Decision Making in Recovery-Oriented Mental Health Care

Abstract

Objective

Patient-centered communication has been linked to patient satisfaction, treatment adherence and outcomes. Shared decision making (SDM) has been advocated as an important and ethically essential aspect of patient-centered care, but SDM has received relatively little attention in mental health care, despite studies indicating that consumers want to be involved in decisions. This is particularly important in a recovery-oriented system, where consumers are active participants in their treatment and rehabilitation. Because medication management is a key component of recovery from severe mental illnesses, this study explores how consumers and providers make decisions in medication management consultations.

Methods

Four providers (3 psychiatrists, 1 nurse practitioner) and 40 consumers with severe mental illness (10 consumers per provider) were recruited from a community mental health center with a recovery-oriented focus. We directly observed 40 medication management appointments. Observations were audio recorded and transcribed. We used emergent thematic analysis to characterize decision making processes.

Results

Providers initiated most decisions, although they often invited consumers to participate in decision making. Decisions initiated by consumers elicited a greater degree of discussion and disagreement, but also frequently resulted in consumers’ preferences prevailing. Consultations generally exhibited more characteristics of person-centeredness than SDM.

Conclusions and Implications for Practice

While we observed a high degree of person-centeredness, SDM was not prevalent. Interventions helping consumers to take greater initiative when working with service providers may be helpful. For example, programs using tools such as peer instruction, internet-based software, and individual case-manager instruction all have shown promise for enhancing SDM in mental health treatment. Further research is needed to determine the degree of SDM in other settings (e.g., with case managers) and the impact of SDM on consumers’ recovery.

Introduction

An increasing body of research confirms the importance of communication in health care and the impact that patient¹-centeredness has on patient satisfaction, adherence, and outcomes. Systematic reviews demonstrate that when patients and providers work together to identify problems, set goals, and make decisions, the result is greater satisfaction and trust, reduced emotional burden, and improved biomedical markers, such as blood sugar and blood pressure control (Rao, Anderson, Inui, & Frankel, 2007; Brown, Stewart, & Ryan, 2003; Stewart et al., 2000). Patient-centeredness has been defined by Mead and Bower (2000) as an approach that considers social and psychological factors in delivering care to an individual, regards the patient as a unique person who assigns personal meaning to his or her illness, and creates a therapeutic alliance in which patient and provider share both power and responsibility. Shared decision making (SDM) has been advocated as an important component of patient-centered care, relating directly to the sharing of power and responsibility. However, in spite of increased emphasis on SDM, the degree to which SDM is accepted—and actually put into practice—is unclear (Braddock, Edwards, Hasenberg, Laidley, & Levinson, 1999; Saba et al., 2006; Whitney, McGuire, & McCullough, 2004). This is especially true for mental health, which typically lags behind the broader medical field in embracing a vision of partnership (Adams & Drake, 2006). The purpose of this study was to explore how decisions are made in mental health consultations, with particular interest in characterizing the degree to which decisions were shared.

Defining Shared Decision Making

While patient-centered care comprises behaviors described above (Mead & Bower, 2000), SDM can be defined as a part of what it means to deliver patient-centered care. While SDM definitions vary, the most widely used definitions conceptualize SDM as an interactive process between at least two parties (patient and provider) in which the sharing of information and opinions occurs, patient preferences and provider responsibilities are discussed, and both parties agree on a course of action (Braddock, Edwards, Hasenberg, Laidley, & Levinson, 1999; Deegan & Drake, 2006; Duncan, Best, & Hagen, 2010; Makoul & Clayman, 2006; Montori, Gafni, & Charles, 2006; Towle & Godolphin, 1999; Whitney, McGuire, & McCullough, 2004). In mental health, SDM has been identified as important for psychiatric rehabilitation and person-centered care (Deegan & Drake, 2006; Drake, Deegan, & Rapp, 2010), and these authors have argued for the enhancement of SDM through tools such as decision aids and electronic decision support programs.

