Towards the future of advance directives in psychiatry

Ben Harman-Jones
Famia Askari
Abubakar Isaac-Momoh
Olubunmi Adebowale
M A Sarfarz
Elizabeth Raybould Centre, Dartford


Initially intended to enhance patient autonomy in times of psychiatric crisis, the origins of the psychiatric advance directive (PAD) can be traced back to the days of the anti-psychiatry movement. In 1982, Thomas Szasz published his vision for the “psychiatric will” (1). Similar to the “living will”, the objective was to allow patients to express their wishes for or against specific forms of treatment in the event of loss of decision-making capacity during a psychiatric crisis.

There has been large variation in PAD implementation internationally in line with the degree to which healthcare professionals (HCPs) are required by law to adhere to their contents. This has been mandatory in parts of the U.S. since 1991 (2), and approximately 50% of all States now allow patients to create legally enforceable PADs (3). The U.K. has been much slower to consider this; Scotland made the legislation in 2003. The position in England is still ambiguous. It was not recognised in the 2007 review of the Mental Health Act (1983), therefore, some interpret the Mental Capacity Act (2005) in such a way that could allow legally-binding PADs (4). As a result of this ambiguity, decisions made in English PADs exist in an advisory context and ultimately HCPs must judge whether to follow them.

The interpretation of the PADs varies considerably. The simplest were those made in the U.S. in the 1990s where legal documents were created by the patient alone (5). Their uptake was very low, which prompted efforts to facilitate completion through supporting material. One study showed a resulting twenty fold increase in completion rate from the use of so-called “facilitated” PADs (6). Practice in the UK has since gone a step further and progressed the concept to the “Joint Crisis Plan” (JCP). Here, patients, HCPs and third parties such as carers agree decisions, representing the most intensive form of PAD yet seen (5).

In this review, we examine how PADs are viewed by the stakeholders who use them, trends in their usage internationally, and then compare these trends to the practice in our own Trust in Kent. We then return to the literature to examine the impact PADs have had on patient outcomes, before finally examining the appropriateness and utility of PAD in modern mental health care.

Current trends in PAD usage

The views of the stakeholders who use PADs have been studied extensively, and there is a clear divide in their expectations of the benefits. Nicaise et al (5) have examined aspects of PAD classification, content and implementation. They view the PAD as a “multistage intervention” rather than a one-off document, due to its varying performance across different objectives, which are of differing relevance to different stakeholders (see Figure 1). This model forms a helpful framework with which to classify PADs, which we will use throughout this review.

They found that patients prefer legally-binding documents, and PADs are viewed positively as an advocacy tool to help facilitate autonomy in their relationship with HCPs. In contrast, HCPs appear more reluctant to subscribe to the same unilaterally beneficial view. Concerns raised include potential reduction in professional autonomy, possible use to refuse all treatment, medico-legal issues, and practical difficulties in implementation. Psychiatrists also point towards an interesting potential dilemmas where the use of PADs may entrench a “worried well” behaviour with less  likelihood of need to use one, whereas sick patients who need one would be less inclined to make one (7). There is also an indication that HCPs view PADs more as a tool to facilitate the care planning process rather than with the original intention of facilitating autonomy.

Figure 1     The “multistage intervention” model aligns the varying capabilities of different forms of PADs against the opportunities they present for different stakeholders. Adapted from Nicaise et al (2013)

Several studies show that, given the choice, one half to two thirds of patients with severe mental illness would like to write a PAD if they were given the help to do so (8-12). So, in the presence of high expectation and desire by the patients to complete a PAD, are they being used? In the U.S., studies show only 4-13% uptake among public sector outpatients (11, 12).

In three audits completed in Kent and Medway NHS and Social Care Partnership Trust (KMPT), from 2009 to 2014, we observed levels of appropriate PAD uptake consistent with that observed in the U.S. (Shanmugham & Picasa, 2009; Adebowale, 2013; Adebowale & Sarfraz, 2014). The results and details of the patients sampled are given in Table 1.





