Introduction In the fall of 2014, we convened a multidisciplinary task force of stakeholders including emergency physicians, radiologists, and a patient representative. The main purposes of this group were to (1) develop a preliminary research agenda for patient centered outcomes research in the context of emergency diagnostic imaging; and (2) to present key questions to guide further research in this area for discussion and prioritization at the consensus conference “Diagnostic Imaging in the Emergency Department: A Research Agenda to Optimize Utilization” on May 12, 2015 in San Diego, CA. In preparation for the consensus conference, the task force conducted 12, 1hour conference calls in which the framework for developing the research agenda and questions for further research were discussed and consensus was achieved using a modified Delphi method. A key question proposed by the committee was the following: What is and is not a patient-centered outcome related to emergency diagnostic imaging? Several outcomes of potential importance to patients include but are not limited to radiation exposure, fear and anxiety associated with imaging, degree of life disruption introduced by the procedure, cost, understanding the results of imaging, the likelihood of an incidental or unexpected finding, and potential downstream investigations triggered by incidental/unexpected findings, among others. When determining whether an outcome is or is not patient-centered, it is informative to consider the working definition of patientcentered outcomes research (PCOR) iteratively formulated and developed by the PatientCentered Outcomes Research Institute (PCORI) after stakeholder and public input that 2 was approved by the Board of Governors on March 5, 2012 (Table 1).1 In the context of diagnostic imaging, intuitively one might consider an outcome such as radiation exposure or an incidental finding to have important ramifications for the patient and therefore be a patient centered outcome. However, note that each of the 4 questions that comprise the definition of PCOR in Table 1 are posed from the patient’s perspective – this implies that the patient’s perspective is elicited, not presumed, by non-consumer healthcare stakeholders. Given these considerations, for an outcome to be considered patientcentered, it should be elicited from and identified by patients as important. Conceptual framework for developing a future research agenda for PCOR in emergency diagnostic imaging When developing a patient-centered framework for determining outcomes in future ED imaging studies, it is important to take a patient’s entire episode of care into consideration – the decision to visit the ED (pre-ED setting), delivery of care while in the ED (in-ED setting), and the patient’s post-discharge experience (post-ED setting). In light of the above considerations and in the context of the proposed framework for PCOR in emergency diagnostic imaging, the task force sought to address two overarching research questions: (1) What outcomes in emergency diagnostic imaging do patients, parents/caregivers and other key stakeholders (emergency clinicians, radiologists, researchers, payers/policy makers, healthcare administrators) consider important? 3 (2) What methods, tools and approaches best facilitate patient engagement in the decision to undergo diagnostic imaging? It was against this backdrop that the task force formulated topics and questions for a future PCOR agenda in emergency diagnostic imaging. This manuscript addresses the pre-ED setting and in-ED setting in detail; the post-ED setting is discussed in a separate manuscript. (reference group 3, manuscript #2) What outcomes in emergency diagnostic imaging do patients, caregivers and other key stakeholders consider important? Pre-ED setting Sarah Jones, a 30 year-old woman with a history of well-controlled migraine headaches, is followed closely by her primary care physician for her symptoms. She takes a beta-blocker daily as a preventative medication and uses an intranasal serotonin receptor agonist to treat acute breakthrough migraines. One summer weekend, while visiting relatives away from her home city, she developed a severe migraine after forgetting to pack and take her migraine medications. Without access to her home treatment regimen, she sought care at the local emergency department. On presenting to the ED, she described her symptoms to the triage nurse, including a sudden onset headache. After a brief consultation with a resident physician, a CT scan of the brain is ordered and performed despite her history of chronic sudden onset headaches and multiple brain imaging tests in the past for her headaches. 4 In this vignette, Mrs. Jones’ decision to seek care in the ED was precipitated not by her desire for additional diagnostic evaluation (in fact she had received many brain imaging tests in the past for her headaches), but rather her wish for acute pain relief. So why did she receive a test she did not need or necessarily want? Her treating ED physician – who did not know Mrs. Jones’s detailed medical history – was concerned about the nature of her headache. When asked, the physician explained: “patients who present to the ED are having, by some self-definition, an emergency. My job is to treat their condition as an emergency, and for Mrs. Jones this meant imaging her brain with a CT scan.” For this patient and countless others, the pathway to the decision to obtain diagnostic imaging in the emergency department begins before arriving in the ED. When confronted with symptoms such as pain, weakness or shortness of breath (all of which can and frequently do trigger a diagnostic imaging test in the ED), patients have choices that often begin at home. They can choose to seek medical care at all or wait to see if the symptoms subside. If they do choose to seek care, they may select immediate evaluation in an ED or choose to be seen in a primary care, urgent care, or specialty care setting. Depending on their mobility and access to transportation, patients may use emergency medical services (EMS) or travel by private transportation to seek medical care. Each of these decisions can lead to alternative downstream diagnostic testing strategies that may or may not be aligned with the patients’ expectations, values or