In order to more clearly specify what constitutes SDM, it is useful to examine Charles and colleagues’ (1997, 1999) model for shared decision making, which describes four essential elements: 1) both patient and provider must be involved; 2) both share information; 3) both express treatment preferences; and 4) a treatment decision is made, and both parties agree to implement the treatment. Expanding on this model, Montori, Gafni and Charles (2006) argued that shared decision making may be understood, and carried out, differently in chronic conditions, as opposed to acute health issues. For example, the majority of illness management in chronic care occurs outside of the provider’s office, in the patient’s family or community context. As well, decisions may be delayed, or modified, from visit to visit until a desired state or outcome is reached, without the delays or modifications leading to harm (Montori, Gafni, & Charles, 2006; Bodenheimer, Lorig, Holman, & Grumbach, 2002). Key in this model of SDM is the centrality of the relationship between patient and provider. Through information exchange and sharing preferences, both parties are afforded the opportunity to come to know each other, building trust and enabling patients to be treated as individuals (Charles, Gafni, & Whelan, 1997; Charles, Gafni, & Whelan, 1999; Montori et al., 2006). Such partnerships become especially important in situations where patients may have difficulty carrying out an agreed-upon treatment plan and need a trusting, supportive provider with whom to discuss challenges and plans for future successes (Montori et al., 2006).

While the models outlined by Charles and colleagues (1997) and Montori and colleagues (2006) present key elements of SDM, Whitney et al. (2003) have argued that the need for SDM is mitigated by the degree of risk accompanying a particular treatment, as well as the treatment’s likelihood of success. The higher the risk of side effects or other problems, and the greater degree of uncertainty regarding the treatment’s success, the more SDM is thought to be necessary.

Shared Decision Making in Community Mental Health and Psychiatric Rehabilitation

SDM is widely advocated for consumers with severe mental illnesses, particularly in a recovery-oriented system where consumers are viewed as actively engaged with their illness management (Drake, Deegan, & Rapp, 2010). Consistent with Whitney et al.’s (2003) assertions, decisions in this context often involve multiple treatment options, many of which have varying degrees of risks and side effects (Lehman et al., 2004; Hamann, Cohen, Leucht, Busch, & Kissling, 2005). Given that severe mental illnesses often require long term treatment and rehabilitation efforts, achieving commitment on the part of the consumer is critical (Montori et al., 2006; Karnieli-Miller & Salyers, 2009). Consistent with this view, several studies have found that consumers with severe mental illnesses, including depression and schizophrenia, are interested in being involved in decisions about their care (Adams, Drake, & Wolford, 2007; Loh, Leonhart, Wills, Simon, & Harter, 2007; Hamann et al., 2005) and may have specific ideas about what shared decision making looks like (Woltmann & Whitley, 2010).

Interest in SDM for people with severe mental illnesses is growing, as evidenced by a SAMHSA toolkit (Curtis et al., 2010), a special issue of the Psychiatric Rehabilitation Journal (2010, Volume 34), and newly developed tools and interventions (Deegan, Rapp, Holter, & Riefer, 2008; Deegan, 2010; Andrews, Drake, Haslett, & Munusamy, 2010). Despite this fact, surprisingly few studies have sought to learn, through direct observation, how decisions are made in consultations with mental health professionals—and whether these decisions are made through a shared process. Such research is essential if we are to understand how better to support consumers’ recovery (Woltmann & Whitley, 2010). Toward this end, we undertook a pilot study to gain firsthand understanding of decision making through direct observation. This particular study focused on medication management appointments, because effective medication management is considered to be a cornerstone of mental health treatment (Lieberman et al., 2005) and is also targeted in new SDM programs (Deegan et al., 2008; Deegan, 2010) as well as in other evidence-based psychosocial interventions (Dixon et al., 2010).

