Treatment of pSVT: A Case for Calcium Channel Blockers

View fullsize

Written by:  Amanda Randolph, MD (NUEM PGY-1) Edited by: Jim Kenny, MD, (NUEM PGY-4) Expert commentary by: Meghan Groth, PharmD - Emergency Medicine Clinical Pharmacist, UMass Memorial Medical Center 

---

The Case

A 37 year-old woman presents to the ED for palpitations. On the monitor, you see her heart rate is 190, but all other vitals are within normal limits. She feels anxious but is otherwise asymptomatic, breathing comfortably on room air. The rest of the physical exam is unremarkable. The patient tells you, “I think it’s my SVT again - I was just here for this last month!”

Her rhythm strip looks something like this:

View fullsize

https://lifeinthefastlane.com/ecg-library/svt/

SVT generally refers to any tachyarrhythmia generated above the His/Purkinje system. For simplicity, the term pSVT in this post will refer to only Atrioventricular Nodal Tachycardia (AVNRT), as it is the most common tachyarrhythmia in patients with normal cardiac structure [1].

---

SVT: Treatment Guidelines

You double-check the current ACLS protocol (2015) for the treatment of pSVT [2]: 

View fullsize

Figure 1. ACLS 2015 guidelines for treatment of AVNRT

This patient is stable, so you try some vagal maneuvers, including carotid massage and Valsalva. You even try the modified Valsalva maneuver you read about in the REVERT trial (straining followed by leg elevation and supine positioning), which is described to have a 43% success rate.

Despite your best efforts, the vagal maneuvers fail, so you ask the nurse to draw up some adenosine. 

At this point, the patient yells, “Absolutely no way! I’m not trying Adenosine - it makes me feel like I’m going to die! There has to be something else.” 

You know Calcium Channel Blockers (CCBs) are recommended as a second line drug if adenosine does not terminate the SVT, or if adenosine is contraindicated. But what does the data say? Is it ever reasonable to jump straight to CCBs?

---

The Problem with Adenosine

Adenosine administration is widely recognized to produce a variety of minor side effects, as listed below4. While not quantified in any studies to date, these “minor” side effects can be extremely traumatic for patients. This distress can have lasting psychological effects that may delay or even prevent patients from seeking care [5].

•  Chest pain (7-40%)
•  Facial flushing (18-44%)
•  Nausea (13%)
• Headache (2-18%)
• Lightheadedness/Dizziness (12%)

The Problem with Calcium Channel Blockers

Current ACC/AHA guidelines give CCBs a class IIa recommendation for use in pSVT [2]. However, most EM practitioners continue to favor Adenosine, in part because of cultural dogma, but also due to concern about inadequate data to regarding the efficacy and safety for calcium channel blocker use. 

One pharmacologic difference between CCBs and Adenosine is the onset of action (100-400 seconds for CCBs compared with 21-34 seconds for Adenosine), which can create a delay to conversion [5]. However, because CCBs are only used in stable patients, this slightly longer onset is unlikely to be clinically significant [6].

More importantly, one of the most feared side effects of Calcium channel blockers is hypotension, as CCBs work by creating negative inotropy and peripheral vasodilation. In one study by Lim et al., the change in blood pressure after administration of adenosine was -2.6/-1.7, compared to -13.0/-8.1 with verapamil [9]. 

Of note, the duration of action is quite long for CCBs (2-5 hours), compared with adenosine (<10 seconds).7 This raises a concern that hypotension and other adverse effects of CCBs may be prolonged. For this reason, CCBs are contraindicated in patients with severe HFrEF.6,7 Additionally, CCBs are relatively contraindicated in patients taking beta blockers, as the combined effect can cause significant bradycardia and even heart block [6].

Theoretically, the use of CCBs via slow infusion instead of IV bolus may reduce the risk of hypotension,8 though there is limited data to support this. One randomized trial by Lim et al. compared the use of adenosine (n = 104) vs slow infusion of verapamil (n = 48) or diltiazem (n = 54), and reported no difference in outcomes between adenosine bolus and slow infusion of verapamil or diltiazem [9].

