American Journal of Emergency Medicine 32 (2014) 233–236
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Patient preferences for testing for pulmonary embolism in the ED using a shared decision-making model☆,☆☆ Brian C. Geyer, MD, PhD, MPH ⁎, Maria Xu, BS, Christopher Kabrhel, MD, MPH Department of Emergency Medicine, Center for Vascular Emergencies, Massachusetts General Hospital, Boston, MA, USA
a r t i c l e
i n f o
Article history: Received 24 September 2013 Received in revised form 8 November 2013 Accepted 8 November 2013
a b s t r a c t Introduction: Shared decision making (SDM) is a process whereby patients and clinicians work together to make informed medical decisions that incorporate patient values. Recent data suggest that, for patients with low pretest probability of pulmonary embolism (PE), doubling the standard D-dimer cutoff may reduce the need for imaging with minimal increase in missed PE diagnoses. We used an SDM approach to determine patient preferences regarding this diagnostic approach. Methods: We prospectively enrolled a consecutive sample of emergency department (ED) patients presenting with chest pain or dyspnea. We provided patients with a standardized description of the diagnostic workup for PE. We also provided image arrays describing the risks of computed tomography in low pretest probability patients and the risks of deferring imaging assuming a D-dimer was less than twice the value normally considered positive. We surveyed patients for their preference to undergo or defer imaging in this scenario. Results: We enrolled 203 ED patients. Mean age was 55 ± 17 years, and 61% were male. Seventy-four patients (37%) elected to defer computed tomography of the pulmonary arteries testing. Patients with a previous PE diagnosis were less likely to defer computed tomography of the pulmonary arteries testing (P = .007). There was no association between the decision to defer testing and age, sex, family history of PE, or self-assessed risk-taking tendency. Conclusions: When presented with a hypothetical scenario, more than one-third of patients deferred imaging for PE based on low clinical probability and a D-dimer less than twice the normal threshold. An SDM approach is acceptable to patients and may decrease imaging for PE. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Shared decision making (SDM) is a structured, patient-centered process that engages the patient and the physician in the exchange of information and selection of a particular pathway for workup or treatment [1]. The key elements of SDM include involvement of both patient and clinician, sharing of information between these parties, taking steps to build a consensus about the preferred approach to a clinical situation, and reaching an agreement about which approach to implement [2]. Shared decision making is an approach that is best suited for situations where there are multiple medically reasonable choices in a clinical situation [3]. The Institute of Medicine has identified patient-centered approaches to clinical encounters as key elements in the provision of quality care in their seminal 2001 policy statement “Crossing the Quality Chasm: A New Health System for the 21st Century” [4]. A 2011
☆ Funding: Massachusetts College of Emergency Medicine. ☆☆ Presentation: This work has not been presented. ⁎ Corresponding author. BWH/MGH Harvard Affiliated Emergency Medicine Residency Department of Emergency Medicine, Boston, MA 02115, USA. E-mail addresses: [email protected] (B.C. Geyer), [email protected] (M. Xu), [email protected] (C. Kabrhel). 0735-6757/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajem.2013.11.019
consensus conference sponsored by the Society for Academic Emergency Medicine entitled “Interventions to Assure Quality in the Crowded Emergency Department” highlighted the relevance of this approach to the realm of emergency medicine by emphasizing the importance of patient satisfaction, patient involvement, and carerelated to patient needs [5]. Specifically, the consensus conference participants point out the importance of patients becoming informed and educated while participating in and guiding decisions about their care in the emergency department (ED). Shared decision making has been shown to be feasible in the ED and, although rare in clinical emergency medicine, has been used successfully in other areas of medicine to reduce antibiotic overuse in acute respiratory infections and increase patient compliance in asthma treatment [1,6,7]. Diagnostic testing for PE, using a combination of D-dimer and contrast-enhanced computed tomography of the pulmonary arteries (CT-PA), is a prototypical opportunity for the incorporation of SDM. Imaging for PE has increased dramatically in recent years [8], and it is now estimated that 1.5% to 2% of all ED patients undergo CT-PA for suspected PE [9,10]. Although sensitive and specific for PE, only 10% of CT-PA studies confirm PE, and the test is not without consequence [10]. Estimates suggest that for every 300 computed tomographic scans performed, patients will develop 1 additional malignancy [11,12]. Furthermore, a recent study demonstrated that the likelihood
234
B.C. Geyer et al. / American Journal of Emergency Medicine 32 (2014) 233–236
