a b s t r a c t

Background: Many patients have unexplained persistent dyspnea after negative computed tomographic pulmo- nary angiography (CTPA). We hypothesized that many of these patients have isolated right ventricular (RV) dys- function from treatable causes. We previously derived a clinical decision rule (CDR) for predicting RV dysfunction consisting of persistent dyspnea and normal CTPA, finding that 53% of CDR-positive patients had isolated RV dys- function. Our goal is to validate this previously derived CDR by measuring the prevalence of RV dysfunction and outcomes in dyspneic emergency department patients.

Methods: A secondary analysis of a prospective observational multicenter study that enrolled patients presenting with suspected PE was performed. We included patients with persistent dyspnea, a nonsignificant CTPA, and for- mal echo performed. right ventricular dysfunction was defined as RV hypokinesis and/or dilation with or with- out moderate to severe tricuspid regurgitation.

Results: A total of 7940 patients were enrolled. Two thousand six hundred sixteen patients were analyzed after excluding patients without persistent dyspnea and those with a significant finding on CTPA. One hundred ninety eight patients had echocardiography performed as Standard care. Of those, 19% (95% confidence interval [CI], 14%-25%) and 33% (95% CI, 25%-42%) exhibited RV dysfunction and isolated RV dysfunction, respectively. Patients with isolated RV dysfunction or overload were more likely than those without RV dysfunction to have a return visit to the emergency department within 45 days for the same complaint (39% vs 18%; 95% CI of the difference, 4%-38%).

Conclusion: This simple clinical prediction rule predicted a 33% prevalence of isolated RV dysfunction or overload. Patients with isolated RV dysfunction had higher recidivism rates and a trend toward worse outcomes.

(C) 2015

Introduction

Dyspnea is a common complaint encountered in the emergency de- partment (ED), accounting for more than 3 million ED visits annually in the United States [1]. Management remains difficult because dyspnea

has many etiologies requiring varied treatment, and clinicians fail to identify the cause in nearly one-half of ED patients [2]. Imaging with computed tomographic pulmonary angiography (CTPA) is commonly used in the ED to evaluate for pulmonary embolism (PE) [3,4], and also provides additional alternative diagnostic information [5].

? Funding sources: National Institutes of Health Grants 5K23HL0774(01-05) (D.M. Courtney), 2R42HL074415-02A1, and 5R42HL074415-03 (J.A. Kline).

?? Prior Presentations: Presented at SAEM May 2014, Dallas, TX.

* Corresponding author. Department of Emergency Medicine, Indiana University School of Medicine, 720 Eskenazi Ave, Fifth Third Faculty Office Building, 3rd Floor Emergency Medicine Office, Indianapolis, IN 46202. Tel.: +1 860 428 0555 (Mobile).

E-mail addresses: [email protected] (F.M. Russell), [email protected] (C.L. Moore), [email protected] (D.M. Courtney), [email protected] (C. Kabrhel), [email protected] (H.A. Smithline), [email protected] (K.E. Nordenholz), [email protected] (P.B. Richman), [email protected] (B.J. O’Neil), [email protected] (M.C. Plewa), [email protected] (D.M. Beam), [email protected] (R. Mastouri), [email protected] (J.A. Kline).

http://dx.doi.org/10.1016/j.ajem.2015.01.026

0735-6757/(C) 2015

Assessment of persistently symptomatic patients after a negative CTPA continues to pose a challenge for clinicians. Kline et al [6] found that 40% of patients who receive one CTPA to evaluate for possible PE underwent a second CTPA, which seldom demonstrated an actionable diagnosis. Because patients who have persistent dyspnea are more like- ly to return for repeat evaluation and undergo repeated nondiagnostic CTPA in the near term, determining the cause of dyspnea may help guide the disposition of ED patients after a negative CTPA.

