|Home | About | Journals | Submit | Contact Us | Français|
Patient- or consumer-centeredness has been recognized as a critical component of quality in primary health care, but is only beginning to be recognized and studied in mental health. Among the first opportunities to be consumer-centered is collaboratively producing an agenda of topics to be covered during a clinic visit. Early agenda setting sets the stage for what is to come and can affect the course, direction, and quality of care.
To study agenda setting practices among 8 prescribers (5 psychiatrists and 3 nurse practitioners) at the beginning of their encounters with 124 consumers diagnosed with schizophrenia spectrum disorders (56%), bipolar disorder (23%), major depression (15%), and other disorders (6%).
We modified an extant agenda setting rubric by adding behaviors identified by a multi-disciplinary team who iteratively reviewed transcripts of the visit openings. Once overall consensus was achieved, two research assistants coded all of the transcripts. Twenty-five transcripts were scored by both raters to establish inter-rater reliability.
We identified 10 essential elements of agenda setting. Almost 10% of visits had no agenda set and only 1 of 3 encounters had partial or complete elicitation of a single concern. Few additional concerns (4%) were solicited and no encounter contained more than 6 essential elements.
Collaborative agenda setting represents a unique opportunity to translate the concept of consumer-centeredness into mental health care. Initial results suggest the rating system is reliable, but the essential elements are not being utilized in practice.
Patient-centeredness has been recognized as a critical component of quality in primary health care (Mead & Bower, 2000) but is only recently beginning to be recognized and studied in mental health care, which has not kept pace with the broader medical field’s focus on partnership (Adams & Drake, 2006). Such an approach, in which consumers become active participants in their own care, is essential to facilitating recovery for persons with severe mental illness (Karnieli-Miller & Salyers, 2010; Torrey & Drake, 2010).
The term “patient-centered care” was first introduced by Levenstein, McCracken, McWhinney, Stewart, and Brown (1986) in a family medicine context and built on the conceptual framework of Engel’s (1977) biopsychosocial model. In the context of mental health services, the equivalent terminology for patient-centered care is consumer-directed or consumer-centered care, which we use hereafter in this paper. Consumer-centered care takes into account individual consumers’ social and psychological needs, regards consumers as unique individuals who assign personal meaning to their illness(es), and fosters the concept of shared power and responsibility between health care providers and consumers (Mead & Bower, 2000).
One area of consumer-centered care that has received a good deal of attention is shared decision making (SDM). The literature on SDM focuses almost exclusively on decisions about treatment options and the extent to which patients are aware and can make informed treatment choices (Joosten et al., 2008; Charles, Gafni, & Whelan, 1997). Far less attention has been paid to other opportunities for being consumer-centered. One such opportunity is the first few moments of a medical or psychiatric visit in which consumers and providers collaboratively discuss the topics or agenda to be covered. These early agenda setting decisions set the stage for what is to come in the visit and can affect the course, direction, and quality of care, including treatment decisions.
The opening moments of many types of social interactions are important. For example, Malcolm Gladwell in his book, Blink (2005), asserts that on the basis of the first 30 seconds of a casual social interaction, raters can predict a variety of outcomes, for example, teaching effectiveness (Ambady & Rosenthal, 1993). Clinical interactions are no different. In studies of non-verbal behavior, researchers have found that the tone of voice used by a clinician early in the visit predicts satisfaction and follow up to treatment recommendations (Milmoe, Rosenthal, Blane, Chafetz, & Wolf, 1967; Roter, Hall, Blanch-Hartigan, Larson, & Frankel, 2011). Likewise, using “thin slice analysis,” a technique for sampling discourse at fixed intervals (typically every 20–30 seconds), researchers have been able to correctly identify physicians who have, and have not, been sued for medical malpractice (Ambady et al., 2002).
Agenda setting can take a variety of forms. For example, physicians can “control” the agenda by asserting what topics will be covered without soliciting input from the patient. Conversely, consumers can sometimes control the agenda by stating the topic(s) they want to cover at the outset of the visit. A consumer who does not wait for a greeting or solicitation of a “reason for the visit” from the physician but launches into the main concern is an example of a consumer-controlled approach to setting the agenda. Finally, agendas can be set collaboratively with each party contributing ideas about what is important to cover in the visit and negotiating whether and when these ideas will be discussed. This style of agenda setting comes closest to being consumer-centered because it is based on shared power and control (Mead & Bower, 2000).