of inducing contrast-induced nephropathy is at least as great as discovering a PE with CT-PA (14% vs 7%), and the development of contrast-induced nephropathy after CT-PA significantly increases the risk of severe renal failure or death from renal failure [13]. Recent studies suggest that varying the threshold defining a positive test based on a patient's pretest probability of PE may be a safe approach to decreasing CT-PA [14,15]. Kabrhel et al [14] estimated this strategy would reduce the use of CT-PA at a ratio of 55 “CT-PA avoided” to one “missed” PE. A subsequent study by Kline et al [15] suggested that the nearly all of the “missed PE” would be small and of limited clinical consequence. However, the acceptable ratio of “CT-PA avoided” to “PE missed” is not known and, because the risks of PE and CT-PA are different, should be determined based on the priorities of the patient. We hypothesized that, when provided with an SDM framework describing the risks and benefits of 2 testing strategies for PE, a substantial proportion of patients will choose to avoid CT-PA based by using a pretest probability-adjusted D-dimer threshold. We performed this study to estimate this proportion and to identify patientlevel factors associated with the decision to undergo or avoid CT-PA. We believe that the adoption of SDM in the diagnostic evaluation of PE has the potential to reduce imaging, cost, and patient risk while preserving patient safety. 2. Methods We performed this study in the ED of an urban university hospital with an annual census of 100 000 visits. We enrolled during randomly selected ED shifts representative of day, evening (07:00-23:00), and weekend shifts. During these shifts, we enrolled consecutive adult (N17 years of age) English-speaking patients presenting with a chief complaint of chest pain or dyspnea. Trained study staff identified patients by screening the electronic ED tracking system in real time and rounding at regular intervals with the ED attending physician during the shift. We included all adult patients who presented with a chief complaint of chest pain or dyspnea. We did not specifically require that a patient be worked up for PE or that PE be in the differential diagnosis. We excluded patients who were hemodynamically unstable or unable to participate in the research study due to altered mental status, intoxication, or limited English capacity. Informed consent was implied by the completion of the study survey. This study was approved by the Human Research Committee of Partners HealthCare. Research staff collected baseline demographic information, medical history (including a prior history of PE or deep vein thrombosis), and vital signs from the patient and their electronic medical record. Using a standard script that took approximately 15 minutes to complete, we presented subjects with information describing the diagnostic evaluation of PE using a combination of pretest probability assessment, D-dimer testing, and CT-PA. Subjects had the option of reading the script or having it read to them by study staff (Figure S3). The diagnostic evaluation presented is consistent with published guidelines and internal guidelines of Partners Healthcare [16]. We then asked subjects to consider a hypothetical scenario in which they have low pretest probability of PE, but the emergency physician is concerned enough about PE to order a D-dimer test, and the D-dimer result is above normal (ie, positive) but less than twice normal. Using the VIDAS D-dimer (Biomerieux, Craponne, France), this corresponds to a value of between 500 ng/mL or higher but less than 1000 ng/mL. Study staff was instructed to frequently ask patients if they had any questions about the material presented in the hypothetical scenario and to address these questions. Furthermore, they were instructed to limit the responses to information contained in the study script. We provided study subjects 2 image arrays of 1000 “similar patients” displaying the risks associated with obtaining or declining a
CT-PA (Supplementary Figures S1–S2). Colored dots were used in image arrays to convey the likelihood of 6 potential outcomes of interest: (1) no adverse effect, (2) mortality from missed PE, (3) radiation-associated malignancy, (4) contrast-induced nephropathy, (5) severe contrast allergy, and (6) iatrogenic injury from treatment for a false-positive CT-PA. Risks were based on the best available literature [11-13]. Each dot color was used once across the pair of image arrays presented to the patient. We asked patients to choose whether, given the hypothetical scenario, they would prefer to have a CT-PA or not. We measured patients' self-perceived risk taking by asking them to draw a hash mark on a 10-cm line between the phrases “I avoid risks” and “I take risks.” We also asked patients to rank the risks presented on the image array in terms of their importance to their decision to accept or decline CT-PA. Clinical staff treating the patient in the ED were blinded to the responses of the patient. 2.1. Power analysis Assuming a 30% rate of CT-PA deferral with a 10% margin of error at P b .05, we needed to enroll 96 patients. We conservatively assumed an enrollment rate of 70%, which would require us to screen 125 to 150 patients. To ensure adequate power, we chose to double the target screening population to 300 patients. 3. Results We screened 308 consecutive patients for participation, enrolling 203 study subjects (Fig.). Mean age was 55 ± 17 years, and 123 (61%) were male. The primary reasons for nonenrollment were an inability to speak English or provide consent. Patient age and sex were similar among enrolled and nonenrolled patients. After reviewing the image arrays, 128 (63%) subjects indicated that they would prefer to have CT-PA, and 74 (37%) of patients indicated that they would decline CT-PA testing if they had a D-dimer above normal but less than twice normal (500-1000 ng/mL) and low clinical probability of PE (Table). The percentage of patients with a previous PE was significantly lower in the group who declined CT-PA testing (0% vs 9.4%; P = .007), reflecting the finding that all patients with previous PE wanted follow-up CT-PA testing. There was no association between age, sex, or family history of PE (Table) and CT-PA preference. There was no significant difference between selfassessed risk taking behavior and CT-PA preference.
Fig. Enrollment scheme and application of exclusion criteria for assessment of SDM process for evaluation of PE in the ED.
