Article

Emergency physician focused cardiac ultrasound improves diagnosis of ascending aortic dissection

a b s t r a c t

Study Objective: Ascending aortic dissection (AAD) is an uncommon, time-sensitive, and deadly diagnosis with a nonspecific presentation. Ascending aortic dissection is associated with aortic dilation, which can be determined by emergency physician Focused cardiac ultrasound (EP FOCUS). We seek to determine if patients who receive EP FOCUS have reduced time to diagnosis for AAD. Methods: We performed a retrospective review of patients treated at 1 of 3 affiliated emergency departments, March 1, 2013, to May 1, 2015, diagnosed as having AAD. All autopsies were reviewed for missed cases. Primary outcome measure was time to diagnosis. Secondary outcomes were time to disposition, misdiagnosis rate, and mortality. Results: Of 386 547 ED visits, targeted review of 123 medical records and 194 autopsy reports identified 32 patients for inclusion. Sixteen patients received EP FOCUS and 16 did not. Median time to diagnosis in the EP FoCUS group was 80 (interquartile range [IQR], 46-157) minutes vs 226 (IQR, 109-1449) minutes in the non-EP FOCUS group (P = .023). Misdiagnosis was 0% (0/16) in the EP FOCUS group vs 43.8% (7/16) in the non-EP FOCUS group (P = .028). Mortality, adjusted for do-not-resuscitate status, for EP FOCUS vs non-EP FOCUS was 15.4% vs 37.5% (P = .24). Median rooming time to disposition was 134 (IQR, 101-195) minutes for EP FOCUS vs 205 (IQR, 114-342) minutes for non-EP FOCUS (P = .27).

Conclusions: Patients who receive EP FOCUS are diagnosed faster and misdiagnosed less compared with patients who do not receive EP FOCUS. We recommend assessment of the thoracic aorta be performed routinely during cardiac ultrasound in the emergency department.

(C) 2015

  1. Introduction

Ascending aortic dissection (AAD) is an uncommon, potentially treatable disease with mortality increasing ~ 1% to 2% per hour from symptom onset to diagnosis and therapy [1-3]. With a prevalence of

2.5 to 3.5 cases per 100000 patient-years, AAD is often misdiagnosed in the emergency department (ED) due to nonspecific signs and symptoms and has been described as “the great masquerader” [3-5]. Al- though contrast-enhanced Computed tomography angiography is the most frequent definitive diagnostic test for AAD, it is not feasible to

Abbreviations: AAD, ascending aortic dissection; CTA, computed tomography angiog- raphy; EP FOCUS, emergency physician focused cardiac ultrasound.

? There were no sources of support for this project, and there are no conflicts of interest

to disclose. No individual was paid to provide manuscript advice, writing, or editing. Only those authors listed had control over the data, how it was analyzed, interpreted, and pre- sented within the manuscript.

?? Presented: Oral Presentation Connecticut College of Emergency Physicians Annual

Meeting. Hartford, CT September 30th, 2015.

* Corresponding author. Joseph R. Pare, MD, Department of Emergency Medicine, Yale University, Yale-New Haven Hospital, 464 Congress Ave, Suite 260, New Haven, CT, 06519. Tel.: +1 203 785 4058; fax: +1 203 785 4580.

E-mail address: [email protected] (J.R. Pare).

perform CTA on all ED patients with symptoms of chest or back pain [5,6]. Point-of-care focused cardiac ultrasound is increasingly used in the ED setting for patients with thoracic complaints; however, current American College of Emergency Physician recommendations for FOCUS do not include routine assessment of the ascending thoracic aorta [5,7]. In 2010, based on cases at our institution where aortic dissection was missed but thoracic aortic pathology was appreciated on quality assurance review, we started to perform routine measurement of the thoracic aortic root.

Aneurysmal dilation of the aorta is frequently present in AAD, with ap- proximately 90% of AAD patients having dilation of the aorta greater than 4 cm at the time of diagnosis [6,8-10]. Emergency physician FOCUS (EP FOCUS) measurements for proximal aortic size have been shown to agree with CTA measurements [11,12]. In addition, ascending aortic dila- tion identified by FOCUS is approximately 70% sensitive for AAD [13].

Our objective was to determine whether patients who received EP FOCUS including evaluation of the thoracic aorta had at least a 1-hour reduction in time to diagnosis for AAD compared with patients who did not receive EP FOCUS and were ultimately diagnosed as having AAD. Secondary objectives were to determine if EP FOCUS reduces mor- tality, misdiagnosis, and time to disposition. It is expected that an

http://dx.doi.org/10.1016/j.ajem.2015.12.005

0735-6757/(C) 2015

abnormal aorta identified on EP FOCUS when used as an early diagnostic tool should improve our study outcome measures.

