What is Therapeutic Necessity?

Introduction

In Herczegfalvy v Austria the European Court of Human Rights ruled that imposed psychiatric treatment without consent did not breach Article 3 of the European Convention of Human Rights (freedom from torture and inhuman or degrading treatment) in situations of medical or therapeutic necessity (Bartlett, 2011: p. 525). In order to grant this warrant to override the usual ethical position of respecting patient autonomy it is necessary to define what medical or therapeutic necessity is. Bartlett, a legal expert, suggested therapeutic necessity should involve “treating the underlying disorder or improving outcomes” in situations of severity (Bartlett, 2011: p. 525) whilst expressing scepticism about psychiatric treatments.

I will argue that Bartlett’s criteria are unnecessarily restrictive as to what a therapeutic objective is (for example can include reducing distress), is overly pessimistic about psychiatric treatment effectiveness, that necessity involves the clinician weighing up several factors and that safeguards exist to oversee clinical decision-making.

1.    Why is Therapeutic Necessity Important?

“A characteristic feature of mental health legislation … is the provision which is made for waiving, in certain cases, that respect for patients’ autonomy, their right not to be treated without their explicit consent, … regarded as a requirement in most fields of medicine.” (Mathews, 2000: p.59)

Psychiatric detention and treatment without consent may be regarded by some as breaching Articles 3 (freedom for torture and inhuman/ degrading treatment),  5 (liberty and security of the person) and 8 (respect for private and family life) of the European Convention of Human Rights (Bartlett, 2012:  352). On the specific issue of enforced treatment without consent the case of Herczegfalvy v Austria is “key” in establishing under which circumstances Article 3 has not been breached (Bartlett, 2011).

“The established principles of medicine are admittedly in principle decisive

in such cases; as a general rule, a measure which is a therapeutic

necessity cannot be regarded as inhuman or degrading. The Court must

nevertheless satisfy itself that the medical necessity has been convincingly

shown to exist.” (Bartlett, 2012: p.365)

Specific advice has been given firstly for necessity (Article 18) and what constitutes therapeutic (Article 19).

“The Recommendation of the Committee of Ministers to Member States concerning the protection of the human rights and dignity of persons with mental disorder, Rec(2004)10,.. sets out different criteria for involuntary treatment and involuntary detention. The substantive recommendations concerning involuntary treatment are as follows: Article 18 – Criteria for involuntary treatment A person may be subject to involuntary treatment only if all the following conditions are met: i. the person has a mental disorder; ii. the person’s condition represents a significant risk of serious harm to his or her health or to other persons; iii. no less intrusive means of providing appropriate care are available; iv. the opinion of the person concerned has been taken into consideration. Article 19 – Principles concerning involuntary treatment 1. Involuntary treatment should: i. address specific clinical signs and symptoms; ii. be proportionate to the person’s state of health; iii. form part of a written treatment plan; iv. be documented; v. where appropriate, aim to enable the use of treatment acceptable to the person as soon as possible.” (Bartlett, 2012: p.374)

Legal justification for enforced psychiatric treatment without consent relies on the condition of therapeutic necessity and split into two further conditions – what is therapeutic intervention and what is necessity. 

2. What is Bartlett’s definition of Therapeutic Necessity?

Bartlett, a legal academic, defines therapeutic (as opposed to therapeutically necessary) as “treatments {which} may appropriately be prescribed which are not therapeutically necessary. Treatments which attack the symptoms of a disorder but not the underlying disease are an obvious example” (Bartlett, 2011: p. 533) as well as “treating the underlying disorder or improving outcomes” (Bartlett, 2011: p. 531). The Code of Practice of the Mental Health Act (1983) requires that appropriate treatment “must be treatment which is for the purpose of alleviating or preventing a worsening of the patient’s mental disorder or its symptoms or manifestations” (Bartlett, 2011: p.533). Whereas Bartlett does not regard lessening symptoms as part of therapeutic necessity – only those treatments that target “underlying disease” or improve outcomes (but not outcomes such as improving symptoms). Bartlett further distinguishes between therapeutic appropriateness and therapeutic necessity with only the latter justifying compulsion (Bartlett, 2011: p.534). To discuss this distinction further:

“that an appropriate treatment is available should not mean that a State should… force people to undergo that treatment.  ‘Medical necessity’ as intended by the Court presumably does not mean that without such treatment the patient will suffer death or serious physical injury, since, at least as regards the neuroleptic medication, it is not obvious that this would have been the result in Herczegfalvy.  The degree to which treatment can be enforced in order to safeguard others in society is likely to prove a controversial point” (Bartlett, 2012 p.366).