Many patients with dyspnea have evidence of coincident right ven- tricular (RV) dysfunction or overload. Right ventricular dysfunction may originate from intrinsic muscle damage caused by ischemia or cardio- myopathies. More commonly, however, RV dysfunction occurs from in- creases in RV afterload, caused by treatable etiologies such as chronic obstructive pulmonary disease, obstructive Sleep apnea, pulmonary ar- terial hypertension, or pulmonary hypertension secondary to left heart disease or chronic thromboemboli [7-13]. Detection of RV dysfunction from pulmonary hypertension is frequently underdiagnosed. Once rec- ognized, it is usually associated with a delay in evaluation and treatment [7,8,14,15]. Patients with underlying RV dysfunction, including RV dila- tion with or without RV hypokinesis, have increased morbidity and mortality [7,12]. In ED patients with dyspnea, RV dilation is an indepen- dent predictor of 1-year mortality [16].

In our derivation article [17], we derived a simple clinical decision rule (CDR) to identify patients with a high probability for isolated RV dysfunction or overload. This rule was derived from both inpatients and ED patients who were all adults older than 17 years and who underwent CTPA for suspected PE. The final rule consisted of persistent dyspnea plus a normal CTPA scan. With a positive rule, 22 (53%) of 41 (95% confidence interval [CI], 37%-69%) patients with transthoracic echocardiography performed had isolated RV dysfunction or overload. In the derivation cohort, when broadening normal CTPA, which exclud- ed patients with underlying chronic diseases such as emphysema or cardiomegaly, to “CTPA without acute process”, the CDR predicted 64 (36%) of 179 (95% CI, 29%-43%) patients to have isolated RV dysfunction or overload. These results were similar to the 30% prevalence of RV dys- function in dyspneic patients reported by Chen et al [16] in the PRIDE study, which examined ED patients with shortness of breath, but was not restricted to those who had CTPA.

Given the importance of RV function on outcomes in a variety of dis- ease states, we believe that early identification and initiation of therapy targeted at the cause of isolated RV dysfunction or overload, if possible, may improve outcome.

The aim of this study was to test the external validity of the previ- ously derived CDR by measuring the prevalence of isolated RV dysfunc- tion or overload and outcomes in ED patients with persistent shortness of breath and CTPA without acute process.

Methods

Study design and setting

This study was a secondary analysis of the Pulmonary Embolism Rule Out Criteria (PERC) database, a prospective observational study in- volving 12 EDs, that collected data on patients presenting to the ED with suspected PE. Institutional review boards approval was obtained at all participating hospitals. The methods from this study have been pub- lished previously [18].

Selection of participants

Patients were enrolled from July 1, 2003, until November 30, 2006. For our main analysis, we included ED patients with persistent short- ness of breath, CTPA without acute process, and formal echocardiogra- phy performed. Persistent dyspnea was defined as patient’s ongoing subjective feeling of shortness of breath at rest while breathing room air, during the index ED visit or hospital stay. “CTPA without acute

process” was defined as a CTPA without evidence of PE, pulmonary infil- trate, thoracic malignancy, pneumothorax, aortic dissection, aortic an- eurysm, or congestive heart failure. Exclusion criteria were patients without a history or complaint of shortness of breath or persistent shortness of breath, patients with a CTPA positive for PE or other acutely significant CT findings, patients with an admission and/or discharge di- agnosis of PE, and patients without an echocardiography performed.

Study protocol

Patients were deemed eligible for enrollment if a diagnostic test for PE was ordered by or under the supervision of a board-certified emer- gency physician. The decision to order this test followed standard care and was based on history and physical examination findings. Diagnostic tests for PE included CTPA scan, pulmonary angiography, and ventilation-perfusion lung scanning. Patients were excluded if they had a known diagnosis of PE found on pulmonary vascular imaging per- formed in the previous 7 days or if they were at risk for loss to follow-up (eg, homeless, international travelers, and prisoners). Study enrollment varied by site and consisted of consecutive enrollment or random selec- tion in 8-hour blocks of time with subsequent medical record review of eligible but unenrolled patients.