  1. Methods
    1. Study design

This was a retrospective cohort analysis of patients treated within a multihospital health care system who were found to have AAD. We adhered to guidelines for retrospective medical record reviews as delineated by Gilbert et al [14] and Kaji et al [15]. To be eligible for enrollment, patients must have been treated in the ED at 1 of 3 affiliated hospitals during a 26-month study period, March 1, 2013, to May 1, 2015. March 1, 2013 was chosen as the starting point because this was when our electronic health record was initiated, allowing us to identify reliably patients with AAD across our hospital system. Patients identified with AAD were analyzed based on those who received EP FOCUS in the ED and those who did not.

Study setting and population

Hospital 1 is a large urban, nonprofit, level 1 trauma academic Tertiary medical center. It is the flagship hospital for this health care system, and its ED has an annual census of approximately 90000 adult patient visits per year. Hospital 2 is a small suburban, nonprofit, community-based freestanding ED. This location has an annual census of approximately 30000 visits per year. Hospital 3 is a large urban, nonprofit, academic community-based hospital and has an annual census of approximately 70000 ED visits. Each of these facilities is a 24-hour health care facility with CTA available for the diagnosis of aortic dissection.

These 3 affiliated sites were selected for inclusion because a single academic emergency ultrasound section is responsible for physician point-of-care ultrasound education and quality assurance. The physicians at these study locations are encouraged to use the “5Es” approach to EP FOCUS with regard to measurement of the proximal aorta on a parasternal long-axis view (Fig. 1) [16]. The 5Es approach is a rapid proto- col recommended to assess for pericardial Effusion, ventricular Ejection function, ventricular Equality (eg, right heart enlargement seen with pulmonary embolus), Exit (proximal aortic size), and Entrance (inferior vena cava size and collapsibility). Evaluation of the aorta using the 5Es protocol is a simple measure using the leading edge technique to obtain a single measure at the most dilated portion of the visualized proximal aorta. All FOCUS examinations including DICOM images are required to

Image of Fig. 1

Fig. 1. Focused cardiac ultrasound aorta measure. Obtained by a parasternal long-axis view, measure of the proximal aorta here shows abnormal aneurysmal dilation of 4.6 cm. Our institution performs echocardiography with the probe pointed to the right shoulder with the marker on the left of the screen [23]. Abbreviations: LA, left atrium; LV, left ventricle; RV, right ventricle.

be added to the EHR. Emergency medicine attendings are responsible for all FOCUS examinations and final interpretations. Templated worksheets for EP FOCUS on our ultrasound documentation and quality assurance system, Q-path (Telexy Healthcare, Port Coquitlam, BC, Canada) require classification of proximal aortic size as normal (b 4 cm), aneurysmal dilation (4-4.5 cm), or aneurysmal (N 4.5 cm). Each study site is also equipped with the same ultrasound machines and probes, the Phillips Sparq (Bothell, Washington). A phased-array S4-2 ultra- sound probe is used for cardiac imaging at these facilities with a fre- quency of 2 to 4 MHz. Point-of-care ultrasound machines are readily available for use at all times at the study site facilities.

Study protocol

The EHR, EPIC ASAP (Verona, Wisconsin) used at all hospitals within this single health care system was queried for patients who had the diagnosis of aortic dissection (determined by International Classification of Diseases, Ninth Revision codes 441.0x and regular expression search using the terms “aorta,” “aortic,” “dissection”) either in the ED or as a discharge diagnosis after hospitalization between March 1, 2013, and May 1, 2015. This query was designed to find all cases of aortic dissec- tion, both Stanford type A and type B, within a patient’s history of diag- noses. To identify cases of missed aortic dissection resulting in death, all hospital autopsy reports during this period were manually reviewed for the diagnosis of AAD as the cause of death. Patients were included if they were older than 18 years and were treated at 1 of 3 affiliated EDs within a month preceding diagnosis for a visit attributed to AAD or during the same hospital visit.

Patients were excluded if they were transferred from a referring facility with a known diagnosis of aortic dissection, if they had a history of acute trauma, if they were not treated in the ED, if medical record re- view did not support the patient having a diagnosis of AAD (eg, improper diagnosis being added to the patient’s medical record or diagnosis being added for a case of AAD that occurred prior to March 1, 2013), if they had a descending aortic dissection defined as distal to the left subclavian artery, or if they were treated at an alternative affiliated nonstudy site. Pa- tients found to have AAD as a cause for death on autopsy were excluded if they arrived to the ED in cardiac arrest. Participants were selected for study enrollment after manual review of the medical record confirmed inclusion criteria and a lack of exclusions. Institutional review board approval was obtained for this study.

Key outcome measures

The primary outcome measure was time to diagnosis as defined by the difference between the time being roomed in an ED treatment area (time stamped in the EHR) to diagnosis of AAD. The time advanced imaging was performed and used as the diagnosis time. Advanced imaging was defined as CTA, magnetic resonance imaging, or transesophageal echocar- diography. Secondary outcome measures included time to disposition, misdiagnosis rate (admitted or discharged with a diagnosis other than AAD), and mortality differences between cohorts. Mortality was adjusted for do-not-resuscitate (DNR) status by excluding patients with a DNR order; this included patients with a previous DNR order, or who were made DNR either prior to or after discovery of the diagnosis.

