Author + information
- Received March 27, 2018
- Accepted March 28, 2018
- Published online September 3, 2018.
- Jeffrey M. Levsky, MD, PhDa,b,∗ (, )
- Linda B. Haramati, MD, MSa,c,
- Daniel M. Spevack, MD, MSb,
- Mark A. Menegus, MDb,
- Terence Chen, MPHa,
- Sarah Mizrachi, BAa,
- Durline Brown-Manhertz, DNPb,
- Samantha Selesny, MDa,
- Rikah Lerer, MDa,
- Deborah J. White, MD, MBAd,
- Jonathan N. Tobin, PhDe,f,
- Cynthia C. Taub, MDb and
- Mario J. Garcia, MDa,b
- aDepartment of Radiology, Division of Cardiothoracic Imaging, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, New York
- bDepartment of Internal Medicine, Division of Cardiology, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, New York
- cDepartment of Internal Medicine, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, New York
- dDepartment of Emergency Medicine, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, New York
- eDepartment of Epidemiology and Population Health, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, New York
- fClinical Directors Network (CDN), New York, New York
- ↵∗Address for correspondence:
Dr. Jeffrey M. Levsky, Department of Radiology, Montefiore Medical Center/Albert Einstein College of Medicine, 111 East 210th Street, Bronx, New York 10467-2490.
Objectives This study sought to compare early emergency department (ED) use of coronary computed tomography angiography (CTA) and stress echocardiography (SE) head-to-head.
Background Coronary CTA has been promoted as the early ED chest pain triage imaging method of choice, whereas SE is often overlooked in this setting and involves no ionizing radiation.
Methods The authors randomized 400 consecutive low- to intermediate-risk ED acute chest pain patients without known coronary artery disease and a negative initial serum troponin level to immediate coronary CTA (n = 201) or SE (n = 199). The primary endpoint was hospitalization rate. Secondary endpoints were ED and hospital length of stay. Safety endpoints included cardiovascular events and radiation exposure.
Results Mean patient age was 55 years, with 43% women and predominantly ethnic minorities (46% Hispanics, 32% African Americans). Thirty-nine coronary CTA patients (19%) and 22 SE patients (11%) were hospitalized at presentation (difference 8%; 95% confidence interval: 1% to 15%; p = 0.026). Median ED length of stay for discharged patients was 5.4 h (interquartile range [IQR]: 4.2 to 6.4 h) for coronary CTA and 4.7 h (IQR: 3.5 to 6.0 h) for SE (p < 0.001). Median hospital length of stay was 58 h (IQR: 50 to 102 h) for coronary CTA and 34 h (IQR: 31 to 54 h) for SE (p = 0.002). There were 11 and 7 major adverse cardiovascular events for coronary CTA and SE, respectively (p = 0.47), over a median 24 months of follow-up. Median/mean complete initial work-up radiation exposure was 6.5/7.7 mSv for coronary CTA and 0/0.96 mSv for SE (p < 0.001).
Conclusions The use of SE resulted in the hospitalization of a smaller proportion of patients with a shorter length of stay than coronary CTA and was safe. SE should be considered an appropriate option for ED chest pain triage (Stress Echocardiography and Heart Computed Tomography [CT] Scan in Emergency Department Patients With Chest Pain; NCT01384448)
- acute chest pain
- admission rate
- coronary CT angiography
- emergency department
- length of stay
- stress echocardiography
Cardiovascular disease accounts for 1 of 3 deaths worldwide, which significantly exceeds any other cause (1,2). The most prevalent cause of cardiovascular mortality is coronary artery disease (2,3). Chest pain is the second most common reason to seek emergency department (ED) treatment in the United States (4). Because of the high prevalence and significant mortality associated with coronary artery disease (3), angina and acute coronary syndrome are often the first diagnostic considerations, and noninvasive cardiac testing is commonly performed. There is avid interest in streamlining ED care to rapidly discharge the large majority of patients with chest pain caused by benign conditions (5).
