Adapting Shared Decision Making for Individuals With Severe Mental Illness

Shared decision making has been encouraged in medicine for decades (1,2) and has recently attracted attention in the mental health field (3). Shared decision making is regarded as an intermediate model between a paternalistic approach (the doctor has all the information and decides alone) and the informed-choice model (the doctor informs the patient , and the patient decides alone). It aims to change the asymmetry between patients and their doctors regarding decisional power and decision-relevant information. It increases patient autonomy without leaving the patient alone in the decisional process (4).

Shared decision making is highly favored by many patients, clinicians, and researchers in mental health settings for various reasons, including ethical, practical, and utilitarian (5–7). In addition, shared decision making may play an essential part in other models, such as patient centeredness or recovery (8). Studies have shown that shared decision making applied in mental health settings yields positive outcomes (3).

Nevertheless, shared decision making is not widely implemented in routine clinical practice (3,9). For the patient group most vulnerable to paternalistic or even coercive treatment—those with severe mental illness, acute psychotic symptoms, and poor insight—and for the most controversial decisions (such as medications), the field is still far from thorough implementation.

Barriers among professionals

Clinicians in somatic medicine cite various barriers to applying shared decision making (10), such as time constraints. Many psychiatrists go even further, classifying specific situations and patients (such as those experiencing an acute psychotic episode) or distinct decisional steps (such as hospital admission) as generally unsuitable for shared decision making (11). Consequently, psychiatrists often act in a paternalistic way or apply pressure to patients (12). Although most mental health professionals acknowledge the advantages of shared decision making (6,10), some consider the approach to be nonbeneficial (11). Furthermore, some psychiatrists are convinced that they already apply shared decision making—although there is evidence that this is often a misconception (13)—and therefore see no need to reconsider their behavior.

However, we believe that the most prominent hurdle is psychiatrists’ fear of using shared decision making in critical situations (those in which making wrong decisions might lead to poor outcomes) or with certain patients (those who lack insight, have poor health literacy, or have impaired decisional capacity) (14). For example, the decision to start antipsychotic treatment of a psychotic patient who lacks insight may cause psychiatrists to fear that they will worsen the patient’s outcome if they do not pressure the patient to take the medication. The psychiatrist may think, “If I leave it up to the patient, he would certainly choose not to initiate treatment. Symptoms would persist or even worsen, and thus I would harm the patient. If I apply pressure and he accepts antipsychotics, he may respond to treatment and likely gain insight. Then he will later be thankful that I proceeded in the way I did.”

Although most psychiatrists will be familiar with such deliberations and will have achieved some (short-term) success by taking this approach, it has several shortcomings. First, it mistakes the physician’s role in shared decision making as a passive one—that is, it confuses shared decision making with the informed-choice model (4). Second, it ignores the experiences of generations of patients who have suffered from pressure and loss of autonomy (15). Third, use of this approach implies that psychiatrists have no other option but to apply pressure when patients are not immediately convinced to take medications. Finally, it ignores evidence that the paternalistic approach leads to poor long-term compliance.

In summary, many psychiatrists may simply lack communication skills—or may not use them—to deal with challenging patients or clinical situations and may thus use a paternalistic approach as the easiest way to solve problems.

Barriers among patients

Many patients are rather passive, creating a barrier to shared decision making in clinical practice (13). Others may not be interested in participation for various reasons (16). Many may not see a need—for example, because they are satisfied with their treatment (17). Others may believe that they already actively participate (13), because they interpret the option of saying no as active participation. In addition, “classical” shared decision making is asking a lot of the patients because it requires skills ranging from communication competence to information processing and evaluation (18). For some psychiatric patients, these barriers may be insurmountable because they are impaired by depressive, negative, or cognitive symptoms. Moreover, patients who have experienced directive decision making or even coercion have often lost the motivation to actively initiate and support shared decision making (Hamann J, Kohl S, McCabe R, et al., unpublished manuscript, 2014). Finally, many patients have low self-esteem regarding their own competencies, for example in regard to health literacy and decision making (19). Some patients may fear negative consequences when they act in a self-confident and self-reliant manner with their psychiatrist.

SDM-PLUS

We believe it is necessary to complement the classical approach to shared decision making with other models, interventions, and communication techniques in order to empower mental health professionals and patients and enable them to use shared decision making. We call this integrative approach SDM-PLUS.

Empowering professionals

We suggest that mental health professionals need to be empowered to reduce their use of pressure in clinical situations. The first step is to ensure that mental health professionals can analyze clinical situations and classify them into three prototypical decision types that imply different decision-making approaches (Figure 1). First, there are the very rare “life-or-death” situations in which a wrong decision (for example, discharging a suicidal patient) imposes an immediate threat to the patient or to others. Second, there are preference-sensitive decisions—those with several more or less equivalent options or in which there is uncertainty about the best way to proceed—that constitute most decisions in mental health care. Finally, there are “best choice” decisions, in which professionals agree on the therapeutic approach but anticipate patient resistance. Undoubtedly, life-or-death situations call for a directive approach; however, shared decision making should be used in all other situations.