preferences. Despite the increasing use of emergency care across all conditions [Tang et al] and despite the increasing use of diagnostic imaging for ED patients [Larson et al], the 5 incidence of patients who seek ED care specifically in order to obtain diagnostic imaging has not been described in detail. Multiple studies have examined the reasons for ED use (for both urgent and non-urgent complaints); common findings include: limited access to a primary care physician, being referred in by a primary care doctor’s office, reduced access to transportation, and the perceived need for urgent treatment. The EMPATH study, a prospective multicenter mixed methods analysis of ED patients, identified that patients actively seek emergency care when they believe they are having a medical emergency and are unable to access or are dissatisfied with outpatient care. [Ragin et al]. Gindi and Jones, using the 2012 National Health Interview Survey (NHIS) described reasons for ED use among children, finding that a closed primary care office (particularly during nights and weekends) was responsible for 75% of recent visits to an ED. [Gindi et al] Richman and colleagues, in a national observational study, noted an association between primary care shortages and ED visits by states. [Richman et al] Cheung et al. found that barriers to timely primary care access such as difficulty getting through on the telephone, limited clinic hours, and lack of transportation were all associated with greater emergency care use for patients with Medicaid.[Cheung et al] Weiss and colleagues, in a survey of adolescents who use the ED, found that PCP referral accounted for 21% of their most recent visits. [Weiss et al]. Qualitative studies have uncovered more nuanced reasons for seeking care in the ED for a variety of types of patients, but few-- if any-have identified diagnostic imaging as a reason. Lawson and colleagues interviewed ED patients with asthma (a condition for which radiography is often overused in ED settings [Knapp et al]) -- finding the perceived need for urgent treatment and inability to access an outpatient provider or medication as the primary drivers of ED use. Howard et al 6 interviewed ED patients with nonurgent complaints and had similar findings, with limited primary care access and PCP referrals cited by many of the respondents.[Howard et al] This body of literature suggests that access to outpatient care, including insurance, transportation, and other factors, may drive emergency care use. Of note, none of these studies identified access to imaging as a specific reason for seeking emergency care. Although ED use is associated with increasing use of diagnostic imaging, the extent to which factors such as barriers to access and referrals in to the ED are driven by the diagnostic needs or preferences of patients and their primary care providers remain uncertain. Specifically, patients present to the ED often because they are sent in or cannot get timely care elsewhere—but how often do they come in specifically because they cannot get access to diagnostic imaging? Moreover, how often do they seek emergency care due to general outpatient access issues, and ultimately receive a diagnostic test that they do not need or want? These questions, pertaining to how patients end-up in emergency care settings and the extent to which these factors lead to downstream diagnostic imaging have yet to be rigorously explored. Research questions from the Patient Perspective: 1. When does my decision to seek care in an emergency department lead to diagnostic imaging? 2. How do I choose whether to seek care in an emergency department versus an urgent care center or outpatient office setting? Does this choice lead to different diagnostic imaging decisions and/or different outcomes for patients like me? 7 3. To what extent do patients like me make decisions about where to seek care based on a desire or expectation of imaging? 4. For patients like me, does the decision to present to the ED for imaging reflect a preference for the least amount of disruption introduced into our lives when interacting with the medical system? 5. How is it that sometimes patients like me end up with imaging tests that we don’t need or want? To what extent are these paths set before I even walk into the healthcare setting? 6. For patients like me, to what degree does the decision to present to the ED for imaging reflect inadequate access to outpatient testing? How frequently do patients like me go to the ED for imaging because we can’t get approval from our insurers for outpatient imaging? How frequently do patients like me present to the ED for imaging because we were referred by another provider (such as a PCP)? 7. If I call EMS and travel by ambulance compared to seeking alternative transportation the hospital, does this impact the likelihood of imaging independent of how sick I am? 8. Within my local region, will some care settings be more likely to perform diagnostic tests on me for the same condition compared to other equivalent types of care settings? In ED-setting 8 Logan Smith, a 10-year old boy, is brought to the ED by his parents after a sledding injury. He was sledding head first, unhelmeted, down a hill when he lost control of the sled and struck his head on a plastic garbage can at the bottom. According to his father, who witnessed the event, he didn’t respond to his name or answer questions for nearly 30 seconds and subsequently began to cry. After being consoled by his father and friends, he got up and began sledding again. After 30 minutes of additional play, he complained that his “tummy” hurt and subsequently vomited. His father called his mother and they both agreed to take Logan to the ED to get “checked out.” After arrival to the ED and waiting 1 hour in the waiting room, Logan is taken to a treatment room and seen by a clinician. The clinician’s pager interrupts the story, which is being told by the father. After stepping out to answer the page from a consulting physician and signing a 72-hour hold for a boarded psychiatric patient, the clinician returns to Logan’s treatment room to complete the history and physical examination. She determines that there are no abnormalities on examination and that Logan is now active and behaving normally. The clinician, after formulating her