All information was obtained from study patients’ EHR. Data elements for hospitalizations associated with the visit for AAD were obtained from the enterprise data warehouse CLARITY (Epic, Verona, Wisconsin). Structured Query Language queries were written to identify and extract desired clinical and demographic information (eg, age, sex, race, and outpatient medications) based on a unique hospital visit number. Data that could not be abstracted using electronic methods were abstracted by manual medical record review (Appendix A: Variable by Abstraction Method).

Manual medical record abstractors included 3 emergency medicine residency-trained physicians (J.R.P., S.T., R.L.). A fourth medical record

reviewer (T.S.) was a second year medical student. All abstractors were trained prior to performing reviews. T.S., as a medical student, was given targeted background reading into the pathology of aortic dissection and performed 10 training studies that were reviewed by the lead study author (J.R.P.) and feedback was given. Weekly meetings were held between medical record abstractors to review coding and progress. The lead study author reviewed all questions about abstrac- tions, and a consensus was made between abstractors about coding disputes. The lead study author reviewed abstracted data weekly to monitor for completeness and accuracy.

Focused cardiac ultrasound interpretations and measurements were ob- tained from Q-path. If the physician did not define the size of the proximal aorta in the generated Q-path report for the FOCUS examination, the indi- vidual stored DICOM images were reviewed to determine if a measure of the proximal aorta was performed but not recorded. If FOCUS did not have an aorta measure available for analysis, the patient was not excluded from the FOCUS group, as there may have been other findings such as peri- cardial effusion or dissection flap seen, which may have affected clinical de- cision making. Time of advanced imaging was manually abstracted from the first image within the radiology Picture Archiving and Communication System SYNAPSE (Fujifilm, Tokyo Japan).

Medical record reviews were based on discretely defined variables for abstraction of data points. Data were entered directly into a standardized worksheet stored on a hospital password-protected on- line database manager, FileMaker Pro (Apple Inc, Santa Clara, California; Appendix B: Data Worksheet). The American Heart Association (AHA) aortic dissection score, based on risk factors, history, and physical examination findings, was calculated to compare pretest probability of AAD between study groups [3].

Interrater reliability and blinding

A second study author reviewed 30% of medical records selected at random to determine reliability for manual medical record abstractions as well as computer abstractions. Although it was not possible to blind reviewers to the presence of AAD as this was an inclusion criteria, the

authors (J.R.P., R.L., S.T., and T.S.) performing medical record abstraction were blinded to the primary outcome measure, time to diagnosis.

An independent physician in the department of radiology with training in body imaging interpretation who was blinded to clinical data, clinical computed tomographic (CT) interpretation, and EP FOCUS performed retrospective measures of the proximal aorta within 4 cm of the aortic valve using images on the radiology Picture Archiving and Communication System. The maximal aortic diameter was mea- sured on axial images (1.25-mm axial CTA or 5-mm axial half-Fourier- acquisition single-shot turbo spin-echo). Coronal images were used to confirm that measurements were made within 4 cm of the aortic valve.

Data analysis

The primary outcome measure was time from being roomed in an ED treatment area to diagnosis by advanced imaging, defined as the time advanced imaging was performed. Power calculation for time to diagnosis was performed using hospital-specific administrative data for average patient length of stay. The average length of stay to disposi- tion for patients at our main facility is 199 minutes with an SD of 45 mi- nutes. Ten patients in each cohort group (EP FOCUS and non-EP FOCUS) were required to detect a 1-hour difference in time to diagnosis with a power of 0.80 and ? of .05.

Data analysis was performed using STATA v13.1 (College Station, Texas). Patient characteristics were evaluated by means with SDs or percentages. Student t tests and Fisher exact tests were performed for continuous data and categorical data as appropriate for comparison between cohort groups. Wilcoxon rank sum tests were performed for nonparametric data including rooming time to diagnosis and rooming time to disposition. For Wilcoxon rank sum tests, median values are pre- sented with interquartile ranges (IQRs). Adjustments were made for multiple hypothesis testing (4) by Holm-Bonferroni correction. Missing data were excluded from subset analysis. A pooled Cohen ? and intraclass correlation coefficient was calculated to assess reliability for manual and electronic medical record abstractions by dual review of 30% of enrollments.

16 Non -FOCUS

16 FOCUS

CTA Ao >4cm 16 *

CTA Ao <4cm 0

CTA No Ao

Measurement: 1

CTA Ao <4cm: 2

CTA Ao >4cm: 13

US >4cm: 10

US <4cm: 1

US without Ao measure: 1

US <4cm: 2

US without Ao

measure: 1

32 Inclusions

Exclusions: Altenative Cause of Death: 192

type A dissection, but arrived in cardiac arrest: 1

Exclusions: Diagnosis known: 4

Dissection during hospitalization: 1

No aortic dissection during study period: 26 Not treated at study site: 38

Type B aortic dissection: 23

Patients by Autopsy Review:

n = 194

Patients by EHR Review:

n = 123

* One patient in this group was discharged, and on return ED visit had FOCUS which correctly identified an aneurysmal aorta

Fig. 2. Flowchart for enrollments. Two patients did not have measures of the aorta available. These patients were still included in FOCUS analysis as findings such as pericardial effusion or visualization of a dissection flap may have affectED management. Abbreviations: Ao, aorta; US, ultrasound.