Determining the optimal modality for chest pain triage is a goal for comparative effectiveness research (6) and has been informed by several recent clinical trials. Early coronary computed tomography angiography (CTA) safely reduces length of stay compared with standard care with functional testing (7–11). Coronary CTA has a distinct speed advantage over stress testing because it is the sole modality considered appropriate in an early-assessment pathway for patients with a normal initial troponin level (12). Coronary CTA reduces radiation exposure compared with radionuclide perfusion imaging (8,10,13). As a result, ED coronary CTA use is increasing rapidly, and all forms of ED functional testing are decreasing (14).
Outstanding concerns regarding coronary CTA use include radiation exposure (10,11,15), increased subsequent noninvasive testing because of detection of intermediate-severity stenosis (7,10,14,16), increased catheterization and coronary revascularization of uncertain benefit (14,16,17), increased downstream clinical resource utilization (14,16), and the burden inherent to incidental findings (18). Each of these concerns is addressed by the alternative use of stress echocardiography (SE) (14,16), a modality that has been assessed in early ED triage (19–22) but is often overlooked. Our goal was to compare the triage efficiency and safety of coronary CTA and SE in a randomized clinical trial.
This study was an open-label, randomized, controlled comparative effectiveness trial between early coronary CTA and early SE in ED chest pain patients with low to intermediate risk who clinically required noninvasive imaging for triage. Permuted block randomization (1:1) was performed using statistical software–generated codes in opaque, sealed, sequentially numbered envelopes for allocation concealment (stored in a locked office and only opened after patient consent was completed by trial coordinators). After randomly assigned imaging, all clinical decisions were made by managing physicians without restriction based on all available data, including the results of randomized testing; there was no pre-specified management algorithm. The study rationale and design were reported previously (23). The trial was monitored by our institutional review board, was HIPAA (Health Insurance Portability and Accountability Act) compliant, was overseen by an independent, extra-institutional data and safety monitoring board, and was registered at ClinicalTrials.gov (NCT01384448).
Consecutive ED patients with acute chest pain or pressure for whom noninvasive testing was requested for clinical management were screened from August 2011 to January 2016 (completion of planned recruitment) at our inner-city medical center during the daytime on weekdays. Patients without known coronary artery disease were potentially eligible if the pre-test probability of significant coronary stenosis was 10% to 90% (24). An initial negative serum troponin level at least 8 h after symptom onset and subsequent resolution of chest pain were required for eligibility. Exclusion criteria were as follows: 1) elevated serum troponin T; 2) electrocardiogram demonstrating acute ischemia or myocardial infarction; 3) hemodynamic instability; 4) unremitting or crescendo chest pain; 5) contraindications to coronary CTA, including renal insufficiency, active asthma, recent contrast administration, poor venous access, dysrhythmia that precluded cardiac gating, and intravenous contrast or other serious allergy; 6) contraindications to SE, including known inability to perform treadmill exercise, severe valvular disease, severe hypertension, and recent beta-blocker pharmacotherapy; and 7) coronary CTA, SE, or cardiac catheterization within the previous 6 months (because subsequent evaluation should be influenced by recent prior testing). All patients provided their own written, informed consent.
Coronary CTA was performed on single-source, 64-detector-row scanners according to established guidelines (25). Heart rate control was achieved by oral followed by intravenous metoprolol tartrate when needed. Coronary calcium scoring was performed; the presence of heavy calcification was assessed in real time and was grounds for termination of the study at the discretion of the monitoring physician. Post-contrast angiography with prospective gating was performed by default. Retrospective gating with tube current modulation was used for patients without adequate response to heart rate control. Tube voltage and current were tailored for dose reduction as allowed by body habitus. Studies were each interpreted by 1 member of a group of fellowship-trained cardiothoracic radiologists with modality-specific advanced training.