Bartlett sets further requirements for therapeutic necessity:

 “likelihood of given result… achieve long-term and sustainable change in … underlying disorder…some sort of substantive threshold of degree of benefit… {when} stabilization of a condition that would otherwise worsen may sometimes be all that is realistic, it is possible… that treatments that will reliably achieve this end might …meet… therapeutic necessity test; but the outcomes of treatment over non-treatment must presumably be real, significant and manifest” (Bartlett, 2011: p. 534).

“Is the degree of likelihood of success relevant–that is, even to prevent death, can it be said that a highly intrusive and painful treatment with a low probability of success is ‘medically necessary’? These and many other questions are left unanswered by the standard articulated… While the treating physician may be best placed to make an initial evaluation of the situation.. he or she is too close to the situation for the process to end with him or her…the State’s right to impose treatment raises a variety of political, legal and social issues as well as medical ones. This is not a purely medical matter, and the process must allow for an independent review of the decision” (Bartlett, 2012: p.367).

Bartlett’s own conclusions on psychiatric treatment are that they do not meet his ideal of knowing based on the patient’s own history of a treatment that has a high degree of success with few adverse effects but that in practice

 “serious mental illness involve a considerable amount of trial and error by clinicians in individual cases…results of this process can be mixed: sometimes, drug combinations are found which largely alleviate the disorder with minimal adverse effects; sometimes drug treatments have minimal beneficial effects, following considerable unpleasant adverse effects. In these circumstances, knowledge of probable outcomes in the specific case may be low, and the above standard would not allow compulsory treatment to be used” (Bartlett, 2011: p. 534)

Further,

“reasonable certainty of diagnosis, therapeutically necessary treatment, and outcome should be required before involuntary treatment is contemplated. If such certainty is absent, it is difficult to see that a human rights infringement is ‘balanced’ by a corresponding benefit (if indeed such a balancing is the correct approach), since the benefit is speculative. The requirement that a therapeutic necessity must be ‘convincingly shown to exist’, as required by Herczegfalvy, suggests a medical consensus on the appropriate treatment in the patient’s case, and disagreements between medical professionals are thus particularly problematic” (Bartlett, 2011: p.535).

Summarising Bartlett’s view is the difference between a treatment being appropriate (such as being indicated for a condition) or available and therapeutic necessity. He does not outline what the grounds for necessity are but does seem to acknowledge it is not necessarily life or death severity but even preventing harm to others is potentially controversial (Bartlett, 2012 p.366). Bartlett expects that to meet therapeutic necessity threshold to justify compulsion that proposed treatments should have a known high probability of affecting the underlying causes of disorder and achieving a substantial clinical benefit (Bartlett, 2011: p. 534). Taking into account the realities of clinical practice in psychiatry as well as adverse effects in treatment Bartlett doubts that psychiatric treatments would meet this threshold (Bartlett, 2011: p. 534 and p. 535) and that independent review of treatment plans by clinicians is necessary when compulsion is used (Bartlett, 2012 p.367).

3. Problems with Bartlett’s interpretation of Therapeutic Necessity

There are several problems with Bartlett’s positions – medical treatment in all of medicine does not necessarily have to address the underlying causes of a disorder to be regarded as therapeutic, psychiatric treatments overlap in effectiveness with many general medical treatments regarded as effective, in general medicine as well as psychiatry treatment often involves uncertainty of outcome, adverse effects are common to all types of medical treatment not just psychiatric treatment, his views on what is necessity may be regarded as overly restrictive to clinicians and finally his values seem to be against psychiatric treatment which may have informed his criteria for therapeutic necessity. I shall start with discussions of what is regarded as therapeutic in medicine before moving onto necessity.

