Article, Cardiology

Prevalence of pulmonary embolism in patients presenting with syncope. A systematic review and meta-analysis

a b s t r a c t

Background: Syncope is a common clinical presentation and establishing an etiology is often challenging. Pulmo- nary embolism (PE) has been thought to be an uncommon cause but a recent report suggested otherwise.

Objective: To establish the prevalence of PE in patients presenting with syncope to the emergency department (ED) and in hospitalized patients.

Methods: We systematically searched Medline, CINAHL, EMBASE, LILACS and Web of Science with relevant key- words and MeSH headings for syncope and PE. Inclusion criteria were patients presenting with syncope to ED or hospitalized due to syncope, and etiologies including PE.

Results: Of 1329 titles and abstracts, 12 (other than Prandoni et al.) met inclusion criteria. Nine studies included 6608 ED patients and 3 included 975 hospitalized patients. The mean age was 62 (95% CI 54-69) for ED patients and 67 (95% CI 64-70) for hospitalized. The pooled estimate of PE prevalence in ED syncope patients was 0.8% (95% CI 0.5-1.3%, I2 = 0%). The pooled estimate of PE prevalence in hospitalized patients was 1.0% (95% CI 0.5-1.9%, I2 = 0). In contrast, the prevalence of PE in Prandoni et al. were 3.8% and 17.3% for ED and hospitalized patients respectively, both significantly higher than in other relevant studies (p b 0.0001).

Conclusion: The estimated prevalence of PE in patients presenting with syncope is low. The Prandoni et al. esti-

mates are significantly higher, suggesting a possible site effect, accrual bias, or investigation strategy. These and the prognostic impact of higher PE prevalence require understanding before changes in practice.

(C) 2017


Syncope is a common clinical problem accounting for 1% to 6% of hospital admissions and 1% to 3% of emergency department (ED) visits [1-4]. There are numerous causes of syncope, some potentially lethal but most relatively benign, and establishing an accurate diagnosis is a common and expensive problem [2,5]. Effective and efficient clinical management of syncope requires knowing the pretest probability of a diagnosis, the diagnostic accuracy of a test, and the impact on clinical outcome.

Among the many possible causes of syncope are pulmonary emboli (PE), which can be associated with recurrences and not infrequently death [6]. Until recently pulmonary emboli were thought to be an un- common cause, but a recent study by Prandoni et al. [7] reported a much higher prevalence than previously suspected. This prompted us to conduct a systematic review and meta-analysis to determine the prevalence of pulmonary embolism in patients presenting with syncope to the emergency department and in hospitalized patients.

* Corresponding author at: University of Calgary, Health Research and Innovation Centre, GAC 82-3280 Hospital Drive NW, Calgary, AB T2N 4Z6, Canada.

E-mail address: [email protected] (Z. Oqab).

Details of the protocol for this systematic review were registered on PROSPERO CRD42017056798 [8].

Search strategy (Appendix A)

We systematically searched Medline, CINAHL, EMBASE, LILACS and Web of Science databases. Search terms combined relevant keywords of fainting, unconsciousness, drop attack, thromboembolism, pulmo- nary infarction and MeSH headings for the concepts: 1) syncope and

2) pulmonary embolism. This was followed by a hand search in refer- ences of included studies for other potential studies.

Eligibility criteria

Studies that included patients presenting with syncope and reported on etiologies including pulmonary embolism were eligible for inclusion. We included all age categories. There were no language, time or patient setting limitations. There was no prior meta-analysis identified and re- view articles were excluded.

0735-6757/(C) 2017

Methods of review

The titles and abstracts of potential papers were independently reviewed by two authors. Full text articles were obtained when there was uncertainty in the studies. Data extraction and assessment of valid- ity were also performed by both reviewers and confirmed by a third re- viewer. Fig. 1 shows the PRISMA diagram.

Quality assessment

The quality of the studies was evaluated independently by two re- viewers. Neither the Newcastle-Ottawa Scale [9] nor the Quality Assess- ment of Diagnostic Accuracy Studies [10] tools were suitable for application to this study. Therefore, we modified an existing scale [11] and summarized the methodological quality of the studies with a series of descriptors. (Supplementary Tables S2 and S3).

Statistical analysis

Data were analyzed using Comprehensive Meta-Analysis software version 3 (Biostat Inc. Englewood, NJ). The studies were divided into two cohorts: 1) patients presenting with syncope to the emergency de- partment, 2) hospitalized Patients with syncope. Pooled odds ratios were calculated using a random-effects model. A two sample z-test of proportions was used to compare the pooled prevalence of the selected studies to Prandoni et al. [7] Heterogeneity was estimated using the I2 statistic in a random effect analysis [12]. The data of Prandoni et al. [7] were not included in the meta-analyses.


