‘A beginning is the time for taking the most delicate care that the balances are correct.’ (From the manual of Muad’Dib by the Princess Irulan, in Dune, by Frank Herbert, 1965)
The long-term care and support of people experiencing mood disorders is a central responsibility of psychiatrists, general practitioners (GPs), and other mental health professionals. This care is based on evidence, the patient’s experiences and the outcomes of treatment to date. One important decision is whether, and if so, when, to discontinue an antidepressant and/or any adjunctive pharmacotherapy. Reference Looi, Allison, Bastiampillai, Kisely, Maguire and Woon1,Reference Malhi, Bell, Szymaniak, Boyce and Looi2
The latest Delphi-informed guidance from the American Society of Clinical Psychopharmacology (ASCP), Reference Goldberg, Swartz, McIntyre, Mago, Malhi and Jha3 by Goldberg and colleagues, on the deprescribing of core psychotropic medications in major depressive disorder and bipolar disorder is therefore a welcome contribution. In this Editorial, we discuss the implications of the ASCP recommendations, while also drawing on the most recent and relevant British, 4 Australian and New Zealand Reference Malhi, Bell, Bassett, Boyce, Bryant and Hazell5 clinical practice guidelines (CPGs) for mood disorders that provide additional advice regarding how deprescribing guidance can be best implemented in clinical practice.
Prescribing and deprescribing in context
At the beginning of diagnosing a mood disorder and/or developing a management plan, there is an inevitable ‘pause’ in clinical practice decision-making wherein the numerous complex factors involved in making a diagnosis and treatment decisions are considered. This is especially the case for a first presentation in primary care. This period, defined as ‘watchful waiting’, entails a longitudinal evaluation of the patient over time while gathering necessary information to make an informed diagnosis and management strategy. Reference Malhi, Bell, Stavdal, Wu, Naylor and Kumari6 It is during this time that psychoeducation should occur, and a discussion regarding deprescribing should be initiated alongside consideration of prescribing options. This should include discussion of psychological therapies, lifestyle measures, projected duration of treatment, the potential for medication side effects, and their management and other adjunctive strategies that can, in combination with medication, promote recovery. Reference Malhi, Bell, Bassett, Boyce, Bryant and Hazell5,Reference Malhi, Bell, Stavdal, Wu, Naylor and Kumari6
There are several reasons why deprescribing should be raised at the outset when deliberating treatment commencement. First, having a clearly defined projected treatment end-point, and determining this from the beginning, facilitates patient engagement and motivation for an adequate intervention trial. Second, discussing deprescribing as an integral component of psychoeducation allows patients to understand the need for a well-considered and clinician-led withdrawal – as opposed to suddenly or randomly stopping medication. It also provides an opportunity to explain the important differences between any initial adverse effects that may occur when starting an antidepressant and withdrawal symptoms that might arise depending on how medication is stopped, and further allows communication of how to differentiate both these sets of symptoms from those which may reflect a recurrence of the illness. Finally, understanding deprescribing equips patients with an appreciation of why it is important to effectively monitor their response to treatments, and especially at the time of initiation and cessation thereof.
Any discussion regarding prescribing and deprescribing should involve personalised collaborative care-planning that encompasses the goals and needs of patients, often broadly referred to as preferences, while noting that options will necessarily vary depending on circumstances. Critically, it is important for clinicians to understand the psychological relationship that patients have with medication and in particular the significance of prescribing and deprescribing, both of which have implications for their identity, sense of self and autonomy. Deprescribing, for example, may raise concerns not only of a relapse in symptoms but of risking a patient’s sense of self by potentially disrupting a hard-won equilibrium, stability in relationships and employment. On the other hand, continuing medication is also an active choice for many, wherein they contend with daily decisions and choose to persist despite adverse effects or other burdens. Thus, the decision to prescribe or not involves a dialogue between the clinician and the patient that accounts for competing uncertainties which can be challenging to accept. It is important therefore to ensure that discussions regarding deprescribing are aimed at optimisation and improving a patient’s functioning, to ensure patients feel supported in their care and minimise a sense of abandonment or giving up on treatment.
Practical approaches to deprescribing
With the understanding that decisions regarding deprescribing are framed by patient experiences of illness and recovery, as well as their expectations and wishes, a few practical aspects of deprescribing can be considered, drawing on advice from the ASCP guidance and extant literature that we have captured broadly in Fig. 1. We have also commented on areas that warrant further investigation and recommendations, noting that the field of psychopharmacology deprescribing is evolving and our understanding is growing as more interest is being taken in this component of management.