  1. Results

There were 386 547 ED patient visits during our study period. A total of 123 patients were identified as having a diagnosis of aortic dissection (Fig. 2). Of those 123, 31 patients were identified to meet criteria for analysis after exclusions. There were 194 autopsy reports available for review during the study period. From autopsy reports, 2 cases of AAD were identified and 1 was excluded because the patient arrived to the ED in cardiac arrest. The other patient was admitted for chest pain. An inpatient noncontrast CT scan showed an ascending aortic aneurysm but did not identify the dis- section flap that was later found on autopsy.

Combining patients from the EHR search and autopsy review, a total

of 32 patients were identified as having AAD meeting the criteria for analysis (Appendix C: Individual Data). There were 16 patients identi- fied in each cohort group, FOCUS and non-EP FOCUS. Of demographic and clinical information, heart rate was the only variable found to be statistically significant between cohort groups (P = .03; Table 1). There was no statistical difference between study location and AHA dissection score between groups. The median time to diagnosis in the FOCUS group was 80 (IQR, 46-157) minutes and 226 (IQR, 109-1449) minutes in the Non-FoCUS group (P = .023; Fig. 3).

Table 1

Demographic and clinical variables by FOCUS cohort

Variable

FOCUS

Non-FOCUS

P

Age (y)

61 (16)

62 (13)

.82

Male

11/16

7/16

.29

Nonwhite

0/16

1/16

1

English speaking

16/16

15/16

1

BMI (kg/m2)

25.9 (6.3)

28.5 (6.4)

.19

Triage ESI

2.1 (0.8)

2.0 (0.6)

.79

DNR

3/16

0/16

.23

Triage vitals

Heart rate (beats/min)

71 (15)

86 (20)

0.03

Systolic BP (mm Hg)

131 (38)

141 (39)

0.47

Diastolic BP (mm Hg)

78 (34)

79 (27)

.93

Respiratory rate (breaths/min)

19 (7)

18 (4)

.57

SpO2 (%)

98 (2)

97 (2)

.15

Temperature (?F)

97.5 (0.6)

97.4 (1.0)

.91

Chest radiograph Wide mediastinum

1/11

1/12

1

Abnormal aortic contour

3/11

4/12

1

Medications

Aspirin

9/16

8/16

1

Anticoagulant

4/16

1/16

.33

Anticholesterol

6/16

7/16

1

Steroids

1/16

1/16

1

Average no. of antihypertensives

2.3 (1.6)

2.4 (2.2)

.93

Average no. of total medications

11.5 (11.6)

12.1 (10.6)

.89

Charlson comorbidity

0.6 (0.7)

1.3 (1.6)

.10

Known aneurysm

4/16

3/16

1

Connective tissue disease

1/16

0/16

1

Previous CABG or AVR

1/16

1/16

1

Inflammatory disease

1/16

3/16

.60

Bicuspid aortic valve

0/16

0/16

1

Hypertension

8/16

13/16

.14

Smoker

6/15

10/16

.29

cocaine use

2/16

1/16

1

ECG

STEMI

1/16

1/15

1

Strain pattern

5/16

1/15

.17

LVH

5/16

4/15

1

Treating hospital

.10

1

12/16

7/16

.15

2

0/16

3/16

.22

3

4/16

6/16

.70

AHA dissection score

1.4 (0.6)

1.1 (0.6)

.27

Values in parenthesis represent SDs. Abbreviations: AVR, aortic valve replacement; BMI, body mass index; BP, blood pressure; CABG, coronary artery bypass graft; ESI, emergency severity index (triage); LVH, left ventricular hyper- trophy; SpO2, peripheral capillary oxygen saturation; STEMI, ST-segment elevation myocardial infarction.

Overall, 93.5% (29/31) of cases of AAD in the study population were found by advanced imaging to have a proximal aorta diameter greater than 4 cm at the time of diagnosis. One patient was unable to be included in analysis because a CTA of the head and neck that identified the type A dis- section was performed for stroke symptoms. Because the entirety of the proximal aorta was not visualized on imaging, a standardized measure of the aorta could not be obtained for inclusion. The size and prevalence of aor- tic aneurysms at the time of diagnosis were similar between groups (Table 2). Mean (SD) aorta size was not statistically different between FOCUS and non-FOCUS patients, 5.45 (0.94) cm and 5.44 (1.44) cm, respectively (P = .98). Of patients in the FOCUS group, 86% (13/15) were found to have a dilated aorta greater than 4 cm. Although not a prima- ry objective of the study upon medical record review for the patients who underwent FOCUS examination, 2 of 16 patients were reported to have a pericardial effusion and 6 of 16 had a dissection flap visualized.

