- Email: [email protected]
Prevalence of pulmonary embolism in patients presenting to the emergency department with syncope
Accepted Manuscript Prevalence of pulmonary embolism in patients presenting to the emergency department with syncope
Alison Frizell, Nicole Fogel, Jacob Steenblik, Margaret Carlson, Joseph Bledsoe, Troy Madsen PII: DOI: Reference:
S0735-6757(17)30639-3 doi: 10.1016/j.ajem.2017.07.090 YAJEM 56877
To appear in: Received date: Revised date: Accepted date:
28 March 2017 3 July 2017 30 July 2017
Please cite this article as: Alison Frizell, Nicole Fogel, Jacob Steenblik, Margaret Carlson, Joseph Bledsoe, Troy Madsen , Prevalence of pulmonary embolism in patients presenting to the emergency department with syncope, (2017), doi: 10.1016/j.ajem.2017.07.090
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Prevalence of Pulmonary Embolism in Patients Presenting to the Emergency Department with Syncope Alison Frizell, MD*, Nicole Fogel, HBSc*, Jacob Steenblik, MPH, MHA*, Margaret Carlson, BS*, Joseph Bledsoe, MD#, Troy Madsen, MD* *University of Utah, Salt Lake City, UT
T
#Intermountain Medical Center, Murray, UT
CR
IP
Meetings: Society for Academic Emergency Medicine Annual Meeting, Orlando, Florida, May 16-19, 2017 Conflicts of Interest: None of the authors has any conflicts of interest to report.
AC
CE
PT
ED
M
AN
US
Word Count: 3065
ACCEPTED MANUSCRIPT Abstract Objectives: A recent study reported a high prevalence of pulmonary embolism (PE) among patients admitted with syncope. We sought to determine whether these findings were validated in our patient population.
T
Methods: We performed a retrospective, secondary analysis of prospectively gathered data
IP
from patients presenting with syncope to an academic emergency department (ED) from
CR
July 2010 to December 2015. We analyzed baseline information from the time of the ED visit,
US
recorded outcomes during the hospital stay, and contacted patients by phone at least 30 days after the ED visit. The primary study outcome was the diagnosis of acute PE in the ED, during
AN
inpatient admission or ED observation unit stay, or by patient report over a 30-day follow-up period.
M
Results: Over the 5.5-year study period, 348 patients with syncope agreed to participate in the
ED
study. 52% of patients were female [95% confidence interval (CI): 46.6-57.4] and the average
PT
age was 48.4 years. Of the enrolled patients, 50.1% (CI: 44.8-55.2) underwent further evaluation for syncope beyond the ED stay: 27% (CI: 22.6-31.9) of patients were admitted to an inpatient
CE
unit for further work-up and 23.9% (CI: 19.7-28.6) of patients were placed in the ED observation unit. The overall rate of PE among patients presenting to the ED with syncope was
AC
1.4% (CI: 0.6-3.3%). 2 patients (0.6%, CI: 0.2-2.1) were diagnosed with a PE while in the ED. None of the patients were diagnosed with a PE during hospital admission or the observation stay associated with the index ED visit. 3 patients (0.9%, CI: 0.3-2.5) reported they had been diagnosed with a PE during the 30 days following their ED visit, two of whom had been admitted to the hospital at the index ED visit but were not diagnosed with a PE at that time. All patients diagnosed with a PE at the time of the ED visit or during the follow-up period were
2
ACCEPTED MANUSCRIPT Pulmonary Embolism Rule Out Criteria (PERC) positive and reported shortness of breath in the ED. Conclusion: In contrast to a previous study, our findings do not support a high rate of PE among
AC
CE
PT
ED
M
AN
US
CR
IP
T
ED patients presenting with syncope.
3
ACCEPTED MANUSCRIPT Introduction Syncope is one of the more common reasons for emergency department (ED) visits, comprising 1-3% of ED visits. [1-3] Syncope is defined as transient loss of consciousness due to global cerebral hypoperfusion characterized by rapid onset, short duration, and spontaneous
T
complete recovery. [4-5] Causes of syncope include neurally mediated syncope such as
IP
vasovagal or situational syncope, syncope from orthostatic hypotension, and cardiac syncope
CR
such as from arrhythmia, structural heart disease or pulmonary embolism (PE). [6]
US
As a potential etiology of syncope, PE carries a high morbidity and mortality, but few studies have evaluated the prevalence of PE in patients with syncope. In an effort to better
AN
understand the prevalence of PE in patients presenting with syncope, one recent study of patients presenting with syncope evaluated all patients for PE as a cause of their syncopal
M
episodes. This study found that nearly one in six patients admitted with first-time syncope had a
ED
PE. [7] The authors’ findings contrasted with several previous studies which noted a much
PT
lower incidence of PE in patients with syncope. [8-17] In our study, we sought to evaluate data from our ED to determine the prevalence of PE
CE
in patients presenting with syncope. We hypothesized that, in contrast to this previous study, our
AC
findings would not support a high rate of PE among ED patients presenting with syncope.
4
ACCEPTED MANUSCRIPT Methods Study Design and Setting: This was a retrospective analysis of a pre-existing data set gathered on patients presenting to the University of (Utah) ED from July 2010 to December 2015. The data set includes
T
prospectively gathered data on patients presenting to the ED with syncope, including baseline
IP
information, outcomes of ED testing and inpatient evaluation, and 30-day follow-up by
CR
telephone and chart review. The University of Utah Institutional Review Board approved the
US
study.
The University of Utah ED is an urban, academic ED located in Salt Lake City with
AN
approximately 40,000 patient visits per year. Clinical protocols do not exist at the University of Utah ED which mandate testing for PE in patients presenting to the ED with syncope, with
M
testing for PE performed at the treating clinician’s discretion. The University of Utah’s ED
ED
observation unit (EDOU) includes a protocol for the evaluation of patients with syncope.
PT
Specific aspects of the EDOU syncope protocol include telemetry monitoring, transthoracic
CE
echocardiogram, and arrangement of outpatient Holter monitoring.
