Pulmonary Embolism: Sensitivity and Specificity

Introduction: Diagnosis of PE and DVT often merge, so diagnostic studies have been listed here for both. Also, see UpToDate article http://www.utdol.com/online/content/topic.do?topicKey=ven_pulm/6608 for details.

For presenting signs/symptoms, this is a great article: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458

Also note that Chronic Thromboembolic Pulmonary Hypertension is a different matter.

[Edit Diagnosis] [Merge dx] [Add prevalence]

Tags: Hematologic Pulmonary Tag this Diagnosis.

Prevalence

Population / Calculator	Prevalence	Comments / Study / Link
Patients suspected of having PE	19.9%	in+outpatients referred for V/Q scan Am J Respir Crit Care Med. 2000 Dec;162(6):2105-8. PMID: 11112122
Well's criteria for PE	0%	Follow this link to a pre-test probability calculator for PE http://www.mdcalc.com/wells-criteria-for-pulmonary-embolism-pe/ Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Wells PS, Anderson DR, Rodger M, Ginsberg JS, Kearon C, Gent M, Turpie AG, Bormanis J, Weitz J, Chamberlain M, Bowie D, Barnes D, Hirsh J. http://www.schattauer.de/index.php?id=1268&pii=th00030416&no_cache=1
More, Edit...

The sensitivity and specificity of findings for Pulmonary Embolism are listed below. See the left navigation bar to change the display.

Pre-Test Probability (Prevalence): %.	Post-Test Probability (Predictive Value): %.	Switch to display mode.
Note that calculating probabilities from more than one** finding is inherently inaccurate because findings are not independent. (For example, using two positive findings that share a common pathogenesis is likely to overestimate the true probability.)

Sensitive and Specific Findings
Finding	Result	Sensitivity	Specificity
[ + ] Leg US	+ 0 -	93%	98%
for DVT by venography. this is "duplex ultrasound". Study: Ann Intern Med. 1989 Aug 15;111(4):297-304. PMID 2667418
[ + ] MRA and MRV	+ 0 -	92%	96%
BUT 52% of patients had technically inadequate results MRA alone had sens = 78% / spec = 99% and was technically inadequate in 25% of patients Study: PIOPED III Ann Int Med 2010 v152(7): 434 http://www.annals.org/content/152/7/434.abstract?etoc
[ + ] CT angiography	+ 0 -	90%	95%
multi-detector CT with CT venography Study: PIOPED II (NEJM 2006;354: 2317)

Specific Findings
Finding	Result	Sensitivity	Specificity
[ + ] V/Q scan	+ 0 -	77%	98%
about 25% of patients had an indeterminate scan and were excluded from this analysis. Study: Re-analysis of PIOPED II: Radiology. 2008 Mar;246(3):941-6. PMID: 18195380
[ + ] Doppler Ultrasound	+ 0 -	29%	97%
for PE Study: Ann Intern Med. 1997 May 15;126(10):775-81. PMID: 9148650
[ + ] Abnormal Cardiac Exam	+ 0 -	22%	88%
including: increased P2, right ventricular lift, or JVD Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] No Typical Symptoms	+ 0 -	12%	87%
no chest pain / hemoptysis / dyspnea as the presenting symptoms presentations were: tachycardia/tachypnea + signs/sxs of DVT, or hypoxemia, or other presentation Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458

Sensitive Findings
Finding	Result	Sensitivity	Specificity
[ + ] Arterial Oxygen, D-dimer, or Respiratory Rate	+ 0 -	96.9%	21.3%
PaO2 < 80, positive D-dimer, OR resp rate > 20 Study: Am J Respir Crit Care Med. 2000 Dec;162(6):2105-8. PMID: 11112122
[ + ] Elevated D-dimer	+ 0 -	95%	25%
positive is generally >500ng/ml In another study, D-dimer had 83% sensitivity and 58% specificity (PMID 11112122) Study: Ann Intern Med 2004 Apr 20;140(8):589-602. PMID: 15096330 Am J Respir Crit Care Med 1999; 160:1043.
[ + ] Arterial Blood Gas or History of DVT	+ 0 -	92.1%	21.1%
increased A-a gradient OR history of DVT/PE Study: Am J Respir Crit Care Med. 2000 Dec;162(6):2105-8. PMID: 11112122
[ + ] Arterial Oxygen or History of DVT	+ 0 -	91.9%	32.4%
PaO2 < 80 OR history of DVT/PE Study: Am J Respir Crit Care Med. 2000 Dec;162(6):2105-8. PMID: 11112122
[ + ] Arterial Blood Gas	+ 0 -	91.9%	14.7%
increased A-a gradient OR PaCO2 < 36 Study: Am J Respir Crit Care Med. 2000 Dec;162(6):2105-8. PMID: 11112122
[ + ] ECG		Sensitivity = 86.5%
any abnormality pooled from reproduced data (91% in this particular case series) Study: Chest. 1997 Mar;111(3):537-43. PMID: 9118684