Clinical data

Treating clinicians collected and recorded data concurrently with routine patient care and prior to the clinician knowing the results of di- agnostic tests. Data were collected on a standardized, Web-based form, with more than 65 data points.

Computed tomographic pulmonary angiography interpretation

The radiology department at each individual institution regulated scanning protocols. All data collected were from final interpretations by a board-certified radiologist. A diagnosis of PE was made if a pulmo- nary arterial filling defect was detected by the interpreting radiologist and identified as an acute PE.

Transthoracic echocardiography

All echocardiograms were performed at the discretion of the clinical care team. All institutions had facilities with Intersocietal Commission for the Accreditation of Echocardiography Laboratories accreditation, and board-certified cardiologists with echocardiography Fellowship training provided final written interpretations of echocardiograms. Ac- cordingly, all reports included estimate of left ventricular (LV) ejection fraction; an estimate of tricuspid regurgitation (TR) graded as none, mild, moderate, or severe; a Doppler measurement of tricuspid jet ve- locity (if obtainable); and a qualitative assessment of RV size (normal or enlarged) and function (normal or hypokinetic). Transthoracic echo- cardiography was categorized into 5 categories including the following:

(1) normal LV and RV size and function, (2) LV ejection fraction (LVEF) less than 45%, (3) isolated moderate to severe TR, (4) isolated RV hypokinesis and dilation, or (5) RV hypokinesis and dilation with mod- erate to severe TR.

Follow-up

Patients were followed up at 45 days postenrollment for outcome. All data, including explicit terms of LV and RV function, were abstracted by trained coordinators and transferred into a Web-based data collec- tion instrument [19]. Direct patient contact was via telephone or mail, supplemented as needed by the medical record and/or a death registry review [18].

Data analysis

We compared means and bivariate frequencies measured in patients with and without echocardiography and in patients with and without isolated RV dysfunction or overload. Means were compared with an

unpaired t test, and bivariate frequencies were compared with a ?² test. Our main analysis was the point estimate of the prevalence of isolated RV dysfunction or overload observed in persistently short of breath pa- tients with a CTPA negative for PE or other significant finding and formal transthoracic echocardiography performed. Right ventricular dysfunc- tion or overload was defined as RV hypokinesis and/or dilation with or without moderate to severe TR. Isolated RV dysfunction or overload was defined as RV dysfunction in the absence of LV systolic dysfunction. We calculated percentages of RV dysfunction and isolated RV dys- function or overload in those patients with a positive decision rule and echocardiography, with 95% CIs from the exact binomial formula (Stats Direct, v 2.7.9, Cheshire, England, UK). Statistical significance

was accepted for P b .05.

Results

A total of 7940 patients were enrolled from the PERC database. Of these, 4784 patients were excluded because they did not meet the criteria for persistent dyspnea or did not have a CTPA scan performed. Five hundred forty patients were then excluded because their CTPA was positive for a PE or other acute process. Of the remaining 2616 pa- tients included for analysis, 198 had formal transthoracic echocardiog- raphy performed. Figure shows a flow diagram of patients meeting the inclusion criteria. Table 1 shows patient characteristics comparing all enrolled patients with and without echocardiography. Overall, pa- tients with echocardiography were older, were admitted to a higher level of care, and had more cardiovascular disease.

Analysis of the 198 patients with echocardiography (Table 2) yielded 38 (19%; 95% CI, 14%-25%) patients with evidence of RV dys- function or overload, including RV dilation and/or hypokinesis with or without moderate to severe TR. One hundred sixty (81%; 95% CI, 74%-

85%) patients had no evidence of RV dysfunction or overload; 53% were female. Compared with patients without RV dysfunction, patients with evidence of RV dysfunction or overload on echocardiography were more likely to be sicker, with 18% vs 7% (95% CI, of the difference, 0.4%- 26%) being admitted to the intensive care unit (ICU), and tended to have a greater history of wheezing and Tobacco use.