AC
Selection of Participants:
We prospectively enrolled patients over the 5.5-year study period from July 2010 through December 2015. Research associates were present in the ED seven days a week between 8 am and midnight. Given that research associates receive undergraduate course credit for their participation, they were not present in the ED on university holidays nor during semester breaks. They identified ED patients with potential syncope based on the triage chief complaint and approached these patients early in the ED stay to obtain consent for study
5
ACCEPTED MANUSCRIPT participation. They then confirmed with patients that they had experienced a recent syncopal event by asking the following question: “Have you passed out (been unresponsive) in the past 24 hours?” Individuals who answered “no” to this question were not included in the study. We excluded patients under the age of 18 years old, prisoners, and individuals who were not capable
T
of providing consent for study participation. Additional reasons not to be included in the
IP
study were patient refusal to participate in the study, failure to report a syncopal event
CR
within the past 24 hours, or inability to complete the survey with the patient due to additional testing procedures or lack of time. Research associates recorded the number of
US
patients screened for study enrollment, but they did not record specific reasons for the
AN
failure to complete the study enrollment and required information.
M
Methods and Measurements:
ED
Research associates administered a standard questionnaire to obtain baseline information at the time of the ED visit, including demographic information, current symptoms, and past
PT
medical history. Research associates also recorded patient vital signs, laboratory and imaging
CE
results, interventions performed in the ED and disposition. We calculated the Pulmonary Embolism Rule-out Criteria (PERC) utilizing patient-reported risk factors and ED vital sign
AC
measurements at the time of ED triage. [18] We collected and managed study data using REDCap, an electronic data capture tool hosted at the University of Utah. [19] REDCap (Research Electronic Data Capture) is a secure, webbased application designed to support data capture for research studies, providing: 1. An intuitive interface for validated data entry; 2. Audit trails for tracking data manipulation and export procedures
6
ACCEPTED MANUSCRIPT 3. Automated export procedures for seamless data downloads to common statistical packages 4. Procedures for importing data from external sources. Information obtained in the baseline survey was not shared with ED providers and these
T
providers may not have even known that their patients had been enrolled in the study. Providers
IP
were blinded to study aims and had no knowledge that we would be looking at PE outcomes
CR
among patients with syncope.
For patients placed in the EDOU or admitted to an inpatient unit, research associates
US
recorded results of testing and patient outcomes. Additionally, research associates attempted to
AN
contact patients by telephone at least 30 days after the ED visit to inquire about any repeat ED visits or hospitalizations, follow up with a primary care provider, and any additional testing
M
performed or diagnoses beyond those identified during the initial encounter. For patients not
ED
contacted by telephone, we reviewed the medical record and noted any significant events or
PT
diagnoses during the 30 days after the ED visit.
CE
Outcomes:
The primary study outcome was the diagnosis of acute PE in the ED, during inpatient
AC
admission or ED observation unit stay, or by patient report over a 30-day follow-up period. We documented the diagnosis of PE based on a positive result of CTPA or VQ scan testing for PE. We utilized both outcomes of ED/inpatient testing and the 30-day follow-up period as we felt that a PE which was clinically relevant in leading to a syncopal event would likely manifest itself during a 30-day window if it had not been identified during the initial ED encounter.
7
ACCEPTED MANUSCRIPT Analysis: We performed data analysis utilizing descriptive statistics with data presented as percentages for categorical variables and means for continuous variables. We evaluated differences between groups of categorical variables utilizing the Pearson’s chi-square test and
T
differences between continuous variables using Student’s t-test. We present results using odds
IP
ratios (ORs) and 95% confidence intervals (CIs), with a p-value less than 0.05 considered
CR
statistically significant. We performed analysis using STATA version 12.0 (StataCorp, College
AC
CE
PT
ED
M
AN
US
Station, TX) and VassarStats Website for Statistical Computation (vassarstats.net).
8
ACCEPTED MANUSCRIPT Results Patient Characteristics: Over the 5.5-year study period, research associates screened 778 potential syncope patients for enrollment in the study. Of these patients, 348 ED patients reported a syncopal
T
event within the preceding 24 hours and agreed to participate in the study. Of the study patients,
IP
52% were female (CI: 46.6-57.4) and the average age was 48.4 years (range: 18-92 years).
CR
Study patients reported their race/ethnicity as White non-Hispanic (79.3%, CI: 74.8-83.2),
US
Hispanic (8.3%, CI: 5.9-11.7), Black (3.2%, CI:1.8-5.6), Native American (2%, CI: 1-4.1), Pacific Islander (0.9%, CI: 0.3-2.5), Asian (0.9%, CI:0.3-2.5), and Other (5.4%, CI: 3.5-8.4).
AN
When asked if they had a previous history of PE or deep vein thrombosis (DVT), 15.8% of patients answered “yes” (CI: 12.3-20). Additional notable baseline clinical characteristics
M
included reported rates of shortness of breath (54.3%, CI: 49.1-59.5), chest pain (49.1%, CI:
PT
ED
43.9-54.4), and calf pain/swelling (17%, CI: 13.4-21.3).
PE Diagnosis During Initial Visit:
CE
We evaluated rates of PE among study patients. While in the ED, 22.7% (CI: 18.6-27.4)
AC
of patients underwent testing for PE: 9.8% (CI: 7.1-13.3) had d-dimer testing while 16.7% (CI: 13.1-21) had CTPA or VQ scan to evaluate for PE (3.8% had a positive d-dimer followed by CTPA or VQ scan, CI: 2.2-6.3). Two patients (0.6%, CI: 0.2-2.1) were diagnosed with a PE in the ED. Of the study patients, 50.1% (CI: 44.8-55.2) underwent further hospital evaluation for syncope beyond the ED stay: 27% (CI: 22.6-31.9) of patients were admitted to an inpatient unit for further work-up and 23.9% (CI: 19.7-28.6) of patients were placed in the ED observation
9
ACCEPTED MANUSCRIPT unit. None of the patients was diagnosed with a PE during hospital admission or the observation stay associated with the index ED visit.