Medication responses and effects (top panel) and deprescribing strategies for mood disorders (bottom panel). The top panel illustrates an example of the temporal relationship between responses to and effects from medications as indicated by the y-axes. The initiation of an antidepressant may coincide with initial adverse effects (first curve) including headache, nausea and gastrointestinal issues, which while acute typically resolve quickly (a matter of days). If only a partial response is achieved from this first antidepressant, then this may be withdrawn and switched to an alternative antidepressant. During this process, antidepressant withdrawal symptoms (ADS) may occur as indicated by the curve and box. These symptoms typically commence a few days after withdrawing an agent and in the case of switching agents, typically resolve as the new agent is titrated. As the individual achieves a full response and recovers, after some time the antidepressant may be withdrawn and again during this time ADS may occur, and/or the individual may experience a relapse of their mood disorder. The bottom panel illustrates the deprescribing strategies that can be implemented for major depression, bipolar disorder and for adjunctive treatments in mood disorders. Each context entails several questions/characteristics (bold) that may be relevant to the patient and under each details the recommended strategies for deprescribing given these factors. AD, antidepressant; GI, gastrointestinal; SSRI, selective serotonin reuptake inhibitor.
Fig. 1 Long description
A diagram illustrating medication responses, effects, and deprescribing strategies for mood disorders. The top panel shows a graph depicting the temporal relationship between responses to and effects from medications. The y-axis represents response and adverse effects, while the x-axis represents time. The graph shows the initiation of an antidepressant, partial response, switching antidepressants, full response, discontinuation of the antidepressant, and potential relapse. Adverse effects such as headache, nausea, gastrointestinal issues, dizziness, vertigo, nervousness/irritability, vivid dreams/nightmares, and diarrhea are indicated. The bottom panel outlines deprescribing strategies for major depression, bipolar disorder, and adjunctive treatments in mood disorders. It includes questions and characteristics relevant to the patient and recommended strategies for deprescribing given these factors.
Antidepressant deprescribing in major depression
Naturally, it is necessary that the strategies that need to be employed allow deprescribing to fit within practical management. The approach of watchful waiting while bedding down the diagnosis and planning management means there is an opportunity to consider longer-term goals, such as deprescribing, which will inevitably occur in patients that respond to treatment. Reference Malhi, Bell, Stavdal, Wu, Naylor and Kumari6 During these early discussions regarding management, patients should be briefed on the potential for discontinuation and withdrawal symptoms, and these should be monitored when switching or tapering off antidepressants – a process that must occur in consultation with their treating doctor. Reference Malhi, Bell, Bassett, Boyce, Bryant and Hazell5 Specifically, the most common symptoms of antidepressant withdrawal should be reviewed and these include dizziness, nausea, vertigo, nervousness and irritability, as well as gastrointestinal symptoms and sleep disturbances. Reference Malhi, Bell, Stavdal, Wu, Naylor and Kumari6 The likely timing of these should also be discussed (Fig. 1).
In practice, a common consideration is ‘when to stop an antidepressant’; and the obvious solution is to ‘stop when the response is inadequate’ and the perceived need for longer-term pharmacotherapy to prevent relapse is deemed low or otherwise manageable without continued prophylactic pharmacotherapy. However, clinically this is seldom clearcut and often the response to medication is only partial, and even an initially promising response may not be sustained. In these less straightforward circumstances, it is important to reconsider the diagnosis, review the medications that have been prescribed to date and, specifically, how they have been administered, and assess the implementation and effects of any additional psychological and other interventions. Clearly, if for instance there are medications with redundant pharmacological modes of action, such as two noradrenergic agents being prescribed for the management of depression, then the clinician should give consideration to gradually ceasing one of these or substituting this strategy altogether with an alternative approach, such as using a broader action antidepressant or determining if augmentation would likely yield better efficacy with satisfactory tolerability (Fig. 1).
Once an antidepressant has been deemed effective, for patients with a remitted nonpsychotic major depression for example, the question of whether to discontinue treatment may be considered once a sustained response has been achieved for 6–12 months, as the risk of relapse after this time is thought to be low. For long-term relapse prevention of major depression, indefinite antidepressant treatment is recommended after three or more lifetime episodes of depression. Also, discontinuation of an antidepressant should be considered if a previously effective agent loses its efficacy despite attempts at dose optimisation (Fig. 1).
Deprescribing in bipolar disorders
In patients with bipolar disorders, there has been longstanding debate about the use of antidepressants, and therefore over the past 2 decades recommendations regarding their prescription have varied considerably. Therefore there has been uncertainty regarding the withdrawal of antidepressants in bipolar disorder and withdrawal of medications overall. This treatment decision is complicated further by the fact that mood stabilisation in bipolar disorder is often a long-term endeavour. Recommendations for withdrawal include stopping or reducing antidepressants in patients with emergent mania or mixed-mood symptoms, or those who have a pattern of rapid cycling bipolar disorder. However, this has to be balanced against the recrudescence of illness and so, as recommended by ASCP guidance, antidepressants should not be stopped in cases of bipolar disorder where cessation has previously led to relapse (Fig. 1). The presence of rapid cycling, manic and mixed-mood features would further encourage antidepressant deprescribing.