gestalt regarding the risk for traumatic brain injury, and taking into consideration perceived parental anxiety, decides to order a head CT to relieve parental anxiety and rule out bleeding in or around the brain with certainty before rushing to the next patient. In this vignette Logan Smith is brought to the ED by his parents for evaluation after a minor head trauma. This scenario highlights several factors potentially influencing the decision to obtain diagnostic imaging in the ED setting. A recent article by Hess and colleagues2 proposes an approach to group each of these factors: (1) factors related to the patient/caregiver; (2) factors related to the treating clinician; (3) factors 9 related to the clinical setting/healthcare delivery system; and (4) strengths and limitations of the scientific evidence base. Factors related to the patient/caregiver might include medical acuity, decision-making capacity, expectations and preferences, willingness to engage in the decision-making process, and potential socioeconomic, cultural, and educational disparities that may impede transformation from an uninformed to an informed preference. Factors related to the clinician influencing decision-making might include cognitive load at the time of the clinical encounter, understanding of the evidence relevant to the decision at hand, perceived equipoise of the decision, ability to educate the patient/caregiver regarding factors relevant to the decision such that they are empowered to apply their values and preferences, and willingness to engage the patient/caregiver in shared decision-making. Factors related to the clinical setting might include time available to communicate with patients, the responsibility of the clinician to balance responsibility for the individual patient versus the population of patients in the ED at that moment in time, and availability of follow-up structures. Factors related to the strength of the scientific evidence base might include the availability of a rigorously developed and validated prediction rule from which reliable estimates of risk can be obtained,3 whether patient-centered decision tools exist to facilitate risk communication, and the degree to which they have undergone testing in trials,4 among other factors. Table 2 outlines the major categories potentially influencing the decision to obtain diagnostic imaging and several factors within each category. Shared decision making (SDM), a collaborative process in which patients and clinicians make health care decisions together, taking into account the best scientific evidence available, as well as the patient’s values and preferences, is being increasingly 10 advocated for in both national health policy and clinical guidelines. (refs ACA, AHA advanced heart failure guidelines). Although few disagree with SDM from a conceptual perspective, application to clinical practice, particularly in a busy, chaotic clinical setting such as the ED, is a challenging prospect. In addition to the factors influencing the decision-making process outlined in Table 2, the clinical scenarios in which SDM might be both feasible and appropriate have not been clearly delineated. Even after a reasonable scenario for SDM has been selected – such as the decision to obtain a head CT in low to moderate risk pediatric minor head trauma in which there exists a well-validated decision rule3 and a decision aid4 to educate and engage parents in the decision-making process – which risk levels are appropriate for SDM, what the effect of SDM might be of engaging low risk patients in whom CT is not indicated, moderate risk patients in whom observation or head CT are both reasonable options, or high risk patients in whom head CT is clearly indicated is another area for further research. Finally, developing patientcentered contextually sensitive approaches to decision making – where on the continuum from informed decision making to paternalism is optimal for a given patient and situation, and how to go about determining an empathic approach to decision-making, is another ripe area for research. Research questions: 1) In a given clinical scenario, what factors influence the decision to obtain advanced imaging? 2) To what degree does each of these factors influence decision-making – and to what degree should each of the factors influence the decision to obtain imaging? 11 3) What affect would alignment of incentives of key stakeholders – patients, clinicians, radiologists, and policymakers/payers – have on diagnostic imaging utilization in the emergency setting? 4) What approaches might effectively align incentives such that the rate of imaging is more closely tailored to both disease risk and risk-informed patient preferences? 5) For a given clinical scenario, what risk levels are appropriate for SDM? 6) What clinical scenarios in the ED setting are most appropriate and feasible to employ a SDM approach? 7) In which clinical scenarios, if any, is taking an assertive, paternalistic approach an optimal contextually sensitive approach to decision making? Post-ED setting: After a patient is discharged from the ED, additional follow-up and care may be needed. A research agenda related to these issues is explored in greater detail elsewhere. (reference manuscript #2) What research methods optimally facilitate the development and assessment of patient-centered outcomes in emergency diagnostic imaging? While there have been tremendous recent advancements in health care treatments and delivery, there have similarly been strides in patient-centered outcomes research (PCOR). Outcomes-driven, patient-centered methods may help improve the ability to 12 understand and balance alternative diagnostic and management strategies for some emergency medicine patients (Gabriel 2012; Lee 2014; Meisel 2012). The historic approach to assessing the contribution of diagnostic imaging to patient management is outlined in Table 3 (left side) (Fryback 1991). This traditional conceptual hierarchical model of imaging efficacy is a helpful tool, beginning with the premise that a valuable diagnostic radiology test should offer both high-quality images that can be reliably interpreted and an accurate diagnosis. However, the hierarchy extends beyond technical and diagnostic accuracy, by considering more broadly how the diagnostic test