The Journal of Emergency Medicine, Vol. 42, No. 2, pp. 220–226, 2012Copyright � 2012 Elsevier Inc.

Printed in the USA. All rights reserved0736-4679/$ - see front matter

doi:10.1016/j.jemermed.2010.05.071

RECEIVED: 7 NoACCEPTED: 18 M

BriefReports

HYPOTENSION IS UNCOMMON IN PATIENTS PRESENTING TO THE EMERGENCYDEPARTMENT WITH NON-TRAUMATIC CARDIAC TAMPONADE

Tushar Kapoor, MD,*† Michael Locurto, MD,* Gino A. Farina, MD,* and Robert Silverman, MD*

*Department of Emergency Medicine, Long Island Jewish Medical Center, New Hyde Park, New York and †Department of EmergencyMedicine, Nassau University Medical Center, East Meadow, New York

Reprint Address: Tushar Kapoor, MD, Department of Emergency Medicine, Nassau University Medical Center,2201 Hempstead Turnpike, East Meadow, NY 11554

, Abstract—Background: Cardiac tamponade is a life-threatening disease in which hypotension is believed to bea common finding. Prior inpatient studies have describednormotensive or hypertensive cases of tamponade; however,because the data were not collected from the EmergencyDepartment (ED), the hemodynamic spectrum may differfromthosepresenting to theED.Objectives:We hypothesizedthat hypotension is uncommon in patients presenting to theED with non-traumatic tamponade. Methods: A retrospec-tive chart review was conducted between January 2002 andDecember 2007 of patients presenting to our ED who weresubsequently diagnosed with cardiac tamponade. Results: Atotal of 34 patients were identified with a diagnosis of tampo-nade. Themean blood pressure onEDarrival was 131/79 mmHg. Upon initial presentation to the ED, 35% (n = 12) ofpatients were hypertensive, 50% (n = 17) were normotensive,and 15% (n = 5) were hypotensive. Of the 5 patients whowerehypotensive on ED arrival, only 2 (6% of all patients)remained hypotensive upon admission to the hospital and be-fore a pericardiocentesis. An average of 995 mL of fluid wasremoved from the pericardium. The chief complaint for themajority of patients in tamponade was shortness of breath(70%); 59%were tachycardic in the ED, and 72%had cardi-omegaly on chest X-ray study. Conclusions: Hypotension isuncommon in patients presenting to the ED with non-traumatic cardiac tamponade. The majority of patients arenormotensive or even hypertensive. Thus, the emergencyphysician should not exclude the diagnosis of tamponadeeven in light of normotension or hypertension. � 2012Elsevier Inc.

vember 2009; FINAL SUBMISSION RECEIVED: 3 Maay 2010

220

, Keywords—cardiac; cardiac tamponade; non-trau-matic; non-traumatic tamponade; medical tamponade;hypotension; hypertension; normotension; blood pressure;blood pressure in tamponade; hypertensive tamponade

INTRODUCTION

Cardiac tamponade is a life-threatening disease process inwhich accumulation of fluid within the pericardial sac re-sults in increased intrapericardial pressure. The pericardialsac can be filled with fluid, pus, blood clots, or gas, and canbe a result ofmedical conditions, trauma, or cardiac rupture(1,2). The most common medical diseases associated withcardiac tamponade include renal failure, connective tissuedisorders, infection, and cancer (1). Due to the elastic prop-erties of the parietal pericardium, cardiac tamponadecanbe further divided into acute (where rapid accumulationof a pericardial effusion can cause hemodynamic compro-mise) or chronic (where pericardial effusion accumulatesmore slowly and larger amounts of fluid generally areneeded to cause tamponade). Presentation can range frommild or vague symptoms such as weakness, anorexia,dysphagia, cough, and feeling faint to severe consequencessuch as pulseless electrical activity and death (2,3). Otherclinical findings commonly reveal dyspnea on exertionand tachypnea, tachycardia, and pulsus paradoxus.

rch 2010;

Triage Blood Pressure Distribution

15%

50%

35%

0%10%20%30%40%50%60%70%80%90%

100%

Triage BP

Hypertensive n = 12Normotensive n = 17Hypotensive n = 5

Average Blood Pressure in ED

6%

76%

18%

0%10%20%30%40%50%60%70%80%90%

100%

Ave ED BP

Hypertensive n = 6Normotensive n = 26Hypotensive n = 2

Figure 1. Blood pressure distribution upon triage andaveraged during ED stay.