Of the 198 patients with an echocardiogram, 115 had normal LVEF, defined as ejection fraction greater than 45%. Thirty-five (30%; 95% CI, 23%-39%) of the patients with normal LVEF had moderate to severe TR without RV dilation. Thirty-eight (33%; 95% CI, 25%-42%) of the 115 patients with normal LVEF had isolated RV dysfunction or overload, 14 with RV dilation and moderate to severe TR and 24 with RV dilation without significant TR.

Table 3 compares outcomes of patients with and without isolated RV dysfunction or overload on echocardiography. Patients with RV dys- function or overload were more likely to have a return visit to the ED within 45 days for the same complaint (39% vs 18%; 95% CI of the differ- ence, 4%-38%) and tended to have more repeat CTPA scans within 45 days (10% vs 5%; 95% CI of the difference, – 4% to 19%) and a higher

mortality rate (16% vs 5%; 95% CI of the difference, – 0.5% to 25%). Also, patients with isolated RV dysfunction or overload were more often discharged with a primary diagnosis of pulmonary hypertension (P b .05) than those patients without evidence of RV dysfunction or overload (Table 3).

Discussion

Unexplained RV dysfunction or overload was found to have a preva- lence of 19% (95% CI, 14%-25%) in persistently dyspneic ED patients with a nonsignificant CTPA. The prevalence of RV dysfunction or overload

Figure. Flow diagram of patients meeting inclusion criteria.

Table 1

Patient characteristics

Demographics

Ethnicity
White	117 (59)	1362 (57)	.45
Black	68 (34)	853 (35)	.791
Other	13 (7)	203 (8)	.368
Vital signs

Echo (n = 198) No echo (n = 2418) P^a

Table 2

Comparison of patients with and without RV dysfunction on echocardiography

RVD (n = 38) No RVD (n = 160) P^a

Demographics

Age (y)	60 +- 17	51 +- 18	b.05
Male	83 (42)	716 (30)	b.05
BMI (kg/m2)	30 +- 9	30 +- 9	.439

Age (y)	63 +- 18	59 +- 17	.193
Male	18 (47)	65 (41)	.449
BMI (kg/m2)	29 +- 7	30 +- 9	.286
Ethnicity
White	25 (66)	92 (58)	.35
Black	10 (26)	58 (36)	.246
Other	3 (8)	10 (6)	.713

A

Highest heart rate (beats/min)	103 +- 26	94 +- 22	b.05
Systolic blood pressure (mm Hg)	128 +- 28	132 +- 25	.058
Respiratory rate (breaths/min)	23 +- 6	22 +- 5	b.05
Oxygen saturation (%)	94 +- 6	96 +- 4	b.05
Temperature (?F)	98 +- 1	98 +- 1	.379
ssociated symptoms Wheezing 36 (18) 345 (14) .795
Chest pain at rest	49 (25)	1134 (47)	b.05
Prior visit for same symptoms	64 (32)	563 (23)	.268

Comorbidities

Chronic lung disease	60 (30)	637 (26)	.867
Tobacco use	84 (42)	899 (37)	.142
Hematologic disease	2 (1)	54 (2)
CHF	16 (8)	53 (2)	b.05
Hypertension	117 (59)	940 (39)	b.05
CAD	46 (23)	267 (11)	b.05
Connective tissue disease	7 (4)	107 (4)	.555
PE or DVT	27 (14)	331 (14)	.285
CKD on HD	7 (4)	28 (1)	b.05
Diabetes	55 (28)	330 (14)	b.05
CVA	19 (10)	96 (4)	b.05
Anxiety	17 (9)	274 (11)	.237
Malignancy	35 (18)	378 (16)	.448
Disposition ICU	19 (10)	48 (2)	b.05
Telemetry bed	121 (62)	633 (26)	b.05
Unmonitored bed	35 (18)	325 (13)	.096
Observation	8 (4)	270 (11)	.002
Discharge	12 (6)	1142 (48)	b.05

Data presented as number (percent) and mean +- SD. Bold numbers are values reaching statistical significance (P b .05).