Evidence suggests that collaboratively setting an agenda at the beginning of a clinic visit increases consumer-centeredness in a number of ways. Consumer (and physician) satisfaction increase (Roter et al., 1997; Williams, Weinman, & Dale, 1998), there is less premature hypothesis testing on the part of the physician (Beckman & Frankel, 1984), consumers feel empowered, and the approach yields more information from which physicians can make appropriate diagnosis and treatment recommendations. In addition, early agenda setting results in fewer “hidden” concerns at the end of the visit, resulting in a more efficient overall visit (Beckman, Frankel, & Darnley, 1985). Finally, agenda setting at the beginning of the visit has been associated with clinical outcomes such as the resolution of chronic headache at one year follow up (The Headache Study Group of the University of Western Ontario, 1986).
Despite the documented benefits of increasing consumer-centered care, agenda setting may be challenging to accomplish in a busy clinical environment where time and throughput are paramount. In particular, issues of time and communication constraints have been highlighted in psychiatric care (Torrey & Drake, 2010). Time pressures may lead physicians to assert control over the visit in the belief that this is the most efficient way to “get the work done.” Likewise, being busy may lead physicians to forego checking for patient comprehension of recommended treatments and other decisions made during the visit (Braddock, Edwards, Hasenberg, Laidley, & Levinson, 1999).
In primary care, evidence of physicians’ exerting early control over the visit agenda comes from studies indicating that physicians interrupt patients within 18–23 seconds of their opening statement in response to a solicitation of agenda items (Beckman & Frankel, 1984; Marvel, Epstein, Flowers, & Beckman, 1999). This research has also documented a statistically significant relationship between early interruption and “late breaking” concerns that are raised at the very end of the visit (Beckman et al., 1985).
Although agenda setting has been explored in primary healthcare (Mauksch, Hillenburg, & Robins, 2001; Brock et al., 2011), little attention has been devoted to this practice in mental health care. In pediatrics and geriatrics, where patients may not be able to speak for themselves or have cognitive impairments, the evidence suggests that greater inclusion of the patient in communication/interaction (both verbal and non-verbal), is associated with more positive outcomes (Greene, Majerovitz, Adelman & Rizzo, 1994; Pantell, Stewart, Dias, Wells & Ross, 1982). Given the widespread acceptance of a recovery-oriented model for persons with severe mental illness, some of whom may have cognitive impairment or difficulty articulating their concerns, one would expect to find many of the elements of agenda setting present in visits between health care providers and consumers with severe mental illness. The purpose of the current study was to examine the extent and quality of agenda setting in a sample of mental health consultations.
All study procedures were approved by the Indiana University Purdue University Indianapolis Institutional Review Board. This was a cross-sectional, secondary analysis of transcribed audio-recorded psychiatric visits combined from three studies of adults in community mental health centers in two Midwestern cities (n=128; Salyers et al., 2012). In two of the samples, consumers were approached sequentially on the day of their visit and were included in the study if they gave consent. The third sample came from a randomized trial of a decision making intervention prior to any intervention. Four transcripts did not capture the beginning of the visit where agenda setting typically takes place, leaving a total sample of 124 for the current analyses. The sample included 8 prescribers (5 psychiatrists and 3 nurse practitioners) and 124 consumers in 3 different community mental health centers. The consumer sample included people with schizophrenia spectrum disorders (56%), bipolar disorder (23%), major depression (15%), and other disorders (6%). The consumer sample was predominantly Caucasian (51%), followed by African-American (44%), or another race (5%). Approximately half were male (51%), and the mean age was 43.2±10.5 years. Data on consumer education level and provider experience were not collected.
To develop a rating system for agenda setting, we began with the rubric developed by Brock et al. (2011). To determine if additional behaviors might be related to agenda setting, a multi-disciplinary team (three doctoral level researchers – one from sociology, health communication, and clinical psychology, respectively; and two research assistants - a doctoral student in clinical psychology and an undergraduate psychology major) used an immersion/crystallization approach to analyzing the data (Crabtree & Miller, 1999). The team read individual transcripts, applied codes, and met to discuss applicability of the concepts to the transcripts. This was an iterative process of individual coding followed by consensus discussions. Once we achieved consensus on the final codebook, two research assistants coded all of the transcripts. The same two research assistants coded 25 of these transcripts in common to establish inter-rater reliability.