SE was performed with an integrated treadmill and cardiac monitoring system according to established guidelines (26). Left ventricular opacification contrast agents were used when necessary for adequate visualization. A full resting echocardiogram was performed; resting wall motion abnormalities or severe valvular disease assessed in real time were grounds for early termination of the study at the discretion of the monitoring physician. The default stressor was Bruce protocol treadmill exercise. Patients found unable to complete the planned exercise protocol received staged dobutamine/atropine infusion. Images were acquired at maximal heart rate and during recovery. Studies were each interpreted by one member of a group of expert cardiologists with modality-specific certification.
The goal of ED noninvasive cardiac testing is to select patients who are safe for immediate discharge and to identify patients in need of further work-up and potential revascularization (5,12). The primary endpoint was hospitalization rate, a measure of efficiency of triage, as determined by physician admission order. The secondary endpoint was length of stay. For discharged patients, ED stay was defined as time from randomization to discharge. Hospital stay, when applicable, was defined as the time from physician admission order to discharge.
Safety outcomes included death, nonfatal major adverse cardiovascular events (myocardial infarction, cerebrovascular accident, and cardiac arrest), complications from noninvasive testing or revascularization, and radiation exposure. Radiation dose from coronary CTA was estimated from dose-length product with a conversion factor of 0.017 mSv/mGy·cm (27). Doses from diagnostic imaging and catheterization were determined with publicly available calculators (28,29).
Clinical resource utilization (hospitalization, ED visit, cardiologist visit, primary care provider visit, changes in pharmacotherapy) was assessed. A post hoc analysis of the clinical reason for initial hospital admission was performed by electronic chart review. The subjective experience of imaging was assessed by questionnaire on a 10-point Likert-type scale (1 is best), in comparison to previous testing, by patient willingness to have the procedure again, and by complaints. The presence or absence of persistent chest pain was assessed at telephone follow-up according to our previously published methods (13).
Outcomes were determined by review of electronic health records and telephone questionnaires at 30 days, 1 year, and study completion (January 2017). Patients without electronic records follow-up and those who were not reached after several attempts by telephone were followed up by contacting their providers’ offices. When we found that care was provided outside our health system, records were requested for review.
Outcomes were assessed by intention to treat (including all recruited participants). We calculated that a sample size of 200 per group had 87% statistical power to detect a reduction of hospital admission from 28% to 15% (23). Proportions were compared with Fisher exact tests. The Kruskal-Wallis H test for independent samples was used for nonparametric data. All tests were 2-tailed and performed at α = 0.05.
Of 400 patients, 201 and 199 were randomized to coronary CTA and SE, respectively (Figure 1). Mean age was 55.0 ± 9.7 years, 170 enrollees (42.5%) were women, and 349 (87.3%) were ethnic minorities (Table 1). The randomly assigned modality was performed in 189 coronary CTA and 188 SE patients. Fourteen coronary CTA studies (7.4%) were terminated as positive after calcium scoring. Six SE studies (3%) were terminated as positive after resting images. Thirty-one SE studies (16%) used dobutamine infusion. A total of 24 SE patients (13%) did not achieve 85% of their maximum predicted heart rate. Clinical management was not dependent on completion of the noninvasive imaging protocol. Follow-up of a minimum of 1 year was complete in 387 patients (97%). One patient died 98 days after randomization. Twelve patients were lost to follow-up (5 in the coronary CTA arm, mean age 50 years, 6 women, all ethnic minorities).
Primary and secondary endpoints
Thirty-nine of 201 coronary CTA patients (19%) and 22 of 199 SE patients (11%) were hospitalized at initial presentation (difference 8%; 95% confidence interval: 1% to 15%; p = 0.026) (Figure 2). Median ED length of stay for discharged patients was significantly longer for the coronary CTA group (5.4 h; interquartile range [IQR]: 4.2 to 6.4 h) than for the SE group (4.7 h; IQR: 3.5 to 6.0 h; p < 0.001). Median hospital length of stay from admission was significantly longer for the coronary CTA group (58 h; IQR: 50 to 102 h) than for the SE group (34 h; IQR: 31 to 54 h; p = 0.002).