Hypotension is Uncommon in Non-traumatic Cardiac Tamponade 221

Beck’s triad (distended neck veins, muffled heart sounds,and hypotension) is the ‘‘classic’’ textbookpresentation (4).

Textbooks of emergency medicine state that the physi-cal examination commonly reveals tachycardia and a lowsystolic arterial pressure (1,5). However, several reportsfrom inpatient populations have found normal or evenelevated blood pressures, suggesting that low bloodpressuremaynot beneeded todiagnosenon-traumatic tam-ponade (6–9). Because emergency physicians often rely onnumeric findings like blood pressure when developinga differential diagnosis, knowledge of the frequency ofhypotension in cases of cardiac tamponade is important,particularly for initial emergency department (ED)evaluation. The goal of this study was to determine thefrequency of hypotension in ED patients diagnosed withnon-traumatic tamponade.

MATERIALS AND METHODS

This study was a retrospective chart review of medicalrecords of adults age 18 years and older presenting to theED of Long Island Jewish (LIJ) Medical Center betweenJanuary 2002 through December 2007. LIJ is a tertiarycare center with an annual ED volume of approximately80,000 patients. Chart screening was done to includepatients in the study population using International Classi-fication of Diseases-9th revision codes for pericardialdiseases where tamponade was suspected in the ED andconfirmed with documented echocardiographic evidenceof non-traumatic tamponade within 24 h of admission, orwith pericardiocentesis. Only cases with documented evi-dence of non-traumatic tamponade were included. Theechocardiographic study definition of tamponade includedany of the following: right atrial, right ventricular diastoliccollapse; abnormal inspiratory decrease of left ventriculardimensions; right atrial compression; abnormal inspira-tory increase in blood flow velocity through the tricuspidvalve and abnormal inspiratory decrease of mitral valveflow velocity; and dilated inferior vena cava with lack ofinspiratory collapse (1,10).

Clinical data were abstracted onto a pre-made datacollection sheet. In the event of disagreement betweentwo of the authors in including or excluding a patient,the final decision was made by a third author.

Cardiomegaly was determined by the attending radiol-ogist’s X-ray study readings. Initial blood pressure wasdetermined during triage. Hypertension was defined byeither a systolic blood pressure$ 140 mmHg or diastolicblood pressure$ 90 mm Hg (11). The National HeartLung and Blood Institute definition of hypotensionwas used: either systolic blood pressure < 90 mm Hg ordiastolic blood pressure < 60 mm Hg (12). The studywas approved by the institutional review board of LongIsland Jewish Medical Center.

RESULTS

A total of 34 patientsmeeting inclusion criteriawere iden-tified, with an average age of 59 years (range 26 to 85years), and 18/34 (53%) were men. The chief complaintfor the majority (70%) was shortness of breath, followedby chest pain (12%), abdominal pain (12%), and weak-ness, syncope, and palpitations (3% each). The cause oftamponade was idiopathic in 33% of patients, cancer in21%, post-surgical in 21%, hemorrhagic in 15%, and9% were due to a connective tissue disorder. A total of32 out of 34 patients had a pericardiocentesis performedon the inpatient service. For the 2 who did not undergopericardiocentesis, one requested only palliative careand the other had early tamponade, was observed, andimproved without invasive intervention. A total of 23patients (68%) had tamponade diagnosed while still inthe ED, and the remainder received the diagnosis within24 h of hospitalization.

Themean blood pressure onEDarrival was 131/79 mmHg. Thirty-five percent (n = 12) of patients were found tohave hypertensive blood pressures, 50% (n = 17) withnormotensive, and 15% (n = 5) with hypotensive bloodpressure (Figure 1). Of the 5 patients who were hypoten-sive on ED arrival, only 2 (6% of all patients) remainedhypotensive upon admission to the hospital and beforea pericardiocentesis. The average blood pressure, defined

Figure 2. Average grouped initial systolic blood pressureof hypertensive, normotensive, and hypotensive cases oftamponade.