affects clinicians’ decision-making, patient management, patient outcomes, and societal benefit. The key concept, as noted in Figure 1, is that not every normal or abnormal imaging result will alter the subsequent diagnostic or therapeutic approach of the clinician. Furthermore, every abnormal finding does not correspond to a therapy that can or should be recommended, based upon available treatments, other acute threats to health, patient priorities, and cost. Furthermore, available therapies that are based on surrogate outcomes often lack direct evidence of benefit for patient-important outcomes (Montori 2012). Additionally, consider a patient who experiences a mild head injury and is found to have a minor traumatic intracranial bleed on neuroimaging. The neuroimaging study will likely result in hospital admission and potentially repeat imaging and thus certainly affect the patient management plan (Level IV evidence); however, as is often the case with small traumatic bleeds, the patient does not require neurosurgical intervention (Jagoda 2009; Kuppermann 2009). Thus, one could argue that neuroimaging did not 13 ultimately affect the patient’s outcome (Level V evidence). In other words, subsets of patients who derive benefit at higher levels of the diagnostic evidence hierarchy are usually less than the subset deriving measurable benefit from lower levels. Unfortunately, clinicians often cannot distinguish in real-time which patients will benefit from imaging and which will not. Importantly, however, the traditional approach does not provide sufficient guidance regarding whether, how, when, or why to address patient-centered outcomes, particularly in chaotic, time-limited acute care situations such as those that occur in emergency medicine. We suggest a modified approach to understanding the ultimate value of emergency diagnostic imaging, one that involves patient-centered outcomes, and integrates patient values and preferences in diagnostic imaging decisions (see Table 3, right side). To ensure that imaging efficacy is coupled with patient-centered care, we must first work as a research community to develop tools that reliably and accurately measure and assess patient-centered outcomes. This must be accomplished across a variety of healthcare settings and include a diversity of patient populations with differing experiences of access to care and patient/provider health literacy and numeracy ability (Carpenter 2014; Griffey 2014). There is a broad spectrum of methodological tools, both qualitative and quantitative, that can be rigorously applied to realize this goal, and to address the topics and questions posed in this article (Fleurence 2015; Gabriel 2012; Meisel 2012; Hickman 2013; Lee 2014)._ 14 Qualitative methods, ideally developed in collaboration with social scientists, have recently been recognized as an essential component of radiology-related patientcentered outcomes research. They may be used to identify outcomes and issues most important to patients (Lee 2014). For example, semi-structured interviews, focus groups, and surveys that engage patients and their caretakers could be used to gain an in-depth understanding of what outcomes are truly important to patients (Robey 2014; Sommers 2012; Karpas 2013). Mello and colleagues recently provided an overview of important considerations for emergency medicine researchers doing survey research (Mello 2013), and the value of qualitative techniques will be enhanced with a commitment to robust methodology (Creswell 2012; Groves 2009; Remler 2011), including the use of welldesigned tools, high-yield sampling strategies, and excellent response rates. At a minimum, such methods can be used to better understand what factors in emergency imaging – whether related to efficient evaluation, anxiety, reassurance, potential adverse effects, incidental findings, cost, or convenience – are noticed and considered most important to patients. If done in a rigorous manner, such qualitative techniques can be coupled with psychometric evaluation to develop validated patientreported outcome measures, as advocated by the National Institutes of Health (NIH)sponsored Patient Reported Outcomes Measurement Information Systems (PROMIS) (PROMIS report; Cella 2007). We distinguish patient-centered outcomes, as described in the introduction, from patient-reported outcomes, which the U.S. Department of Health and Human Services and the Food and Drug Administration define as “any report of the status of a patient’s 15 health condition that comes directly from the patient, without interpretation of the patient’s response by a clinician or anyone else (Health Qual Life Outcomes 2006).” Patient reported outcomes commonly assess patients’ satisfaction with care, the perceived value of treatment, and health-related quality of life (Fleurence 2015; Calvert 2013). For example, if initial investigations determine that anxiety surrounding diagnostic imaging is important to patients – as related to a new diagnosis, an incidental finding, an incomplete understanding of a radiology report, or a decision to pursue a period of observation instead of immediate imaging – researchers could develop and/or use existing scales to assess patient anxiety in this context. Patient reported outcomes with feedback to physicians helps improve patient-physician communication and patient satisfaction (Chen 2013) which may enhance the overall patient experience and facilitate a shared view between a patient and provider regarding care goals. Thus, developing and using valid, reliable, patient-reported outcomes is one important component of advancing patientcentered care (Fleurence 2015). Two principal features of radiology-related patient-centered outcomes research entail an enhanced use of mixed methods techniques and head-to-head comparative trials (Lee 2014). If done effectively and purposively, the combined use of qualitative and quantitative research can lead to more insightful results than either methodology alone (Lee 2014; Albright 2013). Additionally, such an approach can help elucidate why a given recommended practice (e.g., ultrasound for initial evaluation of pediatric appendicitis) is not implemented in a real-world setting and identify ways to overcome barriers to successful implementation (e.g., increased ultrasound technician training and expanded hours of availability) (Lee 2014). 