---

Calcium Channel Blockers vs. Adenosine - The Data

To date, there have been three meta-analyses comparing the efficacy and safety of CCBs to adenosine in patients with pSVT, including a recently published Cochrane review in October 2017.5,6,10 Note that the data described in these studies only refer to the use of Verapamil. Their findings are depicted below (table design inspired by a phenomenal ALiEM post) [8].

View fullsize

A Few Notes on Hypotension after Verapamil:

• None of these metaanalyses specifically reported their definition of hypotension, nor did they clarify whether any of these patients had clinical signs of shock.
• Holdgate and Foo reported two of three hypotensive patients subsequently reverted with adenosine and did not require any other specific treatment for their hypotension (the outcome and interventions for the third case were not reported). 
• The study by Lim et al. using slow infusion of verapamil reported only one patient with clinically significant hypotension, with a drop in blood pressure from 122/81 mmHg to 74/61 mmHg after 7.5 mg of verapamil infusion. This patient’s SVT was terminated by synchronized cardioversion, after which his blood pressure improved to 103/69 mmHg.

---

Case Resolution

After the vagal maneuvers, you give 5mg IV Verapamil. The patient remains stable and converts to sinus tachycardia. She tells you she prefers Verapamil to Adenosine and will be “much less afraid” to come in next time. 

---

Conclusion

Overall, both Adenosine and Verapamil are reasonable choices for termination of SVT. Anecdotally, some patients prefer Verapamil; however, there is limited evidence to support this [6]. Given the current data, physicians should discuss the pros/cons of each drug with the patient and employ shared decision-making when possible. 

---

Take Home Points

•  Start with vagal maneuvers, especially the modified Valsalva
• Adenosine and Verapamil are equally effective for SVT 
  •  Moderate evidence by recent Cochrane review
  •  Class IIa by ACC/AHA
•  Adenosine has a much higher incidence of minor side effects
  • chest pain, facial flushing, nausea, headache, and lightheadedness/dizziness
• Verapamil has a slightly higher risk of hypotension
  • Verapamil: -13/-8 mmHg; Adenosine -2.6/-1.7 mmH
  • Rarely clinically significant - cases reportedly resolved with adenosine or synchronized cardioversion
• Always employ shared decision-making when possible 

---

Expert Commentary

Thank you for your insightful post on this all-too-common conundrum we face in the ED. I think it’s incredibly important to remember, as you point out, that treatment of pSVT in the ED doesn’t have to be a “one size fits all” approach, and that we have more than just adenosine available as a treatment agent.

Most of the data for CCBs in this indication is with verapamil, but I’ve become comfortable recommending diltiazem in its place due to a lower risk of hypotension (see post for reference).

When attempting to mitigate the potential hypotension associated with calcium channel blockers, the study by Lim and colleagues that you mentioned is worth noting in more detail. Rather than the traditional 0.25 mg/kg diltiazem bolus (with 0.35 mg/kg repeat dose), subjects instead received diltiazem at a rate of 2.5 mg/min for up to 20 minutes (max dose 50 mg). This approach can optimize dose, reduce potential for hypotension, and spare the patient that “impending doom” feeling often experienced with adenosine (see further discussion on this here).

There are also some cases when adenosine should not be routinely administered, such as patients with reactive airway disease at risk of bronchospasm. A more thorough review of this topic is presented here but in such cases calcium channel blockers represent a reasonable alternative.

The strategy of using calcium channel blockers for pSVT can perhaps leave providers wanting in terms of the instant gratification that comes with adenosine administration, but agents like diltiazem or verapamil have demonstrated efficacy while avoiding some of the unpleasantries of adenosine.

For me, it comes down to recognizing that adenosine isn’t the only drug we have available for the treatment of pSVT. Calcium channel blockers like diltiazem may be used, and if we decide to try them, we can use different dosing approaches such as the slow bolus method outlined above to reduce some of the potential side effects.