222 T. Kapoor et al.

as the mean of all blood pressure values obtained duringthe ED stay, was 121/75 mm Hg.

Therewere 23 patients with a known history of chronichypertension. Of the 12 patients who, upon triage, hadblood pressures in the hypertensive range, 4 (33%) hadno history of chronic hypertension; of the17 patientswith a normal initial ED blood pressure, 6 (35%) had nohistory of chronic hypertension; and of the 5 hypotensivepatients, 1 (20%) had no history of hypertension.

We also categorized patients as being either hyperten-sive, normotensive, or hypotensive at different points dur-ing their hospital stay (Figure 2). Among patients withhypotension on arrival, the systolic blood pressure im-proved throughout the hospital stay, although it was onlyslightly higher (not statistically significant) on dischargethan on arrival. Normotensive patients on ED arrival hadsimilar systolic blood pressures on hospital discharge,and patients with hypertension on ED arrival had a declinein systolic blood pressure throughout the hospital stay. Thesystolic blood pressure on hospital discharge was lower inpatients who had lower systolic blood pressure on ED ar-rival as compared to thosewith normal or elevated systolicblood pressure on arrival, again suggesting that someinitially lower blood pressures were close to baseline.

The average pre-pericardiocentesis blood pressurewas 127/73 mm Hg, with an average of 975 mL(SD 6 602 mL) of fluid removed from the pericardium.After pericardiocentesis, the average blood pressure was131/70 mm Hg. These data are summarized in Table 1.None of the patients required an emergency pericardio-centesis in the ED; all of the procedures were performedeither in the operating room by cardiothoracic surgeonsor by cardiologists in the catheterization laboratory. Wecompared some of our overall clinical findings to datareported from inpatient studies of non-traumatic tampo-nade; these data are shown in Table 2. Table 3 shows thefrequency of specific echocardiographic diagnostic crite-ria found in those diagnosed with cardiac tamponade.

The exact time between patient arrival in the ED to per-formance of the echocardiogramwas unavailable formost

patients due to either official echocardiogram reports nothaving the time recorded, or the progress note written bythe echocardiographer lacked the time. However, becausethe dates were included, they fell within 24 h of admis-sion. The average time to perform pericardiocentesiswas measured as 30.4 h (SD 6 29.8 h) from ED arrival.Themajority of patients had pericardiocentesis performedwithin 24 h of ED arrival, however, there were 12 patientsin whom the procedure was performed more than 24 hafter ED arrival (range 25–104 h) due to the overall stabil-ity of the patient from early tamponade or trial of medicalmanagement before the repeat echocardiogram thatshowed persistent or worsening effusions.

There were a total of three deaths out of the 34 patientswith tamponade (8.8%). Two of the three had underlyingmalignancies. The cause of death in one was unrelated totamponade because the patient expired from complica-tions of toxic megacolon developing later in the hospitalstay. One patient had both tamponade and a pulmonaryembolism and, upon request from family, only supportivetreatment was given.

DISCUSSION

We found that most patients with cardiac tamponade ofmedical etiology who presented to our ED were eithernormotensive or hypertensive at the time of triage. Con-trary to popular teaching and textbooks of emergencymedicine, only 15% of patients in our study were hypo-tensive upon ED arrival. After ED management (andbefore pericardiocentesis), only 6% remained hypoten-sive. Our findings indicate that normal or elevated bloodpressures do not exclude the diagnosis of tamponade inthe ED, but rather are the more common findings.