Study selection and characteristics

The initial search yielded 1902 articles, and after removing dupli- cates 1329 articles were assessed for titles and abstracts. Fifty-three full text articles were accessed and of these, 38 were excluded due to wrong population involving patients who presented with confirmed PE, 3 were review articles and 12 (other than Prandoni et al. [7]) met the eligibility criteria. There were no randomized controlled trials.

Fig. 1. PRISMA diagram for systematic literature search.

Study participants

The sample sizes ranged from 117 to 2871 patients with a mean study size of 660 (standard deviation (SD) = 721, standard error of mean (SEM) = 200). The weighted mean age of ED patients was 61.5 (95% CI 54.4-68.7, SEM = 3.6) and 49% (SD = 4.8%, SEM = 1.6) were

male. The hospitalized cohort was older with weighted mean age of

67.1 (95% CI 63.6-70.1, SEM = 1.8) and 48.5% (SD = 10.9%, SEM =

4.9) were male.

Study characteristics (Table 1)

There were nine studies involving 6608 ED patients and three stud- ies involving 975 hospitalized patients. Only 4/12 studies reported on the specific methods of diagnosing PE. Vanbrabant et al. [13] performed 13 computed tomography (CT) scans of the chest in patients suspected of having PE and confirmed PE in 3 patients. Neither the reason for or- dering the CT scans nor the clinical presentations of the patients was re- ported. Sarasin et al. [14] reported on 67/650 patients with unknown causes of syncope, of whom 11/67 were suspected to have a PE and 8/ 11 were confirmed using lower limb venous Compression ultrasound, plasma D-dimer measurement, and CT scan. Neither whether the 11 pa- tients underwent all these tests, nor the reasons for ordering the CT scans were reported. Del Rosso et al. [15] diagnosed PE in 5/516 pa- tients; 4 were diagnosed with a CT scan and 1 diagnosed with an echo- cardiogram. Neither the reason for ordering CT scan or echocardiogram nor the clinical presentations of patients was reported. Quinn et al. [16] diagnosed PE in 5/684 patients. PE was confirmed by ventilation perfu- sion (V/Q) scanning, CT of the chest, pulmonary angiography or on au- topsy. It is not reported how many patients were suspected to have PE and underwent confirmatory tests nor their clinical presentations prompting further testing.

Among the hospitalized cohort, Grossman et al. [17] reported on 3 groups of patients; full admissions, 1-day admission and ED observation unit. We combined these and analyzed them as part of hospitalized co- hort. No study systematically screened for PE. The use of CT scans was based on clinical decision-making.

Prevalence of PE in the ED syncope cohort (Fig. 2)

The pooled estimate of PE prevalence in ED syncope cohort was 0.8% (95% CI 0.5-1.3%, I2 = 0, SEM = 0.3). The prevalence of PE in ED patients in the Prandoni et al. [7] study was 3.8%, which is significantly higher than the pooled estimate of the other studies (p b 0.0001). The pooled estimate of PE prevalence in ED studies which reported on methods of diagnosing PE (i.e. CT scans, V/Q scan, echocardiogram etc.) was 1.1% (95%CI 0.7-1.8%, I2 = 5.3, SEM = 0.2).

Prevalence of PE in the hospitalized syncope cohort (Fig. 2)

The pooled estimate of PE prevalence in hospitalized syncope cohort was 1.0% (95% CI 0.5-1.9%, I2 = 0, SEM = 0.2). The prevalence of PE in hospitalized patients with syncope in the Prandoni et al. [7] study was 17.3%, which is significantly higher than other studies (p b 0.0001).

Prevalence of PE in ED and hospitalized cohorts combined (Fig. 2)

The overall pooled estimate of PE prevalence in syncope patients was 0.9% (95%CI 0.6-1.3%, I2 = 46.7%, SEM = 0.2).


We conducted a systematic review that found the prevalence of pul- monary embolism to be b 1% in patients presenting with syncope, confirming it to be an uncommon finding in current practice patterns.