There was 1 patient discharged from the ED who returned for continued symptoms; FOCUS on return measured an aorta of 5.9 cm preced- ing CTA. This patient was included in the calculation for performance of FOCUS to identify an abnormal aorta. There were 2 patients in the FOCUS group who did not have a measurement of the proximal aorta recorded; these were excluded from subset analysis. Focused cardiac ultrasound cor- rectly identified a dilated proximal aorta greater than 4 cm in 10 of 13 patients who had dual imaging available for comparison. Of all cases of AAD that re- ceived FOCUS, an ultrasound measure of the proximal aorta greater than 4 cm identified 71.4% (10/14) of patients with AAD.

Secondary analysis was also performed to evaluate time from rooming to disposition, misdiagnosis rate, and mortality between cohort groups (Table 3). Those in the FOCUS group had a faster, yet nonsignifi- cant rooming time to disposition (median, 134 [IQR, 101-195] minutes vs 205 [IQR, 114-342] minutes; P = .27). The misdiagnosis rate was 0% (0/16) in the FOCUS group and 43.8% (7/16) in the non-FOCUS group (P = .028). Deaths were 5 of 16 in the FOCUS group and 6 of 16 in the non-FOCUS group. Three patients in the FOCUS group were DNR, whereas none of the patients in the non-FOCUS group were DNR. All patients with a DNR order died. Adjustment for patients with DNR in the FOCUS vs non- FOCUS group showed a mortality percentage of 15.4% vs 37.5%, which was not statistically significant (P = .24). For dual abstractions, the calcu- lated pooled ? and intraclass correlation coefficient was 0.93.

  1. Discussion

This is the first study to show that patients receiving EP FOCUS were diagnosed as having AAD faster and misdiagnosed less compared with those who did not undergo EP FOCUS. Because time to diagnosis and treatment of AAD is directly related to mortality, it is likely that perfor- mance of EP FOCUS could improve mortality from AAD. There were no missed cases of aortic dissection for patients who received a FOCUS examination, 0% vs 44% misdiagnosis rate for non-EP FOCUS.

As noted in the Results section, all patients with a DNR were in the FOCUS group and all DNR patients died. After excluding DNR patients for comparison of mortality, there was a notable but not statistical difference (15% vs 38% mortality). Time to disposition was not statisti- cally significant but showed a trend for EP FOCUS patients to have a disposition faster than the non-EP FOCUS group. Although we are un- able to make absolute conclusions regarding the relationship between FOCUS and time to disposition and mortality, we hypothesize that each of these outcome measures, in an appropriately powered study, would demonstrate a benefit for patients who receive EP FOCUS.

Ascending aortic dilation identified by FOCUS is approximately 70%

sensitive for AAD, which outperforms the findings of aortic insufficiency on cardiac auscultation (44%), pulse deficit (19%), chest radiograph with wide mediastinum (62%) or abnormal aortic contour (47%), or electro- cardiogram (ECG) findings (17%-25%) [13,17-20]. Although measures of the proximal aorta can almost always be obtained, visualization of a dissection flap on FOCUS is difficult to detect and therefore less sensitive for AAD [6,13].

Image of Fig. 3

Fig. 3. Time to diagnosis by performance of EP FOCUS. Time to diagnosis measured in minutes. Individual dots represent discrete data points.

All demographic and clinical information reviewed was similar between groups, with the exception of heart rate. Although patients who did not receive EP FOCUS were more tachycardic, it is unlikely that this should have affected the choice to perform FOCUS. In addition, the AHA dissection score was similar between cohort groups. This suggests that patients who received EP FOCUS were similar in risk factors, history, and physical examination findings. The similarity between groups sug- gests that EP FOCUS was performed independently of pretest probability

Table 2

FOCUS and CTA performance

FOCUS performed (n = 15)a No FOCUS (n =

16)

and the likelihood a treating physician would have ordered a diagnostic study for AAD regardless of the EP FOCUS examination.

Emergency physician FOCUS can be performed rapidly at the bedside for patients with concerning symptoms to evaluate the proximal aorta for aneurysm, which is associated with AAD. In the FOCUS group, 6 of the 16 patients had the diagnosis made within 60 minutes, which is a testament to the speed in which FOCUS can be performed. We should also highlight that we use rapid creatinine testing that also allows for expedited CTA testing. Emergency physician FOCUS exposes the patient to no risk and it can be performed faster than routine blood work, chest radiograph, or even ECG in experienced hands. In the initial collaborative paper on FOCUS, the American College of Emergency Physicians and the American Society of Echocardiography described

FOCUS Non-FOCUS P Corrected P

Aorta (Ao) size on CT (cm), mean (SD)

Ao b 4 cm

5.45 (1.43)

2 (13%)

5.44 (0.94)

0

Table 3

Outcome Measures for FOCUS

Ao 4-4.5 cm

2 (13%)