We evaluated inter-rater agreement between the two raters for each of the individual elements of agenda setting. Kappa was not computed because in many cases, consensus occurred 100% of the time, preventing the computation of a value. In addition, for some items there was little variance (because the element was almost always missing). In light of these difficulties, we report percent agreement. We examined the distribution of the item scores, and combined agenda setting items into an index for an overall score.
Through our coding process, we developed 19 elements that were rated as either absent/present or absent/partial/complete (See Table 1). Based on descriptions of agenda setting by Brock et al. (2011), Mauksch et al., (2001), and Marvel et al., (1999), we distinguish between elements that appear most indicative of agenda setting (essential elements) from other elements that are more descriptive of the context (descriptive elements). Percent agreement was above 70% in all cases; for 8 out of 10 essential elements, agreement was at 90% or above.
We created a summary index to examine the number of essential elements present at all in an encounter. That is, if an element was at least partially complete, the encounter was scored as having that element present. As shown in Figure 1, the majority of encounters had evidence of one of the 10 essential elements. This was most commonly establishing rapport (see Table 1). Strikingly, almost 10% had no evidence of any of the agenda setting items, and no encounters had six or more elements present.
Despite the emphasis on consumer-centeredness in health care, agenda setting has received very little attention in the literature, particularly in mental health care. Agenda setting represents the first opportunity in an encounter to be consumer-centered; this sets the stage for the rest of the visit and may create an atmosphere more conducive to collaboration and partnership so critical to effective management of chronic health conditions (Bodenheimer, Lorig, Holman, & Grumbach, 2002; Michie, Miles, & Weinman, 2003).
It could be argued that agenda setting for consumers with severe mental illness might be different for people with severe mental illness than in other populations. For example, patients who are severely depressed might feel “put upon” by multiple attempts to establish an agenda at the beginning of their visits. While this is an empirical question that has not been answered to date and is certainly worthy of further study, the literature on chronic diseases, including depression, in the general medical population suggest that “activated patients” who ask more questions and raise additional concerns have better psychosocial and biomedical outcomes. For example, Greenfield, Kaplan, and Ware (1985) used a simple 20 minute coaching intervention to increase the amount of question asking for patients with hypertension, diabetes, ulcer disease and breast cancer. With the exception of the breast cancer patients who trended toward living longer, the activated patients with hypertension, diabetes, and ulcer disease had significantly better biomedical and functional outcomes. There is also evidence that older patients who are traditionally thought of as being less active and more reticent and who may find it difficult to articulate concerns, can be taught to be more active in their medical encounters with positive results (Cegala, Post & McClure, 2001). Recently, a coaching intervention has been successful for mental health consumers as well (Alegria et al., 2008).
This study presents an initial attempt to capture whether agenda setting is occurring during mental health visits. Our coding rubric was reliable, with good agreement between raters on the presence of different elements of agenda setting. However, overall agenda setting in this sample was quite low. Of particular interest is the very low occurrence of three of the first four elements of agenda setting (orienting the consumer to the day’s visit, eliciting a statement of consumer concerns, and eliciting the full spectrum of concerns at the beginning of the visit), since these elements have been shown to be critical in establishing a partnership between provider and consumer (Fortin, Dwamena, Frankel, & Smith, 2012). In primary health care visits, evidence shows that once interrupted, patients rarely bring up additional problems and concerns at the beginning of the visit (Beckman & Frankel, 1984). Instead, these concerns tend to surface at the end of the visit (Beckman et al., 1985). There is also evidence that full elicitation of patient concerns adds less than a minute to overall visit length and is associated with positive biomedical and psychosocial outcomes (Stewart, Brown, & Weston, 1989). Likewise, patient satisfaction is higher and propensity to sue for medical malpractice in the face of an adverse outcome is lower when patients feel listened to (Beckman, Markakis, Suchman, & Frankel, 1994; Levinson, Roter, Mullooly, Dull, & Frankel, 1997; Mauksch et al., 2001).
