Complex wickedness

Difficult decisions are not new in medicine, often being framed in terms of comorbidity, individual variation or context-dependent factors. What is new is to explicitly differentiate between complicated and complex problems as described in complex system theory. ^{Reference Snowden and Boone6,Reference Rod, Broadbent, Rod, Russo, Arah and Stronks7} Complicated problems, although requiring expertise and careful planning, are relatively predictable and solvable whereas complex problems are dynamic, context-dependent and unpredictable, with outcomes emerging from interactions among multiple interdependent factors. Complex problems therefore necessitate a more holistic outlook, thinking about the entire system rather than isolated factors and acknowledging that small changes in one part of the system might produce unexpected effects elsewhere. Their nature lies in interdependencies, feedback loops and shifting patterns across multiple levels of influence and emergent effects. ^{Reference Öngür and Paulus4} Addressing complex problems requires adaptive strategies that are responsive to these aspects. ^{Reference Öngür and Paulus4} Beyond this, wicked problems, as described by Rittel and Webber, ^{Reference Rittel and Webber5} add another layer: they are not only complex but also inherently value-laden. Stakeholders often hold conflicting perspectives on what the problem is, how to address it and what success looks like. Furthermore, each attempt at solving the problem can have repercussions that cannot be undone, making interventions high-stakes. Wicked problems raise issues of power and questions of legitimacy and values: who gets to define the problem, and whose judgement determines which solutions are acceptable? These dimensions mean that addressing wicked problems requires more than technical expertise: it demands transparency and deliberation to preserve autonomy and justice. ^{Reference Speyer, Zhang, Lysaker and Roe8,Reference Speyer, Lysaker and Roe9}

The complexity of deprescribing

When intervening in complex systems, even well-intentioned recommendations can produce unintended consequences.

For example, current clinical guidelines often conceptualise ‘non-adherence’ in overly simplified terms, presuming a linear and predictable relationship between cause and remedy. ^{Reference McCutcheon, Pillinger, Varvari, Halstead, Ayinde and Crossley10,11} Within this framing, non-adherence is typically treated as a technical or behavioural deviation from the prescribed plan rather than as a complex, context-dependent phenomenon to be understood as often meaningful decision made by the person in treatment. The recommended strategies, such as psychoeducation, adherence monitoring or the use of long-acting injectables, reflect the belief that, if clinicians identify and overcome barriers and apply the gold standard (or evidence-based) intervention, adherence and consequently outcome will improve. This approach implies that the problem lies primarily in the patient’s behaviour or insight, not in the relational, contextual preferences or systemic factors that shape the decision to take medication. In this way, guidelines reduce non-adherence to a matter of compliance and control, reinforcing the notion that it can be solved through standardised, expert-driven action rather than via negotiated understanding.

However, in the real world, people discontinue medication not just once but repeatedly, ^{Reference Hansen, Speyer, Stürup, Hjorthøj, Öngür and Nordentoft12,Reference Wade, Tai, Awenat and Haddock13} and for many different reasons. Viewed through a complexity lens, attempts to solve inherently complex phenomena such as non-adherence through simplified, linear guidelines may generate unintended consequences within the wider system. By privileging standardisation and control, such approaches risk overlooking the dynamic interactions among patients, clinicians and contexts that give rise to patterns of adherence and non-adherence. Interventions based on reductionist models can therefore destabilise other parts of the system by eroding trust, diminishing patient agency or reinforcing patterns of disengagement that they were intended to correct: first, because guidelines are often treated as benchmarks in malpractice evaluations, and deviation from these can be justifiably perceived as professionally risky. This can foster a culture of caution and defensive practice, where strict adherence becomes a form of self-protection. Over time, this defensiveness may erode trust between clinician and patient, reducing the clinician’s willingness to support patient-led decisions – such as dose reduction or alternative care pathways – that fall outside standard protocols. A mechanism intended to ensure safety and accountability thus paradoxically can constrain collaborative, recovery-oriented care.

Second, guideline-driven caution can make clinicians hesitant to support a person’s desire to reduce medication. As a result, some individuals may attempt to taper or discontinue on their own, or to conceal their intentions from their care team. This behaviour increases the risk of adverse outcomes, while the clinician remains legally protected because they have adhered to the established guidelines. In other words, rigid guidelines adherence can unintentionally create a more precarious situation for the patient.

Third, supervised tapering or discontinuation can provide an important opportunity for patients to learn how to make informed decisions about their own care. Denying this opportunity undermines the development of self-management skills and the capacity to understand one’s mental health, potentially reducing autonomy and long-term recovery outcomes.