Although we believe this is the first ED study to eval-uate blood pressure in patients with tamponade, severalin-hospital studies of medical patients have reported tam-ponade blood pressure typically in the normotensive oreven hypertensive range (6–9). Reddy et al. described14 patients with tamponade, none of whom had systolicblood pressure below 100 mm Hg, and 43% of whomhad hypertensive tamponade with systolic bloodpressure above 140 mm Hg (8). Another study by Brownet al. described 6 of 18 patients (33%) with tamponadewho had systolic blood pressure ranging from 150 to210 mm Hg (7). Although none of these studies charac-terized ED patients, the blood pressurewould be expectedto be lower in acutely ill individuals before any treat-ments. However, results from this study do not supportthis assumption because most of the patients turned outto be normotensive or hypertensive.

We found that, in general, blood pressure in patientswho had normal or low values at the time of triage didnot change substantially at the time of hospital discharge.

Table 1. Clinical Data for Patients Studied

PatientAge

(Years) Sex Etiology

FluidRemoved

(mL)Initial ED

BP

Pre-pericardiocentesis

BP

Post-pericardiocentesis

BP ECHO Findings

1 85 M Cancer 2000 82/44 95/49 82/44 RA, RV collapse,RespVar, dilated IVC

2 55 M Post-surgical 2500 88/60 183/115 109/68 CTP3 75 M NA Not

performed92/45 Not

performedNot

performedRA, RV

collapse4 35 M Idiopathic 425 96/64 115/60 110/75 Resp Var5 26 F SLE 600 99/58 124/69 100/55 RV collapse6 54 F Hemorrhagic 1500 105/73 101/73 99/67 RA, RV collapse7 85 F Post-surgical 500 107/62 140/70 160/70 RA collapse, Resp

Var8 41 F Post-surgical/

hemorrhagic1000 108/67 109/68 133/51 Resp Var

9 42 M Hemorrhagic 460 111/80 101/68 97/65 RV collapse, RespVar

10 61 M Cancer 660 112/69 118/60 104/69 RA, RV collapse,Resp Var, dilatedIVC

11 78 F Hemorrhagic 1600 114/84 85/60 103/54 RV collapse12 68 M Cancer 1000 118/73 125/65 113/64 RA, RV collapse,

Resp Var13 52 M Post-surgical 2700 120/80 115/90 110/55 CTP14 74 F Cancer 850 121/67 122/80 122/70 RA, RV collapse,

Resp Var, dilatedIVC

15 60 F Cancer 700 124/74 85/60 115/60 RA collapse, RespVar

16 33 F Connectivetissuedisorder

500 127/95 50/0 186/109 RA collapse, RespVar

17 60 F Post-surgical 400 128/82 134/75 165/75 CTP18 82 F NA Not

performed129/26 Refused

procedureRefusedprocedure

RVcollapse

19 51 F Idiopathic 700 131/76 129/68 135/78 RV collapse,Resp Var

20 38 M Idiopathic 400 131/89 110/63 132/52 RA, RV collapse,Resp Var

21 67 F Idiopathic 800 134/90 111/66 129/75 Resp Var22 69 M Idiopathic 1200 138/68 142/68 154/72 CTP23 57 M Cancer 800 138/81 103/66 163/80 CT surgery

diagnosis in OR24 74 F Idiopathic 800 140/91 140/70 148/102 CT surgery

diagnosis in OR25 34 F Idiopathic 800 140/95 126/70 125/55 RA collapse, Resp

Var26 65 M Fibrinous 400 141/82 140/85 140/85 Resp Var27 60 F Cancer 600 142/100 118/78 112/70 Resp Var28 75 M Post-surgical 800 144/86 165/75 165/65 CTP29 60 M Hemorrhagic 1000 150/100 140/90 189/90 RA collapse30 68 M Idiopathic 800 150/80 159/80 135/85 RA, RV collapse,

dilated IVC31 48 M Idiopathic 700 151/109 112/78 164/109 RA, RV collapse,

Resp Var32 49 M Post-surgical 2000 181/85 143/89 104/64 RA collapse, dilated

IVC33 69 F Idiopathic 600 210/102 203/93 130/60 RA, RV collapse,

Resp Var34 69 M Idiopathic 1400 266/140 233/138 138/55 CTPAve. 59 18 M/16 F 975 131/79 127/73 131/70

ED = emergency department; BP = blood pressure; ECHO= echocardiogram; RA = right atrium; RV = right ventricle; Resp Var = significantrespirophasic variability; CTP = ‘‘Cardiac Tamponade Physiology’’ on report with no other specifics or statement saying ‘‘ECHO evidenceof tamponade’’ with no specifics.