Median follow-up for cardiovascular events was 733 days (IQR: 442 to 1,060 days). Two coronary CTA patients and 1 SE patient died during follow-up at 98, 1,080, and 661 days after recruitment, respectively; all 3 had advanced metastatic cancer diagnosed after recruitment. Nine coronary CTA patients and 6 SE patients had nonfatal major adverse cardiovascular events. No patients in either arm had serious complications from noninvasive imaging. Three coronary CTA patients and no SE patients had serious complications of coronary revascularization (Table 2).
Treatment of coronary disease
Twenty-three coronary CTA patients (12%) and 18 SE patients (9%) underwent coronary catheterization during the follow-up period (p = 0.51). Seven patients in each arm underwent percutaneous interventions. Four coronary CTA patients and no SE patients received coronary bypass grafting. Thirteen coronary CTA and 11 SE patients did not undergo revascularization (Table 2). There were significant increases in lipid-lowering medications and beta-blocker pharmacotherapy in the coronary CTA arm (Table 3).
Subsequent inpatient, ED, outpatient, and noninvasive cardiac imaging resource utilization is shown in Table 3. Coronary CTA identified surgical cardiovascular diagnoses in 3 patients (1 acute aortic intramural hematoma, 1 ascending aortic aneurysm, and 1 stenotic bicuspid aortic valve). SE identified severe mitral insufficiency that led to surgery in 1 patient. Coronary CTA demonstrated significant noncardiovascular diagnoses in 7 patients that might explain their symptoms (3 acute pulmonary emboli, 1 malignant pleural effusion, 1 pneumonia, 1 cholangiocarcinoma, and 1 liver metastases from colon cancer). The numbers of patients admitted for several retrospectively determined clinical rationales were similar in the coronary CTA and SE arms: non–ST-segment elevation myocardial infarction or unstable angina (n = 12 and 10, respectively), completion of a rule-out myocardial infarction protocol (n = 4 and 5, respectively), and assessment of other causes of chest pain (n = 5 and 4, respectively). There were more admissions in the coronary CTA arm than in the SE arm for further assessment of diagnosed coronary artery disease (n = 10 and 1, respectively), non–chest pain complaints (n = 5 and 1, respectively), and pulmonary embolism (n = 3 and 1, respectively).
Median effective dose for patients randomized to and receiving coronary CTA (n = 189) was 6.4 mSv (IQR: 5.3 to 7.8 mSv). SE entailed no ionizing radiation. There were marked, persistent decreases in radiation exposure in the SE arm compared with the coronary CTA arm through the initial ED/hospital work-up, over 1 year, and over the total follow-up period (Table 3).
Subjective patient experience
The experience of initial imaging was assessed subjectively in participants receiving their assigned study; data were available for 181 of 189 coronary CTA patients (96%) and 179 of 188 SE patients (95%). Median scores on a scale of 1 to 10 (1 being highest) were 3 (IQR: 1 to 4) and 2 (IQR: 1 to 3) for coronary CTA and SE, respectively (p = 0.11). SE patients graded their examination more favorably than did coronary CTA patients compared with other previous diagnostic tests (p = 0.05). There was no significant difference between modalities in patient willingness to undergo a repeat examination (p = 0.14).
Complaints or adverse reactions to imaging were reported in 26 of 189 coronary CTA patients (14%) and 12 of 188 SE patients (6.4%; p = 0.03). Seventeen coronary CTA patients (9.0%) and 6 SE patients (3%) complained about examination length or positional discomfort (p = 0.03). No coronary CTA patients and 5 SE patients (3%) complained about chest pain or shortness of breath during imaging (p = 0.03).
A total of 190 coronary CTA and 194 SE patients provided information about persistence of symptoms at 1-month telephone interviews. Fifty-four coronary CTA patients (28%) and 36 SE patients (19%) had continued chest pain or shortness of breath (p = 0.03), which was the same or worse in 21 coronary CTA and 13 SE patients (p = 0.15).