PE Diagnosis During Follow-up:
T
We performed telephone follow-up to evaluate for the diagnosis of PE during the 30
IP
days after the ED visit. Our overall telephone follow-up rate was 68.4% (CI: 63.3-73.1),
CR
and the average time from the ED visit to telephone contact was 56.9 days (CI: 50.5-63.3). For patients whom we could not contact by telephone, we reviewed the EMR for any
US
significant events during the 30 days following the ED visit. We later compared clinical
AN
characteristics of patients contacted by telephone and those who were not contacted by phone to evaluate for potential bias in follow-up data. Patients contacted by phone were
M
similar to those not contacted by phone in gender (percent female: 52.1% vs. 51.8%,
ED
p=0.961), those who were PERC positive (82.3% vs. 75.5%, p=0.109), those who reported a previous PE or DVT (16% vs. 15.5%, p=0.903), and rates of hypoxia (triage oxygen
PT
saturation <95%: 24.8% vs. 20%, p=0.326). Patients who were contacted by phone were
CE
slightly older than those not contacted by phone: 49.8 years vs. 45.4 years (p=0.040). At the time of telephone follow-up, 3 patients (0.9%, CI: 0.3-2.5) reported they had been
AC
diagnosed with a PE during the 30 days following their ED visit for syncope. One patient reported being diagnosed with a pulmonary embolism and myocardial infarction at an outside hospital one week after the ED visit, followed by a two-day hospitalization at this hospital. This patient, who was low risk by Wells’ Criteria (score=1.5), had a negative d-dimer during the index ED visit and was discharged from the ED. A second patient reported a PE approximately three weeks after the ED visit followed by two-day hospitalization. This patient did not have PE
10
ACCEPTED MANUSCRIPT testing in the ED but had been admitted to an inpatient unit for further evaluation of syncope. A third patient reported a pulmonary embolism diagnosed after the ED visit but did not provide additional detail. This patients had a negative CTPA in the ED and had been admitted to an inpatient unit at the time of the ED stay. In none of these cases of self-reported PE were we able
IP
T
to verify the diagnosis of PE utilizing our EMR. [Figure]
CR
Overall Rate of PE Diagnosis:
Based on the PE diagnoses in the ED and the self-reported PE diagnoses during the 30-
US
day follow-up period, our overall rate of PE among patients presenting with syncope was 1.4%
AN
(95% CI: 0.6-3.3%). All patients with PE reported shortness of breath at the time of the ED visit in addition to syncope, and all were PERC positive. Two of the five patients reported calf pain or
M
swelling and one reported a previous history of a PE or DVT. In the cases of the two patients for
ED
whom we had objective results for CTPA performed in the ED, both had large PEs involving the
AC
CE
PT
main pulmonary arteries.
11
ACCEPTED MANUSCRIPT Discussion Our study evaluated data collected at the time of ED visit, during hospitalization, and at a 30-day phone follow-up to determine the prevalence of PE in patients presenting with syncope. We found a low prevalence of PE in our population, with a rate of 1.4% among patients
T
presenting to the ED with a syncopal event in the preceding 24 hours. Two patients were
IP
diagnosed during their ED visit and the other three patients reported being diagnosed with a PE
CR
during the 30 days following their ED visit. Interestingly, no patients were diagnosed with PE
US
while admitted as an inpatient or during their stay in the observation unit. All patients diagnosed with a PE at their ED visit or during the 30-day follow-up period were PERC positive and
AN
reported dyspnea in the ED.
The low prevalence of pulmonary embolism among our patients with syncope contrasts
M
with that reported in a recent study by Prandoni et al. This may be partly due to methodological
ED
differences between our studies. In the Prandoni study, all patients underwent a systematic work-
PT
up for PE regardless of their symptoms. In our study, the work-up for both PE and syncope was at the discretion of the provider. We did not differentiate between patients with first-time or
CE
repeated syncopal events, unlike the Prandoni study. In the Prandoni study, syncope was defined
AC
as “a transient loss of consciousness with rapid onset, short duration (i.e., <1 minute), and spontaneous resolution, with obvious causes such as epileptic seizure, stroke, and head trauma ruled out.” In our study, patients with syncope were identified by a positive response to the question “Have you passed out (been unresponsive) in the past 24 hours?” This may not have excluded trauma or seizure patients as the Prandoni study did. We also included 30-day followup to determine if patients had been diagnosed with PE shortly after their hospitalization. [7] Notably, the Prandoni study was performed in Italian medical centers. The work-up for
12
ACCEPTED MANUSCRIPT PE and diagnosis occurred while patients were admitted to an inpatient unit, not during their ED stay. In contrast, in our study both patients who were diagnosed with PE during their hospitalization were diagnosed in the ED, not during their admission or EDOU stay. It may be that the expectation for the extent of the evaluation in the ED versus inpatient units differs
T
between the medical centers in these two studies. Our findings are more consistent with
IP
previously reported literature regarding the prevalence of pulmonary embolism in patients
CR
presenting with syncope, but still demonstrated a lower prevalence than some of these previous
US
studies. [8-17]
While not all patients underwent a work-up for PE, we did contact 68.4% of patients 30
AN
days after their ED visit to determine if PE was diagnosed in the subsequent time period. Only
M
three patients reported being diagnosed with a PE in the 30 days following their ED visit, two of whom had been admitted to the hospital at their index visit but had not been diagnosed during
ED
that admission. We did not confirm any of these PE diagnoses in our medical record through
PT
chart review, so we were limited in being able to characterize the exact nature of the PE or to corroborate the patient’s report of this new diagnosis. Given this low number of PEs reported
CE
during follow-up, we concluded that the diagnosis of PE was infrequently missed despite the fact
AC
that not all patients underwent a systematic work-up. However, given that we were unable to contact almost 32% of study patients by phone, it is possible that an additional number of patients had PE which we did not note in the medical record review for these patients. Lastly, it is difficult to determine causation between PE and syncope, even in patients where PE was confirmed on imaging. The mechanism of syncope from PE is thought to be due to the sudden obstruction of the most proximal pulmonary arteries which leads to a drop in cardiac output, precipitating cerebral hypoperfusion. [3-4] Therefore, it would seem that only
13
ACCEPTED MANUSCRIPT proximal PEs should result in syncope. Both cases of ED-diagnosed PE in our study were proximal PEs, and in these cases the relationship between the PE and the patient’s syncopal event would seem plausible. However, in the case of the patients who reported PE during the 30-day follow-up period, we are unable to comment on the nature of the PE and the potential
IP
T
relationship to the syncopal event they reported upon arrival to the ED.