B.C. Geyer et al. / American Journal of Emergency Medicine 32 (2014) 233–236 Table Patient demographics and justification for response between patients accepting and declining CT-PA testing in SDM process Decline CT-PA (n = 74) Accept CT-PA (n = 128) P Demographics Age (y) 57.7 Male (%) 56.8 Risk taking score 5.01 PMH PE 0 FH PE 0.014 Justification for response Cancer (%) 27 (36.5) Kidney (%) 24 (32.4) Allergy (%) 3 (4.1) None of the above (%) 20 (14.8) Missed PE (%) – Other dx (%) –
54.2 62.5 5.39 0.0938 0.0547
.151 .212 .319 .007 .149
– – – – 109 (85.2) 19 (14.8)
PMH, past medical history; FH = family history; dx = diagnosis.
Among patients who declined CT-PA testing, the most commonly chosen factors influencing their decision were risk of malignancy (n = 20 [27%]), contrast-induced nephropathy (n = 18 [24%]), and risk of allergic reaction (n = 2 [3%]). Twenty patients (27.0%) declining CT-PA gave a response other than concerns about malignancy, nephropathy, or allergy. The most common of these responses was a belief that CT-PA was unnecessary, independent of the complications. Fourteen patients (20%) did not provide any justification for their decision to defer CT-PA testing. Among patients who accepted CT-PA testing, most patients (n = 109 [85%]) cited concerns about the risk of missing a PE, whereas 19 patients (15%) were principally concerned about the possibility of missing an alternate diagnosis that might be discovered on CT-PA. Although this study was not designed or powered to address a potential change in the ED workup of PE, we were interested in knowing whether any changes in practice arose as a result of our SDM survey. To assess this, we followed the workup of patients in the ED. Of the 203 study participants, 44 had a D-dimer ordered, 10 (23%) of which were between 500 and 1000 ng/mL. Eight of these patients went on to have a CT-PA. Of the 2 patients who did not have a follow-up CT-PA, 1 had severe renal impairment and therefore could not receive intravenous contrast, and 1 was unable to tolerate supine positioning for the test. We found no instances where the hypothetical SDM intervention changed the workup of the study population. 4. Discussion This study provides evidence that more than one-third of patients presenting to the ED with the principal symptoms of PE would defer CT-PA testing in a hypothetical scenario where they have low pretest probability and have a D-dimer less than twice normal. This is the first study to assess patient preferences for this approach and is an initial demonstration of the feasibility of a SDM approach to this question. Previous work by Kabrhel et al [14] showed that adjusting the D-dimer threshold has the potential to significantly reduce CT-PA use, at a ratio of 55 CT-PA avoided for every PE missed. A follow-up study found that more than 90% of PE “missed” using this strategy would be small—segmental or subsegmental—with no residual deep vein thrombosis that would put patients at risk for early PE recurrence [15]. Smaller retrospective studies have also demonstrated no loss in sensitivity with a doubled D-dimer threshold, even when pretest probability is not considered [17]. Ours is the first study to document patient agreement with this approach. Seen on a national scale, the potential benefit of applying an SDM approach to the use of CT-PA is large. Nationally representative studies estimate that among the 110 million patients seen in United
235
States EDs each year, 2.6 million (2.4%) undergo testing for PE, and approximately 1.8 million (70%) of these have low pretest probability. Currently, approximately 1.6 million patients (57%) undergo CT-PA each year [9,10,13]. Our previous work suggests that 15% of D-dimer results are positive, but less than twice normal. Extrapolating these results and those of our survey to national data, we estimate that using an SDM approach to PE imaging could reduce the number of CT-PA by 98 280 annually. 2 600 000 PE Evaluations 0:7 low protest probability 0:15 borderline positive D‐dimer 0:36 decline CT‐PA ¼ 98 280 CT‐PA avoided To serve the broad aims of optimal patient care and good stewardship of health care resources, an approach to reduce CT-PA testing should ideally have direct patient involvement. Shared decision making is an approach that is best suited for situations where there are multiple medically reasonable choices in a clinical situation [3]. Thus, the optimal choice may be dictated by patient values, experiences, and assessment of the risk/benefit relationship of each particular path. This will likely lead to greater variation in care based on the patient's wants and needs, which is an essential indicator of high-quality care [18]. This variation may be driven by patient demographics, particularly age, level of education, disease burden, and the disease entity in question [19,20]. Interventions based on an SDM approach have been used in the ED workup of chest pain as well as in emergency pediatrics [21-23]. 4.1. Limitations We used an SDM approach based on image arrays to rapidly communicate the risks of each diagnostic path proposed in this study. Image arrays have been shown to be easily understandable by patients with low literacy and numeracy, making this format ideal for communicating with patients with diverse levels of education and healthcare experience [24]. However, any study using an SDM approach in a standardized way must decide which information to convey. Although our arrays are consistent with the most recent literature, it is possible that we did not include risks or benefits of CT-PA that patients may have found important. This study involved surveying patients regarding their hypothetical preferences for PE imaging. We enrolled patients with the principal symptoms of PE (chest pain and dyspnea) but did not require that a PE workup be planned or underway. Assessing patient preferences within the safe construct of a hypothetical provides for the safe care of patients while reasonably assessing outcomes. However, it is possible that patients would have answered questions differently than if their decisions were actually being used to decide whether imaging would be performed. The SDM script required approximately 5 minutes to read, and the overall process required about 15 minutes of discussion, although this naturally varied from patient to patient. Importantly, the survey was conducted by research coordinators trained and directly supervised by physician study investigators. Finally, we did not determine whether physicians would agree with or abide by patient preferences. 5. Conclusions When presented with a scenario assuming low clinical probability and a D-dimer less than twice the normal threshold, more than onethird of patients would choose to decline imaging for PE. This study shows that an SDM approach is acceptable to patients and may decrease imaging for PE if incorporated into clinical practice. Empiric proof of this concept is warranted. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ajem.2013.11.019.
