1356 THE NEW ENGLAND JOURNAL OF MEDICINE

MEDICAL PROGRESS

MEDICAL THERAPY OF ACUTE MYOCARDIAL INFARCTION BY APPLICATION OFHEMODYNAMIC SUBSETS (Fint of Two Parts)

JAMES S. FORRESTER, M.D., GEORGE DIAMOND, M.D., KANU CHATTERJEE, M.B., M.R.C.P.,AND H. J. C. SWAN, M.D., PH.D.

I N the broadest sense, current hospital therapyof acute myocardial infarction attempts to pre-

vent, or promptly and effectively treat, the electri-cal and mechanical consequences of regional myocar-dial ischemia or infarction, while preserving jeopar-dized ischemic myocardium. Prevention and effectivetherapy of arrhythmias have largely been realizedthrough the development of coronary-care units. Thegoals of prompt and effective therapy of disorderedcardiac function and preservation of ischemic muscle,although far from achieved, have been greatly ad-vanced in the past several years by the widespread useof hemodynamic monitoring. As a result, new con-cepts concerning the disease itself and new therapieshave been developed. Nevertheless, the systematic ap-plication of this new knowledge in the clinical prac-tice of cardiology in the community has lagged be-hind these advances. This discussion, therefore, hasthree purposes: to describe the method of hemody-namic monitoring; to summarize and interpret thelarge body of clinically relevant information that hasresulted from use of the method; and to apply thisknowledge to the development of a logical approachto therapy of patients with acute myocardial infarc-tion.

~

HEMODYNAMIC MONITORING - THE METHOD

Hemodynamic monitoring uses a balloon-tippedcatheter (Swan-Ganz)1 that is advanced from a pe-ripheral vein to the pulmonary artery at the bedsidewithout the use of fluoroscopy. When inflated, the bal-loon serves to guide the catheter in the direction ofblood flow, through the chambers of the right side ofthe heart into the pulmonary artery, where the bal-loon ultimately impacts in a vessel of equal diameter.In this position, the tip of the catheter no longer re-cords pulmonary arterial pressure, since this waveform is totally damped through the air-filled balloon;rather, it records the downstream pressure, termedpulmonary-capillary pressure. By adding a lumenwith an orifice 30 cm from the catheter tip and im-planting a thermistor in the external surface of thecatheter just proximal to its tip,2~>ne can determinecardiac output by the thermodilution technic. A

From the Division of Cardiology, Cedan-Sinaj Medical Center, and theDepartment of Medicine, University of California at Los Angeles (addressreprint requests to Dr. Forrester at the Division of Cardiology, Cedars-SinaiMedical Center, 4833 Fountajn Ave., Los Angeles, CA 90029).

~~

De<:. 9, 1976

known quantity of cold solution is injected into theright atrium, and the resultant change in pulmonary-artery blood temperature is measured by the thermis-tor. This method therefore measures output of theright side of the heart, which is normally equal to thatof the left in the absence of inti"acardiac shunts. Car-;diac output determined by this method is both accu-rate and highly reproducible.3 Thus, with a singlcatheter in the right side of the heart, the effects 0acute myocardial infarction and its complications up .on right atrial, pulmonary-artery and pulmonary-capillary pressures, and cardiac output can be assessecontinuously (Fjg. 1). Alterations in the magnitudand morphology of these wave forms are of major diagnostic importance in a number of pathologic cardiovascular conditions encountered in the coronarycare unit (Table 1). ~

J-~

INTERPRETATION OF HEMODYNAMIC MEASUREMENi

- MAGNITUDE AND MORPHOLOGY ~

.Right Atrial Pr...ur. ~

Right atrial pressure is equivalent to both the rigventricular pressure in diastole (when the tricuspivalve is open, the right atrium and right ventricle acommon chambers) and to the central venous presure. Reduced right atrial pressure is generally seeondary to hypovolemia. Elevation of pressure hthree major causes: right ventricular failure of any orgin, tricuspi~ regurgitation and pericardial tampoade. In patients with acute myocardial infarction, eevated right atrial pressure most commonly refleright ventricular failure.