The ASCP guidance recommended that it was seldom, if ever, appropriate to aim for medication-free status for patients with bipolar disorder where they have experienced mania (Fig. 1). If the withdrawal of lithium mood stabiliser treatment was to be enacted, then tapering over at least six weeks was recommended when there was no replacement mood stabiliser being considered. Generally speaking, decisions regarding the use of mood stabilisers and their differential benefits and risks are complex. Future guidance might consider specific recommendations for the different medications that are putatively regarded and prescribed as mood stabilisers (i.e. consider deprescribing certain anticonvulsants or other psychotropics meant to treat mood symptoms that fail to produce intended effects).
Deprescribing of adjunctive treatments in major depression or bipolar disorder
It is recommended that antipsychotics are deprescribed if there are adverse effects such as weight gain or tardive dyskinesia, and also potentially after 6 months of stability in nonpsychotic mood disorders (Fig. 1). In addition, it is also recommended that antipsychotics be deprescribed after sustained stability following an acute manic episode, if the antipsychotic was used alongside an antimanic mood stabiliser. However, further research and recommendations are needed as to which antipsychotics have the best benefit-to-risk ratio for mood disorders with and without psychosis, and what the optimal duration of adjunctive antipsychotic medication may be in a mood disorder patient with no history of psychosis or severe mania.
Advice regarding benzodiazepines and benzodiazepine agonists was somewhat clearer and it was recommended that these ideally be deprescribed within 1–3 months to prevent long-term dependence, and low-dose quetiapine used to promote sleep should also be deprescribed if taken for more than 3–6 months (Fig. 1).
What the consensus guidelines provide, and what more is needed
It is important to note that while the ASCP consensus statement provided guidance as per the above clinical scenarios, it also identified a significant proportion of circumstances wherein it was not possible to reach consensus such as, for example, which patients are best suited for deprescribing antidepressants once a sustained response and stability are achieved. Reference Goldberg, Swartz, McIntyre, Mago, Malhi and Jha3 This inability to achieve clear guidance reflects the fact that our understanding of deprescribing is still burgeoning and signifies the need for future research. We therefore highlight some specific areas where we feel additional guidance would be particularly helpful for clinicians and patients to collaboratively and effectively plan long-term care of mood disorders.
First, further research and advice on the recognition of core features indicating relapse of mood disorders versus emergence of symptoms indicative of withdrawal would be beneficial. Progress is being made in this regard, and the timeframe within which symptoms usually occur may help distinguish discontinuation, which typically occurs within the first few weeks of withdrawal from a relapse, which typically takes longer to manifest. Reference Kalfas, Tsapekos, Butler, McCutcheon, Pillinger and Strawbridge7 Moreover, the nature of the symptoms seems to differ, with low mood and suicidal ideation tending not to occur within the context of discontinuation. Reference Kalfas, Tsapekos, Butler, McCutcheon, Pillinger and Strawbridge7 Second, despite methodological challenges, the interaction between deprescribing and psychological treatments merits further research, particularly regarding whether deprescribing benefits people with mood disorders undergoing psychotherapy. And finally, it would be helpful to have further research and guidance on the duration and dosing of adjunctive antipsychotic medication, especially for manic and depressive episodes with or without psychosis.
While the above questions highlight specific clinical issues that require examination, there are also a number of guiding principles that have emerged from ASCP and international resources which should be carried forward into future guidance. For example, while the nature and format of the mode of deprescribing of psychotropic medication is still evolving, and the emerging recommendations focus on various decision points, there remains a need for additional practical information. This could include: tapering schedules for dosage reduction, the utility or necessity of cross-tapers when switching medication and decisions to consider alternative medication (such as temporary conversion to longer half-life serotonergic antidepressants, or as outright replacements). Therefore, future CPGs for mood disorders should update algorithms and pathways to include when and how to deprescribe medications, based on evidence and consensus.