Diagnostic approaches to sy”>Table 1

Study characteristics


Study design






PE Prev alence

Blanc, 2001 [27]

Prospective Observational



57 +- 23



4/454 (0.9%)

Del Rosso, 2008 [15]

Prospective Cohort



63 +- 21


5/516 (1%)

Quinn, 2004 [16]

Prospective Cohort



62 +- 23



5/684 (0.7%)

Sarasin, 2001 [14]

Prospective Cohort



60 +- 23


8/650 (1.2%)

Saravi, 2015 [28]

Prospective Cohort



60 +- 20


2/124 (1.6%)

Sun, 2009 [29]

Retrospective Chart Review



75 +- 11



7/2871 (0.2%)

Sun, 2007 [30]

Prospective Cohort





1/477 (0.2%)

Suzuki, 2004 [31]

Retrospective Chart Review



58 +- 20


4/715 (0.6%)

Vandrabant, 2011 [13]

Retrospective Chart Review



57 +- 22



3/117 (2.5%)

Grossman, 2016 [17]

Prospective Cohort



71 +- 18


3/351 (0.9%)

Cook, 2016 [32]

Prospective Cohort



70 +- 18



4/299 (1.3%)

Sule, 2011 [33]

Retrospective Chart Review



66 +- 17


2/325 (0.6%)

Abbreviations: ED-emergency department, PE-pulmonary embolism, Hosp-hospitalized, (-) not reported.

In contrast, Prandoni et al. [7] reported much higher prevalences: 3.8% and 17.3% for patients in the ED and in hospitalized, respectively.

Current diagnostic approaches to syncope

There is significant variability in the workup of syncope which is in party due to a lack of consensus on a gold standard approach. The chal- lenge to the ED physician, in addition to establishing an etiology, is de- ciding whether patient’s presentation warrants an admission, a specialist referral or Discharge home [1]. Most guidelines agree that his- tory, physical examination and ECG are the best initial diagnostic tests in establishing a diagnosis [1,18,19]. However, if a diagnosis is still not made, considerable variability exists on subsequent testing recommen- dations. Furthermore, while risk stratification is important, no clinical decision rule has been shown to be better than the judgment of the bed- side physician. One of the major limitations of the risk schemes is the in- clusion of all patients regardless of whether or not the underlying co- morbid conditions are associated with syncope [19]. Moreover, the methods used to establish these scores and risk markers from history, physical exam, study end-points, Adverse event rates and time interval between events are very variable between studies [19].

Systematic approach to diagnosing pulmonary emboli in syncope patients

In our studies, CT scans were obtained according to clinical judgment. The patient characteristics, presenting symptoms and reason for ordering the CT scan were not reported. Only Prandoni et al. [7] conducted a protocol-driven systematic search for PE. They screened hospitalized

patients with first episode of syncope for the likelihood of PE using the Simplified Wells score and D-dimer measurement. Patients who had a Wells Score N 4 and/or a positive D-dimer underwent V/Q scan or CT chest. Prandoni et al. [7] obtained a D-dimer in all hospitalized patients and the cut off value for a positive test was between 250 and 500 ng/ml, which is contrary to PE guidelines which recommend using age adjusted D-dimer thresholds or at the minimum a cut of N 500 ng/ml for imaging [20]. Furthermore, 59% of patients underwent an imaging test based on a positive D-dimer alone and a low Simplified Wells score. The specificity of D-dimer test in hospitalized patients is lower than in outpatients or ED patients due to co-morbidities, length of hospitalization and older age af- fecting the levels in the inpatient population [21,22]. Therefore, it is possi- ble that in the Prandoni et al. study a substantially higher number of patients underwent CT scans or V/Q scans because of a lower threshold for ordering confirmatory tests, which led to higher diagnoses of segmen- tal and sub-segmental PEs. Their clinical significance is unclear.

The methods of diagnosing PE are not always reliable. In our review, only 4/12 (Vandrabant [13], Sarasin [14], Del Rosso [15], Quinn [16]) re- ported on the method of diagnosis of PE. Prandoni et al. [7] used CT pul- monary angiography (CTPA) for diagnosis in 74% (72/97) of their PE patients who had high pre-test probability based on their cut offs. CTPA has a high sensitivity and specificity for diagnosing PE but in pa- tients with Low pre-test probability the false positive rate can be as high as 42% [23]. Furthermore, the inter-reader reliability is poor partic- ularly when reading smaller vessels. For example [24] in one study, the rate of discordance among radiologists was 26% in all positive CTPA ex- aminations and increased to 67% where a solitary PE was located in the segmental or sub-segmental arteries. These constituted 33% of PE pa- tients in Prandoni et al. [7].

Fig. 2. Pooled analysis of prevalence of pulmonary embolism in syncope patients compared to data of Prandoni et al. Abbreviations: Hosp-Hospitalized, ED-emergency department.