Clues to the diagnosis of tricuspid insufficiency apericardial tamponade come from inspection of tpressure wave forms. The diagnosis of tricuspid isufficiency can be made in the presence of a "giantwave" in the right atrial tracing (Fig. 2). Pericarditamponade may be suspected in the presence of .

elevated right atrial pressure when diastolic-pressmeasurements from right atrium, right ventricle, pmonary artery and pulmonary capillaries are of equmagnitude and similar contour.

Pulmonary-Artery Pressure

Pulmonary-artery pressure is equal to right ytricular pressure during systole, while the pulmonvalve is open. It is increased in the presence of eitincreased pulmonary vascular resistance (as with

Vol. 295 No. 24 HEMODYNAMIC THERAPY OF MYOCAWIAL INFARCTION - FORRESTER ET AL. 1357

i Ittf .

!,,cl

i

. ., .,

Figure 1. Continuous Pressure Recording as the Pulmonary-Artery Balloon Catheter Is Passed through the Chambers of theRight Side of the Heart.

RA denotes right atrium, RV right ventricle, PA pulmonary artery, arid PCW pulmonary-capillary-wedge pressure (8ee text forI discussion of the Interpretation of these pressures).

existent chronic lung disease or pulmonary embo- plitude-damped version of left atrial pressure. Duringlism) or elevated pulmonary flow (as in ventricular diastole, with an open, non-stenotic mitral valve, theseptal defect). The pulmonary arterial diastolic pres- pulmonary venous bed, left atrium and left ventriclesure is virtually equal to pulmonary-capillary pres- become a common chamber, and pulmonary-capil-sure when pulmonary vascular resistance is normal. lary pressure is then also equal to left ventricular dia-For this reason, pulmonary arterial diastolic pressure stolic pressure. Because the mitral valve begins tois often substituted for pulmonary-capillary pressure, close before the onset of ventricular systole, however,since balloon inflation is not required for its measure- pulmonary-capillary pressure may not be equal to leftment.4-7 When pulmonary vascular resistance is in- ventricular pressure at end-diastole..creased, however, pulmonary arterial diastolic pres- The level of pulmonary-capillary pressure is criticalsure is substantially elevated in comparison to pul- in the assessment of a patient's clinical status for twomonary-capillary pressure. This distinction can be of reasons. In the first place, it is the hydrostatic pres-diagnostic value in the differentiation of pulmonary sure that forces fluids out of the pulmonary vascularembolism from cardiogenic shock, since in pulmo- space into' the interstitial and intra-alveolar spaces.nary embolism, pulmonary arterial diastolic pressure When elevated, therefore, it is the hemodynamicmay be elevated while pulmonary-capillary pressure "cause" of pulmonary congestion. Secondly, cardiac

. is normal, whereas in cardiogenic shock, both pres- function is frequently defined by Starling's law of thesures are elevated and more nearly equal. heart, which states that the strength of cardiac con-

The morphology of the pulmonary arterial pres- traction is proportional to myocardial fiber length (orsure wave is of value in the diagnosis of acute mitral left ventricular volume) at the onset of contraction.regurgitation,"'o in which the tracing is often distort- Since the level of pulmonary-capillary pressure is di-ed by a giant regurgitant wave transmitted backward rectly related to this diastolic fiber stretch, it may bethrough the low-resistance pulmonary vascular bed used as a readily obtained index of left ventricular vol-(Fig. 2). ume. When applied in this manner to construct a car-

diac function curve, it is often termed left ventricular

filling pressure.Pulmonary-Capillary Pre..ura

Since there are no valves betWeen the pulmonarycapillaries and the left atrium and since the pulmo- -In C&Sf:S in which the atrial contribution to ventricular fillins is substan-

1 bed 11 ha' 1 . 1 tial. or in which left ventricular compliance is reduced. left ventricular end-nary vascu ar norma y s a ow resistance, pu - diastolic pressure may exceed mean left ventricular diastolic pressure andmonary-capillary pressure is a phase-delayed and am- pulmonary-capillary presure by as much as 20 mm Hg.