CR
Additional limitations in our study warrant comment. Our study was limited to a single center, and thus our findings may be unique to this center, to the PE and syncope work-ups
US
which are performed at this site, and to the nature of our patient population. While we feel confident in the outcomes related to the ED visit, hospitalization, and the EDOU stay, we were
AN
unable to perform 30-day telephone follow-up with 32% of study patients and may have missed
M
significant events due to this. While we attempted to mitigate this effect through review of the medical records of these individuals to note 30-day events documented in our EMR, we also
ED
recognize that the three individuals who reported PE during the follow-up period did not have
PT
these in our EMR and presumably had these diagnoses outside of our medical system. Patients contacted by phone were slightly older than those whom we were unable to contact, but we
CE
did not otherwise find significant differences between these two groups. Given that these
AC
groups did not differ significantly in their demographics or risk factor profiles, we would not expect a disproportionate number of missed PEs in the group we were unable to contact by phone, but we still may have missed several significant outcomes. The low number of enrolled patients with PE over the 5.5-year study period presents an additional limitation. Our enrollment was limited by the RA coverage hours in the ED and by the study exclusion criteria. Given previous reports suggesting that syncope visits comprise 1-3% of ED visits, and given our ED volume of 40,000 patients per year, we 14
ACCEPTED MANUSCRIPT would expect to have had at least 2,000 patients with syncope in our ED over the enrollment period. RAs enrolled fewer than half of the 778 patients screened, and while we did not systematically note the reasons for exclusion from enrollment, factors ranged from failure to meet enrollment criteria (specifically, a syncopal event in the preceding 24 hours)
T
and patient refusal to participate. The potential for selection bias, related both to the total
IP
number of enrolled patients and to those excluded after initial screening, may have
CR
additionally affected our outcomes.
US
In summary, in contrast to a recent study, our findings do not support a high rate of PE among ED patients presenting with syncope. While our study did not include a systematic work-
AN
up for PE in all patients with syncope, we would expect clinically relevant PE which accounted
M
for syncope to manifest itself during the ED stay, hospitalization, EDOU stay, or 30-day followup period. Patients who had PE and syncope were all PERC positive and reported shortness of
ED
breath, in addition to other signs and symptoms of PE. Our findings do not suggest a need for a
PT
systematic work-up for PE in all patients presenting with syncope, but support a providerdirected work-up based on the full clinical picture of patient symptoms, risk factors, and
AC
CE
objective findings at the time of ED or inpatient evaluation.
15
ACCEPTED MANUSCRIPT References 1. Puppala VK, Dickinson O, Benditt DG. Syncope: classification and risk stratification. J Cardiol 2014;63:171-7.
T
2. Koutkia P, Wachtel TJ. Pulmonary embolism presenting as syncope: case report and review of
IP
the literature. Heart Lung 1999;28:342-7.
CR
3. Keller K, Beule J, Balzer JO, Dippold W. Syncope and collapse in acute pulmonary embolism. Am J Emerg Med 2016;34: 1251-7.
US
4. Moya A, Sutton R, Ammirati F, Blanc JJ, Brignole M, Dahm JB, et al. Guidelines for the diagnosis and management of syncope (version 2009). Eur Heart J 2009;30: 2631–71.
AN
5. Constantino G, Sun BC, Barbic F, et al. Syncope Clinical Management in the Emergency
M
Department: A Consensus From the First International Workshop on Syncope Risk Stratification
ED
in the Emergency Department. Eur Heart J 2016 2016;37(19):1493-8. 6. Strickberger SA, Benson DW, Biaggioni I, et al. AHA/ACCF Scientific Statement on the
PT
evaluation of syncope: from the American Heart Association Councils on Clinical Cardiology,
CE
Cardiovascular Nursing, Cardiovascular Disease in the Young, and Stroke, and the Quality of Care and Outcomes Research Interdisciplinary Working Group; and the American College of
AC
Cardiology Foundation: in collaboration with the Heart Rhythm Society: endorsed by the American Autonomic Society. Circulation 2006;113:316- 27. 7. Prandoni P, Lensing AWA, Prins MH, et al. Prevalence of Pulmonary Embolism among Patients Hospitalized for Syncope. N Engl J Med 2016;375:1524-31. 8. Koutkia P, Wachtel TJ. Pulmonary embolism presenting as syncope: case report and review of the literature. Heart Lung 1999;28:342-7. 9. Sarasin FP, Louis-Simonet M, Carballo D, et al. Prospective evaluation of patients with
16
ACCEPTED MANUSCRIPT syncope: a population-based study. Am J Med 2001;111:177-84. (Duplicate) 10. Jenab Y, Lotfi-Tokaldany M, Alemza- deh-Ansari MJ, et al. Correlates of syncope in patients with acute pulmonary thromboembolism. Clin Appl Thromb Hemost 2015;21:772-6. 11. Keller K, Beule J, Balzer JO, Dippold W. Syncope and collapse in acute pulmonary
T
embolism. Am J Emerg Med 2016;34: 1251-7.
IP
12. Thames MD, Alpert JS, Dalen JE. Syncope in patients with pulmonary embolism. JAMA
CR
1977;238:2509-11.
13. Soteriades ES, Evans JC, Larson MG, et al. Incidence and prognosis of syncope. N Engl J
US
Med 2002;347:878-85. (duplicate)
AN
14. Theilade J, Winkel BG, Holst AG, Tfelt-Hansen J, Svendsen JH, Haunsø S. A nation- wide, retrospective analysis of symptoms, comorbidities, medical care and autopsy findings in cases of
M
fatal pulmonary embolism in younger patients. J Thromb Haemost 2010;8:1723-9.
ED
15. Castelli R, Tarsia P, Tantardini C, Pantaleo G, Guariglia A, Porro F. Syncope in patients with pulmonary embolism: comparison between patients with syncope as the presenting symptom of
PT
pulmonary embolism and patients with pulmonary embolism without syncope. Vasc Med
CE
2003;8:257-61.
16. Saravi M, Ahmadi Ahangar A, Hojati MM, et al. Etiology of syncope in hospitalized
AC
patients. Caspian J Intern Med 2015;6:233-7. 17. Blanc JJ, L'Her C, Touiza A, Garo B, et al. Prospective evaluation and outcome of patients admitted for syncope over a 1 year period. Eur Heart J. 2002 May;23(10):815-20. 18. Kline JA, Courtney DM, Kabrhel C, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008: 6(5): 772-80 19. Harris PH, Taylor R, Thielke R, et. al. Research electronic data capture (REDCap) - A
17
ACCEPTED MANUSCRIPT metadata-driven methodology and workflow process for providing translational research
AC
CE
PT
ED
M
AN
US
CR
IP
T
informatics support. J Biomed Inform. 2009 42(2):377-81.