Meghan Groth, PharmD

Emergency Medicine Clinical Pharmacist 

UMass Memorial Medical Center

---

[Peer-Reviewed, Web Publication]   Randolph A,  Kenny J (2018, May 28 ). Treatment of pSVT: A case for calcium channel blockers.  [NUEM Blog. Expert Commentary by Groth, M]. Retrieved from http://www.nuemblog.com/blog/PSVT

---

Posts you may also enjoy

---

References

1. Burns, E., Supraventricular Tachycardia (SVT), in Life in the Fast Lane, M. Cadogan and C. Nickson, Editors. 2012.
2. Page R, Joglar J, Caldwell M, et al. 2015 ACC/AHA/HRS Guideline for the Management of Adult Patients With Supraventricular Tachycardia: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol. 2016;67(13):e27-e115.
3. Appelboam, A., et al., Postural modification to the standard Valsalva manoeuvre for emergency treatment of supraventricular tachycardias (REVERT): a randomised controlled trial. The Lancet, 2015. 386(10005): p. 1747-1753.
4. Adenosine. Lexi-Drugs. Lexicomp. Wolters Kluwer Health, Inc. Riverwoods, IL.  Available at:  http://online.lexi.com.  Accessed November 12, 2017.
5. Delaney, B., J. Loy, and A.-M. Kelly, The relative efficacy of adenosine versus verapamil for the treatment of stable paroxysmal supraventricular tachycardia in adults: a meta-analysis. European Journal of Emergency Medicine, 2011. 18(3): p. 148-152.
6. Alabed S, Sabouni A, Providencia R, Atallah E, Qintar M, Chico TJA. Adenosine versus intravenous calcium channel antagonists for supraventricular tachycardia. Cochrane Database of Systematic Reviews 2017, Issue 10. Art. No.: CD005154. DOI: 10.1002/14651858.CD005154.pub4.
7. [Peer Reviewed, Web Publication] S. Brubaker and B. Long (2017 Feb 1). Treatment of Refractory SVT: Pearls and Pitfalls. [EmDocs.net, Expert Commentary by A. Koyfman]. Retrieved from http://www.emdocs.net/treatment-refractory-svt-pearls-pitfalls/
8. [Web Publication] S. Rappaport and M. Groth (2016 Mar 3).  Calcium channel blockers for stable SVT: A first line agent over adenosine? [AliEm Blog]. Retrieved from https://www.aliem.com/2016/03/calcium-channel-blockers-stable-svt-alternative-to-adenosine/ 
9. Lim S, Anantharaman V, Teo W, Chan Y. Slow infusion of calcium channel blockers compared with intravenous adenosine in the emergency treatment of supraventricular tachycardia. Resuscitation. 2009;80(5):523-528.
10. Holdgate, A. and A. Foo, Adenosine versus intravenous calcium channel antagonists for the treatment of supraventricular tachycardia in adults. The Cochrane Library, 2006. 

The PESIT Trial: Do All Syncope Patients Need a PE Workup?

Written by: Alex Ireland, MD (NUEM PGY-2) Edited by: Josh Zimmerman, MD,  (NUEM class of 2017) Expert Commentary by: D. Mark Courtney, MD

---

Syncope is defined as a transient loss of consciousness and postural tone caused by cerebral hypoperfusion. This chief complaint is familiar to most emergency physicians given it accounts for more than 1-2 million patient visits in the US annually.[1]  The differential diagnosis is broad and crosses multiple organ systems, including, but not limited to, cardiovascular, neurovascular,  and hemorrhagic/hematologic causes.  Pulmonary embolism (PE) is one cardiovascular cause that is considered to be infrequent. This is because when associated with syncope, it is often made more clinically apparent by signs and symptoms such as dyspnea, chest pain, tachycardia, hypotension, and hypoxemia.  The American Heart Association consensus statement on the evaluation of syncope gives little attention to diagnostics for PE, instead focusing on arrhythmia, ischemia, and structural heart disease. [2]  We know from a San Francisco Syncope Rule validation trial that when adverse events occur in the minority of patients with generalized syncope, they are related mostly to cardiac arrhythmias, followed by MI- not PE. [3] Previous studies have cited the prevalence of pulmonary embolism as a cause of syncope in hospitalized patients as low as 1.6%. [4]  But are we underestimating and under-diagnosing this potentially lethal condition? The PESIT study sought to systematically assess the prevalence of pulmonary embolism in patients admitted for syncope. [5] 

---

“Prevalence of Pulmonary Embolism among Patients Hospitalized for Syncope. Paolo Prandoni, M.D., Ph.D., Anthonie W.A. Lensing, M.D., Ph.D., Martin H. Prins, M.D., Ph.D., Maurizio Ciammaichella, M.D., Marica Perlati, M.D., Nicola Mumoli, M.D., Eugenio Bucherini, M.D., Adriana Visonà, M.D., Carlo Bova, M.D., Davide Imberti, M.D., Stefano Campostrini, Ph.D., and Sofia Barbar, M.D., for the PESIT Investigators*. N Engl J Med 2016; 375:1524-1531. October 20, 2016”

---

What was the PESIT trial? 