Critical psychiatrists offer two models of prescribing goals – one a “disease-centred model” where medication reverses hypothesised mechanisms of disease producing symptoms -similar to Bartlett’s definition of therapeutic - or a “drug-centred model” where a drug is regarded as always harmful and producing many effects of which some may be regarded by the patient as beneficial such as blunting of painful emotions (Yeomans et al, 2015) – which would not meet Bartlett’s criteria. A more sophisticated view of therapeutic interventions in psychiatry has three potential goals: “alleviation of symptoms without affecting the disease course; prevention of progression, but not curative …; and curative or preventive …” (Lieberman et al, 2019: p.794

Sackett - the prime mover behind the “Evidence Based Medicine movement” - along with his co-authors outlined a list of objectives of treatments (Sackett et al, 1991: p. 189) which are relevant to discussions of therapeutic intentions in general medicine as whole as well as psychiatry in particular. Cure as an objective should be applied to circumstances such as a reversal of disease process but in layman terms often refers to situations where the symptoms are eliminated or a disease mechanism is compensated for (such as insulin to replace the body’s own insulin or benzodiazepines are used to greatly reduce the symptoms of alcohol withdrawal). Prevent recurrence refers to situations where treatment is aimed at preventing a return of the medical condition (or maintenance antipsychotics in schizophrenia) and can be expanded to preventing a first occurrence of a condition (such as vaccination to prevent infections) (Huda, 2019: p.80). Limiting deterioration are where interventions are given to prevent further problems such as the use of plasters and surgical fixation to prevent displacement of fractures or skills training to prevent loss of psychosocial functioning in severe mental illness. Many medical conditions are treated to prevent known complications associated with the condition such as the use of antihypertensives in high blood pressure to prevent cardiovascular disease. Relieving distress or symptoms is an important therapeutic goal as it is often distress or other unpleasant experiences labelled as symptoms that bring people to seek out medical help. Examples include painkillers for pain or antidepressants in depression.

These therapeutic goals regarded as legitimate in medicine as a whole and applied to psychiatry in particular are clearly broader than the narrow definition of Bartlett of affecting the underlying cause of the disorder. It should also be noted at this point that treatments used in general medicine do not always affect the underlying cause of the disorder either (Huda, 2019: p.291-292). Although outcomes such as preventing suicide and homicide may be accepted as outcomes by Bartlett they are thankfully relatively infrequent so it is difficult to have studies large enough to demonstrate proven benefit in achieving these outcomes. This should not be used as a reason to rule out psychiatric treatment unless that is one’s intention a priori as other treatment goals are regarded as legitimate in medicine.

In terms of establishing “substantive threshold of degree of benefit”(Bartlett, 2011: p.534) for psychiatric treatments, there is an overlap in effectiveness between medication used in psychiatry and general medicine in achieving therapeutic objectives (Leucht et al, 2012 and Leucht et al, 2015). What is the threshold in establishing substantial benefit? The psychiatric treatment with the largest effectiveness magnitude is electroconvulsive therapy (Huda, 2019: p.284) so if a threshold for substantial benefit is set that excludes antidepressants, antipsychotics and mood stabilisers then the only treatment option left under Bartlett’s criteria for enforced treatment would be electroconvulsive therapy – surely a drastic conclusion and potentially harmful when the excluded treatments may be more appropriate choices for the particular patient.

Bartlett’s point about trial and error in deciding treatments also applies in general medicine given the overlap in effectiveness with psychiatric and general medical treatments. There is often no absolute certainty about treatment in general medicine either. Number Needed to Treat (NNT) is a metric used in the medical literature: it indicates how many people need to be treated with the specified treatment before one person achieves the target significant clinical benefit compared to those given placebo or other control treatment. The NNT for psychiatric treatments overlaps with those of treatments regarded as effective in general medicine (Huda, 2019: p. 285-286). This suggests that there is a similar degree of therapeutic uncertainty in general medicine and psychiatry. As for certainty of diagnosis, there are similar degrees of diagnostic agreement in psychiatry compared to general medicine (Huda, 2019: p.120-143). Finally, problematic adverse effects are frequently found in treatments used in general medicine (Huda, 2019: 288-289). One argument that could be used by Bartlett is that he is discussing enforced treatments not treatments in everyday practice but the comparison presented here is relevant – that treatments in psychiatry are similar in some important respects to those used in general medicine and fulfil therapeutic criteria. The question of whether they should be enforced then rests on the necessity.

An article by Curtice summarised several of the crucial issues when it comes to necessity:

“Margin of appreciation …clinical decisions that are proportional, therapeutically necessary and in keeping with accepted clinical practice are very unlikely to be outside this margin…

Proportionality Clinical intervention needs to balance the severity of the effect of the intervention with the severity of the presenting clinical problem, i.e. be a proportionate response to a clinical scenario.