16 While qualitative techniques can help establish an improved understanding of outcomes important to patients, quantitative methods essential to comparative effectiveness research (CER) are needed to advance patient-centered care in emergency imaging (Flourence 2015; Lee 2014; Meisel 2012; Montori 2012; Hickman 2013; Remler 2011). Outcome-driven, patient-oriented CER could include controlled studies, experiments, quasi-experiments, randomized trials, secondary database analyses, and systematic reviews to determine what work best between multiple approaches to diagnostic imaging. For example, because any decision regarding medical imaging has its own potential benefit/harm profile, comparing the influence of different risk communication strategies on outcomes such as patient knowledge, understanding, decision quality, and resource use would be invaluable. Similarly, experimental testing of well-designed decision-aids is needed for imaging decisions, especially where clinical equipoise exists. Incorporating patient/caregiver preferences and values into the decision to image versus observe for certain pediatric patients after minor head injury, for example, could facilitate improvement of outcomes most important to those patients/caregivers.4 PCORI consistently highlights rigorous CER as essential to advance and improve patient-centered care (Flourence 2015; Hickman 2013). Moreover, multiple opportunities exist for CER involving diagnostic imaging in emergency medicine (Lee 2014; Meisel 2012; Montori 2012). The final stage of all research is dissemination and implementation (Knowledge Translation) of high quality, reliable clinical data to diverse healthcare settings. A research agenda related to Knowledge Translation developed in another breakout session is described elsewhere (reference AEM CC KT manuscript). 17 All of the methodological approaches mentioned here would benefit from stakeholder engagement, with true patient and community partnership throughout the entire research process (Flourence 2015; Lee 2014; Hickman 2013; Wells 2009). This approach will have many long-term benefits, including helping to identify what outcomes are most valued by patients and how best to disseminate such information to the public (Flourence 2015). It may also be worthwhile for researchers to collaborate with groups such as PCORI (http://www.pcori.org/), the Informed Medical Decision Foundation (http://www.informedmedicaldecisions.org/), the Preventing Overdiagnosis movement (http://www.preventingoverdiagnosis.net/), and the Lown Institute Right Care Alliance (http://lowninstitute.org/take-action/join-the-right-care-alliance/), given their ongoing dedication towards similar efforts. Finally, while we have focused largely on engaging patients in research aimed to advance patient-centered care with respect to imaging, stakeholders at the provider and health systems levels must also be engaged. Emergency physicians will play a critical role in the success of these efforts, and helping them overcome perceived obstacles with respect to over-imaging and patient engagement will be required (Kanzaria 2015a, Kanzaria 2015b). However, input must also be sought from radiologists, consulting and referring clinicians, and hospital leadership in the form of adequate infrastructure for health information technology and care coordination. Research Questions 1. When considering the impact of imaging on patient management, how can the traditional (Fryback + Thornbury) frameworks for diagnostic 18 efficacy accurately and reliably incorporate and measure patient-centered care in the acute care setting? 2. Can we use patient-reported outcome assessment tools, developed and validated in other settings, for diagnostic imaging in the emergency department/acute care setting? 3. How can we use patient-reported outcome tools to advance patientcentered care? 4. How can comparative effectiveness research be used to advance patientcentered care in emergency imaging when comparing imaging modalities that are both already considered efficacious? 5. Which stakeholders need to be engaged to enhance patient-centered outcomes research on emergency diagnostic imaging? 6. What is an effective approach to identifying and engaging key stakeholders? 7. How can patient and community stakeholders be better engaged in patient-centered research efforts on diagnostic imaging in the emergency department? 8. What is the optimal role of the patient or advocate in decision-making (e.g., on obtaining imaging or not, on the selection of a given diagnostic imaging study etc.) in the ED? 9. How can the NIH framework for dissemination and implementation be adapted for diagnostic imaging in the emergency department so the patient’s voice is heard through this process? 19 10. How can uncertainty be communicated when there is limited evidence about the best clinical evaluation strategy? 11. When limited evidence exists, how should this impact the decision to engage patients in share decision-making around diagnostic imaging in the ED? Specifically, should patients be engaged in shared decisionmaking when there is limited evidence on the best diagnostic approach? If so, how can we accomplish this? 12. Given the uncertain external and internal validity of prognostic models, when can shared decision-making investigators be sufficiently confident in model-derived prognostic estimates to incorporate these estimates into shared decision-making studies of patient-centered care? 13. When improved evidence becomes available, how should patient decision-tools be updated? Can the older/inferior models be adapted, or would new decision- tools need to be created? 14. How would updating the decision-aids with more refined or reliable models impact the effectiveness of decision tools? Would these updates compromise the prior patient-centered diagnostic efficacy research that was based upon the inferior model? What research methods facilitate the development and assessment of tools and approaches to enhance risk/benefit communication and patient engagement? 