Method

Direct observations were conducted at a community mental health center in a medium-sized Midwestern city serving children and adults with severe mental illnesses at 200% of poverty level or below. The agency organizes treatment teams around the assertive community treatment (ACT) model and offers services to consumers who have a prior hospitalization history, homelessness, and/or incarceration. The first four prescribers contacted, who served adults, agreed to participate in the study (three psychiatrists and one nurse practitioner). A research assistant recruited 10 consumers per provider. During recruitment, agency staff introduced the assistant, who described the study and reviewed the informed consent statement. Consumers consented to be audio-taped during their medication check, and also completed a brief packet of surveys, and were paid $10. Providers, who were not paid, allowed us to audiotape the visit and supplied diagnoses and ratings of substance abuse and medication adherence. All procedures were approved by the university’s Institutional Review Board.

We approached 43 consumers to recruit a sample of 40 (a 93% participation rate). Two declined citing lack of time, and one declined for personal reasons. Participants had a mean age of 43.5 ±15.2 years, were predominantly Caucasian (31, 78%), and most had at least a high school diploma (28, 70%). About half were female (21, 53%). Diagnoses included schizophrenia/schizoaffective disorder (25, 63%), major depression (8, 20%), bipolar disorder (4, 10%), or other (3, 8%). Six (15%) had a co-occurring substance abuse disorder. Most were being served on state-certified ACT teams (29, 73%).

The 40 observations (one per consumer) were audio-taped, transcribed verbatim, reviewed for accuracy, de-identified, and imported into Atlas-ti for qualitative analysis. Using emergent thematic analysis, three team members listened to audio and reviewed transcripts to identify broad themes related to decision making. After analyzing several transcripts together as a team, team members then independently coded transcripts to identify types of decisions, who initiated them, how the final decision was made (e.g., mutual discussion, disagreement, or passive acceptance), and whose preferences were ultimately reflected in the final decision. Team members met weekly or bi-weekly as a group to discuss findings, resolve discrepancies, and refine coding. During analysis, we noted both decisions meeting the criteria outlined in Charles et al.’s (1997, 1999) model, and also instances in which possible decisions were discussed but no resolution was reached in the session. In addition, the length of each of the 40 consultations was recorded.

Results

We identified 132 decisions in 40 visits. The decisions fell into three domains: medication decisions (n=52), the consumer’s follow-up appointment (n=32), and “other” (e.g., whether a consumer should obtain lab tests or attend a recommended class; n=48).

For the majority of decisions, the provider initiated the decision making process (See Table 1). When consumers initiated, the decision most often began with a question (e.g., “Are you still wanting to take me off meds?”) or a concern (e.g., “I haven’t really heard any voices. I have had sort of like, it’s sort of like someone’s … talking to me, but I’m not getting clear words.”)

Table 1.

Decision-making in psychiatric visits

Type of Decision
	Medications	Next Appointment	“Other”	Total
Who initiates decisions
Provider	81%	100%	88%	88%
Consumer	13%	0	12%	10%
3rd Party	6%	0	0	2%
Decision-making process
Disagreement	2%	0	14%	6%
No Disagreement	62%	81%	72%	70%
Negotiation	36%	19%	14%	24%
Whose preferences are reflected in the final decision
Provider	61%	79%	57%	63%
Consumer	13%	14%	18%	15%
Both	9%	0	0	4%
3rd Party	6%	0	2%	3%
No Final Decision	11%	7%	22%	15%
Appointment Length
		Range	Mean	Median
Consumer-Initiated		9.5–30 min	19.8	19.9
Provider-Initiated		9–32 min	19.0	18.4

Outright disagreement was rare. The following exchange illustrates one such instance:

Provider: Are you talking about this stuff in group?

Consumer: I didn’t go to group last week.

Provider: I think that’d be good grist for the mill, for you to bring up in the group and get some feedback from the others. Be good for you to at least talk about it there.

Consumer: I know, but I don’t like crying and all that.

Provider: Well, some say it’s good to get it out.

Consumer: You know how miserable it is just to keep thinking that you’re coming forward and then find that you’re going backwards?

Provider: Well, crying when you talk about it doesn’t necessarily mean that you’re going backwards. It’s just holding it in and not addressing it may be going backwards more than addressing it and talking about it in your therapy. That would be an effort to make forward steps in your life. So, anyway, I would encourage you to do that.

Consumer: Okay.