236
B.C. Geyer et al. / American Journal of Emergency Medicine 32 (2014) 233–236
References [1] Flynn D, et al. Engaging patients in health care decisions in the emergency department through shared decision-making: a systematic review. Acad Emerg Med 2012;19(8):959–67. [2] Charles C, Whelan T, Gafni A. What do we mean by partnership in making decisions about treatment? BMJ 1999;319(7212):780–2. [3] Whitney SN, McGuire AL, McCullough LB. A typology of shared decision making, informed consent, and simple consent. Ann Intern Med 2004;140(1):54–9. [4] Institute of Medicine (U.S.). Committee on Quality of Health Care in America. Crossing the quality chasm: a new health system for the 21st century. Washington, D.C: National Academy Press; 2001. xx, 337 p. [5] Pham JC, et al. Interventions to improve patient-centered care during times of emergency department crowding. Acad Emerg Med 2011;18(12):1289–94. [6] Legare F, et al. Training family physicians in shared decision-making to reduce the overuse of antibiotics in acute respiratory infections: a cluster randomized trial. CMAJ 2012;184(13):E726–34. [7] Wilson SR, et al. Shared treatment decision making improves adherence and outcomes in poorly controlled asthma. Am J Respir Crit Care Med 2010;181(6): 566–77. [8] Winer-Muram HT, et al. Suspected acute pulmonary embolism: evaluation with multi-detector row CT versus digital subtraction pulmonary arteriography. Radiology 2004;233(3):806–15. [9] Kline JA, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6(5):772–80. [10] Raja AS, et al. Effect of computerized clinical decision support on the use and yield of CT pulmonary angiography in the emergency department. Radiology 2012; 262(2):468–74. [11] Einstein AJ, et al. Radiation dose to patients from cardiac diagnostic imaging. Circulation 2007;116(11):1290–305. [12] Hurwitz LM, et al. Radiation dose from contemporary cardiothoracic multidetector CT protocols with an anthropomorphic female phantom: implications for cancer induction. Radiology 2007;245(3):742–50.
[13] Mitchell AM, et al. Prospective study of the incidence of contrast-induced nephropathy among patients evaluated for pulmonary embolism by contrastenhanced computed tomography. Acad Emerg Med 2012;19(6):618–25. [14] Kabrhel C, et al. Potential impact of adjusting the threshold of the quantitative Ddimer based on pretest probability of acute pulmonary embolism. Acad Emerg Med 2009;16(4):325–32. [15] Kline JA, et al. D-dimer threshold increase with pretest probability unlikely for pulmonary embolism to decrease unnecessary computerized tomographic pulmonary angiography. J Thromb Haemost 2012;10(4):572–81. [16] Fesmire FM, et al. Critical issues in the evaluation and management of adult patients presenting to the emergency department with suspected pulmonary embolism. Ann Emerg Med 2011;57(6):628e75–52e75. [17] Vossen JA, et al. Clinical usefulness of adjusted D-dimer cut-off values to exclude pulmonary embolism in a community hospital emergency department patient population. Acta Radiol 2012;53(7):765–8. [18] Sepucha KR, Fowler Jr FJ, Mulley Jr AG. Policy support for patient-centered care: the need for measurable improvements in decision quality. Health Aff (Millwood) 2004 Suppl Variation: p. VAR54-62. [19] Deber RB, et al. Do people want to be autonomous patients? Preferred roles in treatment decision-making in several patient populations. Health Expect 2007; 10(3):248–58. [20] Say R, Murtagh M, Thomson R. Patients' preference for involvement in medical decision making: a narrative review. Patient Educ Couns 2006;60(2):102–14. [21] Karpas A, Finkelstein M, Reid S. Parental preference for rehydration method for children in the emergency department. Pediatr Emerg Care 2009;25(5): 301–6. [22] Kline JA, et al. Randomized trial of computerized quantitative pretest probability in low-risk chest pain patients: effect on safety and resource use. Ann Emerg Med 2009;53(6):727e1–35e1. [23] Hess EP, et al. The chest pain choice decision aid: a randomized trial. Circ Cardiovasc Qual Outcomes 2012;5(3):251–9. [24] Galesic M, Garcia-Retamero R, Gigerenzer G. Using icon arrays to communicate medical risks: overcoming low numeracy. Health Psychol 2009;28(2):210–6.