9, 1916

nto thel1onary-herrnis-: of theI to thatts. Car-h accu-I single:rects ofons up-

,ry-cap-LSsessedgnitudeajor di-~ic car-ronary-

EMENTS

le righticuspidide areIS pres-lly sec-are hasmy ori-impon-ion, el-reflects

ley and. of the:pid in-giant V.cardiale of anressureIe, pul-.f equal

1t ven-nonaryr eitherith co-

lHE NEW ENGLAND JOURNAL OF MEDICINE ~ 9, 1976USB

Table 1. Normal and Abnormal Intracardlac Pressures..

AUATInID IN WHICH

boca.AIID

R V failure; PE;

COPD; tricuspid-valve abnormal!-ity; pericardia!tamponade.

Systolic:t resistance: PE;

COPD;t I1ow: VSD;

Diutolic:t re8iatanc:c: PE;

capo;All c:aUICS of t

PCP.LV failuremitral valve ~tamponade1 LV compliuce:

hypertrophy;infan:tion.

RA

PA

0-1

IS - JOs:l2f

S-I2tPCP

ORA -- ,... -. PA , ...,. IIV riPt -ndo. P£ , ,n IrIc8opid i_lIId8ncJ. MII8iIraI iMoolllcil8cJ. copo cIIro8ic u...

"--"'1 ~ VSD _lricu18. .rec.. LV 18ft w:atrido, " PCP ary-capillary '

t AIIIooooP cIi.1c8I 01""1_,, at appro.imately .1 "'III H..... , ,.. UJIII8I' IiIDiI of io 11 - .....

increase

- -

,Vol. 295

er than pulmonary-capillary pressure when perfor-mance oC the right side oC the heart is impaired: acute-ly, with pulmonary embolism; or chronically, withchronic obstructive pulmonary disease. Conversely,pulmonary-capillary pressure will be substantiallygreater than right atrial pressure when leCt ventricu-lar dysCunction predominates, as in acute myocardialinfarction.16 Thus, although elevation oC right atrialpressure remains a useful index oC the magnitude oCright-sided heart Cailure, it must be recognized thatflorid pulmonary edema may occur in the presence oCa normal right atrial pressure (or central v.enous pres-sure) in patients with acute myocardial--infarc-tion. 17.\1

The pulmonary-capillary wave Corm, like that oC theright atrium, is bifid and oC low magnitude. Acutemitral regurgitation alters this wave Corm in the samemanner as tricuspid regurgitation alters right atrialpressure: the diagnostic ugiant V wave" oC this dis-order is a direct reflection oC the marked increase inleft atrial pressure caused by a rapid increase in leftatrial volume during ventricular systole (Fig. 2).

s..c.K: P"nuNS

T1: aiant v waw:("w:ntricularized"wave form);

tamponade: "djutolicplateau." with para-doxical inspiratoryriae.

MI: rctrOiradc Vwive;

VSD: wide pulaepnIIIUrc;

tampoaadc: narrowpUR preaIIrc;

COPD: marked rcapi-ratory fluctuation.

CARDIJOUTPl

L/mil

figura 3-

The tr8CI.. ."..

iahed iJjtance 81tiaI1y t(tioo.

The I

ponentiisticaU,disordealreadysecondprodt1e:cbaractclrcula1miscalccurve r:outputleft aiomolW1sessedblood I

systemcalcuJaheart t

. ........

Althwitb amtIUII0fII/1ul.tuallyure; I:

preuemonafradiol.tion 01to20r

MI: pant V ave;tamponade: "diastolic

plateau," with para.doJlical in.piratoryrUe;

COPD: m...kcdrapi-ratory nlKtualion.