20 SDM was initially conceptualized as a single event in time, applied to a one-time decision (e.g., total knee surgery) and a reasonable amount of time (i.e. days, weeks) to engage in the process and make a decision. A decade ago, Montori and colleagues focused on SDM in chronic care and highlighted the importance of the patient-clinician partnership as a prerequisite for SDM in this setting.6 Inherent to decision making in this context was the opportunity to use the experience of outcomes from prior choices or other new information to revise decisions. Recently, SDM has been applied in the ED setting, where time is of the essence and the importance of quickly establishing a patient-clinician partnership are crucial to the decision-making process and patient care. (Pitts, Health Affairs) The extent to which this newer conceptualization of SDM can be appropriately measured with instruments developed for previous concepts of SDM is currently unknown. SDM should ensure evidence-based, patient-centered, high quality decisions. Decision making quality outcomes hence reflect the extent to which patients are knowledgeable about the current decision at stake, are clear about their values and preferences in this context, are meaningfully engaged in the decision making process, and are comfortable with the decision taken.7,8 Scholl and colleagues proposed a framework mapping decision making quality according to decisional needs/ antecedents (prior to the encounter), decision-making process (during the encounter), and decisional outcome (post encounter/decision).9 Table 4 includes examples of validated measures associated with each of these categories. 21 To date, there is still no consensus as to the best decision-making quality outcome or set of outcome measures. In a review of 35 SDM trials, Kryworuchko and colleagues identified 35 unique primary outcome measures and eight different instruments, highlighting challenges in the selection of relevant high quality outcome measures.20 The most recent review of SDM instruments by Scholl and colleagues reported 28 measures that had undergone psychometric testing.9 This review revealed the poor quality of reporting of psychometric evaluation of these available instruments, limited validation of most, and the need for a consistent measurement framework in the field. Moreover, despite the patient-centered nature of SDM, patients and their clinicians have not participated - to the best of our knowledge - in the development of currently available patient-centered SDM outcome measures. These findings suggest poor agreement about optimal SDM outcome measures and dissatisfaction with the ability of available measures to validly measure SDM, let alone SDM in the fast-paced setting of the emergency department. To date, two of the most used scales in SDM remain the Decisional Conflict Scale, 13 which assesses a patient’s decisional conflict defined as personal uncertainty about which course of action to take when choice among competing options involves risk, regret, or challenge to personal life values, and the OPTION scale, a third observer scale that assesses the extent to which clinicians engage patients in the decision making process.15{Couet, 2013 #6904} The recent 2014 Cochrane Collaborative systematic review on patient decision aids including 115 randomized trials reported that the use of decision aids increases 22 patients’ knowledge [mean difference (MD) 13.34 out of 100; 95% CI 11.17 to 15.51] and the accuracy of risk perception (RR 1.82; 95% CI 1.52 to 2.16) and decreases the proportion of patients who were passive in decision making (RR 0.66; 95% CI 0.53 to 0.81), or who decided to undergo major elective surgery in favor of more conservative options (RR 0.79; 95% CI 0.68 to 0.93).8 Another systematic review by Flynn et al. found only 5 studies that aim to engage patients in the decision-making process in the ED. Of those, only one trial evaluated a decision aid for the emergency department, and there were no decisions aids for engaging patients in the decision to undergo imaging.21 This trial reported increased patients’ knowledge [effect size, MD 0.67; 95% CI, 0.34-1.0] and an increase in engaged decision making process [effect size, MD 19.6; 95% CI, 1.6-21.6]. In summary, the proliferation and variation of measures, and the largely untested use of these measures in emergency settings provide significant opportunity for methods development for patient-centered outcomes in emergency care research. Moreover, because choices to obtain diagnostic imaging remain core decisions in emergency care, patient-centered outcome measure development and validation that focus on acute care diagnostic testing will doubtless add significant value to research and patient care in this arena. Research Questions 23 1. Among the existing measures used to determine decisional needs, the quality of the decision-making process itself and decisional outcomes, which are potentially applicable for use in PCOR conducted in the ED setting? 2. What is the validity of the existing PCOR measures when applied in the ED context? 3. Which of the applicable existing decision-making measures require refinement (or development of new suitable measures) to optimally measure decisionmaking in the ED setting? 4. What key components of decision-making around diagnostic imaging in the ED setting are inadequately captured by existing applicable measures? 5. What patient-reported outcomes applicable to emergency diagnostic imaging are inadequately captured by existing decision-making measures? 6. How do we incorporate additional patient reported outcomes that go beyond decisional outcomes? 24 Table 1. Definition of Patient-centered Outcomes Research (PCOR) proposed by the Patient Centered Outcomes Research Institute (PCORI) after considerable patient, stakeholder, and public input and approved by the PCORI Board of Governors on March 5, 2012. PCOR helps people and their caregivers communicate and make informed healthcare decisions, allowing their voices to be heard in assessing the value of healthcare options. This research answers 4 patient-centered questions, such as: 1. “Given my personal characteristics, conditions, and preferences, what should I expect will happen to me?” 2. “What are my options, and what are the potential benefits and harms of those options?” 