4 (25%)

Ao N 4.5 cm

Other FOCUS findings

11 (73%)

12 (75%)

value a

Pericardial effusion

2/16

– Dissection missed

0/16

7/16

.007

.028

Dissection flap seen on FOCUS

6/16

– Died

5/16

6/16

Aorta size on CT

FOCUS correctly identified

Died (no DNR)

2/13

6/16

.24

.24

measure

Time to disposition, median

134 (101-

205 (114-

.14

.27

Ao b 4 cm

1/1 (100%)

– (IQR)

195)

342)

Ao N 4 cm

10/13 (77%)

– Time to diagnosis, median

80 (46-

226 (109-

.0075

.023

a One FOCUS patient did not have aorta measure due to CTA of the head and neck that (IQR) 157) 1449)

did not visualize the proximal aorta for standardized calculations. a Holm-Bonferroni correction.

several primary diagnostic goals: assessment of pericardial effusion, left ventricular function, right ventricular strain, and volume status [7]. Although assessment of thoracic aortic root was mentioned in passing regarding chest pain, we believe that based on our results, ED patients with thoracic symptoms should also undergo routine assessment of the ascending thoracic aorta, as described in the 5Es of EP FOCUS [16].

Limitations

The primary limitation to this study is the potential for selection bias for performance of EP FOCUS based on some unmeasured variable such as physician gestalt for the presence of potential thoracic aortic pathol- ogy. Although we made an attempt to extract all available data to com- pare groups and found them to be similar (with the exception of heart rate), this is a limitation of a retrospective study. We also specifically calculated the AHA dissection score and found that it was similar be- tween groups. This score takes into account high-risk clinical factors, such as Classic symptoms, historical risk factors, and physical examina- tion findings. This suggests that the choice to perform FOCUS and defin- itive imaging was not biased by patient presentation. Although the number of patients (32) is relatively small, it would be difficult and re- source intensive to conduct a prospective study that identified these many AADs.

Although our study was adequately powered to answer our primary outcome measure, the study was not powered to specifically answer our secondary outcome measures. Even so, it is compelling that none of the patients who received FOCUS were misdiagnosed, vs 7 of 16 misdiagno- ses in the non-FOCUS group.

We used time of advanced imaging as our time to diagnosis. The Radiology reports often did not comment on the time a medical team was notified of the radiographic findings. We also were unable to use the time the radiology report was finalized as the diagnosis time, as there were some radiology reports that were not finalized until the following day. Therefore, time of imaging was the most re- liable and objective measure of the time to diagnosis and was avail- able for all study patients.

Although we did review all autopsy reports during this period, there are likely many deceased patients who did not have an autopsy performed. It is unclear how many of these cases may have had AAD and how this could affect our results. A specific concern would be patients evaluated in the ED and discharged who did not undergo autopsy. There is, however, little that can be done to address this limitation.

We were able to reliably identify patients with AAD using computer abstraction, but we were unable to extract the number of patients who presented with thoracic complaints to determine the proportion of these patients that received EP FOCUS. At our institution, we encourage the liberal use of ultrasound, as there were 10873 FOCUS examinations performed during the study period. Of these studies, we were unable to abstract the number of FOCUS aorta measures greater than 4 cm to de- termine how this may affect the use of CTA. Taylor et al [11] have previ- ously published that FOCUS is 95% specific for aortic dilation, so we suspect that the number of ultrasounds leading to CTA that were then found to have normal aorta measures was low. It is, however, more like- ly that FOCUS would lead to the identification of aneurysmal aortas without dissection. We used a single measurement at the most dilated portion of the aorta on a parasternal long-axis view. There is likely var- iability in how much of the proximal can be evaluated by transthoracic echocardiography. Additional work should be completed to determine how this variability should be incorporated into our findings. Alterna- tively, anatomic variances such as bicuspid aortic valves area known to be associated with a higher incidence of dilated aortic roots, and it is unclear how our findings should be applied to these populations [21,22]. Further investigation is needed to address these limitations and determine the cost-effectiveness of this testing methodology.

  1. Conclusion

This investigation has shown that time to diagnosis for AAD is significant- ly lower for patients who undergo EP FOCUS examination. In addition, pa- tients who received FOCUS were less likely to have the diagnosis of dissection missed in the ED. Although we found a trend toward improved outcomes, a study with larger numbers would be required to confirm the presence (or absence) of a Mortality benefit for EP FOCUS in patients with symptoms of aortic dissection. Additional research should be performed to determine if adjustments for aortic size on FOCUS should be incorporated into clinical decision making for patients with suspected aortic dissection. Because FOCUS is a rapid, noninvasive test that may expedite diagnosis of a life-threatening condition, we recommend that evaluation of the aorta be performed on patients who have symptoms suggestive of aortic dissection.