On the positive side, the providers in our sample did attempt to create rapport with consumers, a communication behavior that is key to creating a safe and welcoming patient-centered atmosphere (Frankel & Stein, 2001). Beyond creating rapport, the visits we studied reflect a model of care in which consumers had little say in setting the agenda for what they might want to accomplish in their visit. It is especially notable that two of every three consumers were not asked if they had concerns they wanted to discuss. Although 31 (25%) of consumers brought up concerns prior to elaborating on a topic, this left 51 (41%) who may not have expressed, or been directly asked about, additional agenda items. Furthermore, only 11% of consumers explicitly stated that all of the agenda items they wanted to discuss were brought up prior to elaborating on a topic. These physician-focused visits may represent a more traditional paternal style of communication in which the provider assumes control over the agenda on behalf of the consumer, a style described by Szasz and Hollender (1956) and typical of doctor-patient relationships in the 1950’s.
The introduction of the biopsychosocial model in 1977, patient-centered care in 1985, and the Institute of Medicine Report, Crossing the Quality Chasm in 2001, all describe a partnership model of the medical encounter that contrasts sharply with the traditional physician-centered model. The consumer-centered model is one in which consumers are active participants in their own care and where individual goals and values are actively sought and respected. Using consumer-centered care as a “gold standard” for high quality healthcare (a finding of the Institute of Medicine in its 2001 report), our findings suggest that there is a significant improvement opportunity for professionals who provide health care to consumers with severe mental illness when it comes to agenda setting. Teaching agenda setting skills to physicians can be done in as little as 3.5 hours, with evidence that skills are put into practice with positive results (Rodriguez et al., 2008). Moreover, once learned, engaging in agenda setting typically adds less than a minute to the average ambulatory visit (Stewart et al., 1989).
Caution should be observed in understanding our results. First, this was a convenience sample of a small number of mental healthcare providers who were participating in three different studies not originally designed for this purpose. As a result, we were unable to examine the data in a more sophisticated manner (i.e., examining provider, consumer, and clinic effects and their interactions). Such a design is an important direction for future research. Second, the visits we studied were return visits presumably for medication management. As a result, consumers and providers may have specific, preconceived ideas about the content of such visits, which one could argue would preclude the need for agenda setting. However, because these are regular check-ins with consumers, this is still an important opportunity for the provider and consumer to identify any new or persisting concerns that might not be part of a regular medication check. Moreover, consistent with a recovery-oriented model of care for consumers with mental illnesses, efforts to welcome the consumer’s involvement in discussions about his or her health are essential if consumers are to take ownership of their illness self-management.
Also consistent with a recovery-oriented model, shared decision making is receiving increasing attention in mental health care (Adams & Drake, 2006; Adams, Drake, & Wolford, 2007; Deegan, 2010; Deegan & Drake, 2006; Deegan, Rapp, Holter, & Riefer, 2008; Matthias, Salyers, Rollins, & Frankel, 2012). However, we would argue that setting an agenda at the beginning of the visit is the first opportunity for consumers and providers to make decisions together (about what to discuss in the encounter), and establishes the foundation for a collaborative visit in which consumers’ opinions and input are welcomed, thereby creating an optimal environment for shared decision making when treatment decisions must be made (Matthias, Salyers, and Frankel, 2013). Thus, agenda setting may be an important prerequisite for effective communication, such as shared decision making, during the rest of the visit. Future research should further explore the relationship between agenda setting and shared decision making. While we had the opportunity to examine such a relationship, the low variability seen in our sample precluded statistical analysis of these relationships.
Agenda setting in mental health care has received very little attention, despite calls and evidence for explicitly identifying consumers’ concerns as integral to providing effective care. Closing gaps in knowledge about the effect(s) of agenda setting on outcomes and satisfaction with care represents an opportunity and perhaps obligation to identify the best and most effective practices in care for consumers with severe mental illness.
Funding for this study was provided by grant R24 MH074670 from the National Institute of Mental Health, by the U.S. Department of Veterans Affairs Health Services Research and Development Service (CDA 10-034 and HSP 04-148), and by the Kansas Department of Social and Rehabilitation Services. The authors thank the providers and consumers from the Adult and Child Center, Indianapolis, Indiana; Wyandot Center, Kansas City, Kansas; and Bert Nash Community Mental Health Center, Lawrence, Kansas.