Threshold of severity to engage Article 3 Ill-treatment must attain a minimum level of severity; assessment of this minimum is relative. ...” (Curtice, 2008: p. 390)

The concept of necessity requires balancing several factors to demonstrate that the therapeutic intervention is “proportionate response to a clinical scenario” (Curtice, 2008: p.390). These factors include “severity” of the problem (degree of symptoms, degree of impairment and probability of improvement if no treatment is offered as well as probability and nature of consequent risks) and “severity” of proposed treatment (the nature of the treatment, the probability and nature of adverse effects). There is also the question of probability of the treatment being effective but this also needs balanced with the consequences of not treating the problem – the worse the prognosis the more likely a low chance of success would be regarded as justified by necessity. Examples from clinical practice are now discussed.

An elated manic patient may wish to continue in this state (though there are often periods of mixed moods including depressed mood) and not see any problem with spending large amounts of money or reckless impulsive decisions that may cause problems such as sexual activity with people other than their spouse. The clinician may decide that the priority is to reduce the manic symptoms which should also have the benefit of preventing complications in the person’s life by their disinhibited behaviour. (One patient told me “you spoilt my fun” when I had reduced their manic symptoms). A depressed patient who is suicidal may not agree with the doctor for starting treatment with the therapeutic goal of reducing symptoms and preventing occurrence of suicide: they may believe they deserve to suffer and die because of guilt, they may think treatment is pointless because of depressive pessimism, they may prioritise harm avoidance of treatment side-effects A patient with what the doctor regards as delusions or hallucinations may believe in the truth of what the doctor describes as psychotic symptoms. They may want to have any associated distress reduced but in consonant with their beliefs/ hallucinations (such as the persecution ending from persecutors). Apart from reducing distress the doctor may wish to prevent the occurrence of events – such as self-harm or aggression- that may be associated with the psychosis.

Weighing up the potential harms with treatment with the consequent risks of not treating is easier for certain risks – if the patient is a higher risk to others or self and harder for others – such as for continued presence of what seems mild-moderate distress or impaired functioning. Deciding on necessity depends on how one values these different factors. Sometimes the conclusion of weighing up these factors is that no treatment is indicated. This type of weighing up these factors is the heart of clinical practice with the additional complication of over-riding autonomy.  Although Bartlett expresses the view that decisions about enforced treatment are more than medical decisions (Bartlett, 2012: p.367) these decisions take place within the context of a legal framework with limits on medical decision-making, often with guidance such as the Code of Practice in England and Wales to provide the principles that should guide the values used to underpin decision-making  with supervisory procedures such as appeals and Tribunals to oversee the use if compulsory treatments. Bartlett focusses on the right of autonomy but ignores other rights such as the right to life.

Bartlett’s own values are at a minimum highly sceptical about psychiatric treatments. He puts forwards his own view that “benefits of psychiatric medication are not necessarily clear-cut, and the risks can be significant” (Bartlett, 2011: p.517) despite the overlap in effectiveness and adverse effects with treatments used in general medicine as discussed above. Contrary to his own claim to want a “balanced reflection” (Bartlett 2011; p. 517) barely mentions any benefits but instead discusses at length adverse effects including a claim that antipsychotics kills more people that it saves which is contradicted by the research evidence that antipsychotics reduce mortality (Tiihonen et al, 2009). This illustrates the perils of relying on balanced accurate evidence from someone advocating a legal argument.

Bartlett describes psychiatric medication as “mind-altering drugs” (Bartlett, 2011: p.535) – a highly value-laden term that makes the medication sound frightening but just highlights conceptual confusion. After all, psychiatric medication is used to benefit people with mental health conditions whom are regarded as having unpleasant, distressing or impaired functioning of mind– hence the therapeutic aim would to be alter the state of mind in a beneficial way surely? Which begs the question what does Bartlett think psychiatric treatment should alter? The feet? These values may inform Bartlett’s views that only reversing the underlying causes of the condition or better (unspecified) outcomes but not reducing symptoms should be regarded as valid therapeutic necessity to justify enforced treatment. However, given most psychiatric treatments are justified on clinical grounds for reducing symptoms and does not reverse underlying mechanisms of mental disorders (many of which are unknown) (see discussion above) Bartlett’s proposals -perhaps influenced by his values towards psychiatric treatment - would rule out of bounds most psychiatric treatments as fulfilling therapeutic necessity criteria.