Abbreviations: BMI, body mass index; CAD, coronary artery disease; CHF, congestive heart failure; Chronic lung disease, chronic obstructive pulmonary disease, asthma, and intersti- tial lung disease; CKD, chronic kidney disease; CVA, cerebrovascular disease; DVT, deep venous thrombosis; HD, hemodialysis.

^a P values from unpaired t test or ?² test.

increased to 33% (95% CI, 25%-42%) when excluding patients with abnormal LVEF.

Prior research has looked to define the prevalence of RV dysfunction in ED patients presenting with dyspnea. The PRIDE study looked to de- termine the use of N-terminal pro-brain natriuretic peptide (NT- proBNP) compared with echocardiographic parameters in ED patients with shortness of breath. In the PRIDE study, 134 (23%) of 599 dyspneic patients had an echocardiogram. Similar to our study, patients who re- ceived an echocardiography in the PRIDE study had more severe symp- toms and were more likely to have a history of heart disease. Interestingly, 30% of these patients had changes in RV fractional area or RV dilation, 20% had RV hypokinesis, and 30% had moderate or severe TR [16].

Our derivation article found that patients with a positive CDR (consisting of persistent dyspnea and normal CTPA) and echocardiogra- phy performed as routine management exhibited a prevalence of isolat- ed RV dysfunction or overload of 53% (95% CI, 37%-69%). This number decreased to a prevalence of 36% (95% CI, 29%-43%) when broadening the definition of normal CTPA (which excluded patients with underly- ing chronic diseases such as emphysema) to CTPA without acute pro- cess. This latter definition more accurately reflects the definition of negative CTPA used in this article.

Both the PRIDE and derivation studies show a similar prevalence of isolated RV dysfunction or overload, comparable to the current work,

Associated symptoms

Wheezing	10 (26)	26 (16)	.148
Chest pain at rest	8 (21)	41 (26)	.557
Prior visit for same symptoms	12 (32)	52 (33)	.913

Comorbidities

Chronic lung disease	13 (34)	47 (29)	.809
Tobacco use	20 (53)	64 (40)	.156
CHF	12 (32)	37 (23)	.165
Hypertension	23 (61)	94 (59)	.841
CAD	13 (34)	33 (21)	.074
PE or DVT	7 (18)	20 (13)	.339
CKD on HD	2 (5)	5 (3)	.521
Diabetes	10 (26)	45 (28)	.822
Anxiety	3 (8)	14 (9)	.865
Malignancy	7 (18)	28 (18)	.893
Disposition ICU	7 (18)	12 (8)	.039
Telemetry bed	23 (60)	101 (63)	.765
Unmonitored bed	4 (11)	31 (19)	.198
Observation	1 (3)	7 (4)	.623
Discharge	3 (8)	9 (6)	.598

Abbreviations: BMI, body mass index; CAD, coronary artery disease; CHF, congestive heart failure; Chronic lung disease, chronic obstructive pulmonary disease, asthma, and intersti- tial lung disease; CKD, chronic kidney disease; DVT, deep venous thrombosis; HD, hemo- dialysis; RVD, RV dysfunction.

^a P values from unpaired t test or ?² test. Data presented as number (percent) and mean+- SD.

but comprise a different heterogeneous population and subset of data. Our study differs from the PRIDE study, however, in that we further risk stratified patients using CTPA to rule out PE or other significant find- ings. Furthermore, because LV failure is a leading cause of RV dysfunc- tion [20], we examined the prevalence of isolated RV dysfunction or overload by excluding patients with an abnormal LVEF. In the derivation study, we defined isolated RV dysfunction or overload by excluding pa- tients with both systolic and/or diastolic LV dysfunction. However, in the present work, we were unable to exclude patients with LV diastolic