Negative Findings
Finding	Result	Sensitivity	Specificity
[ + ] No Typical Symptoms	+ 0 -	12%	87%
no chest pain / hemoptysis / dyspnea as the presenting symptoms presentations were: tachycardia/tachypnea + signs/sxs of DVT, or hypoxemia, or other presentation Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Chest Pain	+ 0 -	17%	79%
non-pleuritic Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Pleuritic Chest Pain or Hemoptysis	+ 0 -	44%	44%
all comers. slightly lower rate in patients with PE but no prior cardiopulmonary disease Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Pleuritic Chest Pain	+ 0 -	47%	41%
Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458

Poorly Diagnostic Findings
Finding	Result	Sensitivity	Specificity
[ + ] Calf or Thigh Swelling	+ 0 -	39%	80%
Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Homan's Sign	+ 0 -	33%	79%
for DVT by venography. See http://www.turner-white.com/pdf/hp_mar01_homan.pdf for a full description. Briefly: "The patient�s knee should be in the flexed position. The examiner should forcibly and abruptly dorsiflex the patient�s ankle and observe for pain in the calf and popliteal region, which constitutes a positive sign". Study: Angiology. 1969 Apr;20(4):219-23. PMID 5778113
[ + ] Signs of DVT	+ 0 -	47%	77%
in calf or thigh: edema, erythema, tenderness, or palpable cord Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Calf or Thigh Pain	+ 0 -	42%	75%
Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Dyspnea	+ 0 -	36%	74%
dyspnea as the only presenting symptom Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Abnormal Lung Exam	+ 0 -	37%	64%
including: crackles, wheezes, friction rub, etc Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] pro BNP	+ 0 -	60%	62%
>1.25pmol/L Study: J Thromb Haemost. 2006 Mar;4(3):552-6. Epub 2005 Dec 23. PMID 16405522
[ + ] Tachypnea	+ 0 -	57%	53%
RR > 20 Study: Am J Med. 2007 Oct;120(10):871-9. PMID: 17904458
[ + ] Arterial Blood Gas	+ 0 -	84.2%	27.4%
increased A-a gradient Study: Am J Respir Crit Care Med. 2000 Dec;162(6):2105-8. PMID: 11112122
[ + ] Tachycardia		Sensitivity = 60%
pooled from reproduced data - sensitivity = 36-90% depending on the case series. In this particular case series, it was not the most prevalent ECG finding; rather, T wave inversion was. In one other series, it was 90% prevalent (equivalent to TWI), and in PMID 123074, it is the most prevalent at 69% while TWI is only 42%. In this study, it was only present in 'massive' PE (defined by clot burden on PA cath). Study: Chest. 1997 Mar;111(3):537-43. PMID: 9118684
[ + ] T-Wave Inversion		Sensitivity = 49%
anterior leads V1-V4 pooled from reproduced data - sensitivity ranges 42-89%; in this study it was the most prevalent ECG sign Study: Chest. 1997 Mar;111(3):537-43. PMID: 9118684
[ + ] Low ECG Voltage		Sensitivity = 29%
Study: Chest. 1997 Mar;111(3):537-43. PMID: 9118684
[ + ] S1 Q3 T3 Pattern		Sensitivity = 27.5%
S wave in I, Q wave in III, and inverted T in III pooled from reproduced data - sensitivity ranges 12-50% in different series Interestingly, despite the classical description of this feature as representing right heart strain, it was not predictive of massive PE in this case series. Study: Chest. 1997 Mar;111(3):537-43. PMID: 9118684
[ + ] Right Bundle Branch Block		Sensitivity = 22%
Study: Chest. 1997 Mar;111(3):537-43. PMID: 9118684