Hypotension is Uncommon in Non-traumatic Cardiac Tamponade 223

Table 2. Comparison of Frequency of Signs and Symptoms in the Diagnosis of Cardiac Tamponade

Presenting PatientCharacteristics

This Study(n = 36)

Guberman et al.,1981 (9) (n = 56)

Levine et al.,1991 (6) (n = 50)

Cooper et al.,1995 (14) (n = 30)

Gibbs et al.,2000 (15) (n = 46)

Dyspnea 79% - 83% 87% 90%Cardiomegaly 72% 95% 68% - 100%Tachycardia 59% 77% 74% 65% 88%Hypotension* 15% 36% 14% 30% 24%

* Hypotension is defined as systolic blood pressure < 100 mm Hg in the Guberman, Levin, Cooper, and Gibbs studies (6,9,14,15). Whenthis same criterion is applied to our data, 17.6% of patients would be categorized as hypotensive.

224 T. Kapoor et al.

However, when the initial blood pressure was elevated,the average blood pressure was decreased at the time ofhospital discharge. The finding of an elevated blood pres-sure believed to be due to tamponade has been previouslyreported. In one study, elevated blood pressure and pe-ripheral vascular resistance were present but decreasedwithin 15 min of pericardiocentesis (7).

Although not specific, a number of characteristics canhelp diagnose cardiac tamponade, including dyspnea,tachycardia, pulsus paradoxus, elevated jugular venouspressure, and cardiomegaly on chest radiograph (13).In addition, a history of cancer, recent cardiac surgery,being on anticoagulants, and electrocardiographic changessuch as the presence of low voltage may help in thediagnosis (6,9,14,15). In our study population, dyspnea,tachycardia, and cardiomegaly on chest radiograph werethe most likely findings to be present. In our study, thesefindings were based on retrospective data review and it ispossible that the frequency of some symptoms and signswere underestimated. Ruling out pericardial tamponadebased on non-specific findings is challenging, and an echo-cardiogram is usually needed to assist with the diagnosis.

The pericardial pressure curve indicates that in theearly phase of tamponade (given the elastic property ofthe pericardial sac), patients may have normal or elevatedblood pressures. In traumatic tamponade, rapid filling ofa relatively undistensible pericardial sac may result inan unstable patient. In medical cases, tamponade may re-sult from slower accumulation of pericardial fluids, andlow blood pressure may be a much later finding. Becauseuntreated cases of tamponade are believed to be associatedwith poor outcomes, identifying these cases before severe

Table 3. Frequency of Echocardiographic (ECHO)Diagnostic Criteria Found

ECHO CriteriaNumberof Cases Percent

Right atrial diastolic collapse 17 53%Right ventricular diastolic collapse 15 47%Respirophasic variability 18 56%Dilated inferior vena cava 4 13%Dilated inferior vena cava with

minimal respiratory collapse3 9%

hemodynamic compromise occurs is the challenge for theclinician.

Limitations

One limitation of this study was the study design beinga retrospective chart review, and thus data, especiallyon history and physical examination, were based oncharting and taken from a single tertiary care center.We did come across two cases of non-traumatic tampo-nade that were direct transfers from other hospitals tothe coronary care unit, thus bypassing the ED; theywere not included in the study. Furthermore, becauseonly cases that were diagnosed within 24 h of ED presen-tation were analyzed, there may have been cases of hypo-tensive or hypertensive tamponade in the ED diagnosedlater in the hospital course. Such missed cases in theED might have caused us to underestimate or overesti-mate the number of patients with hypotension.

CONCLUSIONS

Most ED patients with cardiac tamponade of medicaletiologywill have normal or elevatedblood pressures onar-rival. The emergency physician must rely on patient riskfactors, presenting complaints, and other objective findingswhen considering the diagnosis of cardiac tamponade.