Hospitalization and length of stay
This randomized comparative effectiveness trial provides the first comparison of coronary CTA and SE in ED chest pain patients and shows a statistically significant 8% reduction of hospitalization for patients triaged with SE compared with coronary CTA. The trial population is notable for ethnic diversity and a high burden of cardiovascular risk factors, which is an understudied group yet increasingly the norm of urban practice. With SE, median ED length of stay from randomization for those discharged was significantly reduced by 13% (0.7 h), and median hospital stay was significantly reduced by 41% (24 h). The increased triage efficiency of SE was safe in terms of major adverse cardiovascular events, repeat ED visits, and repeat hospitalizations at intermediate-term follow-up (median 2 years; 100% at 30 days; 97% at 1 year).
One previous trial showed reduced hospitalizations for ED triage by functional testing (particularly stress radionuclide perfusion imaging) compared with coronary CTA (7). Two larger trials, however, showed marked increases in ED discharge (as opposed to continued observation or hospital admission) with early coronary compared with standard care (9,10). A significant minority of standard care patients in these prior trials were not imaged. The trials also used different noninvasive imaging modalities and chest pain observation units. These factors could contribute to the differences between their results and those of the current study.
In contrast to our results, previous randomized trials unanimously showed decreased length of stay with coronary CTA compared with standard care (7–11). This is most likely the direct result of performing coronary CTA earlier in the diagnostic algorithm than standard care noninvasive testing (7–10). In our study, coronary CTA and SE were performed at the same point in the diagnostic work-up (23) and recruited patients during weekday, daytime hours, which differs from some practices. Our observation of decreased hospitalization length of stay in the SE arm is of uncertain cause. One possibility is that cardiac catheterization was delayed in patients in the coronary CTA arm to avoid rapidly repeated contrast administrations.
Subsequent catheterization and revascularization
We found no significant difference between coronary CTA and SE in subsequent catheterization and revascularization, with the majority of patients catheterized in both arms not undergoing coronary intervention. Thus, despite current standard noninvasive testing, there is a disappointingly high rate of catheterization without subsequent intervention, as observed on a larger scale retrospectively (30). Although a very large multicenter trial of outpatients showed a modest decrease in catheterization without intervention with coronary CTA as opposed to functional testing (which our trial was not powered to detect), it also showed increased overall rates of catheterization, percutaneous intervention, and bypass surgery (31). We also observed a nonsignificant trend toward more bypass surgery with coronary CTA. Increased revascularization in coronary CTA patients has been confirmed by meta-analyses (17,32) and is of undetermined benefit.
Patients undergoing coronary CTA received significantly more anti-lipid and beta-blocker pharmacotherapy, and there was a trend toward more new aspirin prescriptions in the coronary CTA arm than in the SE arm. This might be explained by a greater proportion of patients with demonstrated coronary artery disease and a larger number of patients admitted with more extensive work-ups. A previous ED trial comparing coronary CTA with functional testing showed no difference in medical management (9); however, an outpatient multicenter trial involving the addition of coronary CTA to standard care showed marked changes in pharmacotherapy (33). A subgroup analysis of the latter trial (34) and a subsequent meta-analysis of several trials (32) showed a significant decrease in nonfatal myocardial infarction in patients undergoing coronary CTA, which might be related to improved guideline-directed medical treatment.
Unsurprisingly, SE markedly reduced ionizing radiation exposure during the initial work-up, an effect that persisted over complete follow-up. This is consistent with previous trials comparing coronary CTA and functional testing (10,11,31). Increasing population radiation exposure is a challenge. Although noninvasive imaging for acute chest pain is pervasive, we remain cognizant that some retrospective data (35,36) and expert consensus documents (37) advocate for no initial testing; this would represent the ultimate reduction in radiation exposure and utilization. When imaging is indicated, a deep understanding of the capabilities of modalities without ionizing radiation has the potential to improve the radiation footprint of diagnostic strategies in numerous fields. On the other hand, we found that coronary CTA provided diagnoses of additional noncardiac causes of pain, which are not assessable by SE. This is frequently the case, because computed tomography provides far more comprehensive anatomic information than alternative modalities. Although it provides some additional diagnoses, the availability of more anatomic data also results in incidental findings, some of which are overdiagnoses (18).