Pulmonary emboli and outcomes in syncope

It is postulated that the main mechanisms of PE causing syncope in- clude acute right ventricular failure, Bezold-Jarisch type reflex and dys- rhythmias associated with PE [25]. Not all PEs present with syncope, and the location of the PE is relevant to clinical outcomes (including synco- pe) and has prognostic value. In a prior study, only main pulmonary ar- tery and lobar artery PEs were associated with syncope [25,26]. However, this association is seen only in small number of patients with PE, in other instances patients with large pulmonary artery PE do not present with syncope [25,26]. The studies in our review did not ex- plicitly report the characteristics of the PE. In the Prandoni et al. [7] study, about 1 in 3 of diagnosed PE was segmental or sub-segmental which are clinically unlikely to cause syncope. Furthermore, the location of the PE and not the Clot burden is associated with increased risk for all- cause mortality and morbidity, proximal clots being more significant [26]. Therefore, it is unclear whether the presence of the PE in the Prandoni et al. [7] study was a direct cause of syncope, or simply an in- nocent bystander. It is also not clear whether they had prognostic significance.


Our study has several limitations. The clinical presentations and pa- tient characteristics triggering the investigations for PE were not explic- itly listed in the studies. Therefore, a direct comparison between Prandoni et al. [7] and clinical decision making approach cannot be made. Given the wide range of causes of syncope, inferring a specific eti- ologic cause from test findings is difficult and uncertain. The study is subject to bias due to clinical heterogeneity among the included studies, including differences in medical decision making leading to the evalua- tion of possible PE as well as baseline characteristics of the population enrolled. Furthermore, it is unclear what follow up was done for patients discharged from the ED to ensure they did not have a PE. Our results are subject to the limitation of meta-analyses which include the aggregation of data from different studies with variable methods, patient population and baseline characteristics.


At this point the clinical implications of the Prandoni study are un- clear. We do not know if the higher prevalence of PE was due to system- atic screening, or whether the PE caused syncope, or simply resulted from co-morbidities and hospitalization, or indeed whether they had prognostic significance. This is important not only due to treatment of patients with oral anticoagulants and exposure to bleeding risks, but also due to potential for missing a serious underlying cause. However, the magnitude of the Prandoni et al. findings cannot be ignored and re- quires further assessment.



Conflicts of interest

None of the authors have any conflict, financial or otherwise.

Appendix A. Medline search strategy

exp. Syncope/
  • (syncope or faint* or unconscious* or drop attack*).kw,tw.
  • 1 or 2
  • exp. Pulmonary Embolism/.
  • (pulmonary emboli* or PE or thromboemboli* or pulmonary infarct*).kw,tw.
  • 4 or 5
  • 3 and 6
  • remove duplicates from 7
  • Supplementary data

    Supplementary data to this article can be found online at http://dx.