Further, the individual situation will determine the best approach to shared decision making. For preference-sensitive decisions, the classical approach is advised in mental health settings. This emphasizes describing options, communicating risks and benefits, and identifying patient preferences (20). More challenging are best-choice decisions because neither the classical approach, which requires some agreement between patient and provider, nor the paternalistic approach, which increases patient resistance, will work if, for example, a patient with poor insight refuses antipsychotic treatment during an acute psychotic episode. In these situations, we suggest use of communication strategies, such as motivational interviewing (19,21) and techniques developed by the Harvard Negotiation Project (22). Motivational interviewing is already used in mental health settings. The Harvard Negotiation Project, developed at Harvard Law School, constitutes a method for negotiating mutually satisfactory agreements. The aim is to reach win-win solutions that go beyond compromising, even in very difficult negotiations (23). Communication techniques used in motivational interviewing and in the Harvard Negotiation Project overlap—not only with each other but also with shared decision making (24). Thus it is possible to identify a set of common communication skills that facilitate shared decision making in best-choice decisions. We refer to these skills as SDM-PLUS.

In the situation described above in which a psychotic patient is ambivalent about or refuses medications, a set of these techniques can be used. First, the provider analyzes the situation (rule out life-or-death situation) and concludes that the patient’s participation is the goal. The provider also concludes that classical shared decision making would not be effective as an initial approach because of the patient’s resistance and decides that SDM-PLUS techniques will be more effective, flagging the situation as a best-choice scenario. Second, the provider should actively start the negotiation process with the patient (as suggested by the Harvard Negotiation Project). The negotiation process includes strategies such as taking time-outs, identifying one’s own interests (that is, the option that is seen as the best choice by the psychiatrist), and developing an alternative plan (identifying other potential solutions if the patient does not accept the best-choice option). This step, although it seems self-evident, is one of the crucial elements of successful negotiation and is referred in literature as BATNA (Best Alternative To Negotiated Agreement) strategy (23).

Third, throughout the course of the consultation, strategies adapted from motivational interviewing should be implemented to avoid confrontation, including listening reflectively, evoking change talk, and exploring common ground (19,21). In addition, supplementary use of some techniques of classical shared decision making (such as assisting patients to identify or develop their individual preferences) may be helpful in reaching mutual decisions.

Thus we argue that SDM-PLUS techniques are necessary in clinical situations in which there is a lack of agreement between the patient and the mental health provider, because such agreement is the basis for using the classical approach to shared decision making. SDM-PLUS does not replace the principles of shared decision making (4) but rather serves as a platform to implement shared decision making in clinical situations involving patients with severe mental illness. Practical training in SDM-PLUS techniques is needed for mental health professionals (psychiatrists, nurses, social workers, and psychologists). Because many professionals have some experience with the techniques described above (24), the training should focus on analyzing clinical situations and applying communication strategies in routine clinical care.

Empowering patients

At all times it is necessary to empower patients in order to facilitate their participation in decision making. Ideally, the goal is to enable patients to demand their preferred level of participation. Existing interventions should be used to create a “participatory atmosphere” that helps to motivate, empower, and enable patients to participate in shared decision making. These interventions include social skills training that addresses communication skills in the context of mental health care to facilitate “self-directed recovery” (25) and psychoeducation (26) to increase self-esteem and health literacy. Health literacy can also be improved by use of decision support systems that educate patients to prepare for certain decisions (27). In addition, some programs specifically address patients’ activation, motivation, and skills to facilitate shared decision making (28,29). Finally, implementation of shared decision making can be promoted through incorporation of consumer-driven approaches and peer support that address shared decision making directly (30) or indirectly by addressing recovery or related fields (31). To create a participatory atmosphere, we believe that collaboration between the various professions (for example, nurses, psychiatrists, and social workers) is crucial for patient empowerment.

Limitations

SDM-PLUS has limitations. First, the methods that we propose should be integrated into shared decision making are not complete. A variety of alternative methods and techniques can enrich shared decision making and foster its implementation (32). Second, some aspects of SDM-PLUS seem paternalistic at first sight—for example, when we suggest that clinicians (not patients) decide whether a decision is preference sensitive. However, we view patient participation as the clear goal in all decisional situations. Therefore, decision analysis by the clinician is not intended to prevent some patients from engagement but rather to guide clinicians in implementing patient participation most effectively. Finally, to reduce complexity, our approach does not address the role of patients’ families and the influence of patients’ sociocultural background, although we acknowledge both the importance of involving family members and the need to be responsive to cultural characteristics.

Conclusions

Shared decision making has found its way into mental health care to a limited extent only, and especially “challenging” patients do not yet benefit from this approach. To foster implementation of shared decision making for this patient group, we suggest an integrative approach to shared decision making—SDM-PLUS—for the mental health field. We hope that our proposal will serve as a basis for further discussion about wider implementation of shared decision making in the mental health field.