Figure 1. Inclusion and Exclusion Flowchart

    The PESIT trial was a cross-sectional study of patients older than 18 years of age who were hospitalized for a first episode of syncope. This was a multicenter trial, taking place in 11 hospitals (2 academic, 9 nonacademic) in Italy. There were several important exclusion criteria to this study including patients on anticoagulation, pregnant patients, and those with previous syncopal episodes.   Most importantly, however, the study excluded all patients discharged and solely focused on those admitted for an inpatient evaluation (Figure 1). 

    All included patients were subjected to a standardized history aimed at suggesting the cause of syncope and identifying risk factors for pulmonary embolism (Figure 2). They then underwent mandatory chest radiography, electrocardiography, arterial blood gas testing, and routine blood testing, including a D-dimer assay. Further diagnostic workup for causes other than pulmonary embolism was not standardized and varied between patients based on attending physician discretion.

View fullsize

Figure 2. Standardized History Questions

View fullsize

    The presence or absence of PE was assessed with a validated algorithm based on pretest clinical probability and the result of the D-dimer assay. The pretest clinical probability was defined according to the simplified Wells score, which classifies PE as being “likely” or “unlikely” (Table 1). In patients who had a high pretest clinical probability, a positive D-dimer assay, or both, computed tomographic pulmonary angiography or ventilation-perfusion lung scanning was performed.

---

Outcomes

    The primary outcome of the PESIT trial was to measure the prevalence of pulmonary embolism in patients admitted to the hospital for syncope. Secondarily, the thrombotic burden was assessed by measuring the most proximal location of embolus or the percentage of perfusion defect area.

    In 58.9% (339/560) of patients, PE was ruled out by low pre-test probability and a negative D-dimer. Of the remaining patients, 42.2% (97/229) had a pulmonary embolism. In the entire cohort, the prevalence of PE was 17.3% (97/560). Thus, the authors concluded that nearly one of every six patients hospitalized for a first episode of syncope has a pulmonary embolism (Figure 3, Table 2).

Figure 3. Results Flowchart

---

Strengths and Weaknesses of the Study

    The major strength of the PESIT trial is the systematic approach with which PE's were diagnosed. ALL admitted patients underwent consideration of and testing for PE based on risk factors and pretest probability. Even if, for example, the initial ECG suggested an arrhythmia as the cause of syncope, if they did not rule out for PE, they went on to get a scan. This ensured a complete and thorough investigation. Furthermore, their results were internally valid, as the prevalence of PE was consistently 15-20% across all 11 centers.

View fullsize

    Additionally, the population studied is also fairly representative of those who most emergency medicine physicians would admit to the hospital with syncope. Reasons for admission included trauma, severe comorbidities, failure to identify a cause in the ED, and a high probability of cardiac syncope. While we don’t commonly site the EGSYS score that they used, the components such as palpitations, heart disease, and an exertional component to syncope, are all considerations that factor into our clinical gestalt for admission.

    A glaring problem with this study is the significant variation in diagnostic workup beyond PE. Other potential explanations for syncope were much more likely to be undetermined in those with confirmed embolism (see table 2). Because further workup was left to the discretion of individual physicians, alternative and perhaps more causative causes of syncope may have been under-diagnosed or underreported.

    Furthermore, this study has limited external validity when used to assess our ED population as a whole, given the exclusion criteria of all patients discharged from the ED. When recalculating the prevalence based on all patients who presented to the ED, only 1 in 26 (97/2584 = 3.8%) patients were diagnosed with a PE, far fewer than their conclusive 1 in 6. A major contributing factor is likely age. As expected, the mean age of patients discharged was much younger at 54 years (range 16 to 79). These patients are much less likely to develop PE based on currently available decision rules such as the PERC Rule.