More common was negotiation to arrive at a mutually agreed upon decision. These instances were characterized by both parties openly sharing opinions and/or concerns. Negotiation occurred most frequently in medication discussions. Just over a third of all medication decisions were arrived at through a process of information sharing and expression of treatment preferences (See Table 1). The following passage illustrates this shared process:

Provider: Well, we had talked about [medication] as a possible option to take at night, but as you know, that would involve the blood draws every week.

Consumer: Uh-huh.

Provider: And, of course, my ulterior motive is that I think might help some of those other symptoms you have of schizophrenia. Where are you with that?

Consumer: Schizophrenia…for the past few weeks, I’m feeling better. It’s let up and so…knock on wood. [shared laughter] ‘Cause I don’t want it to come back. It’s just kind of leaving.

Provider: Okay. You want to just not rock the boat right now then?

Consumer: Yeah.

Provider: Put this [medication] option on the back burner?

Consumer: Yeah.

Provider: Okay. Well, I’m glad it’s lettin’ up.

In this passage, the provider first gives information about a new medication, then expresses an opinion that the medication may also help with the consumer’s other schizophrenia symptoms. The consumer responds by sharing information about the improvement of the schizophrenia symptoms and their preference (that the provider picks up on) not to change anything at present. In the end, both parties reach an agreement not to “rock the boat.”

In terms of the final outcome, consumers agreed to providers’ stated preferences about two thirds of the time. The remainder of the decisions resulted in either consumers’ preferences being reflected in the final decision, or in no final decision being reached during the session. The only category in which both the consumer’s and provider’s stated preferences were reflected in the final decision was in decisions about medications; however, they comprised a minority of all medication decisions (See Table 1).

Decision making was also observed to be different depending on who initiated the process. When providers brought up a decision for deliberation, disagreement rarely occurred. On the other hand, when consumers began the process, disagreement occurred a third of the time. Interestingly, however, when the consumer initiated the decision making process, the consumer’s preferences were more frequently reflected in the final decision, either fully or through some negotiation. When the process was provider-initiated, the provider’s preference most frequently prevailed.

With a general overview of all decisions in mind, we now discuss in detail each of the three observed domains of decisions: medication decisions, decisions about the consumer’s follow-up appointment, and “other” decisions.

Medication Decisions

In most cases the provider initiated discussion regarding medication decisions (See Table 1). This was often accomplished through an agenda-setting question posed to the consumer, such as, “How are you feeling about that change we made in your medicine?” or “You okay with leaving all your doses and the medicines the same but just taking them later like this?” In other cases, a decision was initiated by the provider making a suggestion: “I would also start you…on a little bit of [medication] just to see if that doesn’t help with the shakiness in your hands and maybe some of the muscle stiffness in your back.” In some cases when the provider invited consumer participation in a decision, he/she explicitly left the final decision to the consumer: “I’ll do whatever you want to do. If you want to keep the [medication] as our main medicine, we could try it for some more months.”

When the consumer initiated discussion about medication, it was often a request. In the following exchange, the consumer requests a change in medication, to which the provider is initially resistant, but accedes to after the consumer provides a rationale:

Consumer: I wonder if you would add [medication] to my morning meds. You used to do that.

Provider: A, we may have tried that earlier and B, it’s not like it will help any because what really matters with drugs like [medication] is that you get it all in you every day. So you’ve been getting 40 mg every day. So just changing it around, taking part of it in the morning and part of it at night is unlikely to make a difference.

Consumer: I’m just having a little more stress than I normally do. There are more things that stress me out. I have more things stressing me, more uncertainty, which is not good.

Provider: Well we can try it if you want and see if it helps.

In another case a consumer initiated a decision by bringing up a problem with a medication.

Consumer: Um, I don’t need the…[medication]. It’s not working, so…

Provider: You wanna just bag it then?

Consumer: Yeah. Work on it on my own.

Provider: Okay.

Decisions about Follow-Up Appointments

Decisions associated with the consumer’s follow-up appointment were the most lopsided, with providers initiating all decisions with little or no disagreement and/or discussion. In most cases the preference voiced by the provider was the final decision.