3. “What can I do to improve the outcomes that are most important to me?” 4. “How can clinicians and the care delivery systems they work in help me make the best decisions about my health and health care?” To answer these questions, PCOR: Assesses the benefits and harms of preventive, diagnostic, therapeutic, palliative, or health delivery system interventions to inform decision-making, highlighting comparisons and outcomes that matter to people; Is inclusive of an individual's preferences, autonomy, and needs, focusing on outcomes that people notice and care about such as survival, function, symptoms, and health-related quality of life; Incorporates a wide variety of settings and diversity of participants to address individual differences and barriers to implementation and dissemination; and Investigates (or may investigate) optimizing outcomes while addressing burden to individuals, availability of services, technology, and personnel, and other stakeholder perspectives. 25 Table 2. Factors potentially influencing the decision to obtain diagnostic imaging in a given patient in the emergency department setting.* Patient factors Acuity Decision-making capacity Willingness to engage in the decision-making process Socioeconomic, ethnic, and educational disparities Clinician factors Willingness to engage patients in shared decision-making Cognitive load at that moment in time Perceived equipoise of the decision Understanding of the evidence relevant to the decision at hand Ability to rapidly and succinctly recall and communicate the benefits and trade-offs of the decision from a patient-centered perspective Ability to create a safe environment so patients feel comfortable asking questions regarding their management options System/healthcare delivery context factors Population responsibility Undifferentiated symptoms: focus on diagnosis and disposition Availability of follow-up structures Time available to communicate with patients Strengths and limitations of the scientific evidence base Strength of the evidence establishing the benefits and trade-offs of the available management options Is the evidence base established on outcomes that matter to patients? Availability of robust risk prediction tools to guide decision-making Availability of decision aids to facilitate shared decision-making Approaches to risk communication that best facilitate knowledge transfer and accurate risk perception *Modified from Hess et al.2 26 Table 3. Comparison of Traditional and Patient-Centered Model of Imaging Efficacy Traditional Approach to Diagnostic Imaging Efficacy (Frybach 1991) Level 1 – Technical Efficacy Is the technical quality of images high? Example: An interpretable cranial computed tomography (CT) scan can be obtained without excessive radiation exposure to the patient. Level 2 – Diagnostic Accuracy Efficacy Level 3 – Diagnostic Thinking Efficacy Example unchanged since this is a technical and not patient-centric issue, however cannot proceed to subsequent levels without this level of efficacy: An interpretable cranial computed tomography (CT) scan can be obtained without excessive radiation exposure to the patient. Does the imaging technique have excellent diagnostic test characteristics? Example: The cranial CT distinguishes intracranial hemorrhage from no intracranial hemorrhage in trauma patients, but whether this diagnostic accuracy changes subsequent clinical actions or outcomes is not the issue. Example unchanged since this is a technical and not patient-centric issue, however cannot proceed to subsequent levels without this level of efficacy: The cranial CT distinguishes intracranial hemorrhage from no intracranial hemorrhage in trauma patients, but whether this diagnostic accuracy changes subsequent clinical actions or outcomes is not the issue. Does the imaging result produce a change in the physician’s diagnostic thinking? Do patient-centered outcomes, as well as patient values and preferences, influence the physician’s approach to diagnostic imaging? How does the imaging study affect the patient management plan? Example: The subset of patients identified by the cranial CT (who would otherwise not have been identified) have an available therapy or action that would have been neglected or less likely were they not identified by the cranial CT, but whether the intervention improves outcomes is not the issue. Level 5 – Patient Outcome Efficacy Is the technical quality of images adequate to lead to a diagnosis? Does the imaging technique have excellent diagnostic test characteristics? Example: The cranial CT identifies subsets of patients with traumatic intracranial hemorrhage that existing, traditional, often cheaper and more readily available diagnostic tests like history, physical exam, and clinician gestalt would not already identify, but whether this new knowledge creates actionable steps is not the issue. Level 4 – Therapeutic Efficacy Patient-Centered Approach to Diagnostic Imaging Efficacy Can we measure the effect of imaging on the patient outcome? Example: In contemplating whether or not to order the cranial CT to identify subsets of patients with traumatic intracranial hemorrhage that existing, traditional, often cheaper and more readily available diagnostic tests like history, physical exam, and clinician gestalt would not already identify, test-ordering clinicians incorporate patient values, preferences, and outcome goals. How do patient-centered outcomes, as related to available therapies based on diagnostic imaging, influence the patient management plan? Example: In evaluating whether or not to order the cranial CT, test-ordering and treating clinicians contemplate the subset of patients who may not otherwise be identified, and these clinicians incorporate patient-centered outcomes in contemplating next steps with any CT finding before ordering the CT. Can we measure the effect of imaging on patientcentered outcomes? 