Cardiac Output

Diminished cardiac output is often the result of sev-eral mechanisms operating simultaneously. By far themost common factors are decreased contracting mus-cle mass,I'.20 which may be either infarcted or isch-emic but noncontractile,21.12 and cardiac arrhyth-mias.23.24 Less frequently, specific defects of cardiacstructure diminish cardiac output, often in associa-tion with a substantial reduction in contracting mus-cle mass. These mechanical derects include mitral in-sufficiency and ventricular septal defect. In addition, awide variety of noncardiac factors, including dimin~

toPA pcw

~Vol. 295 No. 24 HEMODYNAMIC TIiERAPV OF MYOCARJ)IAL INFARCTIONc.9, 1916

:l perfor-d: acute-Hy, withnversely,Itantiallyventricu-,acardial;ht atrialrtitude ofized thatesence of

ECG

CARDIACOUTPUT

L/mlnous pres-("'infarc-

lat oftheAcute'e

the same,ht atrialthis dis-

crease in

Figure 3. Cardiac-Output Curve from a Patient with Ventricu-lar Septal Defect.

The tracing exhibits a secondary "recirculation peak," whichis typical of left-to-right shunts (see text for discussion of

mechanism).lie in left~. 2).

ished intravascular volume, increased vascular resis-tance and metabolic factors, may contribute substan-tially to diminished cardiac output in acute infarc-tion. .

The normal cardiac output curve has a smooth ex-ponential downslope. This morphology is character-.istically distorted by ventricular septal defect.15 In thisidisorder, large quantities of cooler blood that haveralready traversed the pulmonary circuit reappear a~second time at the pulmonary-artery sampling site,,producing a second, diagnostic "recirculation peak"characteristic of left-to-right shunts (Fig. 3). This re-circulation may cause' current bedside computers tomiscalculate true cardiac output grossly, and the

i curve must be calculated by hand. The result is that: output of the right side of the heart equals flow in the: left side of the heart plus shunt flow. The ratio of pul-f monary to systemic flow ("shunt ratio") may be as-lliessed by determination of the oxygen saturation of: blood from the right atrium, pulmonary artery and a. systemic artery. Output of the left side of the heart isr calculated by dividing output of the right side of thei heart by this ratio. 15

lit of sev-~y far theing mus-t or isch-

arrhyth-f cardiacI associa-:ing mus-l1li tral in-ddition, aLg dimin=

pcw

Relation of Hemodynamics to Clinical State

Although the hemodynamic changes that occurwith acute myocardial infarction are numerous, in-creased pulmonary-caPillary pressure and diminished cardiacoutput are the final common pathways for production oC vir-tually all the cliniCal signs and symptoms oC heart Cail-ure. Increased pulmonary-capillary pressure is ex-pressed clinically as the signs and symptoms oC pul-monary congestion. Thus, as the pressure increases,radiologic manifestations oC acute pulmonary conges-tion ordinarily Collow a specific pattern26 asfollows~ 18to 20 mm Hg, onset of pulmonary congestion; 20. to 25

iciency.lI;essure isd pressureis the left

rade to theand cap iI-flated, the)re clearly,this posi-

ufficiency;he diagno-lary-arterypressure.

FORRESTER E1' AL. tJ59

mm Hg, moderate congestion; 25 to 30 mm Hg, se-vere congestion; and >30 mm Hg, onset of pulmo-nary edema.