Details of patients sampled

Patients in the community under enhanced CPA

Patients in the community under enhanced CPA

Patients admitted across 2 inpatient wards

Number of patients sampled




Proportion of patients with a documented PAD




Table 1

Data from three audits (Shanmugham & Picasa, 2009; Adebowale, 2013; Adebowale & Sarfarz, 2014) examining PAD usage across Kent and Medway NHS and Social Care Partnership Trust.
Care Programme Approach (CPA)

These consistent findings raise the question of why PAD completion rate is so low. A lack of staff and patient awareness is the first observation; data from our Trust shows that 79% of inpatients surveyed did not know about writing a PAD. Among inpatient mental health nurses, only 49% had received formal training on their use, and only 69% were aware of the Trust policy. The policy (2014) recommends that all patients should be supported to develop a PAD as part of the CPA process.

While these findings may go some way to explain this low PAD uptake, we believe this is a myopic explanation. We now concentrate on variations in how the benefits of PADs make them more suited to enhancing certain parts of care than others – an issue we feel goes much further to explaining the low rate of PAD uptake observed internationally.

The benefits of PADs

A Cochrane review examined two UK-based randomised controlled trials comparing outcomes in adults with severe mental illness depending on PAD usage (13). The authors concluded that the use of PADs led to no significant difference in care in terms of hospital admission or use of outpatient services, or compliance with treatment, indicating a lack of efficacy in facilitating autonomy. However, those using PADs required less contact with social workers, and had lower numbers of violent incidents. This is interesting because the types of PADs used in the constituent trials were f-PADs and JCPs; consistent with Nicaise et al (5) that these more intensive forms of PAD are more conducive to enhancing the therapeutic alliance between patients and HCPs, rather than facilitating patient autonomy (see Figure 1).

Similarly, in 2006 Swanson et al reported that patients using f-PADs had a greater working alliance with HCPs than those not using them, and they were more likely to believe their mental health services were meeting their needs (6). It was also later reported that f-PADs reduced the chance of a patient undergoing a coercive intervention by 50% (14).

It is therefore clear that PADs have the potential to produce benefits beyond the original intention of facilitating autonomy, a finding consistent with the multistage intervention model (Figure 1).

Ethical framework of PADs

The use of advance directives in psychiatry was spawned from the use of similar documents in physical health care; specifically, their use to communicate prior decisions at the end of a person’s life. There are, however, crucial and large differences in the circumstances affecting decisions made in end-of-life (EOL) care and those made in the treatment of an acute deterioration in mental state. These considerations are not necessarily reflected in PADs, the ethical foundations of which are still firmly rooted in EOL care.

The differences between decision making in EOL care and in psychiatric crisis are explored in Table 2. From an HCP perspective, the value of the advance directive is generally to permit them to act in a patient’s best interests when the patient is no longer able to decide for themselves. In the context of EOL care, “best interests” would usually be to not perform aggressive treatment, such as cardiopulmonary resuscitation (CPR), because the chances of a return to baseline are low (15). In direct contrast, “best interests” in psychiatric crisis may actually be to perform aggressive treatment, such as electroconvulsive therapy (ECT), because the chances of a return to baseline may be high (15)

Data from KMPT showed that out of the 6 PADs where contents were available, 3 spoke about treatment preference, and 3 spoke about treatment refusal. These findings show a trend towards anecdotal evidence from international practice that PADs are more often used to refuse treatment than to provide consent (16). There is clearly more to be done to ensure the decisions made in PADs are aligned to assisting HCPs in providing patient-centred care during psychiatric crisis.

There is evidence that the way in which advance decisions are interpreted varies considerably depending on the patient’s reasoning behind their decisions. One survey of HCPs showed that only 22% would follow a decision to refuse all medication due to paranoid delusions about them, whereas 72% would follow the same decision made instead due to concerns about side-effects (17). Due to the nature of mental illness, there is a therefore a disproportionately high chance that PADs are overridden compared to other advance directives.

Table 2 summarises the differences in decision making between EOL care and psychiatric crisis, which together may contribute to a high likelihood of override in psychiatric crisis compared to EOL care (15).


EOL care

Psychiatric crisis

Common dilemma facing clinicians

Not providing aggressive treatment that is unlikely to return a patient to their baseline (e.g. CPR)

Providing aggressive treatment that is likely to return a patient to their baseline (e.g. ECT)

Prior patient experience to inform decision

Not possible


Decision between capacitous/


Easy: Communication likely to be lost at same time as other faculties

Hard: Communication likely to be retained after other faculties are lost, leading to possibility of outward communication of disordered thought

Likelihood of override of advance decision




The original need that PADs were introduced to meet – to facilitate the autonomy of patients upon loss of capacity in a psychiatric crisis – is underserved by the way in which PADs are commonly being used in practice. As we have shown, the uptake of PADs is much lower than originally intended, and this experience is consistent with practice in KMPT. There are a number of potential reasons for this, and unsurprisingly, failings in training and awareness are frequently deferred to as recommendations for improvement. While this is welcome, it is our view that the reasons for low PAD uptake are far more pervasive and difficult to disrupt.

