There are more choices for place of end-of-life cancer care due to shifts in care to the community, better understanding of the clinical course of cancers, equipment portability, pharmacology advances, and consumer expectations. Place of care has various meanings to patients and families and represents more than a particular geographic location . Options can include hospice, private residence, nursing home, continuing complex care facility, homeless shelter, and/or hospital. For patients with good symptom control, instrumental support, and a predictable course of illness, decisions regarding place of care are often based on values and expectations [2–5].
Patients frequently experience decisional conflict (personal uncertainty about the best course of action) when considering place for end-of-life care. Personal preferences are weighed against practical considerations and concern for others [2, 3, 5, 6], thus contributing to decisional conflict. Other modifiable factors such as knowledge gaps, unrealistic expectations about outcomes, lack of clarity about what matters most, and feeling pressured to choose a particular option exacerbates the decisional conflict . Unresolved decisional conflict can lead to decisional delay or reversal, dissatisfaction, regret, and blaming the provider [8, 9]. Failure to elicit the valued priorities of terminally ill patients and their families can result in missed opportunities, decreased quality of life, unwelcome interventions, and increased risk for complicated bereavement for survivors [3, 10–12].
It is generally known that decisional conflict can be reduced with decision support interventions such as decision aids and nurse coaching [7, 13]. However, there have been few studies evaluating either of these interventions for decisions at the end-of-life. Although patient decision aids may be useful for some common discrete crossroads decisions with standardized options and outcomes, there may be more payoffs in focusing on coaching interventions that can be applied broadly to care management at the end-of-life. While end-of-life decisions are highly values-sensitive, they also bear a strong resemblance to chronic condition management decisions, which focus on situation monitoring, priority setting, and implementation [14–16].
There is evidence that influencing practitioners' knowledge and attitudes about communication and decision support can strengthen subsequent decision support practices, thus matching care planning to patient preferences and avoiding the use of non-valued interventions [17–20]. Nevertheless few studies have empirically examined the impact of a theoretically informed, decision support training intervention on the quality of decision support knowledge and practices provided by practitioners  and none have been undertaken within the context of palliative care practice.
Patients and families want clinicians to listen to their views and preferences  and standard palliative practice calls for patient inclusion in care planning [22, 23]. Most patients with advanced cancer want full information and the majority wish to participate actively in decision making . Although seriously ill hospital patients want to discuss end-of-life issues, their preferred decision making role varies and is difficult to predict , highlighting the need for active and regular assessment of patients' decision making needs. However, clinicians may avoid raising uncomfortable topics [26–28] and often lack skills and confidence in helping patients in non-directive ways .
Systematic review findings confirm that training clinicians in patient centered approaches is an effective strategy for increasing patients understanding of the evidence and implications . For instance, decision coaching by nurses has helped to foster an informed use of resources and avoid the over-use of interventions that patients don't value in urology care  and in gynecological care . Practitioners, such as nurses and other professional care coordinators, through their trusted and frequent interactions with patients are well positioned to elicit and explore decisional needs such as decisional conflict and related factors (e.g.: knowledge, values clarity, and support) . Practitioners can then coach patients with information, values clarification, support and links to resources. While a key challenge is the need to strengthen practitioners' decision support skills  training interventions have been shown to markedly improve the quality of practitioners' decision support skills for other clinical problems .
Accountability to quality patient outcomes and fiscal responsibility confirms the need to practically and pragmatically address patients' end-of-life decision making needs. Systematic and rigorous evaluation of an evidence-based intervention designed to improve the quality of practitioners' decision support could illuminate best practices for decision support and advance the fields of shared decision making, patient-practitioner communication, palliative care, and ultimately improve the lives of those who are living with a terminal illness and their families.
Multifaceted interventions show promise in influencing professional behavior change [34–37]. An exploratory study to identify target variables, choose and refine interventions, and establish their theoretical basis prior to large scale effectiveness trials is a sound research approach [38–40]. The paucity of empirical inquiry, in this area to date, warrants an exploratory study to describe the constant and variable components of a potential intervention and a feasible protocol for intervention delivery.
Guiding Theoretical Models
As the study aims to influence decision support behavior an understanding of factors that can be partially modified, such as attitudes and perceptions of norms that drive actions, is required to inform intervention messages and enhance the potential for behavior change. The pragmatic and conceptual focus of the study also requires an empirically proven, clinically relevant decision support framework to guide intervention content. The Theory of Planned Behavior (TPB) [41, 42] and the Ottawa Decision Support Framework (ODSF)  fit these criteria. The former predicts the likelihood of behavior change while the latter provides a three step path to optimize quality decision support. The TBP [40, 44–46] and the ODSF [3, 47, 48] are relevant to nursing and have performed well in numerous health-related studies.
Briefly, the TPB proposes that the strength of the intention to change is the primary determinant of actual behavior change. This intention is determined by: (1) a person's attitude to the new behavior (strength of perceived advantages and disadvantages); (2) the extent of self-perceived social pressure to perform or not to perform the new behavior, and (3) degree of perceived control over being able to perform the behavior .
According to the ODSF, decision support interventions can be tailored to address modifiable determinants of decisional conflict (knowledge; outcome expectations; values clarity ; support factors) resulting in better quality decisions that are informed and consistent with patients' values . In this study, the ODSF will guide decision support skill acquisition interventions and measures of change.
Clinician-identified factors associated with practice change: utility, strong evidence basis, and flexibility to acknowledge the individuality of patients  fit well with the ODSF. Historically, the ODSF has been well received by clinicians [3, 33, 48], has shown robustness in randomized control studies [7, 33] and is predicated on a patient-centered approach which recognizes the unique context, circumstance and patient characteristics situated within the decision support encounter. The TPB and ODSF have guided the study design in the following ways:
1. The predictive potential of theory facilitates selection of training intervention components which offers the best probability of success [36, 52].
2. The TPB offers a useful lens to examine attitudes, beliefs and perceived control for engaging in decision support practices and will inform the selection of key messages attached to intervention strategies.
3. Mapping the intervention components onto TPB variables provides a useful template to ensure multiple targets of behavior change are addressed and is consistent with knowledge dissemination best practices [35, 53].
4. The ODSF describes a well defined approach to quality decision support provision and provides a practical vehicle to structure components of a knowledge and skill building intervention.
5. Availability of a broad inventory of theoretically grounded and empirically validated, reliable, time tested tools operationalizing ODSF constructs provides a rigorous platform to inform study interventions and will strengthen the trustworthiness of study findings.