The earliest change associated with an increasein pulmonary-capillary pressure is redistribution offlOW27-)o to the upper lobes of the lung. As pressure isfurther increased, fluid passes from th~ capillaries tothe perivascular and interstitial tissue, resulting indiminished clarity of the borders of medium-sizedpulmonary vessels and the development of perihilarhate.2I,]o")2 As pressure is increased still further, fluidmoves to the perialveolar space, resulting in the ap-pearance of radiolucent grapelike clusters surroundedby radiodense fluid called the "peri-acinar rosette. "28,3\

Finally, with very high pulmonary-capillary pressure,rosettes coalesce, and other signs are accentuated, re-sulting in the radiologic appearance of acute pulmonaryedtmlJ.33

In a similar manner, the signs and symptoms ofperipheral hypoperfusion to the brain, kidney andskin are the clinical expression of reduced cardiac out-put and become increasingly severe as cardiac outputis reduced. Thus, normalized cardiac output. relatesto peripheral hypoperfusion as follows34: 2.7 to 4.3liters per minute per square meter, normal range;2.2 to 2.7, subclinical depression; 1.8 to 2.2, onsetof clinical hypoperfusion; and <1.8, cardiogenicshock.

These correlations of pulmonary-capillary pressureand cardiac index with clinical state are shown in Fig-ure 4. -rhe single value for pulmonary-capillary pres-sure and cardiac index that most precisely separatespatients with and without pulmonary congestion andperipheral hypoperfusion are 18 mm Hg and 2.2 litersper minute per square meter respectively.34 Converse-ly, on the basis of clinical criteria, one can predictwhether cardiac index and pulmonary-capillary pres-sure are abnormal in slightly more than 80 per cent ofpatients. There are two major limitations. of clinicalprediction of the hemodynamic state. The first is thatapproximately one fourth of the patients designatedclinically as not having hypoperfusion have a cardiacindex of <2.2 liters per minute per square meter.Most cases of this error occur in patients who haveclinical evidence of pulmonary congestion. The sec-ond important clinical limitation is failure to recog-nize elevation of pulmonary-capillary pressure (>18npn Hg), which occurs in approximately 15 per centof patients. Thus, clinical criteria can be used to re-flect the range of cardiac hemodynamics accurately inthe majority of cases, and the most common clinicalerrorS involve underdiagnosis of hemodynamic de-

pression.

*T 0 normalize for differences in cardiac output due to body size, cardiacoutput is divided by body-surface area (in square meters) to obtain cardiacindex. Body-surface area is determined from the hei&ht and weight of thepatient with use of published tables.

~

":r.r

THE NEW ENGLAND JOURNAL OF MEDICINE Dec. 9, 19761360

Figure ... Relation between Altered Cardiac Hemodynamicsand the Clinical Signs of Heart Failure.

The left panel shows the relation between cardiac index andperipheral hypoperfusion. and the right panel that betweenpulmonary-capillary pressure and pulmonary congestion.Solid dots indicate the presence of clinical findings of pe-ripheral hypoperfusion in the left panel, and pulmonary con-gestion in the right panel. The data suggest that signs of pe-ripheral hypoperfuslon begin to become apparent at a cardi-ac-Index level of approximately 2.2 liters per minute persquare meter and that clinical pulmonary congestion beginsto become apparent at a pulmonary-capillary pressure of ap-

proximately 18 mm Hg. Mean ~1 S.D. are shown.

Differences between Clinical and HemodynamicAssessment - How Discrepancies May Occur

Although there are a number of mechanisms bywhich the physical examination may be misleadingabout the state of cardiac function, several are note-worthy. Table 2 shows four cues, each illustratingone mechanism by which gross discrepancies be-tween the clinical and hemodynamic evaluations ofcardiac performance may occur. The fU'st patient, ex-amined soon after intravenous administration of furo-

Table 2. Discrepancies between Clinical and HemodynamicStates. .CUlac:AL BAl. CI PC, PoaIIuI MICHANIIM

"An

Modente "LV RaIea, collpRioa by 2.6 12 "PhaMI.." afterfailure" x-ray study diuretic therapy

Moderate "LV Dyspnea. diapborelia 3.1 IS Noncardiacfailure" 5.. S,

Scvere"LV Ra'" pulmoury 3.6. 21 Dec:reuoddiutolicfailure" - edema by II-ray compliaDcc

studyMild "LV S" normal by x-ray 2.2 30 ChrooIc failure with

failure" study substantial com-Jl8ll8&tioa

°L v Ieft .-rIcIo. CI (100 "'-I-frat). at. PCP, ".-ap-iIIary ,.- (100 - If8).