Table 3

Patient outcomes with normal LVEF

RVD (n = 38)		No RVD (n = 77)	Pa
Mortality	6 (16)	4 (5)	.117
Return visits	15 (39)	14 (18)	.013
Repeat CTPA Final diagnosis	4 (10)	4 (5)	.290
COPD	8 (21)	8 (10)	.120
Asthma	1 (3)	3 (4)	.727
ILD	2 (5)	2 (3)	.463
Pneumonia	5 (13)	8 (10)	.659
CHF	7 (18)	12 (16)	.730
MI	1 (3)	1 (1)	.607
Sepsis	1 (3)	0 (0)	.152
Chest pain	3 (8)	14 (18)	.143
Pulmonary HTN	9 (24)	0 (0)	b.05
Other	10 (26)	37 (48)	.025

Data presented as number (percent). Other included dyspnea, valvular heart disease, syn- cope, Acute bronchitis, pleural effusion, cancer, atrial dysrhythmias, anemia, Hypertensive emergency, scleroderma, sleep apnea, renal failure, and pericarditis. Abbreviations: CHF = congestive heart failure; COPD, chronic obstructive pulmonary disease; CTPA, Computed tomography pulmonary angiography; HTN, hypertension; ILD, interstitial lung disease; MI, myocardial infarction; RVD, RV dysfunction.

^a P values from unpaired t test or ?² test.

dysfunction because this secondary analysis lacked echocardiographic information on LV Diastolic function. Nonetheless, the prevalence of iso- lated RV dysfunction or overload is similar in the derivation and valida- tion samples (36% vs 33%, respectively).

Prior literature has shown that patients with RV dysfunction or over- load have worse outcomes. In patients with left heart failure or COPD, RV dysfunction or overload is associated with increased acute exacerba- tions and increased Hospital readmissions compared with patients with normal RV function [7,21,22]. Patients with pulmonary hypertension complicated by RV dysfunction generally have a significantly poorer prognosis and decreased longevity [12,13,15].

This study corroborates findings in the derivation study showing a trend toward higher rates of return visits with repeat CTPA scanning among patients with isolated RV dysfunction or overload. In the present study, patients with isolated RV dysfunction or overload had statistically significant higher rates of ICU admissions and return visits within 45 days for the same complaint. They also exhibited higher mortality rates and higher rates of repeat Chest imaging within 45 days, but these outcomes did not reach statistical significance. Patients with RV dysfunction or overload were more likely to be diagnosed at hospital discharge with pulmonary hypertension than those patients without evidence of RV dysfunction on echocardiography. These results support the need for a clinical pathway to refer rule-positive patients for screen- ing with echocardiography followed by evaluation in a multidisciplinary dyspnea clinic.

Limitations

This study analyzed a subset of patients from a previously collected sample where ordering an echocardiogram was not standardized but was ordered at the discretion of the treating physician. The patients may have had characteristics or Severity of symptoms not easily identi- fied by a secondary analysis, which persuaded clinicians to order the echocardiogram. Also, this was a multi-institution study with a lot of in- trinsic variation, especially with availability of echocardiography among these institutions. Therefore, there was some selection bias in the pa- tients who had an echocardiogram performed, as these patients were older and sicker.

Findings on echocardiography were divided into 5 categories as listed in the Methods section. It is possible that some patients included in the moderate to severe TR subgroup may have also had evidence of RV dysfunction and not recorded as such. As previous literature has shown, moderate to severe TR has a high correlation with RV dysfunc- tion and/or dilation [23]. Thus, the prevalence of RV dysfunction may be underestimated and may actually be larger than 19%. We also de- fined isolated RV dysfunction or overload on echocardiography as evi- dence of RV dysfunction without concurrent LV systolic dysfunction. Because this was a secondary analysis of previously collected data, we lacked echocardiographic information on LV diastolic function, as this was not recorded in the original data set. In light of this, the prevalence of isolated RV dysfunction may actually be less than 33%. Lastly, there were no sample size calculations because this was a secondary analysis. We had a small sample size of 198 patients, which is the result of a sub- set analyses from a much larger data set and therefore have larger CIs. A prospective study of echocardiogram findings in all cases of persistent dyspnea and negative CTPA might find a different prevalence of isolated RV dysfunction or overload with smaller CIs.

