P atient centeredness is a distinct dimension of healthcare quality. Patient-centered care is respectful of and responsive to individual patient preferences, needs, and values in ways that guide treatment decisions.1 Translating patient centeredness into routine care is challenging. Citing evidence that patients often do not want a role in treatment decision making, many initially questioned the utility of engaging patients or eliciting their preferences.2,3 It is not surprising that many subjects enrolled in older studies of medical decision making would endorse passive decision making styles, given the paucity of tools to facilitate meaningful patient or caregiver engagement. When such tools were available, they were often overly complex and technical with little appreciation for health literacy, numeracy, or cognitive biases. The purpose of these tools was to facilitate decision making for clinicians rather than explore or uncover patient preferences, values, and goals. In essence, these earlier studies attempted to measure patient participation, experiences, and preferences within the context of clinician-centric medical encounters.
Articles see p 353, 360, 368The results described by Dunlay et al 4 in this issue of Circulation Cardiovascular Quality and Outcomes are illustrative of these early attempts in understanding patient centeredness within the context of a clinician-centered healthcare environment. Their study sought to better understand the endof-life preferences of patients with heart failure and how those preferences changed during the disease course and, in particular, during hospitalizations for acute exacerbation. 5 The focus on do-not-resuscitate (DNR) orders as a surrogate is indicative of our limited capacity to clearly understand, elicit, and document end-of-life preferences. DNR orders are rudimentary measures of patient values and preferences for end-of-life care; orders are often placed with little or no previous discussions and are largely for the purpose of clinician decision making. 5,6 Dunlay et al 4 found that changes in DNR status in patients with heart failure are associated with acute heart failure exacerbations and with predictors of mortality (eg, age, functional impairment, malignancy, and comorbid chronic obstructive pulmonary disease). The disconnect between DNR orders and advance care planning is evident by the finding that patients with class 3 or 4 heart failure symptoms were less likely to have DNR orders in place on study enrollment. 4 The results of this study highlight the lack of standardized, patient-centered measures of advanced care planning. In contrast, the presence of DNR orders has surprisingly powerful effects on clinicians' decision making that go far beyond the simple intent of documenting the patient's preferences for resuscitation. In a previous study, Beach and Morrison 7 found that the presence of a DNR order was associated with significantly lower willingness to draw blood cultures, place central lines, or provide blood transfusions among physicians. T...