In addition, it is critical to include the suggestion that deprescribing should be discussed at the outset of treatment in future guidance. This is because it informs patients that pharmacotherapy is intended to be a focused strategy for designated symptoms and for a specific duration, and that it is based on the outcomes of treatment. This can help to engage patients as decision-making partners in gauging the evolution of treatment success and relapse risk. The process of deprescribing antidepressants also provides an opportunity to discuss broader issues such as functional recovery and the psychological impact of the illness as well as the anticipation and prevention of recurrences. This facilitates discussion of psychological therapies, lifestyle measures and other adjunctive non-pharmacological interventions that are important in the management of mood disorders: as treatments during acute episodes of illness, to provide prophylaxis and reduce the risk of recurrence. Reference Malhi, Bell, Stavdal, Wu, Naylor and Kumari6
The future of clinical guidance
Traditionally, CPGs have been regarded as a valuable resource for psychiatrists, trainees and other health professionals, as they have a number of advantages including improving the quality and consistency of patient care having based their recommendations on reliable and up-to-date evidence, thereby promoting effective treatments in clinical practice, as well as their use as educational or professional development resources. However, CPGs also have limitations compared with other approaches used to synthesise evidence. Reference Kisely, Herrman, Chamoli, Hemachandra, Lawn and Bajurny8 These include the potential for bias, as recommendations may depend on who was involved in their development and how consensus was reached. Experts’ views, their academic or demographic backgrounds as well as received training might all play a role in shaping their decisions and pose limitations to reproducibility of the findings and their generalisability to diverse clinical settings or subspecialties. Reference Goldberg, Swartz, McIntyre, Mago, Malhi and Jha3 Recent years have seen the advent of alternatives including meta-reviews, umbrella reviews and overviews of systematic reviews, and while exact definitions vary, all share the common advantages of reproducibility, transparency and the possibility of greater standardisation. Reference Kisely, Herrman, Chamoli, Hemachandra, Lawn and Bajurny8 These alternatives may therefore be preferable for evidence synthesis and appraisal, while CPGs are more appropriate for implementation guidance. Reference Quek, Reus Perello, Lee, Abraham, Adams and Almeida9 Indeed, as CPGs typically bring together a broad range of expertise, they are perhaps best suited to addressing general clinical questions that also depend on contextual factors such as access. For example, more knowledge is needed to help determine when psychotherapy or other psychosocially based treatments may be preferable to pharmacotherapies for certain patients in managing potentially chronic or recurrent conditions such as depression, and here CPGs may be particularly useful in providing guidance.
In the process of implementing prescribing and deprescribing guidance, each patient and their clinician will need to consider how the evidence and advice, such as it is, is applied in their best interests to effectively manage their illness pharmacologically and facilitate access to non-pharmacological and adjunctive interventions. These decisions need to be considered from the beginning, with input from all parties, to ensure the balance of risks and benefits is maintained.
Acknowledgements
J.F.G. and G.S.M. were part of the Delphi process and authors on the ASCP article referenced in this Editorial. G.S. is currently enrolled as a Doctor of Philosophy student.
Author contributions
J.C.L.L., S.K. and G.S.M. all contributed to the conceptualisation and original draft. E.B., K.S., G.S., J.F.G. and A.N. edited the manuscript. S.K. and E.B. developed the figure. A.N. contributed as a person with lived experience.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
Declaration of interest
G.S.M. is the College Editor for the Royal College of Psychiatrists and the Editor-in-Chief of BJPsych but did not take part in the review or decision-making process of this paper. J.C.L.L. is the Editor-in-Chief of Australasian Psychiatry. S.K. is the Editor-in-Chief of ANZJP. He is a member of BJPsych editorial board but did not take part in the review or decision-making process of this paper. This paper reflects the views of the authors, not the journal or any other organisation that the authors are affiliated with. J.C.L.L. has received funding as a consultant on a US National Institutes of Health grant. E.B. has received joint grant funding from the University of Sydney and the National Taiwan University (Ignition Grant), grant funding from The North Foundation and grant funding from the Greek Young Matrons’ Associations (GYMA). S.K. has received grant or research support from the Australian Research Council, the National Health and Medical Research Council and Cancer Australia. K.S. and G.S. received grant funding from GYMA. J.F.G. has been a consultant to Abbvie, Alvogen, Bristol Myers Squibb, Genomind, Luye Pharmaceuticals, Neumora and Otsuka; has served on the speakers’ bureau for Alkermes, Axsome, Bristol Myers Squibb, Johnson & Johnson and Vanda Pharmaceuticals; and has received royalties from American Psychiatric Publishing and Cambridge University Press. G.S.M. has received support from Wellcome for GALENOS, grant or research support from the National Health and Medical Research Council, Australian Rotary Health, NSW Health, the American Foundation for Suicide Prevention, the Ramsay Research and Teaching Fund, Elsevier, AstraZeneca, Janssen-Cilag, Lundbeck, Otsuka and Servier; and has been a consultant for AstraZeneca, Janssen-Cilag, Lundbeck, Otsuka and Servier. He is the recipient of an investigator-initiated grant from Janssen-Cilag (PoET Study), joint grant funding from the University of Sydney and National Taiwan University (Ignition Grant) and grant funding from The North Foundation.
eLetters
No eLetters have been published for this article.