In contrast to advance directives in EOL care, which are well-suited to communicating the refusal of unnecessarily aggressive treatment; to be useful in psychiatry, PADs must be able to communicate consent regarding a necessarily aggressive treatment. The assumption that the same ethical framework that advance directives for EOL care are built on is equally appropriate for psychiatric crisis is profoundly misguided. High potential override rates (reasons for which are detailed in Table 2) confirm that even when a PAD is made, there is a low likelihood it will be used, hence giving insight into why perhaps the efficacy of PADs in facilitating autonomy is low. Further, perhaps HCPs assume this, explaining why uptake is also low.

This disparity is contradictory, because patients with mental health problems are likely to have a high level of insight into what it is like to suffer a deterioration, so are in a much stronger position to give informed consent than anyone is about EOL care (after all, who can claim to have experienced CPR?). Yet why are we willing to accept such EOL refusals as “competent”, and not accept equally (if not more) competent permissions in psychiatry? Patients are also generally interested in using advance decisions to aid HCP decision making, rather than increase complexity of care by refusing treatment (5). Clearly, the desire to improve communication is present on both sides – and we believe that uptake and outcomes can be improved if delivery methods are modernised.

We identified evidence demonstrating that more modern and intensive forms, namely the f-PAD and JCP, are more suited as “multistage interventions”, rather than being one-shot directives (5). The use of f-PADs and JCPs has been shown to correlate with lower social worker contact and lower numbers of violent acts (13) – evidence that the more intensive PADs (which are most widely used) may be more suited to enhancing the therapeutic alliance and improving partnership working (Figure 1). Swanson et al (6, 14) confirmed such benefits of f-PADs, with a striking reduction in the chance of coercive intervention.

Based on these observations, it is our view that the potential of the PAD in modern mental health care goes far beyond its original goal. What was originally intended to be a simple intervention is now used in practice to add layers of nuance to the CPA – as is the case in our own Trust. However, we also recognise the reasons why the PAD movement began, and believe it would be wrong for the specific goal of enhanced autonomy in psychiatry to be lost in this merger. Instead, we believe the concept may be used to both empower patients and assist clinicians simultaneously if the ethical uniqueness of psychiatric crisis is defined, recognised, and adopted, as we explain in the final section.


Firstly, we recommend fully embracing the best practice communication principles used in the intensive forms of PADs (f-PADs and JCPs) by fully incorporating them into CPA protocols. In our Trust this would be best-placed as part of the existing enhanced CPA process (where certain aspects of PADs have already been incorporated). The potential benefits of this would be to improve the quality of the therapeutic alliance between HCPs and patients, and also to improve the efficiency of mental health services by improving partnership working.

Secondly, we are keen to preserve the potential for PADs to maximise patient autonomy. However, in order to be clinically valuable and meaningful in psychiatry, any such process must recognise the uniqueness of the challenges in psychiatric crisis. A solution may benefit from inspiration from the so-called “Ulysses contract” widely described in the literature, named after the Greek myth, where Ulysses ordered his men to tie him to the mast of his ship and not allow him to succumb to the Sirens’ song, which would have meant unknowingly throwing himself to his death.

Next steps

Improve ability of CPA to enhance therapeutic alliance and partnership working

    1. To review the elements of our Trust’s existing enhanced CPA protocol and identify inclusion and exclusion of elements of the intensive PADs
    2. To recommend changes to enhance its potential to fully deliver the best practice communication principles promoted by intensive PADs
    3. Any changes will be systematically communicated to HCPs, patients and carers throughout the Trust, along with this review, explaining reasoning and expected benefits
    4. To perform a retrospective audit 6 months after changes are made to determine if uptake rates have improved
    5. To perform a snapshot cross-sectional study to understand the views of stakeholders involved in use of the enhanced CPA protocol

Explore the appetite for a dedicated process to plan for psychiatric crisis that is clinically useful and unlikely to be overridden