18
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
Figure: Flow chart for the diagnosis of PE among ED patients with syncope
19
Alison Frizell, Nicole Fogel, Jacob Steenblik, Margaret Carlson, Joseph Bledsoe, Troy Madsen PII: DOI: Reference:
S0735-6757(17)30639-3 doi: 10.1016/j.ajem.2017.07.090 YAJEM 56877
To appear in: Received date: Revised date: Accepted date:
28 March 2017 3 July 2017 30 July 2017
Please cite this article as: Alison Frizell, Nicole Fogel, Jacob Steenblik, Margaret Carlson, Joseph Bledsoe, Troy Madsen , Prevalence of pulmonary embolism in patients presenting to the emergency department with syncope, (2017), doi: 10.1016/j.ajem.2017.07.090
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Prevalence of Pulmonary Embolism in Patients Presenting to the Emergency Department with Syncope Alison Frizell, MD*, Nicole Fogel, HBSc*, Jacob Steenblik, MPH, MHA*, Margaret Carlson, BS*, Joseph Bledsoe, MD#, Troy Madsen, MD* *University of Utah, Salt Lake City, UT
T
#Intermountain Medical Center, Murray, UT
CR
IP
Meetings: Society for Academic Emergency Medicine Annual Meeting, Orlando, Florida, May 16-19, 2017 Conflicts of Interest: None of the authors has any conflicts of interest to report.
AC
CE
PT
ED
M
AN
US
Word Count: 3065
ACCEPTED MANUSCRIPT Abstract Objectives: A recent study reported a high prevalence of pulmonary embolism (PE) among patients admitted with syncope. We sought to determine whether these findings were validated in our patient population.
T
Methods: We performed a retrospective, secondary analysis of prospectively gathered data
IP
from patients presenting with syncope to an academic emergency department (ED) from
CR
July 2010 to December 2015. We analyzed baseline information from the time of the ED visit,
US
recorded outcomes during the hospital stay, and contacted patients by phone at least 30 days after the ED visit. The primary study outcome was the diagnosis of acute PE in the ED, during
AN
inpatient admission or ED observation unit stay, or by patient report over a 30-day follow-up period.
M
Results: Over the 5.5-year study period, 348 patients with syncope agreed to participate in the
ED
study. 52% of patients were female [95% confidence interval (CI): 46.6-57.4] and the average
PT
age was 48.4 years. Of the enrolled patients, 50.1% (CI: 44.8-55.2) underwent further evaluation for syncope beyond the ED stay: 27% (CI: 22.6-31.9) of patients were admitted to an inpatient
CE
unit for further work-up and 23.9% (CI: 19.7-28.6) of patients were placed in the ED observation unit. The overall rate of PE among patients presenting to the ED with syncope was
AC
1.4% (CI: 0.6-3.3%). 2 patients (0.6%, CI: 0.2-2.1) were diagnosed with a PE while in the ED. None of the patients were diagnosed with a PE during hospital admission or the observation stay associated with the index ED visit. 3 patients (0.9%, CI: 0.3-2.5) reported they had been diagnosed with a PE during the 30 days following their ED visit, two of whom had been admitted to the hospital at the index ED visit but were not diagnosed with a PE at that time. All patients diagnosed with a PE at the time of the ED visit or during the follow-up period were
2
ACCEPTED MANUSCRIPT Pulmonary Embolism Rule Out Criteria (PERC) positive and reported shortness of breath in the ED. Conclusion: In contrast to a previous study, our findings do not support a high rate of PE among
AC
CE
PT
ED
M
AN
US
CR
IP
T
ED patients presenting with syncope.
3
ACCEPTED MANUSCRIPT Introduction Syncope is one of the more common reasons for emergency department (ED) visits, comprising 1-3% of ED visits. [1-3] Syncope is defined as transient loss of consciousness due to global cerebral hypoperfusion characterized by rapid onset, short duration, and spontaneous
T
complete recovery. [4-5] Causes of syncope include neurally mediated syncope such as
IP
vasovagal or situational syncope, syncope from orthostatic hypotension, and cardiac syncope
CR
such as from arrhythmia, structural heart disease or pulmonary embolism (PE). [6]
US
As a potential etiology of syncope, PE carries a high morbidity and mortality, but few studies have evaluated the prevalence of PE in patients with syncope. In an effort to better
AN
understand the prevalence of PE in patients presenting with syncope, one recent study of patients presenting with syncope evaluated all patients for PE as a cause of their syncopal
M
episodes. This study found that nearly one in six patients admitted with first-time syncope had a
ED
PE. [7] The authors’ findings contrasted with several previous studies which noted a much
PT
lower incidence of PE in patients with syncope. [8-17] In our study, we sought to evaluate data from our ED to determine the prevalence of PE
CE
in patients presenting with syncope. We hypothesized that, in contrast to this previous study, our
AC
findings would not support a high rate of PE among ED patients presenting with syncope.
4
ACCEPTED MANUSCRIPT Methods Study Design and Setting: This was a retrospective analysis of a pre-existing data set gathered on patients presenting to the University of (Utah) ED from July 2010 to December 2015. The data set includes
T
prospectively gathered data on patients presenting to the ED with syncope, including baseline
IP
information, outcomes of ED testing and inpatient evaluation, and 30-day follow-up by
CR
telephone and chart review. The University of Utah Institutional Review Board approved the
US
study.
The University of Utah ED is an urban, academic ED located in Salt Lake City with
AN
approximately 40,000 patient visits per year. Clinical protocols do not exist at the University of Utah ED which mandate testing for PE in patients presenting to the ED with syncope, with
M
testing for PE performed at the treating clinician’s discretion. The University of Utah’s ED
ED
observation unit (EDOU) includes a protocol for the evaluation of patients with syncope.
PT
Specific aspects of the EDOU syncope protocol include telemetry monitoring, transthoracic
CE
echocardiogram, and arrangement of outpatient Holter monitoring.