When exceptions occurred, the provider initiated the discussion by asking the consumer when he/she wanted to return, and the consumer made the decision, as can be seen here:

Provider: How many weeks?

Consumer: Probably…I would say about three weeks, just in case.

Provider: Yeah. You may need some more support through all this.

Consumer: Yeah, just…you know, even just to see your face, it might help me, you know.

Provider: (laughing) All right. I will see you in three weeks.

Although scheduling a follow-up appointment may not always lend itself to a shared process, this excerpt illustrates that at times such a process can be used to strengthen and reinforce the therapeutic relationship (Charles, Gafni, & Whelan, 1997, 1999).

“Other” Decisions

Decisions were categorized as “other” when they did not fit into either of the above categories. Examples included basic issues, such as whether the consumer’s weight or blood pressure should be measured and whether certain laboratory tests should be performed. This category also included decisions specifically related to consumers’ rehabilitation efforts: whether a consumer should attend a support group on drug addiction, encouraging a suicidal consumer to attend group therapy, talking to a case manager about issues with scheduling other medical appointments, and steps to be taken to resolve personal issues in the consumer’s life, such as Medicaid issues, domestic problems, and financial concerns.

In spite of the variation in this domain, providers most often initiated the decision making process, although, interestingly, providers made the final decision just over half the time. In the remaining cases, either no decision was made in the session, or the consumer’s preference was reflected in the final decision (See Table 1). An example of a provider-initiated discussion, with no final decision being reached in the session, is seen in the following excerpt:

Provider: I may have mentioned this to you before. We might want to look at…if this has been going on for, like your wife has been saying, for years, you might be benefitted by having what’s called a sleep study.

Consumer: Yeah.

Provider: And just, long story short, it’s where you go in an evening and they, they hook you to, it’s called an EEG machine. It just kind of monitors what your brain is doing when you’re trying to sleep, and some people just have actual disorders of being able to fall asleep because something within the nervous system is not functioning quite right.

Consumer: Yeah.

Provider: So that might be something we need to look at because like I said before, you’re on a goodly amount of meds that should probably for most of us be putting us horizontal.

Consumer: Yeah.

Without any resolution, the provider moves on to a discussion of the consumer’s adherence to his medication regimen, and participation in the sleep study is not revisited in this consultation.

Discussion

The purpose of this pilot study was to gain an initial understanding of how decisions are made in medication management appointments for adults with severe mental illness and to identify the types of decisions made. As a result, this research provides a useful typology to understand decision making in recovery-oriented medication consultations. We identified three domains of decisions: those related to medications, the consumer’s follow-up appointment, and “other” decisions.

In most observed decisions, both parties shared opinions or concerns, and frequently arrived at an agreed-upon decision. Outright disagreement was rare. In spite of this, most of the observed decisions fell short of what constitutes SDM as defined by Charles et al. (1997, 1999) and Montori et al. (2006). For example, in most cases both parties contributed to discussion, and frequently both expressed opinions or preferences. Less common were the sharing of technical information (e.g., a full discussion of available choices and potential outcomes, relating research evidence to the consumer’s particular symptoms, and discussing benefits vs. side effects—see Montori et al., 2006) and mutual agreement to a course of action. In fact, consumers’ preferences were infrequently reflected in the final decision (see Table 1). This phenomenon may be due, in part, to the fact that this study was cross-sectional and many of the observed discussions might have been continuations from previous appointments, or might have been revisited in subsequent appointments not observed. Indeed, Montori and colleagues (2006) contend that a distinguishing characteristic of SDM in chronic care is that decisions typically do not have to be made immediately and may be deferred without resulting in harm for the patient or consumer. Although observed decisions did not frequently fit all criteria for SDM, we did witness a high degree of person-centeredness and consumer involvement in many of the exchanges between consumers and providers.

Providers initiated the majority of decisions. Often, however, the initiation included an invitation for the consumer to offer an opinion, rather than a direct expression of the provider’s own preferences. Invitations—directly requesting input from the other person—appear to be one way to involve consumers in decision making. These provider invitations are noteworthy in light of recent research (Goss et al., 2008), using the OPTION scale (a standardized scale to code SDM) to code 80 consultations, which found that psychiatrists made minimal attempts to involve consumers in decisions. While their methodology differed from ours, this disparity nonetheless suggests that further investigation is warranted.