REFERENCES

1. Niemann JT. The cardiomyopathies, myocarditis, and pericardialdisease. In: Tintinalli JE, Kelen GD, Stapczynski JS, eds. Emergencymedicine: a comprehensive study guide. 6th edn. New York:McGraw-Hill Companies, Inc.; 2004:383–5.

2. Spodick DH. Acute cardiac tamponade. N Engl J Med 2003;349:684–90.

3. Shabetai R. Pericardial effusion: haemodynamic spectrum. Heart2004;90:255–6.

4. Beck CS. Two cardiac compression triads. JAMA 1935;104:714–6.5. Jouriles NJ. Pericardial and myocardial disease. In: Marx JA,

Hockberger RS, Walls RM, et al., eds. Rosen’s emergency medi-cine: concepts and clinical practice. 6th edn. Philadelphia, PA:Mosby Inc.; 2006:1285–6.

6. Levine MJ, Lorell BH, Diver DJ, Come PA. Implications ofechocardiographically assisted diagnosis of pericardial tamponadein contemporary medical patients: detection before hemodynamicembarrassment. J Am Coll Cardiol 1991;17:59–65.

Hypotension is Uncommon in Non-traumatic Cardiac Tamponade 225

7. Brown J, MacKinnon D, King A, Vanderbrush E. Elevated arterialblood pressure in cardiac tamponade. NEngl JMed 1992;327:463–6.

8. Reddy PS, Curtiss EL, O’Toole JD, Shaver JA. Cardiac tamponade:hemodynamic observations in man. Circulation 1978;58:265–72.

9. Guberman BA, Fowler NO, Engel PJ, GueronM, Allen JM. Cardiactamponade in medical patients. Circulation 1981;64:633–40.

10. Fowler NO. Cardiac tamponade: a clinical or an echocardiographicdiagnosis? Circulation 1993;87:1738–41.

11. Chobanian AV, Bakris GL, Black HR, et al. National High BloodPressure Education Program Coordinating Committee. Seventhreport of the Joint National Committee on Prevention, Detection,Evaluation, and Treatment of High Blood Pressure. Hypertension2003;42:1206–52.

12. National Heart Lung and Blood Institute. Diseases and conditionsindex: hypotension. Available at: http://www.nhlbi.nih.gov/health/dci/Diseases/hyp/hyp_whatis.html. Accessed September 25, 2008.

13. Roy CL, Minor MA, Brookhart MA, Choudhry NK. Does thispatient with pericardial effusion have cardiac tamponade? JAMA2007;297:1810–8.

14. Cooper JP, Oliver RM, Currie P, Walker JM, Swanton RH. How dothe clinical findings in patients with pericardial effusions influencethe success of aspiration? Br Heart J 1995;73:351–4.

15. Gibbs CR, Watson RD, Singh SP, Lip GY. Management of pericar-dial effusion by drainage: a survey of 10 years’ experience in a citycentre general hospital serving a multiracial population. PostgradMed J 2000;76:809–13.

226 T. Kapoor et al.

ARTICLE SUMMARY

1. Why is this topic important?Tamponade is a challenging diagnosis to make in the

Emergency Department (ED) because patients may pres-ent with a variety of signs and symptoms. Some cliniciansmight eliminate the possibility of tamponade due to nor-mal or hypertensive blood pressure readings, which maybe an erroneous assumption.2. What does this study attempt to show?

This study attempts to show that hypotension is presentin only a minority of patients presenting to the ED withtamponade. It also attempts to show that the majority ofpatients with tamponade in the ED are normotensive orhypertensive.3. What are the key findings?

Most patients with cardiac tamponade of medical etiol-ogy who presented to our ED were either normotensive orhypertensive at the time of triage. Only 15% of patientswere hypotensive upon ED arrival. The chief complaintfor the majority of patients in tamponade was shortnessof breath, and the majority also had cardiomegaly on chestX-ray study.4. How is patient care impacted?

These findings show that clinical suspicion of tampo-nade should not decrease due to either normal blood pres-sure readings or hypertension. Therefore, in the ED therewill be a greater chance of making the diagnosis of tam-ponade in patients who are not hypotensive.