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Patient preferences for testing for pulmonary embolism in the ED using a shared decision-making model☆,☆☆ Brian C. Geyer, MD, PhD, MPH ⁎, Maria Xu, BS, Christopher Kabrhel, MD, MPH Department of Emergency Medicine, Center for Vascular Emergencies, Massachusetts General Hospital, Boston, MA, USA
a r t i c l e
i n f o
Article history: Received 24 September 2013 Received in revised form 8 November 2013 Accepted 8 November 2013
a b s t r a c t Introduction: Shared decision making (SDM) is a process whereby patients and clinicians work together to make informed medical decisions that incorporate patient values. Recent data suggest that, for patients with low pretest probability of pulmonary embolism (PE), doubling the standard D-dimer cutoff may reduce the need for imaging with minimal increase in missed PE diagnoses. We used an SDM approach to determine patient preferences regarding this diagnostic approach. Methods: We prospectively enrolled a consecutive sample of emergency department (ED) patients presenting with chest pain or dyspnea. We provided patients with a standardized description of the diagnostic workup for PE. We also provided image arrays describing the risks of computed tomography in low pretest probability patients and the risks of deferring imaging assuming a D-dimer was less than twice the value normally considered positive. We surveyed patients for their preference to undergo or defer imaging in this scenario. Results: We enrolled 203 ED patients. Mean age was 55 ± 17 years, and 61% were male. Seventy-four patients (37%) elected to defer computed tomography of the pulmonary arteries testing. Patients with a previous PE diagnosis were less likely to defer computed tomography of the pulmonary arteries testing (P = .007). There was no association between the decision to defer testing and age, sex, family history of PE, or self-assessed risk-taking tendency. Conclusions: When presented with a hypothetical scenario, more than one-third of patients deferred imaging for PE based on low clinical probability and a D-dimer less than twice the normal threshold. An SDM approach is acceptable to patients and may decrease imaging for PE. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Shared decision making (SDM) is a structured, patient-centered process that engages the patient and the physician in the exchange of information and selection of a particular pathway for workup or treatment [1]. The key elements of SDM include involvement of both patient and clinician, sharing of information between these parties, taking steps to build a consensus about the preferred approach to a clinical situation, and reaching an agreement about which approach to implement [2]. Shared decision making is an approach that is best suited for situations where there are multiple medically reasonable choices in a clinical situation [3]. The Institute of Medicine has identified patient-centered approaches to clinical encounters as key elements in the provision of quality care in their seminal 2001 policy statement “Crossing the Quality Chasm: A New Health System for the 21st Century” [4]. A 2011
☆ Funding: Massachusetts College of Emergency Medicine. ☆☆ Presentation: This work has not been presented. ⁎ Corresponding author. BWH/MGH Harvard Affiliated Emergency Medicine Residency Department of Emergency Medicine, Boston, MA 02115, USA. E-mail addresses: [email protected] (B.C. Geyer), [email protected] (M. Xu), [email protected] (C. Kabrhel). 0735-6757/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajem.2013.11.019
consensus conference sponsored by the Society for Academic Emergency Medicine entitled “Interventions to Assure Quality in the Crowded Emergency Department” highlighted the relevance of this approach to the realm of emergency medicine by emphasizing the importance of patient satisfaction, patient involvement, and carerelated to patient needs [5]. Specifically, the consensus conference participants point out the importance of patients becoming informed and educated while participating in and guiding decisions about their care in the emergency department (ED). Shared decision making has been shown to be feasible in the ED and, although rare in clinical emergency medicine, has been used successfully in other areas of medicine to reduce antibiotic overuse in acute respiratory infections and increase patient compliance in asthma treatment [1,6,7]. Diagnostic testing for PE, using a combination of D-dimer and contrast-enhanced computed tomography of the pulmonary arteries (CT-PA), is a prototypical opportunity for the incorporation of SDM. Imaging for PE has increased dramatically in recent years [8], and it is now estimated that 1.5% to 2% of all ED patients undergo CT-PA for suspected PE [9,10]. Although sensitive and specific for PE, only 10% of CT-PA studies confirm PE, and the test is not without consequence [10]. Estimates suggest that for every 300 computed tomographic scans performed, patients will develop 1 additional malignancy [11,12]. Furthermore, a recent study demonstrated that the likelihood
234
B.C. Geyer et al. / American Journal of Emergency Medicine 32 (2014) 233–236