    We commonly use tools such as the PERC Rule or the Wells score to quantify our pretest probability because they include characteristics known to be associated with PE. As seen in the clinical characteristics described in table 2, patients diagnosed with pulmonary embolism had a high prevalence of symptoms commonly associated with PE, such as hypoxemia and tachycardia, or clinical manifestations of DVT. They were also far more likely to have a history of previous venous thromboembolism or to have active cancer. A large proportion of PESIT patients diagnosed with PE presented exactly as we would expect patients with a PE to present. These  data strengthen support for current clinical practice, which for most physicians means only entering the diagnostic pathway for pulmonary embolism when history and physical exam suggests it as a potential diagnosis.

    Lastly, the PESIT trial spends significant effort quantifying radiographic burden of pulmonary embolism across all their positive cases. As mentioned earlier, all patients were evaluated for PE, regardless of clinical gestalt. However, we must remember that degree of radiographic filling defect does not necessarily correlate with clinically significant pulmonary embolism. In 40% of patients with PE, the extent of pulmonary vascular obstruction was at the segmental or subesegmental level or was less than 25% of total lung area. Some of these may have been clinically insignificant, not likely to have caused the syncopal event, and perhaps were discovered incidentally. A better predictor for the severity of PE might have incorporated factors such as heart rate, systolic blood pressure < 100 mmHg, elevated BNP, and elevated troponins, which are used in the simplified PE Severity Index and a subsequently developed composite score to predict mortality from PE. [6,7}

---

Take Home Message

    In summary, Pulmonary embolism is an important “must-not miss” cause of syncope, but it is likely an uncommon diagnosis in patients who pass out, recover, and appear well – the main way that most patients in the US with syncope present. Overall, 1 in 26 patients who present to the ED with first time syncope may have a PE. A large portion of these may be clinically insignificant and not causative of syncope. Those with symptoms and signs of PE are more likely to have a significant PE and should be evaluated as such. Thus, we should not change our current clinical practice based on the results of the PESIT trial.

---

Expert Commentary

Internal validity:

Dr. Ireland’s review of the Prandoni study reviews it’s methods and results quite well.  He also points out the important ways in which we as consumers of literature should address a study.  The first is with respect to internal validity (how well the authors measured what they in fact intended to measure).  This takes into account potential for bias, or systematic ways in which error could be introduced into the measurement.  Dr. Ireland does not seem to identify many problems with this study from that standpoint.   In general I agree with this, if the goal of measurement of this study was to exhaustively test all patients who are admitted largely at the discretion of their doctors after a first event of syncope.  If so, then the methods of this paper seem to have resulted in reasonable internal validity.  However if the goal of the paper is to identify the prevalence of symptomatically significant PE among ALL patients with first time PE, then it is quite possible that there is bias due to the methods of inclusion.  Specifically, the fact that only admitted patients were included, and those admitted had a fairly large degree of comorbid conditions may have resulted in a measurement of PE that is higher than what would be expected for in general “syncope.”

External validity:

The second key question for a consumer of medical literature to ask is, simplistically, “are these patients like mine?”  The answer to this is probably no.  The patients included are different than a typical US ED "undifferentiated passing out patient" not just because they are Italian, but also because they had a high prevalence of symptoms to suggest PE such as history of PE/DVT,  or history of malignancy. In the US, we would probably not consider many of these patients to be “undifferentiated syncope.” Rather, we would consider them to be possible symptomatic PE patients and simply test them.  Dr. Ireland points this out in his evaluation of table 2.  It is not a novel finding that many of the patients with these comorbidities, signs, and symptoms went on to have PE when tested (Courtney Annals of EM Annals of Emergency Medicine 2010;55(4):307–315).

Another way to examine generalizability is to examine the course of these syncope patients and ask if this is similar to the course that would be taken in the US. Of the 2584 ED syncope patients in this study, 1867 were discharged. Are we discharging 72% of our syncope patients?  Whether or not we should be is another question.  However,  it is likely that the US ED environment has a much lower threshold for admission for syncope than the Italian setting, similar to the way that the US ED environment has a much lower threshold for testing for PE than the European setting.  Therefore, it is highly likely that, at least to some extent, these Italian syncope patients are more ill than the average US ED syncope patient.  This is supported by the elderly median age in the Prandoni study of 80….meaning half of all their patients were over 80 years of age!  Also note that of the 717 remaining patients not discharged, a further 157 were excluded.  So this sample really is a unique selected group…..making it difficult to generalize this study to the average US ED patient with syncope.  