We observed the greatest amount of discussion and negotiation for medication decisions. This observation is consistent with the notion that multiple medication options are often available, and psychiatric medications carry varying degrees of risk and side effects that should be discussed with consumers to assure the medication is the best fit for the consumer (Lehman et al., 2004; Hamann, Cohen, Leucht, Busch, & Kissling, 2005). In contrast to medication decisions, consumer involvement occurred least frequently with decisions about follow-up appointments, where the provider most often made the final decision. This may be because, in many cases, consumers have a previously agreed upon interval between appointments, and consumers (and providers) may not perceive this as a particularly important decision. Indeed, Whitney and colleagues (2003) would argue that such a decision carries little risk or uncertainty, and thus does not require a shared process. However, it could also be argued that consumer preference still ought to be elicited in order to be consistent with a recovery-oriented approach.

Although consumers were frequently involved in treatment discussions, providers’ preferences still tended to dominate, and in many cases decisions were made without any disagreement or negotiation between parties. This pattern suggests that SDM was less than optimal in this sample, and suggests a need for training providers and patients in shared decision making skills. In primary care, shared decision making interventions aimed at patients (Bieber et al., 2006) and providers (Sullivan, Leigh, & Gaster, 2006) have shown promise with respect to increasing satisfaction and coping for both parties. In mental health, programs and tools are just beginning to be developed to enhance SDM (Alegria et al., 2008; Andrews et al., 2010; Deegan et al., 2008; Deegan, 2010).

Decision making played out differently when initiated by the consumer. These decisions more often resulted in disagreement than provider-initiated decisions. At the same time, such decisions were associated with greater consumer involvement in the decision making process, ultimately leading to greater sharing of information and preferences, which in turn resulted in more instances of the consumer’s preference being reflected in the final course of action. Because so few decisions were initiated by consumers, it is difficult to draw any definitive conclusions from this trend. Nevertheless, this finding suggests that a greater degree of participation by both parties occurs when consumers initiate the decision making process, and corroborates the work of Young et al. (Young, Bell, Epstein, Feldman, & Kravitz, 2008), who observed the same phenomenon in primary care patients with depression. This is also consistent with interventions designed to help consumers take a greater role in the process (Alegria et al., 2008).

This study also illustrates that, even in cases where no final decision is reached and SDM does not occur, engaging in some elements of Charles et al.’s (1997, 1999) model of SDM has benefits. For example, in an excerpt quoted in the Results section, the provider tried to convince the consumer to make a commitment to talk about her issues in group therapy. Although she was resistant and the provider dropped the discussion, this attempt to reach a decision nonetheless prompted both parties to talk about potentially important issues related to the consumer’s life, treatment, and recovery. Thus, even in instances where SDM does not occur, person-centered communication that incorporates some elements of shared decision-making, has potential to direct communication toward a recovery orientation and enhance the consumer-provider relationship.

In addition, it is noteworthy that, even though these were medication management visits, we observed many instances of decisions that were not about medications. In particular, the “other” category comprised a number of decisions directly related to rehabilitation (e.g., support group attendance, case management issues, financial and domestic problems). This is important because it is possible that engaging in communication in which decisions are discussed and shared leads to a greater degree of person-centeredness, in which consumers and providers can move beyond medications and discuss other areas critical to recovery. Additionally, the occurrence of such discussions in these visits further demonstrates the importance and applicability of assessing and teaching shared decision making outside of medication management, in other key areas of mental health treatment.