of inducing contrast-induced nephropathy is at least as great as discovering a PE with CT-PA (14% vs 7%), and the development of contrast-induced nephropathy after CT-PA significantly increases the risk of severe renal failure or death from renal failure [13]. Recent studies suggest that varying the threshold defining a positive test based on a patient's pretest probability of PE may be a safe approach to decreasing CT-PA [14,15]. Kabrhel et al [14] estimated this strategy would reduce the use of CT-PA at a ratio of 55 “CT-PA avoided” to one “missed” PE. A subsequent study by Kline et al [15] suggested that the nearly all of the “missed PE” would be small and of limited clinical consequence. However, the acceptable ratio of “CT-PA avoided” to “PE missed” is not known and, because the risks of PE and CT-PA are different, should be determined based on the priorities of the patient. We hypothesized that, when provided with an SDM framework describing the risks and benefits of 2 testing strategies for PE, a substantial proportion of patients will choose to avoid CT-PA based by using a pretest probability-adjusted D-dimer threshold. We performed this study to estimate this proportion and to identify patientlevel factors associated with the decision to undergo or avoid CT-PA. We believe that the adoption of SDM in the diagnostic evaluation of PE has the potential to reduce imaging, cost, and patient risk while preserving patient safety. 2. Methods We performed this study in the ED of an urban university hospital with an annual census of 100 000 visits. We enrolled during randomly selected ED shifts representative of day, evening (07:00-23:00), and weekend shifts. During these shifts, we enrolled consecutive adult (N17 years of age) English-speaking patients presenting with a chief complaint of chest pain or dyspnea. Trained study staff identified patients by screening the electronic ED tracking system in real time and rounding at regular intervals with the ED attending physician during the shift. We included all adult patients who presented with a chief complaint of chest pain or dyspnea. We did not specifically require that a patient be worked up for PE or that PE be in the differential diagnosis. We excluded patients who were hemodynamically unstable or unable to participate in the research study due to altered mental status, intoxication, or limited English capacity. Informed consent was implied by the completion of the study survey. This study was approved by the Human Research Committee of Partners HealthCare. Research staff collected baseline demographic information, medical history (including a prior history of PE or deep vein thrombosis), and vital signs from the patient and their electronic medical record. Using a standard script that took approximately 15 minutes to complete, we presented subjects with information describing the diagnostic evaluation of PE using a combination of pretest probability assessment, D-dimer testing, and CT-PA. Subjects had the option of reading the script or having it read to them by study staff (Figure S3). The diagnostic evaluation presented is consistent with published guidelines and internal guidelines of Partners Healthcare [16]. We then asked subjects to consider a hypothetical scenario in which they have low pretest probability of PE, but the emergency physician is concerned enough about PE to order a D-dimer test, and the D-dimer result is above normal (ie, positive) but less than twice normal. Using the VIDAS D-dimer (Biomerieux, Craponne, France), this corresponds to a value of between 500 ng/mL or higher but less than 1000 ng/mL. Study staff was instructed to frequently ask patients if they had any questions about the material presented in the hypothetical scenario and to address these questions. Furthermore, they were instructed to limit the responses to information contained in the study script. We provided study subjects 2 image arrays of 1000 “similar patients” displaying the risks associated with obtaining or declining a
CT-PA (Supplementary Figures S1–S2). Colored dots were used in image arrays to convey the likelihood of 6 potential outcomes of interest: (1) no adverse effect, (2) mortality from missed PE, (3) radiation-associated malignancy, (4) contrast-induced nephropathy, (5) severe contrast allergy, and (6) iatrogenic injury from treatment for a false-positive CT-PA. Risks were based on the best available literature [11-13]. Each dot color was used once across the pair of image arrays presented to the patient. We asked patients to choose whether, given the hypothetical scenario, they would prefer to have a CT-PA or not. We measured patients' self-perceived risk taking by asking them to draw a hash mark on a 10-cm line between the phrases “I avoid risks” and “I take risks.” We also asked patients to rank the risks presented on the image array in terms of their importance to their decision to accept or decline CT-PA. Clinical staff treating the patient in the ED were blinded to the responses of the patient. 2.1. Power analysis Assuming a 30% rate of CT-PA deferral with a 10% margin of error at P b .05, we needed to enroll 96 patients. We conservatively assumed an enrollment rate of 70%, which would require us to screen 125 to 150 patients. To ensure adequate power, we chose to double the target screening population to 300 patients. 3. Results We screened 308 consecutive patients for participation, enrolling 203 study subjects (Fig.). Mean age was 55 ± 17 years, and 123 (61%) were male. The primary reasons for nonenrollment were an inability to speak English or provide consent. Patient age and sex were similar among enrolled and nonenrolled patients. After reviewing the image arrays, 128 (63%) subjects indicated that they would prefer to have CT-PA, and 74 (37%) of patients indicated that they would decline CT-PA testing if they had a D-dimer above normal but less than twice normal (500-1000 ng/mL) and low clinical probability of PE (Table). The percentage of patients with a previous PE was significantly lower in the group who declined CT-PA testing (0% vs 9.4%; P = .007), reflecting the finding that all patients with previous PE wanted follow-up CT-PA testing. There was no association between age, sex, or family history of PE (Table) and CT-PA preference. There was no significant difference between selfassessed risk taking behavior and CT-PA preference.
Fig. Enrollment scheme and application of exclusion criteria for assessment of SDM process for evaluation of PE in the ED.