27 Level 6 – Societal Efficacy Example: The newly identified subset of patients who receive the intervention derive some measurable benefit, but this benefit might be a surrogate measure or of lesser importance to patients. Example: In both a) evaluating whether or not to order the cranial CT, and in b) obtaining a CT to identify the subset of patients who receive the intervention and will derive some measurable benefit, clinicians contemplate and ascertain patient-centric outcomes like pain, disability, and mortality. Can we analyze societal costs and benefits of diagnostic imaging technology? Can we analyze both individual and societal costs, harms, and benefits of diagnostic imaging technology? Example: The subset of patients who are newly identified, receive the subsequent intervention, and derive measurable (albeit not necessarily patientcentric) benefit ultimately also benefit society in terms of less lost work-days, less healthcare dollars spent, or reduced disability. Example: The subset of patients who are newly identified, receive the subsequent intervention, and derive measurable patient-centric benefit ultimately also benefit society in terms of less lost work-days, less healthcare dollars spent, or reduced disability. 28 Table 4. Validated measures of decisional quality according to decisional needs/ antecedents (prior to the encounter), the decision-making process (during the encounter), and decisional outcome (post encounter/decision). Decisional needs/ antecedents Control Preference Scale10 Preparation for decision making scale11 Decision self-efficacy Scale12 Decision-making process Decisional Conflict Scale13 Knowledge14 OPTION Scale15 SDMQ-916 Decisional outcome Decisional Regret Scale17 Satisfaction with decision Scale18 COMRADE scale19 29 Figure 1: Hierarchy of Diagnostic Efficacy Venn Diagram. Diagnostic tests (whether history, physical exam, labs or imaging) attain different levels of efficacy and impact different proportions of patients depending upon the outcome. In radiology, the first level of efficacy is the ability to obtain and interpret images. All subsequent levels of efficacy must attain this first level, but not all patients in whom Level I efficacy are attainable will benefit from higher levels. This Venn diagram depicts the concept that each higher level of efficacy described in Table 3 loses a portion of patients from the preceding lower level of efficacy. How many patients lose benefit as higher levels of efficacy are attained is unknown and probably varies from study-to-study and across populations. Need to obtain permission to reproduce this Figure. Level I Images Obtainable & Reliably Interpreted Level II Images Accurately Rule-In or Rule-Out Diagnosis Level III Image Results Alter Subsequent Physician Diagnostic and/or Therapeutic Options Level IV Available & Acceptable Therapy with Any Measurable Outcomes for Subset Now Diagnosed Who Would Otherwise Be Missed Level V Patient-Centric Therapeutic Benefits Measurable Level VI Societal Level Benefits Measurable 30 Figure 2: The National Institute of Health Framework for Dissemination and Implementation Research 31 References 1. Patient Centered Outcomes Research. (Accessed April 27, 2015, at http://www.pcori.org/content/patient-centered-outcomes-research.) 2. Hess EP, Grudzen CR, Thompson R, Raja AS, Carpenter CR. Shared Decision Making in the Emergency Department: Respecting Patient Autonomy When Seconds Count. Academic emergency medicine : official journal of the Society for Academic Emergency Medicine 2015;In press. 3. Kuppermann N, Holmes JF, Dayan PS, et al. Identification of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study. Lancet 2009;374:1160-70. 4. Hess EP, Wyatt KD, Kharbanda AB, et al. Effectiveness of the head CT choice decision aid in parents of children with minor head trauma: study protocol for a multicenter randomized trial. Trials 2014;15:253. 5. Makoul G, Clayman ML. An integrative model of shared decision making in medical encounters. Patient Educ Couns 2006;60:301-12. 6. Montori VM, Gafni A, Charles C. A shared treatment decision-making approach between patients with chronic conditions and their clinicians: the case of diabetes. Health expectations : an international journal of public participation in health care and health policy 2006;9:25-36. 7. Sepucha KR, Borkhoff CM, Lally J, et al. Establishing the effectiveness of patient decision aids: key constructs and measurement instruments. BMC Med Inform Decis Mak 2013;13 Suppl 2:S12. 8. Stacey D, Legare F, Col NF, et al. Decision aids for people facing health treatment or screening decisions. The Cochrane database of systematic reviews 2014;1:CD001431. 9. Scholl I, Koelewijn-van Loon M, Sepucha K, et al. Measurement of shared decision making - a review of instruments. Z Evid Fortbild Qual Gesundhwes 2011;105:313-24. 10. Degner LF, Sloan JA, Venkatesh P. The Control Preferences Scale. The Canadian journal of nursing research = Revue canadienne de recherche en sciences infirmieres 1997;29:21-43. 11. Graham ID, O'Connor AM. Preparation for Decision Making Scale - User Manual. 1996. 12. O'Connor AM. Decision Self Efficacy Scale - User Manual. 1995. 13. O'Connor AM. Validation of a decisional conflict scale. Medical decision making : an international journal of the Society for Medical Decision Making 1995;15:25-30. 14. O'Connor A. User Manual _ Knowledge [Document on the Internet]. Ottawa: Ottawa Hopsital Research Institute; 2000. 15. Elwyn G, Hutchings H, Edwards A, et al. The OPTION scale: measuring the extent that clinicians involve patients in decision-making tasks. Health expectations : an international journal of public participation in health care and health policy 2005;8:34-42. 16. Kriston L, Scholl I, Holzel L, Simon D, Loh A, Harter M. The 9-item Shared Decision Making Questionnaire (SDM-Q-9). Development and psychometric properties in a primary care sample. Patient education and counseling 2010;80:94-9. 32 17. Brehaut JC, O'Connor AM, Wood TJ, et al. Validation of a Decision Regret Scale. Medical Decision Making 2003;23:281-92. 18. Holmes-Rovner M, Kroll J, Schmitt N, et al. Patient satisfaction with health care decisions: the satisfaction with decision scale. Med Decis Making 1996;16:58-64. 19. Edwards A, Elwyn G, Hood K, et al. The development of COMRADE--a patient-based outcome measure to evaluate the effectiveness of risk communication and treatment decision making in consultations. Patient Educ Couns 2003;50:311-22. 20. Kryworuchko J, Stacey D, Bennett C, Graham ID. Appraisal of primary outcome measures used in trials of patient decision support. Patient Educ Couns 2008;73:497-503. 21. Hess EP, Knoedler MA, Kline J, et al. Effect of a Decision Aid on Patient Knowledge, Patient Satisfaction, Safety, and Resource Use in Low-Risk Emergency Department Patients with Chest Pain: A Randomized Trial. Acad Emerg Med 2010;18.
© Copyright 2024