Appendix A. Variables by abstraction method

Computer abstracted Manually abstracted

Vitals DNR

Heart rate Chest radiograph findings

Systolic BP Charlson comorbidity

Diastolic BP Hx of aneurysm

Respiratory rate Hx of connective tissue disease

SpO2 Hx of CABG or AVR

Temperature Hx of inflammatory disease

BMI Hx of bicuspid aortic valve

Triage ESI Hypertension

Medications Smoker

Aspirin Cocaine use

Anticoagulant ECG findings

Anticholesterol Time of imaging

Steroids Type of dissection

Average no. of antihypertensives Diagnosis missed in ED Average no. of total medications Died

Rooming time FOCUS performed

Language Patient discharged from ED with return

Age Dissection seen on FOCUS

Sex FOCUS aorta size

Race FOCUS ejection fraction interpretation Rooming time to EHR disposition FOCUS effusion seen

Type of advanced imaging performed Previous diagnosis

Cardiac arrest prior to arrival Previous aneurysm size Treated hospital

AHA dissection score

Abbreviations: AVR, aortic valve replacement; BMI, body mass index; BP, blood pressure; CABG, coronary artery bypass graft; EHR, electronic health record; ESI, emergency severity index (triage); Hx, history; SpO2, peripheral capillary oxygen saturation.

Appendix B. Manual abstraction tool

Appendix C. Individual data for outcome measures

Case no.

EP FOCUS

performed

Time to diagnosis (min)

Time to disposition (min)

ED

diagnosis

DNR

Died

1

No

232

352

Dissection

No

No

2

No

220

363

Dissection

No

No

3

No

522

142

COPD

No

No

4

No

2139

590

AMS

No

No

5

No

60436

100

CVA

No

No

6

No

80

153

Dissection

No

Yes

7

No

138

207

Dissection

No

No

8

No

759

127

Neuropathy

No

Yes

9

No

37

79

Dissection

No

No

10

No

49

94

Dissection

No

No

11

No

207

292

Dissection

No

No

12

No

154

331

Dissection

No

Yes

13

No

256

280

Dissection

No

No

14

No

7496

444

Back pain

No

Yes

15

No

46

56

STEMI

No

Yes

16

No

11048

203

Chest pain

No

Yes

17

Yes

192

205

Dissection

No

No

18

Yes

111

185

Dissection

No

No

19

Yes

25

101

Dissection

No

No

20

Yes

177

286

Dissection

No

No

21

Yes

14

140

Dissection

Yes

Yes

22

Yes

45

175

Dissection

Yes

Yes

23

Yes

72

105

Dissection

No

No

24

Yes

41

75

Dissection

Yes

Yes

25

Yes

47

120

Dissection

No

No

26

Yes

137

152

Dissection

No

No

27

Yes

54

48

Dissection

No

Yes

28

Yes

178

257

Dissection

No

Yes

29

Yes

66

100

Dissection

No

No

30

Yes

88

62

Dissection

No

No

31

Yes

94

128

Dissection

No

No

32

Yes

293

459

Dissection

No

No

Abbreviations: AMS, altered mental status; COPD, chronic obstructive pulmonary disease; CVA, cerebral vascular accident; STEMI, ST-segment elevation myocardial infarction.

References

  1. Braverman AC. Acute aortic dissection: clinician update. Circulation 2010;122: 184-8. http://dx.doi.org/10.1161/CIRCULATIONAHA.110.958975.
  2. Harris KM, Strauss CE, Eagle KA, Hirsch AT, Isselbacher EM, Tsai TT, et al. Correlates of delayed recognition and treatment of Acute type A aortic dissection: the Interna- tional Registry of Acute Aortic Dissection (IRAD). Circulation 2011;124:1911-8. http://dx.doi.org/10.1161/CIRCULATIONAHA.110.006320.
  3. Hiratzka LF, Bakris GL, Beckman JA, Bersin RM, Carr VF, Casey DE, et al. 2010 ACCF/ AHA/AATS/ACR/ASA/SCA/SCAI/SIR/STS/SVM guidelines for the diagnosis and man- agement of patients with thoracic aortic disease. J Am Coll Cardiol 2010;55: e27-129. http://dx.doi.org/10.1016/j.jacc.2010.02.015.
  4. Elefteriades JA, Barrett PW, Kopf GS. Litigation in nontraumatic Aortic diseases–a tempest in the malpractice maelstrom. Cardiology 2008;109:263-72. http://dx.doi. org/10.1159/000107790.
  5. Diercks DB, Promes SB, Schuur JD, Shah K, Valente JH, Cantrill SV. Clinical policy: critical issues in the evaluation and management of adult patients with suspected

acute nontraumatic thoracic aortic dissection. Ann Emerg Med 2015;65:32-42.e2. http://dx.doi.org/10.1016/j.annemergmed.2014.11.002.