    1. To invite HCPs, patients and carers from our region to a series of focus groups to discuss the issue and idea
    2. To discuss the limitations of current approaches, as detailed in this review, and to gain input on their views towards a new process that embraces the challenges of psychiatric crisis


  1. Szasz TS. The psychiatric will: A new mechanism for protecting persons against” psychosis” and psychiatry. American Psychologist. 1982 Jul;37(7):762.
  2. The Patient Self-Determination Act (PDSA) 1990. United States Congress.
  3. Davis JK. How to justify enforcing a Ulysses contract when Ulysses is competent to refuse. Kennedy Institute of Ethics Journal. 2008;18(1):87-106.
  4. Weller PJ. “Advance Directives and the Translation of Human Rights Principles in Mental Health Law – Towards a Contextual Analysis”, paper presented at the Quarterly Mental Health Consumer, Carer and Community Forum of the Australian Capital Territory Mental Health Community Coalition, Canberra, 12 May 2008.
  5. Nicaise P, Lorant V, Dubois V. Psychiatric advance directives as a complex and multistage intervention: a realist systematic review. Health & social care in the community. 2013 Jan;21(1):1-4.
  6. Swanson JW, Swartz MS, Elbogen EB, Van Dorn RA, Ferron J, Wagner HR, McCauley BJ, Kim M. Facilitated psychiatric advance directives: a randomized trial of an intervention to foster advance treatment planning among persons with severe mental illness. American Journal of Psychiatry. 2006 Nov;163(11):1943-51.
  7. Swanson JW, Tepper MC, Backlar P, Swartz MS. Psychiatric advance directives: An alternative to coercive treatment? Psychiatry. 2000 May 1;63(2):160-72.
  8. Backlar P, McFarland BH, Swanson JW, Mahler J. Consumer, provider, and informal caregiver opinions on psychiatric advance directives. Administration and Policy in Mental Health and Mental Health Services Research. 2001 Jul 1;28(6):427-41.
  9. Noble LM, Douglas BC. What users and relatives want from mental health services. Current Opinion in Psychiatry. 2004 Jul 1;17(4):289-96.
  10. Srebnik DS, Russo J, Sage J, Peto T, Zick E. Interest in psychiatric advance directives among high users of crisis services and hospitalization. Psychiatric Services. 2003 Jul;54(7):981-6.
  11. Swanson JW, Swartz MS, Hannon MJ, Elbogen EB, Wagner HR, McCauley BJ, Butterfield MI. Psychiatric advance directives: A survey of persons with schizophrenia, family members, and treatment providers. International Journal of Forensic Mental Health. 2003 Apr 1;2(1):73-86.
  12. Swanson J, Swartz M, Ferron J, Elbogen E, Van Dorn R. Psychiatric advance directives among public mental health consumers in five US cities: prevalence, demand, and correlates. Journal of the American Academy of Psychiatry and the Law Online. 2006 Jan 1;34(1):43-57.
  13. Campbell LA, Kisely SR. Advance treatment directives for people with severe mental illness. The Cochrane database of systematic reviews. 2009(1):CD005963.
  14. Swanson JW, Swartz MS, Elbogen EB, Van Dorn RA, Wagner HR, Moser LA, Wilder C, Gilbert AR. Psychiatric advance directives and reduction of coercive crisis interventions. Journal of Mental Health. 2008 Jan 1;17(3):255-67.
  15. Olsen DP. Ethically Relevant Differences in Advance Directives for Psychiatric and End-of-Life Care. Journal of the American Psychiatric Nurses Association. 2016 Jan;22(1):52-9.
  16. Hobbs P, Hobbs P. The limitations of advance directives and statements in mental health. Australasian Psychiatry. 2007 Jan 1;15(1):22-5.
  17. Wilder CM, Elbogen EB, Swartz MS, Swanson JW, Van Dorn RA. Effect of patients’ reasons for refusing treatment on implementing psychiatric advance directives. Psychiatric Services. 2007 Oct;58(10):1348-50.
  18. Adebowale O & Sarfraz M. A. Audit 715/14. 2014.
  19. Mental Capacity Act 2005 (UK).
  20. Mental Health (Care and Treatment) Act 2003 (Scot), ss 275, 276.
  21. Mental Health Act 1983 (UK).
  22. Pisaca M & Adebowale. Audit 589/12. 2013.
  23. Shanmugham S, Pisaca M. Audit 190/08. 2009.