Conclusions

In an independent sample, RV dysfunction or overload has a preva- lence of 19% in patients with persistent shortness of breath with a non- significant CTPA. In the subset of patients with preserved LVEF, the prevalence of isolated RV dysfunction or overload increases to 33%. Among patients selected for echocardiography based on standard clini- cal judgment, this CDR predicts significant risk for isolated RV

dysfunction or overload, higher recidivism, and worsened outcomes. These data suggest a need for wider use of echocardiography to screen this subset of patients for RV dysfunction or overload and a need for spe- cialist referral to identify treatable causes of RV dysfunction or overload. Future goals will be to prospectively validate this rule in the ED setting, with a next step of using bedside ultrasound to further risk stratify these patients.

Conflicts of Interest

Christopher L. Moore has a $4000 honorarium/consulting fee from Philips to work on developing an online tutorial for bedside echo in PE.

D. Mark Courtney is on the Janssen pharmaceuticals advisory board and has had previous grant support from the National Institutes of Health for PE research. Christopher Kabhrel has grant funding from the NIH, Stago Diagnostics and Siemen’s Healthcare; is a consultant for Genentech and Janssen pharmaceuticals. Jeffrey A. Kline has grant funding from the National Institutes of Health and Ikaria; is a consultant for Genentech, Stago Diagnostics, and Janssen pharmaceuticals; and owns in CP Diagnos- tics LLC. Kristen E. Nordenholz, MD, has performed unrestricted research with Alere, Boehringer Ingelheim, and Genentech. The following authors declare no financial disclosures or relationships: Frances M. Russell, MD; Michael C. Plewa, MD; Peter B. Richman, MD; Howard A. Smithline, MD; Brian J O’Neil, MD; Daren M. Beam, MD; and Ronald Mastouri, MD.

Author Contributions

J.K. conceived the study. C.L.M., D.M.C., K.E.N., M.C.P., P.B.R., H.A.S.,

B.J.O., C.K., and D.M.B. were involved in recruitment of participants in the study, gathering data, and quality control. J.K. and F.M.R. were re- sponsible for the literature search, organizing the data prior to analysis, and data analysis. F.M.R. drafted the manuscript, and all authors were significantly involved in subsequent revisions. F.M.R. and J.K. take re- sponsibility for the manuscript as a whole.