  1. Meredith EL, Masani ND. Echocardiography in the emergency assessment of acute aortic syndromes. Eur J Echocardiogr 2009;10:i31-9. http://dx.doi.org/10.1093/ ejechocard/jen251.
  2. Labovitz AJ, Noble VE, Bierig M, Goldstein SA, Jones R, Kort S, et al. Focused cardiac ultrasound in the emergent setting: a consensus statement of the American Society of Echocardiography and American College of Emergency Physicians. J Am Soc Echocardiogr 2010;23:1225-30. http://dx.doi.org/10.1016/j.echo.2010.10.005.
  3. Neri E, Barabesi L, Buklas D, Vricella LA, Benvenuti A, Tucci E, et al. Limited role of aortic size in the genesis of acute type A aortic dissection. Eur J Cardiothorac Surg 2005;28:857-63. http://dx.doi.org/10.1016/j.ejcts.2005.10.013.
  4. Pape LA, Tsai TT, Isselbacher EM, Oh JK, O’Gara PT, Evangelista A, et al. Aortic diameter N =5.5 cm is not a good predictor of type A aortic dissection: observations from the International Registry of Acute Aortic Dissection (IRAD). Circulation 2007; 116:1120-7. http://dx.doi.org/10.1161/CIRCULATIONAHA.107.702720.
  5. Cozijnsen L, Braam RL, Waalewijn RA, Schepens MAAM, Loeys BL, van Oosterhout MFM, et al. What is new in dilatation of the Ascending aorta? Review of current lit- erature and practical advice for the cardiologist. Circulation 2011;123:924-8. http:// dx.doi.org/10.1161/CIRCULATIONAHA.110.949131.
  6. Taylor RA, Oliva I, Van Tonder R, Elefteriades J, Dziura J, Moore CL. Point-of-care focused cardiac ultrasound for the assessment of thoracic aortic dimensions, dila- tion, and aneurysmal disease. Acad Emerg Med 2012;19:244-7. http://dx.doi.org/ 10.1111/j.1553-2712.2011.01279.x.
  7. Nazerian P, Vanni S, Morello F, Castelli M, Ottaviani M, Casula C, et al. Diagnostic per- formance of focused cardiac ultrasound performed by emergency physicians for the assessment of ascending aorta dilation and aneurysm. Acad Emerg Med 2015;22: 536-41. http://dx.doi.org/10.1111/acem.12650.
  8. Nazerian P, Vanni S, Castelli M, Morello F, Tozzetti C, Zagli G, et al. Diagnostic perfor- mance of emergency transthoracic focus cardiac ultrasound in suspected acute type A aortic dissection. Intern Emerg Med 2014;9:665-70. http://dx.doi.org/10.1007/ s11739-014-1080-9.
  9. Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996; 27:305-8.
  10. Kaji AH, Schriger D, Green S. Looking through the retrospectoscope: reducing bias in emergency medicine chart review studies. Ann Emerg Med 2014;64:292-8. http:// dx.doi.org/10.1016/j.annemergmed.2014.03.025.
  11. Kennedy Hall M, Coffey EC, Herbst M, Liu R, Pare JR, Andrew Taylor R, et al. The “5Es” of emergency physician-performed focused cardiac ultrasound: a protocol for rapid identification of effusion, ejection, equality, exit, and entrance. Acad Emerg Med 2015;22:583-93. http://dx.doi.org/10.1111/acem.12652.
  12. Hagan PG, Nienaber CA, Isselbacher EM, Bruckman D, Karavite DJ, Russman PL, et al. The International Registry of Acute Aortic Dissection (IRAD): new insights into an old disease. JAMA 2000;283:897. http://dx.doi.org/10.1001/jama.283.7.897.
  13. Evangelista A, Avegliano G, Aguilar R, Cuellar H, Igual A, Gonzalez-Alujas T, et al. Impact of contrast-enhanced echocardiography on the diagnostic algorithm of acute aortic dis- section. Eur Heart J 2010;31:472-9. http://dx.doi.org/10.1093/eurheartj/ehp505.
  14. Khandheria BK, Tajik AJ, Taylor CL, Safford RE, Miller FA, Stanson AW, et al. Aortic dissection: review of value and limitations of two-dimensional echocardiography in a six-year experience. J Am Soc Echocardiogr 1989;2:17-24.
  15. Kodolitsch Y, Krause N, Spielmann R, Nienaber CA. Diagnostic potential of combined transthoracic echocardiography and x-ray computed tomography in suspected aor- tic dissection. Clin Cardiol 1999;22:345-52.
  16. Hahn RT, Roman MJ, Mogtader AH, Devereux RB. Association of aortic dilation with regurgitant, stenotic and functionally normal bicuspid aortic valves. J Am Coll Cardiol 1992;19:283-8.
  17. Gurvitz M, Chang R-K, Drant S, Allada V. Frequency of aortic root dilation in children with a bicuspid aortic valve. Am J Cardiol 2004;94:1337-40. http://dx.doi.org/10. 1016/j.amjcard.2004.07.130.
  18. Moore C. Current issues with emergency cardiac ultrasound probe and image con- ventions. Acad Emerg Med 2008;15:278-84. http://dx.doi.org/10.1111/j.1553- 2712.2008.00052.x.

Leave a Reply

Your email address will not be published. Required fields are marked *