-

Vol. 29S

semide, had clinically moderate pulmonary conges-tion, but normal pulmonary-capillary pressure. Thedifference in clinical and hemodynamic evaluationreflected the rapid reduction in pulmonary-capiUarypressure to within the normal range after furosemide,with persistence of rales and x-ray pianges. In suchcases there may be a phase lar' or as much as 48hours between hemodynamic stabilization and resolu-tion of abnormal physical and radiologic signs. Thesecond patient, with associated chronic obstnKtiwpulmonary dis,ase, was clinically diagnosed ,as hav-ing moderate pulmonary congestion because of dysp-nea and rales. Hemodynamics, however, were withinthe normal range, indicating that the patient's physi-cal signs were extracardiac in origin. The third pa-tient, with long standing hypertension, had clinical-ly severe pulmonary congestion on examination andradiographic evidence of pulmonary edema. Hemo-dynamic measurements. revealed a high normal car-diac output and a markedly increased pulmonary-capillary pressure. The normal cardiac output withelevated pulmonary-capillary pressure in this pa-tient, therefore, was probably secondary to left ven-tricular hypertrophy and ,eduad wntricula, comfJli-tmt:e.l2-u The fourth patient, with long standing cImnt-ic compensated heart failure, was asymptomatic, butwas considered to have minimal left ventricular fail-ure on the basis of an audible third heart sound. He-modynamics, however, revealed substantial reduc-tion of cardiac output and a marked elevation in pul-monary-capillary pressure. In such a case, selective:peripheral vasoconstriction may mask the presence of ;low cardiac output, and thickening of the pulmonary !vascular walls can allow the lung to tolerate a sub- 'stantial elevation in pulmonary-capillary pressure.without clinical congestion. Thus, a substantial diJ- Iparity between clinical and hemodynamic evalua- ~tion secondary to such com/JmSalory mlClaanisms can;

CNIDIAC .L/INn'U.

occur.

A..e..ment of Left Ventricular Pump Performance

The second use of pulmonary-capillary pressureand cardiac index, beyond determining the magni-tude of pulmonary congestion and peripheral hypo-perfusion, is for assessing left ventricular pump per-formance. Figure 5 plots cardiac index VI. pulmo-nary-capillary pressure (the Starling relation) in 200patients with acute myocardial infarction at the timof admission.)4 From these graphic data, several im-portant conclusions become apparent. First of all,there is a wide degree of variability in left ventricularperformance in patients with. acute myocardial in-farction. Secondly, the level of cardiac performance atth,. tilT1f' of Msoitalization bears a direct relation to in-'h~~pi~~ ~~rt;lity~nas defined by the highly impor-jtant difference in mortality rates established by a car-Idiac index of 2.2 lite" per minute per square meter,and a pulmonary-capillary pressure of 18 rom Hg.J

I

295 No. 24 HEMODYNAMIC THERAPY OF MYOCARDIAL INFARCTION - FORRESTER ET AL.

conges- jlIre. The I'aluation :capillary ,osemide, .In such

:h as 48d resolu-~ns. The)bstructive,as hav-oC dysp-

re within:'s physi-:hird pa-clinical-

tion and. Hemo-mal car-monary-put withthis pa-I left verI-

camPli-og chron-ktic, butular fail-ilnd. He"lJ reduc-n in pul-selectiveesence of

Llmonary:e a sub-pressureI1tial dis-: evalua-urns can

Lastly, approximately one third of the patients hadnormal levels of left ventricular performance duringthe time they were having an acute myocardial in-farction. Thus, the first important principle in the ap-proach to therapy of disordered cardiac function: acutemyocardial infarction cannot be viewed as a single functionaldisorder; rather it is a series of subsets.