AC
Selection of Participants:
We prospectively enrolled patients over the 5.5-year study period from July 2010 through December 2015. Research associates were present in the ED seven days a week between 8 am and midnight. Given that research associates receive undergraduate course credit for their participation, they were not present in the ED on university holidays nor during semester breaks. They identified ED patients with potential syncope based on the triage chief complaint and approached these patients early in the ED stay to obtain consent for study
5
ACCEPTED MANUSCRIPT participation. They then confirmed with patients that they had experienced a recent syncopal event by asking the following question: “Have you passed out (been unresponsive) in the past 24 hours?” Individuals who answered “no” to this question were not included in the study. We excluded patients under the age of 18 years old, prisoners, and individuals who were not capable
T
of providing consent for study participation. Additional reasons not to be included in the
IP
study were patient refusal to participate in the study, failure to report a syncopal event
CR
within the past 24 hours, or inability to complete the survey with the patient due to additional testing procedures or lack of time. Research associates recorded the number of
US
patients screened for study enrollment, but they did not record specific reasons for the
AN
failure to complete the study enrollment and required information.
M
Methods and Measurements:
ED
Research associates administered a standard questionnaire to obtain baseline information at the time of the ED visit, including demographic information, current symptoms, and past
PT
medical history. Research associates also recorded patient vital signs, laboratory and imaging
CE
results, interventions performed in the ED and disposition. We calculated the Pulmonary Embolism Rule-out Criteria (PERC) utilizing patient-reported risk factors and ED vital sign
AC
measurements at the time of ED triage. [18] We collected and managed study data using REDCap, an electronic data capture tool hosted at the University of Utah. [19] REDCap (Research Electronic Data Capture) is a secure, webbased application designed to support data capture for research studies, providing: 1. An intuitive interface for validated data entry; 2. Audit trails for tracking data manipulation and export procedures
6
ACCEPTED MANUSCRIPT 3. Automated export procedures for seamless data downloads to common statistical packages 4. Procedures for importing data from external sources. Information obtained in the baseline survey was not shared with ED providers and these
T
providers may not have even known that their patients had been enrolled in the study. Providers
IP
were blinded to study aims and had no knowledge that we would be looking at PE outcomes
CR
among patients with syncope.
For patients placed in the EDOU or admitted to an inpatient unit, research associates
US
recorded results of testing and patient outcomes. Additionally, research associates attempted to
AN
contact patients by telephone at least 30 days after the ED visit to inquire about any repeat ED visits or hospitalizations, follow up with a primary care provider, and any additional testing
M
performed or diagnoses beyond those identified during the initial encounter. For patients not
ED
contacted by telephone, we reviewed the medical record and noted any significant events or
PT
diagnoses during the 30 days after the ED visit.
CE
Outcomes:
The primary study outcome was the diagnosis of acute PE in the ED, during inpatient
AC
admission or ED observation unit stay, or by patient report over a 30-day follow-up period. We documented the diagnosis of PE based on a positive result of CTPA or VQ scan testing for PE. We utilized both outcomes of ED/inpatient testing and the 30-day follow-up period as we felt that a PE which was clinically relevant in leading to a syncopal event would likely manifest itself during a 30-day window if it had not been identified during the initial ED encounter.
7
ACCEPTED MANUSCRIPT Analysis: We performed data analysis utilizing descriptive statistics with data presented as percentages for categorical variables and means for continuous variables. We evaluated differences between groups of categorical variables utilizing the Pearson’s chi-square test and
T
differences between continuous variables using Student’s t-test. We present results using odds
IP
ratios (ORs) and 95% confidence intervals (CIs), with a p-value less than 0.05 considered
CR
statistically significant. We performed analysis using STATA version 12.0 (StataCorp, College
AC
CE
PT
ED
M
AN
US
Station, TX) and VassarStats Website for Statistical Computation (vassarstats.net).
8
ACCEPTED MANUSCRIPT Results Patient Characteristics: Over the 5.5-year study period, research associates screened 778 potential syncope patients for enrollment in the study. Of these patients, 348 ED patients reported a syncopal
T
event within the preceding 24 hours and agreed to participate in the study. Of the study patients,
IP
52% were female (CI: 46.6-57.4) and the average age was 48.4 years (range: 18-92 years).
CR
Study patients reported their race/ethnicity as White non-Hispanic (79.3%, CI: 74.8-83.2),
US
Hispanic (8.3%, CI: 5.9-11.7), Black (3.2%, CI:1.8-5.6), Native American (2%, CI: 1-4.1), Pacific Islander (0.9%, CI: 0.3-2.5), Asian (0.9%, CI:0.3-2.5), and Other (5.4%, CI: 3.5-8.4).
AN
When asked if they had a previous history of PE or deep vein thrombosis (DVT), 15.8% of patients answered “yes” (CI: 12.3-20). Additional notable baseline clinical characteristics
M
included reported rates of shortness of breath (54.3%, CI: 49.1-59.5), chest pain (49.1%, CI:
PT
ED
43.9-54.4), and calf pain/swelling (17%, CI: 13.4-21.3).
PE Diagnosis During Initial Visit:
CE
We evaluated rates of PE among study patients. While in the ED, 22.7% (CI: 18.6-27.4)
AC
of patients underwent testing for PE: 9.8% (CI: 7.1-13.3) had d-dimer testing while 16.7% (CI: 13.1-21) had CTPA or VQ scan to evaluate for PE (3.8% had a positive d-dimer followed by CTPA or VQ scan, CI: 2.2-6.3). Two patients (0.6%, CI: 0.2-2.1) were diagnosed with a PE in the ED. Of the study patients, 50.1% (CI: 44.8-55.2) underwent further hospital evaluation for syncope beyond the ED stay: 27% (CI: 22.6-31.9) of patients were admitted to an inpatient unit for further work-up and 23.9% (CI: 19.7-28.6) of patients were placed in the ED observation
9
ACCEPTED MANUSCRIPT unit. None of the patients was diagnosed with a PE during hospital admission or the observation stay associated with the index ED visit.