Therapeutic objectives apart from affecting underlying causes pf disorder or preventing outcomes – such as reducing distress or limiting deterioration – are regarded as legitimate in general medical practice. Psychiatric treatments overlap in effectiveness with treatments in general medicine and setting too high a threshold to justify enforced treatment would leave only electroconvulsive therapy as the only compulsory treatment option. Uncertainty of treatment response also occurs in general medicine. Necessity involves a balancing of the likelihood and nature of benefits and harms of proposed treatment versus the likely outcomes and risks of the patient’s clinical state. There are various laws, guidance and supervisory procedures to oversee the medical decision making. Bartlett’s statements on psychiatric treatments suggest his values are against medication which influence his criteria for therapeutic necessity.

4. Conclusion

Bartlett set out restrictive criteria for therapeutic necessity over-riding autonomy and justifying enforced treatment for mental disorder – affecting underlying causes of the disorder and /or affecting outcomes apart from reducing symptoms as well as requiring lack of uncertainty over benefits and expressed scepticism over psychiatric treatment in terms of effectiveness and adverse effects. He argued these are not solely medical decisions.

I argued therapeutic necessity is a conjunction of two terms – what is therapeutic and what is necessity. Therapeutic involves a broad series of potential therapeutic goals such as preventing occurrence of negatively perceived outcomes or relieving distress (Sackett et al, 1991: 189) rather than Bartlett’s narrow definition. Psychiatric treatments are in many ways similar to treatments in general medicine in terms of effectiveness, uncertainty over benefit and propensity to adverse effects. Bartlett’s scepticism over psychiatric treatments seem to be underpinned by values biased against psychiatric medication. Necessity involves judgements by clinicians as to the current severity and risk of adverse outcomes of the patient’s current clinical situation and that of the proposed treatment. These judgments and actions by clinicians are limited by statute with guidance on how to implement the laws and supervision of clinical practice for example by review tribunals.

References

Bartlett, P., 2011. ‘The necessity must be convincingly shown to exist’: Standards for compulsory treatment for mental disorder under the Mental Health Act 1983. Medical law review, 19(4), pp.514-547.

Bartlett, P., 2012. Chapter 14 Rethinking Herczegfalvy: the Convention and the control of psychiatric treatment in Brems, E. ed., 2012. Diversity and European human rights: rewriting judgments of the ECHR. Cambridge University Press.

Bloch, S.E. and Chodoff, P.E., 1991. Psychiatric ethics. Oxford University Press

Curtice, M., 2008. Article 3 of the Human Rights Act 1998: implications for clinical practice. Advances in Psychiatric Treatment, 14(5), pp.389-397.

Emanuel, E.J. and Emanuel, L.L., 1992. Four models of the physician-patient relationship. Jama, 267(16), pp.2221-2226

Fulford, K.W.M., Smirnov, A.Y.U. and Snow, E., 1993. Concepts of disease and the abuse of psychiatry in the USSR. The British Journal of Psychiatry, 162(6), pp.801-810

Hope, T., 2004. A very short introduction to medical ethics.Oxford University Press, Oxford

Huda, A.S., 2019. The Medical Model in Mental Health: An Explanation and Evaluation. Oxford University Press, Oxford.

Lieberman, J.A., Small, S.A. and Girgis, R.R., 2019. Early detection and preventive intervention in schizophrenia: from fantasy to reality. American Journal of Psychiatry, 176, pp.794-810

Leucht, S., Hierl, S., Kissling, W., Dold, M. and Davis, J.M., 2012. Putting the efficacy of psychiatric and general medicine medication into perspective: review of meta-analyses. The British Journal of Psychiatry, 200(2), pp.97-106.

Leucht, S., Helfer, B., Gartlehner, G. and Davis, J.M., 2015. How effective are common medications: a perspective based on meta-analyses of major drugs. BMC medicine, 13(1), p.253.

Matthews, E., 2000. Autonomy and the psychiatric patient. Journal of applied philosophy, pp.59-70.

Sackett D.L., Haynes R.B., Guyatt G.H. & Tugwell P., 1991 Clinical Epidemiology: A Basic Science for Clinical Medicine (Second Edition) Boston/Toronto/London. Little, Brown and Company

Tiihonen, J., Lönnqvist, J., Wahlbeck, K., Klaukka, T., Niskanen, L., Tanskanen, A. and Haukka, J., 2009. 11-year follow-up of mortality in patients with schizophrenia: a population-based cohort study (FIN11 study). The Lancet, 374(9690), pp.620-627.

Yeomans, D., Moncrieff, J. and Huws, R., 2015. Drug-centred psychopharmacology: a non-diagnostic framework for drug treatment. BJPsych Advances, 21(4), pp.229-236.