Patients rated their experience with SE more favorably than coronary CTA compared with previous testing. There were significantly more complaints and minor adverse reactions in coronary CTA patients. These findings are at variance with our previous observations comparing coronary CTA and radionuclide perfusion imaging (13). An explanation for the difference is the low rate of pharmacological stressor use with SE in the current study (16%) compared with the previous trial, in which the majority of patients undergoing nuclear stress testing received intravenous agents (38).
Primarily, this trial was a single-center study of moderate size, which limits generalizability. Nevertheless, it was adequately powered for the studied outcomes, and several observations were highly statistically significant. In addition, the 2-year median follow-up far exceeds that of the majority of ED studies. Furthermore, the population analyzed consisted of ethnic minorities of low socioeconomic status, who remain understudied (39) and underserved. Application of the results of this study to other settings is dependent on the availability of imaging services and modality-specific expertise, which varies by institution. We recruited during the daytime on weekdays, when both examinations could be performed and interpreted by experts. The management of patients who present “after hours” was not assessed by our study; such patients would likely have significantly longer lengths of stay. Second, the primary outcome depended on the clinical decision to hospitalize patients, not on a strict algorithm. This pragmatic approach more closely resembles that of real practice. Similarly, hospitalization length of stay depends on numerous factors beyond the initial imaging modality. Third, identification of downstream clinical events is limited to our healthcare system and patient interviews; however, this limitation should not bias the results toward either modality. This trial was not powered to detect differences in major adverse cardiovascular events, which might not be detected by even the largest imaging studies to date in any setting (31). Fourth, our diagnostic approach did not incorporate high-sensitivity troponin assays (40), which are not clinically available in our health system; use of this biomarker would reduce the number of patients requiring any imaging. The potential decrease in imaging that can be achieved and the resultant population characteristics of those to be imaged depend on which of the multiple published triage protocols is used (41). Importantly, the optimum management to improve outcomes of patients with low but detectable troponin levels remains controversial (42–44) and would significantly impact this calculation. We also did not assess the effects of advances in functional imaging and coronary CTA, most notably methods for determination of stenosis significance such as computed tomography–based fractional flow reserve (45), stress perfusion (46), and sequential testing protocols (47,48). Technological improvements are continuous and can be significant in settings where these approaches are available. Finally, pre-test probability of severe coronary disease was determined by an historical method, which is an overestimation in current practice; several alternative methods have been described more recently and deserve consideration (49). As a result, the studied cohort was relatively low risk, and this caveat must be considered carefully before these findings are applied in higher-risk patients.
In our busy urban ED setting, SE was safe and effective in chest pain triage compared with coronary CTA. SE resulted in the discharge of a significantly higher proportion of patients with significantly shorter lengths of stay, was safe at intermediate-term follow-up, and provided a better patient experience. Coronary CTA patients, however, received more pharmacotherapy for coronary artery disease than SE patients, and there was a trend toward increased coronary bypass surgery in the coronary CTA arm. These interventions are of uncertain benefit and require further study. Stress echocardiography, a modality that entails no ionizing radiation exposure, should be considered an appropriate option for ED chest pain triage.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: SE is a safe and effective triage method for acutely symptomatic ED chest pain patients that has the potential to reduce hospital admission, length of stay, and radiation exposure.
TRANSLATIONAL OUTLOOK: Further studies are needed to elucidate the best respective roles for the available imaging modalities and when imaging might safely be avoided in ED chest pain triage.
The authors thank the study coordinators Dr. Elana Clark, Dr. Samuel Friedman, Kiara Genao, Dr. Esther Hirschhorn, Joshua Jay, Yosefa Schoor, and Shayna Vega. The authors acknowledge the contribution of the study’s Data and Safety Monitoring Board members, Drs. Michael Farkouh (chair), Gregory Pearson, James Godbold, and Jennifer Liu.