PE Diagnosis During Follow-up:
T
We performed telephone follow-up to evaluate for the diagnosis of PE during the 30
IP
days after the ED visit. Our overall telephone follow-up rate was 68.4% (CI: 63.3-73.1),
CR
and the average time from the ED visit to telephone contact was 56.9 days (CI: 50.5-63.3). For patients whom we could not contact by telephone, we reviewed the EMR for any
US
significant events during the 30 days following the ED visit. We later compared clinical
AN
characteristics of patients contacted by telephone and those who were not contacted by phone to evaluate for potential bias in follow-up data. Patients contacted by phone were
M
similar to those not contacted by phone in gender (percent female: 52.1% vs. 51.8%,
ED
p=0.961), those who were PERC positive (82.3% vs. 75.5%, p=0.109), those who reported a previous PE or DVT (16% vs. 15.5%, p=0.903), and rates of hypoxia (triage oxygen
PT
saturation <95%: 24.8% vs. 20%, p=0.326). Patients who were contacted by phone were
CE
slightly older than those not contacted by phone: 49.8 years vs. 45.4 years (p=0.040). At the time of telephone follow-up, 3 patients (0.9%, CI: 0.3-2.5) reported they had been
AC
diagnosed with a PE during the 30 days following their ED visit for syncope. One patient reported being diagnosed with a pulmonary embolism and myocardial infarction at an outside hospital one week after the ED visit, followed by a two-day hospitalization at this hospital. This patient, who was low risk by Wells’ Criteria (score=1.5), had a negative d-dimer during the index ED visit and was discharged from the ED. A second patient reported a PE approximately three weeks after the ED visit followed by two-day hospitalization. This patient did not have PE
10
ACCEPTED MANUSCRIPT testing in the ED but had been admitted to an inpatient unit for further evaluation of syncope. A third patient reported a pulmonary embolism diagnosed after the ED visit but did not provide additional detail. This patients had a negative CTPA in the ED and had been admitted to an inpatient unit at the time of the ED stay. In none of these cases of self-reported PE were we able
IP
T
to verify the diagnosis of PE utilizing our EMR. [Figure]
CR
Overall Rate of PE Diagnosis:
Based on the PE diagnoses in the ED and the self-reported PE diagnoses during the 30-
US
day follow-up period, our overall rate of PE among patients presenting with syncope was 1.4%
AN
(95% CI: 0.6-3.3%). All patients with PE reported shortness of breath at the time of the ED visit in addition to syncope, and all were PERC positive. Two of the five patients reported calf pain or
M
swelling and one reported a previous history of a PE or DVT. In the cases of the two patients for
ED
whom we had objective results for CTPA performed in the ED, both had large PEs involving the
AC
CE
PT
main pulmonary arteries.
11
ACCEPTED MANUSCRIPT Discussion Our study evaluated data collected at the time of ED visit, during hospitalization, and at a 30-day phone follow-up to determine the prevalence of PE in patients presenting with syncope. We found a low prevalence of PE in our population, with a rate of 1.4% among patients
T
presenting to the ED with a syncopal event in the preceding 24 hours. Two patients were
IP
diagnosed during their ED visit and the other three patients reported being diagnosed with a PE
CR
during the 30 days following their ED visit. Interestingly, no patients were diagnosed with PE
US
while admitted as an inpatient or during their stay in the observation unit. All patients diagnosed with a PE at their ED visit or during the 30-day follow-up period were PERC positive and
AN
reported dyspnea in the ED.
The low prevalence of pulmonary embolism among our patients with syncope contrasts
M
with that reported in a recent study by Prandoni et al. This may be partly due to methodological
ED
differences between our studies. In the Prandoni study, all patients underwent a systematic work-
PT
up for PE regardless of their symptoms. In our study, the work-up for both PE and syncope was at the discretion of the provider. We did not differentiate between patients with first-time or
CE
repeated syncopal events, unlike the Prandoni study. In the Prandoni study, syncope was defined
AC
as “a transient loss of consciousness with rapid onset, short duration (i.e., <1 minute), and spontaneous resolution, with obvious causes such as epileptic seizure, stroke, and head trauma ruled out.” In our study, patients with syncope were identified by a positive response to the question “Have you passed out (been unresponsive) in the past 24 hours?” This may not have excluded trauma or seizure patients as the Prandoni study did. We also included 30-day followup to determine if patients had been diagnosed with PE shortly after their hospitalization. [7] Notably, the Prandoni study was performed in Italian medical centers. The work-up for
12
ACCEPTED MANUSCRIPT PE and diagnosis occurred while patients were admitted to an inpatient unit, not during their ED stay. In contrast, in our study both patients who were diagnosed with PE during their hospitalization were diagnosed in the ED, not during their admission or EDOU stay. It may be that the expectation for the extent of the evaluation in the ED versus inpatient units differs
T
between the medical centers in these two studies. Our findings are more consistent with
IP
previously reported literature regarding the prevalence of pulmonary embolism in patients
CR
presenting with syncope, but still demonstrated a lower prevalence than some of these previous
US
studies. [8-17]
While not all patients underwent a work-up for PE, we did contact 68.4% of patients 30
AN
days after their ED visit to determine if PE was diagnosed in the subsequent time period. Only
M
three patients reported being diagnosed with a PE in the 30 days following their ED visit, two of whom had been admitted to the hospital at their index visit but had not been diagnosed during
ED
that admission. We did not confirm any of these PE diagnoses in our medical record through
PT
chart review, so we were limited in being able to characterize the exact nature of the PE or to corroborate the patient’s report of this new diagnosis. Given this low number of PEs reported
CE
during follow-up, we concluded that the diagnosis of PE was infrequently missed despite the fact
AC
that not all patients underwent a systematic work-up. However, given that we were unable to contact almost 32% of study patients by phone, it is possible that an additional number of patients had PE which we did not note in the medical record review for these patients. Lastly, it is difficult to determine causation between PE and syncope, even in patients where PE was confirmed on imaging. The mechanism of syncope from PE is thought to be due to the sudden obstruction of the most proximal pulmonary arteries which leads to a drop in cardiac output, precipitating cerebral hypoperfusion. [3-4] Therefore, it would seem that only
13
ACCEPTED MANUSCRIPT proximal PEs should result in syncope. Both cases of ED-diagnosed PE in our study were proximal PEs, and in these cases the relationship between the PE and the patient’s syncopal event would seem plausible. However, in the case of the patients who reported PE during the 30-day follow-up period, we are unable to comment on the nature of the PE and the potential
IP
T
relationship to the syncopal event they reported upon arrival to the ED.