Dr. Jeff Kline and I explored the possible external generalizability of this report in a re-examination of the PERC dataset which included 7940 patients from 12 US emergency departments, all of whom had  symptoms prompting testing for PE.  Among 466 PE positive patients, 6.6% reported syncope, while among 7474 PE negative patients 6.0% reported syncope (95% CI for difference, -1.7 - 3.0).  This suggested syncope was not a predictor of PE. We also noted in the Prandoni study a mean age of 75 and a high prevalence of main pulmonary artery clot (42%), something we have not found in US studies of undifferentiated PE patients where median percent obstruction was 9%.

Bottom line: 

In elderly syncope patients with some combination of tachycardia, tachypnea, hypotension, active cancer, and perhaps especially those without a clear suspected cause of syncope, PE should be a consideration that warrants testing.  Perhaps this should be considered even when patients do not have the more traditional symptoms of PE such as dyspnea or chest pain.  However, we would caution clinicians NOT to interpret this study as rationale for widespread testing on all or nearly all US ED syncope patients.  The outcome of such a simplistic interpretation of this study would undoubtedly result in further radiation and contrast burden and harms for our patients.

D. Mark Courtney, MD

Associate Professor of Emergency Medicine and Medical Social Sciences, Northwestern Emergency Medicine

---

Posts You May Also Enjoy

---

How to cite this post

[Peer-Reviewed, Web Publication] Ireland A,  Zimmerman J (2017, Sep 27). The PESIT Trial: Do All Patients Need a Syncope Workup?  [NUEM Blog. Expert Commentary By Courtney DM]. Retrieved from http://www.nuemblog.com/blog/PESIT

---

Resources

1. Syncope Evaluation in the Emergency Department Study (SEEDS): A Multidisciplinary Approach to Syncope Management. Win K. Shen, Wyatt W. Decker, Peter A. Smars, Deepi G. Goyal, Ann E. Walker, David O. Hodge, Jane M. Trusty, Karen M. Brekke, Arshad Jahangir, Peter A. Brady, Thomas M. Munger, Bernard J. Gersh, Stephen C. Hammill and Robert L. Frye. Circulation. 2004;110:3636-3645.

2. AHA/ACCF Scientific Statement on the evaluation of syncope: from the American Heart Association Councils on Clinical Cardiology, 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 Cardiology Foundation: in collaboration with the Heart Rhythm Society: endorsed by the American Autonomic Society. Strickberger SA, et. Al. American Heart Association Councils on Clinical Cardiology, Cardiovascular Nursing, Cardiovascular Disease in the Young, and Stroke; Quality of Care and Outcomes Research Interdisciplinary Working Group. American College of Cardiology Foundation.; Heart Rhythm Society.; American Autonomic Society. Circulation. 2006 Jan 17;113(2):316-27. No abstract available. Erratum in: Circulation. 2006 Apr 11;113(14):e697.

3. Prospective validation of the San Francisco Syncope Rule to predict patients with serious outcomes. Quinn J, McDermott D, Stiell I, Kohn M, Wells G.. Ann Emerg Med. 2006 May;47(5):448-54.

4. Etiology of syncope in hospitalized patients. Saravi M, Ahmadi Ahangar A, Hojati MM, Valinejad E, Senaat A, Sohrabnejad R, Khosoosi Niaki MR. Caspian J Intern Med. 2015 Fall;6(4):233-7.

5. Prevalence of Pulmonary Embolism among Patients Hospitalized for Syncope. Prandoni P, Lensing AW, Prins MH, Ciammaichella M, Perlati M, Mumoli N, Bucherini E, Visonà A, Bova C, Imberti D, Campostrini S, Barbar S; PESIT Investigators. N Engl J Med. 2016 Oct 20;375(16):1524-1531.

6. Jiménez D, Aujesky D, Moores L, et al. Simplification of the pulmonary embolism severity index for prognostication in patients with acute symptomatic pulmonary embolism. Arch Intern Med 2010; 170:1383.

7. Jiménez D, Kopecna D, Tapson V, et al. Derivation and validation of multimarker prognostication for normotensive patients with acute symptomatic pulmonary embolism. Am J Respir Crit Care Med 2014; 189:718.

---