B.C. Geyer et al. / American Journal of Emergency Medicine 32 (2014) 233–236 Table Patient demographics and justification for response between patients accepting and declining CT-PA testing in SDM process Decline CT-PA (n = 74) Accept CT-PA (n = 128) P Demographics Age (y) 57.7 Male (%) 56.8 Risk taking score 5.01 PMH PE 0 FH PE 0.014 Justification for response Cancer (%) 27 (36.5) Kidney (%) 24 (32.4) Allergy (%) 3 (4.1) None of the above (%) 20 (14.8) Missed PE (%) – Other dx (%) –
54.2 62.5 5.39 0.0938 0.0547
.151 .212 .319 .007 .149
– – – – 109 (85.2) 19 (14.8)
PMH, past medical history; FH = family history; dx = diagnosis.
Among patients who declined CT-PA testing, the most commonly chosen factors influencing their decision were risk of malignancy (n = 20 [27%]), contrast-induced nephropathy (n = 18 [24%]), and risk of allergic reaction (n = 2 [3%]). Twenty patients (27.0%) declining CT-PA gave a response other than concerns about malignancy, nephropathy, or allergy. The most common of these responses was a belief that CT-PA was unnecessary, independent of the complications. Fourteen patients (20%) did not provide any justification for their decision to defer CT-PA testing. Among patients who accepted CT-PA testing, most patients (n = 109 [85%]) cited concerns about the risk of missing a PE, whereas 19 patients (15%) were principally concerned about the possibility of missing an alternate diagnosis that might be discovered on CT-PA. Although this study was not designed or powered to address a potential change in the ED workup of PE, we were interested in knowing whether any changes in practice arose as a result of our SDM survey. To assess this, we followed the workup of patients in the ED. Of the 203 study participants, 44 had a D-dimer ordered, 10 (23%) of which were between 500 and 1000 ng/mL. Eight of these patients went on to have a CT-PA. Of the 2 patients who did not have a follow-up CT-PA, 1 had severe renal impairment and therefore could not receive intravenous contrast, and 1 was unable to tolerate supine positioning for the test. We found no instances where the hypothetical SDM intervention changed the workup of the study population. 4. Discussion This study provides evidence that more than one-third of patients presenting to the ED with the principal symptoms of PE would defer CT-PA testing in a hypothetical scenario where they have low pretest probability and have a D-dimer less than twice normal. This is the first study to assess patient preferences for this approach and is an initial demonstration of the feasibility of a SDM approach to this question. Previous work by Kabrhel et al [14] showed that adjusting the D-dimer threshold has the potential to significantly reduce CT-PA use, at a ratio of 55 CT-PA avoided for every PE missed. A follow-up study found that more than 90% of PE “missed” using this strategy would be small—segmental or subsegmental—with no residual deep vein thrombosis that would put patients at risk for early PE recurrence [15]. Smaller retrospective studies have also demonstrated no loss in sensitivity with a doubled D-dimer threshold, even when pretest probability is not considered [17]. Ours is the first study to document patient agreement with this approach. Seen on a national scale, the potential benefit of applying an SDM approach to the use of CT-PA is large. Nationally representative studies estimate that among the 110 million patients seen in United
235
States EDs each year, 2.6 million (2.4%) undergo testing for PE, and approximately 1.8 million (70%) of these have low pretest probability. Currently, approximately 1.6 million patients (57%) undergo CT-PA each year [9,10,13]. Our previous work suggests that 15% of D-dimer results are positive, but less than twice normal. Extrapolating these results and those of our survey to national data, we estimate that using an SDM approach to PE imaging could reduce the number of CT-PA by 98 280 annually. 2 600 000 PE Evaluations 0:7 low protest probability 0:15 borderline positive D‐dimer 0:36 decline CT‐PA ¼ 98 280 CT‐PA avoided To serve the broad aims of optimal patient care and good stewardship of health care resources, an approach to reduce CT-PA testing should ideally have direct patient involvement. Shared decision making is an approach that is best suited for situations where there are multiple medically reasonable choices in a clinical situation [3]. Thus, the optimal choice may be dictated by patient values, experiences, and assessment of the risk/benefit relationship of each particular path. This will likely lead to greater variation in care based on the patient's wants and needs, which is an essential indicator of high-quality care [18]. This variation may be driven by patient demographics, particularly age, level of education, disease burden, and the disease entity in question [19,20]. Interventions based on an SDM approach have been used in the ED workup of chest pain as well as in emergency pediatrics [21-23]. 4.1. Limitations We used an SDM approach based on image arrays to rapidly communicate the risks of each diagnostic path proposed in this study. Image arrays have been shown to be easily understandable by patients with low literacy and numeracy, making this format ideal for communicating with patients with diverse levels of education and healthcare experience [24]. However, any study using an SDM approach in a standardized way must decide which information to convey. Although our arrays are consistent with the most recent literature, it is possible that we did not include risks or benefits of CT-PA that patients may have found important. This study involved surveying patients regarding their hypothetical preferences for PE imaging. We enrolled patients with the principal symptoms of PE (chest pain and dyspnea) but did not require that a PE workup be planned or underway. Assessing patient preferences within the safe construct of a hypothetical provides for the safe care of patients while reasonably assessing outcomes. However, it is possible that patients would have answered questions differently than if their decisions were actually being used to decide whether imaging would be performed. The SDM script required approximately 5 minutes to read, and the overall process required about 15 minutes of discussion, although this naturally varied from patient to patient. Importantly, the survey was conducted by research coordinators trained and directly supervised by physician study investigators. Finally, we did not determine whether physicians would agree with or abide by patient preferences. 5. Conclusions When presented with a scenario assuming low clinical probability and a D-dimer less than twice the normal threshold, more than onethird of patients would choose to decline imaging for PE. This study shows that an SDM approach is acceptable to patients and may decrease imaging for PE if incorporated into clinical practice. Empiric proof of this concept is warranted. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ajem.2013.11.019.