CR
Additional limitations in our study warrant comment. Our study was limited to a single center, and thus our findings may be unique to this center, to the PE and syncope work-ups
US
which are performed at this site, and to the nature of our patient population. While we feel confident in the outcomes related to the ED visit, hospitalization, and the EDOU stay, we were
AN
unable to perform 30-day telephone follow-up with 32% of study patients and may have missed
M
significant events due to this. While we attempted to mitigate this effect through review of the medical records of these individuals to note 30-day events documented in our EMR, we also
ED
recognize that the three individuals who reported PE during the follow-up period did not have
PT
these in our EMR and presumably had these diagnoses outside of our medical system. Patients contacted by phone were slightly older than those whom we were unable to contact, but we
CE
did not otherwise find significant differences between these two groups. Given that these
AC
groups did not differ significantly in their demographics or risk factor profiles, we would not expect a disproportionate number of missed PEs in the group we were unable to contact by phone, but we still may have missed several significant outcomes. The low number of enrolled patients with PE over the 5.5-year study period presents an additional limitation. Our enrollment was limited by the RA coverage hours in the ED and by the study exclusion criteria. Given previous reports suggesting that syncope visits comprise 1-3% of ED visits, and given our ED volume of 40,000 patients per year, we 14
ACCEPTED MANUSCRIPT would expect to have had at least 2,000 patients with syncope in our ED over the enrollment period. RAs enrolled fewer than half of the 778 patients screened, and while we did not systematically note the reasons for exclusion from enrollment, factors ranged from failure to meet enrollment criteria (specifically, a syncopal event in the preceding 24 hours)
T
and patient refusal to participate. The potential for selection bias, related both to the total
IP
number of enrolled patients and to those excluded after initial screening, may have
CR
additionally affected our outcomes.
US
In summary, in contrast to a recent study, our findings do not support a high rate of PE among ED patients presenting with syncope. While our study did not include a systematic work-
AN
up for PE in all patients with syncope, we would expect clinically relevant PE which accounted
M
for syncope to manifest itself during the ED stay, hospitalization, EDOU stay, or 30-day followup period. Patients who had PE and syncope were all PERC positive and reported shortness of
ED
breath, in addition to other signs and symptoms of PE. Our findings do not suggest a need for a
PT
systematic work-up for PE in all patients presenting with syncope, but support a providerdirected work-up based on the full clinical picture of patient symptoms, risk factors, and
AC
CE
objective findings at the time of ED or inpatient evaluation.
15
ACCEPTED MANUSCRIPT References 1. Puppala VK, Dickinson O, Benditt DG. Syncope: classification and risk stratification. J Cardiol 2014;63:171-7.
T
2. Koutkia P, Wachtel TJ. Pulmonary embolism presenting as syncope: case report and review of
IP
the literature. Heart Lung 1999;28:342-7.
CR
3. Keller K, Beule J, Balzer JO, Dippold W. Syncope and collapse in acute pulmonary embolism. Am J Emerg Med 2016;34: 1251-7.
US
4. Moya A, Sutton R, Ammirati F, Blanc JJ, Brignole M, Dahm JB, et al. Guidelines for the diagnosis and management of syncope (version 2009). Eur Heart J 2009;30: 2631–71.
AN
5. Constantino G, Sun BC, Barbic F, et al. Syncope Clinical Management in the Emergency
M
Department: A Consensus From the First International Workshop on Syncope Risk Stratification
ED
in the Emergency Department. Eur Heart J 2016 2016;37(19):1493-8. 6. Strickberger SA, Benson DW, Biaggioni I, et al. AHA/ACCF Scientific Statement on the
PT
evaluation of syncope: from the American Heart Association Councils on Clinical Cardiology,
CE
Cardiovascular Nursing, Cardiovascular Disease in the Young, and Stroke, and the Quality of Care and Outcomes Research Interdisciplinary Working Group; and the American College of
AC
Cardiology Foundation: in collaboration with the Heart Rhythm Society: endorsed by the American Autonomic Society. Circulation 2006;113:316- 27. 7. Prandoni P, Lensing AWA, Prins MH, et al. Prevalence of Pulmonary Embolism among Patients Hospitalized for Syncope. N Engl J Med 2016;375:1524-31. 8. Koutkia P, Wachtel TJ. Pulmonary embolism presenting as syncope: case report and review of the literature. Heart Lung 1999;28:342-7. 9. Sarasin FP, Louis-Simonet M, Carballo D, et al. Prospective evaluation of patients with
16
ACCEPTED MANUSCRIPT syncope: a population-based study. Am J Med 2001;111:177-84. (Duplicate) 10. Jenab Y, Lotfi-Tokaldany M, Alemza- deh-Ansari MJ, et al. Correlates of syncope in patients with acute pulmonary thromboembolism. Clin Appl Thromb Hemost 2015;21:772-6. 11. Keller K, Beule J, Balzer JO, Dippold W. Syncope and collapse in acute pulmonary
T
embolism. Am J Emerg Med 2016;34: 1251-7.
IP
12. Thames MD, Alpert JS, Dalen JE. Syncope in patients with pulmonary embolism. JAMA
CR
1977;238:2509-11.
13. Soteriades ES, Evans JC, Larson MG, et al. Incidence and prognosis of syncope. N Engl J
US
Med 2002;347:878-85. (duplicate)
AN
14. Theilade J, Winkel BG, Holst AG, Tfelt-Hansen J, Svendsen JH, Haunsø S. A nation- wide, retrospective analysis of symptoms, comorbidities, medical care and autopsy findings in cases of
M
fatal pulmonary embolism in younger patients. J Thromb Haemost 2010;8:1723-9.
ED
15. Castelli R, Tarsia P, Tantardini C, Pantaleo G, Guariglia A, Porro F. Syncope in patients with pulmonary embolism: comparison between patients with syncope as the presenting symptom of
PT
pulmonary embolism and patients with pulmonary embolism without syncope. Vasc Med
CE
2003;8:257-61.
16. Saravi M, Ahmadi Ahangar A, Hojati MM, et al. Etiology of syncope in hospitalized
AC
patients. Caspian J Intern Med 2015;6:233-7. 17. Blanc JJ, L'Her C, Touiza A, Garo B, et al. Prospective evaluation and outcome of patients admitted for syncope over a 1 year period. Eur Heart J. 2002 May;23(10):815-20. 18. Kline JA, Courtney DM, Kabrhel C, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008: 6(5): 772-80 19. Harris PH, Taylor R, Thielke R, et. al. Research electronic data capture (REDCap) - A
17
ACCEPTED MANUSCRIPT metadata-driven methodology and workflow process for providing translational research
AC
CE
PT
ED
M
AN
US
CR
IP
T
informatics support. J Biomed Inform. 2009 42(2):377-81.
18
ACCEPTED MANUSCRIPT
AC
CE
PT
ED
M
AN
US
CR
IP
T
Figure: Flow chart for the diagnosis of PE among ED patients with syncope
19