The ‘wickedness’ of deprescribing

The ‘wickedness’ of deprescribing lies in the divergent goals, values and perspectives of those involved. Normative tensions arise from value-laden judgements rather than from purely technical considerations:

First, both continuing and discontinuing medication involve inherent risks, but of different kinds: ongoing treatment may produce long-term adverse effects whereas discontinuation may increase the likelihood of relapse or destabilisation. Determining which risk is more acceptable is not a purely clinical decision – it reflects underlying judgements about well-being, quality of life and the individual’s tolerance for uncertainty. These judgements are shaped by values, cultural norms and assumptions about what constitutes a ‘good’ outcome in mental healthcare.

Second, clinicians must navigate the tension between respecting individual autonomy and safeguarding broader societal or relational interests. Supporting a patient’s preference to reduce or stop medication involves weighing their personal goals against potential consequences for others, including family members, caregivers or the public, in contexts where safety may be implicated. This balancing act highlights that decision-making in psychiatry is never solely about the individual patient – it is embedded within a network of ethical and social responsibilities.

Third, assessments of insight and decision-making capacity are inherently interpretive and context-dependent. Questions such as ‘insight into what?’ and ‘judged by whom?’ reveal that competence is not a neutral clinical measure but a socially and culturally mediated construct. Judgements about a patient’s competence are not neutral, but reflect normative assumptions about acceptable risk, responsibility and conformity to prevailing social and medical expectations. In practice, this means that evaluations of competence often reflect the values of the system or clinician as much as the clinical capacity, reasoning or ‘insight’ of the patient.

These tensions underscore that deprescribing is not merely a clinical or technical task, but rather a fundamentally ethical and value-driven challenge, highlighting the limits of prescriptive guidelines.

Clinical guidelines addressing complex wickedness

We believe there are myriads of clinical situations that vary in their degree of complexity and wickedness. We do not seek to reinforce a rigid binary distinction between types of problems, but rather to depict clinical challenges as existing along a spectrum of complexity and wickedness. Within this spectrum, some problems are relatively straightforward whereas others, like deprescribing, lie at the extreme, requiring navigation of uncertainty, conflicting values and unpredictable outcomes. To remain relevant, clinical guidelines must engage with both complexity and wickedness rather than reduce them to binary directives. Addressing complexity requires a transdisciplinary epistemic foundation that shares the power to define problems and solutions among stakeholders, rather than leaving guidelines development solely to professionals. Addressing wickedness calls for explicit recognition of the philosophical assumptions and normative factors embedded in decision-making.

To support these complex decisions, we propose a shift in how guidelines are framed and used. Rather than defaulting to risk containment or adherence-focused responses, clinicians should approach requests about discontinuation by exploring the patient’s perspective and engaging in collaborative deliberation about the risks, benefits and alternatives. ^{Reference Speyer, Ustrup and Ødegaard14} First, a clinician’s primary role is to inform patients; accordingly, guidelines should focus on providing clear, evidence-based information about the potential risks and benefits of continuing or discontinuing medication, rather than issuing prescriptive recommendations. Second, guidelines should offer tools to support ethical reasoning, helping clinicians recognise and navigate the normative tensions inherent in deprescribing, including balancing individual autonomy, patient safety and broader societal obligations. Third, shared decision-making should be reframed as shared deliberation, emphasising an ongoing, iterative process rather than a one-off determination, allowing decisions to evolve in response to changing patient circumstances and preferences. Finally, clinicians should have access to up-to-date knowledge and practical guidance on tapering strategies, including dose-reduction schedules and monitoring protocols, to maximise safety and minimise the risk of relapse or adverse effects.

In rapidly evolving areas, guidelines require more frequent revision. Developing guidelines as living documents, continuously updated as evidence evolves and clinical contexts shift, is one structural response to this challenge, ensuring that recommendations do not harden into outdated standards applied to problems that they no longer adequately represent. Furthermore, guidelines developed solely within professional and academic communities embed a particular epistemic standpoint as the authoritative standard against which all clinical behaviour is evaluated, systematically marginalising the experiential knowledge of those most directly affected. Addressing complex wickedness therefore requires that service users, carers, clinicians and researchers participate as genuine co-producers. Finally, meaningful deliberation presupposes equal access to knowledge. If guidelines content remains intelligible only to specialists, the conditions for genuine shared decision-making are undermined from the outset. Accessibility is therefore not a presentational consideration but a matter of epistemic justice, a prerequisite for ensuring that all stakeholders can engage as informed participants in the decisions that govern their care.