    1. Sheldon RS, Morillo CA, Krahn AD, et al. standardized approaches to the investiga- tion of syncope: Canadian cardiovascular society position paper. Can J Cardiol 2011;27:246-53.
    2. Sun BC, Emond JA, Camargo Jr CA. Direct Medical costs of syncope-related hospital- izations in the United States. Am J Cardiol 2005;95:668-71.
    3. Kapoor WN. Evaluation and management of the patient with syncope. JAMA 1992; 268:2553-60.
    4. Brignole M, Menozzi C, Bartoletti A, et al. A new management of syncope: prospec- tive systematic guideline-based evaluation of patients referred urgently to general hospitals. Eur Heart J 2006;27:76-82.
    5. Developed in collaboration w, European Heart Rhythm A, Heart Failure A, et al. Guidelines for the diagnosis and management of syncope (version 2009): the task force for the diagnosis and management of syncope of the European Society of Car- diology (ESC). Eur Heart J 2009;30:2631-71.
    6. Konstantinides SV, Torbicki A, Agnelli G, et al. ESC guidelines on the diagnosis and management of acute pulmonary embolism. Eur Heart J 2014;35:3033-69 3069a- 3069k.
    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. Oqab ZSR. What is the prevalence of pulmonary embolism in patients presenting with syncope? A systematic review and meta-analysis. PROSPERO 2017:CRD42017056798; 2017.
    9. Wells G, Shea B, O’Connell D, et al. The Newcastle-Ottawa scale (NOS) for assessing the quality of nonrandomized studies in meta-analysis. The Ottawa Health Research Institute: Ottawa, Ontario; 2011.
    10. Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003;3:25.
    11. Pournazari P, Oqab Z, Sheldon R. Diagnostic value of neurological studies in diagnos- ing syncope: a systematic review. Can J Cardiol 2017.
    12. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta- analyses. BMJ: Br Med J 2003;327:557-60.
    13. Vanbrabant P, Van Ouytsel V, Knockaert D, Gillet JB. diagnostic yield of syncope in- vestigation (initiated) in the emergency department: a pilot study. Acta Clin Belg 2011;66:110-5.
    14. Sarasin FP, Louis-Simonet M, Carballo D, et al. Prospective evaluation of patients with syncope: a population-based study. Am J Med 2001;111:177-84.
    15. Del Rosso A, Ungar A, Maggi R, et al. Clinical predictors of Cardiac syncope at initial evaluation in patients referred urgently to a general hospital: the EGSYS score. Heart 2008;94:1620-6.
    16. Quinn JV, Stiell IG, McDermott DA, Sellers KL, Kohn MA, Wells GA. Derivation of the San Francisco Syncope Rule to predict patients with Short-term serious outcomes. Ann Emerg Med 2004;43:224-32.
    17. Grossman AM, Volz KA, Shapiro NI, et al. Comparison of 1-day emergency depart- ment observation and inpatient ward for 1-day admissions in syncope patients. J Emerg Med 2016;50:217-22.
    18. The Task Force for the Diagnosis and Management of Syncope of the European Soci- ety of Cardiology (ESC). Guidelines for the diagnosis and management of syncope (version 2009). Eur Heart J 2009;30:2631-71.
    19. Shen W-K, Sheldon RS, Benditt DG, et al. ACC/AHA/HRS guideline for the evaluation and Management of patients with syncope: a report of the American College of Car- diology/American Heart Association Task Force on clinical practice guidelines, and the Heart Rhythm Society. Circulation 2017;2017.
    20. Raja AS, Greenberg JO, Qaseem A, et al. Evaluation of patients with suspected acute pulmonary embolism: best practice advice from the clinical guidelines committee of the American college of physicians. Ann Intern Med 2015;163:701-11.
    21. Schrecengost JE, LeGallo RD, Boyd JC, et al. Comparison of diagnostic accuracies in outpatients and hospitalized patients of D-dimer testing for the evaluation of suspected pulmonary embolism. Clin Chem 2003;49:1483-90.
    22. Brotman DJ, Segal JB, Jani JT, Petty BG, Kickler TS. Limitations of D-dimer testing in unselected inpatients with suspected venous thromboembolism. Am J Med 2003; 114:276-82.
    23. Stein PD, Fowler SE, Goodman LR, et al. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med 2006;354:2317-27.
    24. Hutchinson BD, Navin P, Marom EM, Truong MT, Bruzzi JF. Overdiagnosis of pulmo- nary embolism by pulmonary CT angiography. Am J Roentgenol 2015;205:271-7.
    25. 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 pre- senting symptom of pulmonary embolism and patients with pulmonary embolism without syncope. Vasc Med 2003;8 (257-61).
    26. Vedovati MC, Becattini C, Agnelli G, et al. Multidetector CT scan for acute pulmonary embolism: embolic burden and clinical outcome. Chest 2012;142:1417-24.
    27. Blanc J-J, L’her C, Touiza A, Garo B, L’her E, Mansourati J. Prospective evaluation and outcome of patients admitted for syncope over a 1 year period. Eur Heart J 2002;23: 815-20.
    28. Saravi M, Ahmadi Ahangar A, Hojati MM, et al. Etiology of syncope in hospitalized patients. Caspian J 2015;6:233-7.
    29. Sun BC, Derose SF, Liang LJ, et al. Predictors of 30-day serious events in older patients with syncope. Ann Emerg Med 2009;54:769-78 (e1-5).
    30. Sun BC, Hoffman JR, Mangione CM, Mower WR. Older age predicts short-term, seri- ous events after syncope. J Am Geriatr Soc 2007;55:907-12.
    31. Suzuki M, Hori S, Nakamura I, Soejima K, Aikawa N. Long-term survival of Japanese patients transported to an emergency department because of syncope. Ann Emerg Med 2004;44:215-21.
    32. Cook OG, Mukarram MA, Rahman OM, et al. Reasons for hospitalization among emergency department patients with syncope. Acad Emerg Med 2016;18:18.
    33. Sule SS, Palaniswamy C, Aronow WS, et al. Etiologies of syncope in 325 consecutive patients hospitalized for syncope. Circulation: cardiovascular quality and outcomes conference: quality of care and outcomes research in cardiovascular disease and stroke, 4. ; 2010.

    Leave a Reply

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