References

1. Pitts SR, Niska RW, Xu J, Burt CW. National Hospital Ambulatory Medical Care Sur- vey: 2006 emergency department summary. National health statistics reports; 2008 1-38.
2. Kline JA, Shapiro NI, Jones AE, Hernandez J, Hogg MM, Troyer J, et al. Outcomes and radiation exposure of emergency department patients with chest pain and shortness of breath and ultraLow pretest probability: a multicenter study. Ann Emerg Med 2013;63:281-8.
3. Samad Z, Hakeem A, Mahmood SS, Pieper K, Patel MR, Simel DL, et al. A meta- analysis and systematic review of computed tomography angiography as a diagnos- tic triage tool for patients with chest pain presenting to the emergency department. J Nucl Cardiol 2012;19:364-76.
4. Gallagher MJ, Raff GL. Use of multislice CT for the evaluation of emergency room pa- tients with chest pain: the so-called “triple rule-out”. Catheter Cardiovasc Interv 2008;71:92-9.
5. Richman PB, Courtney DM, Friese J, Matthews J, Field A, Petri R, et al. Prevalence and significance of nonthromboembolic findings on chest computed tomography angi- ography performed to rule out pulmonary embolism: a multicenter study of 1,025 emergency department patients. Acad Emerg Med 2004;11:642-7.
6. Kline JA, Courtney DM, Beam DM, King MC, Steuerwald M. Incidence and predictors of repeated computed tomographic pulmonary angiography in emergency depart- ment patients. Ann Emerg Med 2009;54:41-8.
7. Zakir RM, Al-Dehneh A, Maher J, Saric M, Berkowitz RL. Right ventricular failure in patients with preserved ejection fraction and diastolic dysfunction: an underrecognized clinical entity. Congest Heart Fail 2007;13:164-9.
8. Lang IM, Plank C, Sadushi-Kolici R, Jakowitsch J, Klepetko W, Maurer G. Imaging in pulmonary hypertension. J Am Coll Cardiol Img 2010;3:1287-95.
9. Rao S, Tate DA, Stouffer GA. Hemodynamic findings in severe tricuspid regurgitation. Catheter Cardiovasc Interv 2013;81:162-9.
10. Chaouat A, Weitzenblum E, Krieger J, Oswald M, Kessler R. Pulmonary hemodynam- ics in the obstructive sleep apnea syndrome. Results in 220 consecutive patients. Chest 1996;109:380-6.
11. Dumitrascu R, Tiede H, Eckermann J, Mayer K, Reichenberger F, Ghofrani HA, et al. Sleep apnea in precapillary pulmonary hypertension. Sleep Med 2013;14:247-51.
12. van de Veerdonk MC, Kind T, Marcus JT, Mauritz GJ, Heymans MW, Bogaard HJ, et al. Progressive right ventricular dysfunction in patients with pulmonary arterial hyper- tension responding to therapy. J Am Coll Cardiol 2011;58:2511-9.
13. Vonk-Noordegraaf A, Haddad F, Chin KM, Forfia PR, Kawut SM, Lumens J, et al. Right heart adaptation to pulmonary Arterial hypertension: physiology and pathobiology. J Am Coll Cardiol 2013;62:D22-33.
14. Piazza G, Goldhaber SZ. The acutely decompensated right ventricle: pathways for di- agnosis and management. Chest 2005;128:1836-52.
15. Maron BA, Choudhary G, Khan UA, Jankowich MD, McChesney H, Ferrazzani SJ, et al. Clinical profile and underdiagnosis of pulmonary hypertension in US veteran pa- tients. Circ Heart Fail 2013;6:906-12.
16. Chen AA, Wood MJ, Krauser DG, Baggish AL, Tung R, Anwaruddin S, et al. NT- proBNP levels, echocardiographic findings, and outcomes in breathless patients: results from the ProBNP Investigation of Dyspnoea in the Emergency Depart- ment (PRIDE) echocardiographic substudy. Eur Heart J 2006;27:839-45.
17. Kline JA, Russell FM, Lahm T, Mastouri R. Derivation of a screening tool to iden- tify patients with right ventricular dysfunction or tricuspid regurgitation after negative Computerized tomographic pulmonary angiography. Pulm Circ 2015; 27 [in press].
18. Kline JA, Courtney DM, Kabrhel C, Moore CL, Smithline HA, Plewa MC, et al. Pro- spective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008;6:772-80.
19. Kline JA, Johnson CL, Webb WB, Runyon MS. Prospective study of clinician- entered research data in the emergency department using an Internet-based system after the HIPAA Privacy Rule. BMC Med Inform Decis Mak 2004;4:17.
20. Haddad F, Doyle R, Murphy DJ, Hunt SA. right ventricular function in cardiovascular disease, part II: pathophysiology, clinical importance, and management of right ven- tricular failure. Circulation 2008;117:1717-31.
21. Chaouat A, Naeije R, Weitzenblum E. Pulmonary hypertension in COPD. Eur Respir J 2008;32:1371-85.
22. Wells JM, Washko GR, Han MK, Abbas N, Nath H, Mamary AJ, et al. Pulmonary arte- rial enlargement and acute Exacerbations of COPD. N Engl J Med 2012;367:913-21.
23. Mutlak D, Aronson D, Lessick J, Reisner SA, Dabbah S, Agmon Y. Functional tricuspid regurgitation in patients with pulmonary hypertension: is pulmonary artery pres- sure the only determinant of regurgitation severity? Chest 2009;135:115-21.