236
B.C. Geyer et al. / American Journal of Emergency Medicine 32 (2014) 233–236
References [1] Flynn D, et al. Engaging patients in health care decisions in the emergency department through shared decision-making: a systematic review. Acad Emerg Med 2012;19(8):959–67. [2] Charles C, Whelan T, Gafni A. What do we mean by partnership in making decisions about treatment? BMJ 1999;319(7212):780–2. [3] Whitney SN, McGuire AL, McCullough LB. A typology of shared decision making, informed consent, and simple consent. Ann Intern Med 2004;140(1):54–9. [4] Institute of Medicine (U.S.). Committee on Quality of Health Care in America. Crossing the quality chasm: a new health system for the 21st century. Washington, D.C: National Academy Press; 2001. xx, 337 p. [5] Pham JC, et al. Interventions to improve patient-centered care during times of emergency department crowding. Acad Emerg Med 2011;18(12):1289–94. [6] Legare F, et al. Training family physicians in shared decision-making to reduce the overuse of antibiotics in acute respiratory infections: a cluster randomized trial. CMAJ 2012;184(13):E726–34. [7] Wilson SR, et al. Shared treatment decision making improves adherence and outcomes in poorly controlled asthma. Am J Respir Crit Care Med 2010;181(6): 566–77. [8] Winer-Muram HT, et al. Suspected acute pulmonary embolism: evaluation with multi-detector row CT versus digital subtraction pulmonary arteriography. Radiology 2004;233(3):806–15. [9] Kline JA, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6(5):772–80. [10] Raja AS, et al. Effect of computerized clinical decision support on the use and yield of CT pulmonary angiography in the emergency department. Radiology 2012; 262(2):468–74. [11] Einstein AJ, et al. Radiation dose to patients from cardiac diagnostic imaging. Circulation 2007;116(11):1290–305. [12] Hurwitz LM, et al. Radiation dose from contemporary cardiothoracic multidetector CT protocols with an anthropomorphic female phantom: implications for cancer induction. Radiology 2007;245(3):742–50.
[13] Mitchell AM, et al. Prospective study of the incidence of contrast-induced nephropathy among patients evaluated for pulmonary embolism by contrastenhanced computed tomography. Acad Emerg Med 2012;19(6):618–25. [14] Kabrhel C, et al. Potential impact of adjusting the threshold of the quantitative Ddimer based on pretest probability of acute pulmonary embolism. Acad Emerg Med 2009;16(4):325–32. [15] Kline JA, et al. D-dimer threshold increase with pretest probability unlikely for pulmonary embolism to decrease unnecessary computerized tomographic pulmonary angiography. J Thromb Haemost 2012;10(4):572–81. [16] Fesmire FM, et al. Critical issues in the evaluation and management of adult patients presenting to the emergency department with suspected pulmonary embolism. Ann Emerg Med 2011;57(6):628e75–52e75. [17] Vossen JA, et al. Clinical usefulness of adjusted D-dimer cut-off values to exclude pulmonary embolism in a community hospital emergency department patient population. Acta Radiol 2012;53(7):765–8. [18] Sepucha KR, Fowler Jr FJ, Mulley Jr AG. Policy support for patient-centered care: the need for measurable improvements in decision quality. Health Aff (Millwood) 2004 Suppl Variation: p. VAR54-62. [19] Deber RB, et al. Do people want to be autonomous patients? Preferred roles in treatment decision-making in several patient populations. Health Expect 2007; 10(3):248–58. [20] Say R, Murtagh M, Thomson R. Patients' preference for involvement in medical decision making: a narrative review. Patient Educ Couns 2006;60(2):102–14. [21] Karpas A, Finkelstein M, Reid S. Parental preference for rehydration method for children in the emergency department. Pediatr Emerg Care 2009;25(5): 301–6. [22] Kline JA, et al. Randomized trial of computerized quantitative pretest probability in low-risk chest pain patients: effect on safety and resource use. Ann Emerg Med 2009;53(6):727e1–35e1. [23] Hess EP, et al. The chest pain choice decision aid: a randomized trial. Circ Cardiovasc Qual Outcomes 2012;5(3):251–9. [24] Galesic M, Garcia-Retamero R, Gigerenzer G. Using icon arrays to communicate medical risks: overcoming low numeracy. Health Psychol 2009;28(2):210–6.