This study was supported by American Heart Association Scientist Development Grant 11SDG7380006. The funder had no role in the study design, trial conduct, data collection/analysis/interpretation, manuscript preparation, or in the decision to submit the manuscript for publication. All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- computed tomography angiography
- emergency department
- interquartile range
- stress echocardiography
- Received March 27, 2018.
- Accepted March 28, 2018.
- 2018 American College of Cardiology Foundation
- ↵World Health Organization. The top 10 causes of death. Available at: www.who.int/mediacentre/factsheets/fs310/en/. Accessed December 27, 2017.
- Benjamin E.J.,
- Virani S.S.,
- Callaway C.W.,
- et al.
- ↵Centers for Disease Control and Prevention, National Center for Health Statistics. National Hospital Ambulatory Medical Care Survey: 2014 Emergency department summary tables. Available at: www.cdc.gov/nchs/data/nhamcs/web_tables/2014_ed_web_tables.pdf. Accessed December 27, 2017.
- Amsterdam E.A.,
- Kirk J.D.,
- Bluemke D.A.,
- et al.,
- on behalf of the American Heart Association Exercise,
- Cardiac Rehabilitation, and Prevention Committee of the Council on Clinical Cardiology,
- Council on Cardiovascular Nursing, and Interdisciplinary Council on Quality of Care and Outcomes Research
- Hlatky M.A.,
- Douglas P.S.,
- Cook N.L.,
- et al.
- Goldstein J.A.,
- Gallagher M.J.,
- O’Neill W.W.,
- Ross M.A.,
- O’Neil B.J.,
- Raff G.L.
- Goldstein J.A.,
- Chinnaiyan K.M.,
- Abidov A.,
- et al.,
- CT-STAT Investigators
- Rybicki F.J.,
- Udelson J.E.,
- Peacock W.F.,
- et al.
- Morris J.R.,
- Bellolio M.F.,
- Sangaralingham L.R.,
- et al.
- Nudi F.,
- Lotrionte M.,
- Biasucci L.M.,
- et al.
- Hlatky M.A.,
- Iribarren C.
- Trippi J.A.,
- Lee K.S.,
- Kopp G.,
- et al.
- Heitner J.F.,
- Klem I.,
- Rasheed D.,
- et al.
- Abbara S.,
- Blanke P.,
- Maroules C.D.,
- et al.
- ↵Radiation risk calculator. Available at: www.xrayrisk.com/calculator/calculator.php. Accessed December 27, 2017.
- ↵Radiation Dose Assessment Resource (RADAR) medical procedure radiation dose calculator. Available at: www.doseinfo-radar.com/RADARDoseRiskCalc.html. Accessed December 27, 2017.
- Foy A.J.,
- Dhruva S.S.,
- Peterson B.,
- Mandrola J.M.,
- Morgan D.J.,
- Redberg R.F.
- Williams M.C.,
- Hunter A.,
- Shah A.S.,
- et al.,
- SCOT-HEART Investigators
- Sandhu A.T.,
- Heidenreich P.A.,
- Bhattacharya J.,
- Bundorf M.K.
- Reinhardt S.W.,
- Lin C.J.,
- Novak E.,
- Brown D.L.
- National Institute for Health and Care Excellence
- Levsky J.M.,
- Travin M.I.,
- Haramati L.B.
- Dedic A.,
- Lubbers M.M.,
- Schaap J.,
- et al.
- Twerenbold R.,
- Boeddinghaus J.,
- Nestelberger T.,
- et al.
- Roos A.,
- Bandstein N.,
- Lundbäck M.,
- Hammarsten O.,
- Ljung R.,
- Holzmann M.J.
- Januzzi J.L. Jr..,
- Suchindran S.,
- Coles A.,
- et al.
- Bonaca M.P.
- Lubbers M.,
- Coenen A.,
- Kofflard M.,
- et al.
- Rozanski A.,
- Berman D.S.