As with any research, this study has some limitations. First, the majority of consumers in this study had schizophrenia and were on assertive community treatment (ACT) teams, which serve the most severely disabled and disadvantaged consumers. Our findings may in part reflect the fact that these particular ACT programs embraced a recovery-oriented practice of partnering with consumers, which is consistent with the ideals of person-centered care (Mead & Bower, 2000; Salyers & Tsemberis, 2007). This may also explain, in part, why our findings are inconsistent with Goss and colleagues (2008), who observed very little collaboration in the psychiatric consultations they studied. Second, because we observed one visit per consumer, we did not have the opportunity to ask each party what their views of the interaction were or to determine whether the decisions were actually implemented. Similarly, we were not able to assess the development of relationships over time or the length of the relationship between the two parties. This information could be useful because the manner in which consumers and providers engage in decision making might look different depending on the stage of the relationship (e.g., initial consultation versus a check-up after years of contact) or of the illness (e.g., extreme symptom exacerbation versus stability or growth). Given that people who experience severe mental illnesses often require longer treatment and rehabilitation efforts, it is likely that some of the observed decisions had been discussed in prior consultations, and will continue to be discussed in the future (Montori et al., 2006). This study will inform future research that will follow consumers and providers over time in order to see a fuller spectrum of decision making over the course of multiple consultations. As with other qualitative studies, our findings were not designed to generalize beyond the providers and consumers we observed. However, findings from this study will be useful to inform future studies of larger populations and that may also expand beyond medication management into areas such as case management.

This study adds to the growing literature on decision making in the treatment of severe mental illness and extends the applicability of Charles et al.’s (1997, 1999) and Montori et al.’s (Montori et al., 2006) models of SDM to the domain of mental health care. Relatively few studies have examined decision making in mental health via direct observation and analysis of how consumers and providers influence each other (Hassan, McCabe, & Priebe, 2007). This mutual influence is particularly important to examine in light of research suggesting that participation in medical visits is complex, dependent upon both parties and how they respond to each other (Street, Gordon, Ward, Krupat, & Kravitz, 2005; Young et al., 2008). Importantly, the findings from this study illustrate that even unsuccessful attempts at SDM have the potential to benefit the consumer and strengthen the patient-provider relationship. Moreover, the findings from this study suggest that, while we see evidence of person-centered care and consumer involvement, we see less evidence that SDM occurred in these consultations.

Although this study focused on medication management visits, our findings have important implications for the broader recovery-oriented mental health community. Greater knowledge of how decision making occurs with prescribers can potentially translate to other domains of mental health care, such as case management, as case managers seek to give consumers as much control as possible over the services they receive (Woltmann & Whitley, 2010). Direct observations of decision making, such as those in the current study, coupled with studies capturing consumers’ perspectives and understanding of decision making (e.g., Woltmann & Whitley, 2010), have the potential to enhance clinicians’ understanding of SDM and how best to achieve this goal. For example, a common technique observed in this study was the clinician explicitly inviting the consumer’s opinion—a tool that can effectively translate beyond medication management into a broader array of mental health services. In a similar vein, tools are being developed to help consumers and clinicians with shared decision making related to goal setting, supported employment, smoking cessation, substance abuse, and weight management (Andrews et al., 2010).

Although studies indicate that consumers with severe mental illness desire to take part in treatment decisions and in most cases are legally and mentally able to do so (Grisso & Appelbaum, 1995; Hamann et al., 2005), providers may be resistant to this practice without evidence of effectiveness (Chinman & Allende, 1999). Providers may view SDM as time consuming, and thus not feasible (Legare et al., 2010), and some research outside of mental health supports this view (Charles et al., 1999; Gotler et al., 2000). However, Loh and colleagues (Loh et al., 2007) conducted a shared decision making intervention in primary care with individuals who were depressed and found that sharing decisions did not require additional consultation time. In our own data, consultations in which the consumer initiated the decision making process, (i.e., those with the greatest degree of consumer involvement) were very close in length to consultations without consumer-initiated decisions, although our sample size was relatively small (See Table 1). The question of appointment length, clearly a barrier to SDM for some (Legare et al., 2010), still needs to be resolved. Future research is particularly important because, outside of mental health, SDM is associated with better health outcomes and greater satisfaction (Greenfield, Kaplan, & Ware, 1985). In the mental health field, SDM programs are just beginning to be developed and represent an important avenue for making positive changes that embrace a recovery-oriented system for consumers.