are not trying hard enough. If we use more medications and at higher doses, we will be able to achieve more benefit. The problem of “residual” symptoms is to be fixed by more and more medications of the same kind. These clinicians and researchers sometimes don’t appreciate that the problem may be that the wrong kinds of medications are being used for the wrong kinds of diagnoses. Or, as a related problem, the diagnoses, often based on DSM, may not be biologically valid enough to improve markedly with certain drug classes (8). In other words, the underlying assumption behind this fallacy is that our diagnostic system is legitimate and we just need to get better treatments for those diagnoses. This assumption itself is questionable.

4. Polypharmacy is good.

Even if remission is not set as the goal, there is the belief that multiple medications will be more effective than any single medication. This approach, called polypharmacy, has been criticized in medical tradition going back to the 19th century, but in recent years it has seen a resurgence in psychiatry. One cannot assume that combinations of medications are more effective than monotherapy. Sometimes they are, but sometimes they are not, and the adjudication has to rely on clinical research that proves the matter one way or the other. This is an empirical question, not a logical one.

5. Make one change at a time.

Although laudable in principle, this approach can lead to an overly slow process of finding the right medication or medication combination in an individual. One can go too fast or too slowly in practice. The rationale for making one change at a time is not really a clinical rationale. A basic principle of bioethics is that the primary purpose of clinical practice is to improve the patient’s clinical state, not to obtain knowledge. The latter is a goal for research. Knowing what is doing what is a knowledge question, and really belongs to the world of research. Getting patients better is the first goal of the clinician. Even if two changes are made at the same time, or three, what really matters is whether patients improve. Later, when patient are doing well, clinicians can remove one medication from the mix, to see if improvement persists.

6. Four to eight weeks is a sufficient trial of a drug.

The opposite problem to the fifth fallacy is the problem of going too fast. This is the case when multiple medications are tried one after the other, for one to two months in duration. This belief that one to two months is the timeframe for a medication trial is based on clinical research studies, often for depressive episodes, where acute efficacy is being studied. Clinical practice, in contrast, typically involves long-term treatment, not just treatment for 1 to 2 months. Almost always, then, what is relevant for clinical practice is long-term maintenance prevention of symptoms or episodes of illness, not just its acute treatment. As discussed in the second fallacy, one cannot presume that there will be long-term maintenance prevention benefit with a drug which is only been proven effective for acute short-term symptoms.

7. Dose low, go slow.

This fallacy again relates to going too slow. The first part of this belief has to do with the idea of starting a low dose with every patient. This is not reasonable because many patients present with severe acute symptoms that require medium to higher doses for immediate symptom control. For instance in a severe outpatient acute manic state, medium to high doses of a dopamine blocker should be given quickly so as to prevent possible hospitalization.

8. Always taper a drug.

The view that medication should always be tapered is restrictive to an excessive degree. There are many medications which do not have withdrawal syndromes, nor any known clear common risks from discontinuation, such that tapering is unnecessary.

9. Treat the most severe symptom.

It is common for clinicians to rely on the concept of “target symptoms.” Frequently, the target symptom is the most severe symptom that a patient experiences. This fallacy comes back to the problem of taking a symptom-oriented approach to treatment, as opposed to a disease-oriented approach. It would be like focusing on fever as the most severe symptom of an infection, rather than trying to treat the underlying disease. One could give plenty of acetaminophen, but only modest symptomatic benefit would be achieved without an antibiotic. This scenario may be similar when multiple or higher doses of monoamine agonists are given for anxiety or depressive symptoms, but the underlying affective illness is not treated. It is the case with medical diseases that the most severe symptoms are not related directly to the underlying disease. Many medical diseases have no symptoms at all, such as hypertension or cancer. The severity of symptoms therefore does not correlate at all with the disease, nor does it accurately direct which kind of drug should be selected.

10. Always incorporate the patient’s preference.

In contemporary medicine, it is no longer acceptable to be paternalistic with one’s patients. Yet clinicians are presumed to possess such specialized knowledge, and are held accountable for the application of that knowledge both ethically and legally. Thus, clinicians need to be aware that they are expected to have knowledge that is not possessed by the patient, and they need to determine how they can apply that knowledge in a way that is respectful of the patient’s individual liberties, but without equalizing their relationship in a way that is neither factually true nor legally protected.

Conclusions

All clinicians have a philosophy of psychiatry. They accept certain concepts, usually based on what they were taught in their training, and then they implement those concepts in practice. Academics speak of “evidence” but often these pre-existing concepts have more impact on practice than data from studies. Half of those concepts are false, and half are true, and it is important to which half is which.

REFERENCES

1. Pickering GW. The purpose of medical education. Br Med J. Jul 21 1956;2(4985):113-6. doi:10.1136/bmj.2.4985.113
2. Ghaemi SN. Clinical Psychopharmacology: Principles and Practice. Oxford University Press; 2019.
3. Ghaemi SN. Toward a Hippocratic psychopharmacology. Can J Psychiatry. Mar 2008;53(3):189-96. doi:10.1177/070674370805300309
4. Ghaemi SN. DSM-5 and the miracle that never happens. Acta Psychiatr Scand. Jun 2014;129(6):410-2. doi:10.1111/acps.12263
5. Ghaemi SN. Symptomatic versus disease-modifying effects of psychiatric drugs. Acta Psychiatr Scand. 2022;in press
6. Barroilhet SA, Ghaemi SN. When and how to use lithium. Acta Psychiatr Scand. Sep 2020;142(3):161-172. doi:10.1111/acps.13202
7. Thase ME, Jonas A, Khan A, et al. Aripiprazole monotherapy in nonpsychotic bipolar I depression: results of 2 randomized, placebo-controlled studies. Journal of clinical psychopharmacology. 2008;28(1):13-20.
8. Ghaemi SN. Taking disease seriously in DSM. World Psychiatry. Oct 2013;12(3):210-2. doi:10.1002/wps.20082