To this end, a number of classification methods forpatients with acute myocardial infarction have beenproposed. Two basic approaches have been used.35-4\The first, of which the best known is probably themethod of Killip, defines clinically relevant syndromes(such as pulmonary edema and cardiogenic shock),whereas the second method uses multiple clinical orhemodynamic criteria, appropriately weighted, toestablish a numerical probability of survival. Bothapproaches are useful for establishing prognosis ingroups of patients, but neither method relates theclinical presentation to its specific hemodynamic de-terminants, nor are they directly based on the level ofcardiac performance. These methods, therefore, haveonly limited relevance to therapeutic decision mak-

ing.A more broadly relevant subset classification relates

the clinical presentation to specific hemodynamic de-

pressure~ magni-'al hypo-lmp per-.pulmo-1) in 200the time..eral im-It of all,ntricularrdial in-:nance ation to in-if impor-by a car-re metermm Hg.

Table 3. Hemodynamic Basis of Clinical Presentation.

~

a.u1S INCUAUD PuLMONARY- DtIcuAIIDCAI'ILLA&Y Pussuu CuDw: IMDU

Subjective Dyspnea

Objective Ra1eI, x-ray evidence

of conpltion

1361

terminants of left ventricular pump performance (Ta-ble 3). Clinical subsets are defined by the manifesta-tions of pulmonary congestion (reflecting increasedpulmonary-capillary pressure) and peripheral hypo-perfusion (reflecting decreased cardiac index).].4 Onthe basis of these findings, four clinical s'!bsets (C) aredefined: CI, no pulmonary congestion or peripheralhypoperfusion; CII, pulmonary congestion withouthypoperfusion; CIII, peripheral hypoperfusion with-out congestion; and CIV, both hypoperfusion andcongestion. These clinical subsets are of substantialvalue in assessing short-term prognosis. Althoughclinically uncomplicated patients virtually always sur-vive hospitalization (1 per .cent mortality), mortalityincreases to approximately 10 per cent when pulmo-nary congestion is present and to 60 per cent whenboth pulmonary congestion and peripheral hypoper-fusion are detectable.

Since pulmonary congestion is due to increased pul-monary-capillary pressure, and peripheral hypoper-fusion to depressed cardiac index, four comparable he-

Table 4. Mortality Rates In Clinical and Hemodynamic Sub-sets.

Suan ~...v I'UlPlII!aAL " MORTALITYCOI'IOaTlO"- HVPOPlUUSIoNt

a.JIQC4L H8I1ODYN4MlC

39

2351

.II.160

modynamic subsets are established by the specific lev-els of cardiac index and pulmonary-capillary pres-sure described earlier in the discussion of clinical andhemodynamic correlations. The comparable hemo-dynamic subsets (H» therefore, are defined as fol-lows: HI, pulmonary-capillary pressure <18 mm Hg,and cardiac index >2.2 liters per minute per squaremeter; HII, pulmonary-capillary pressure >18 mmHg, and cardiac index >2.2 liters per minute persquare meter; HIlI, pulmonary-capillarypressure<18 mmHg, and cardiac index <2.2 liters per min-ute per square meter; and HIV, pulmonary-capillarypressure >18 mm Hg, and cardiac index <2.2 litersper minute per square meter. Clinical subset classifi-cation correctly predicts into which hemodynamicsubset a patient will fall in approximately 70 per centof the cases. Because of this high correlation, the mor-tality rates in the four hemodynamic subsets are com-parable to those in clinical subsets (Table 4). Moststriking is the fact that the presence of peripheral hy-poperfusion, diagnosed either by clinical evaluation orby a cardiac index <2.2 liters per minute per squaremeter, is associated with an increase of approximate-ly 10 times in mortality rate.

Obtundation, fatigue

~~

'IHENEW1362

REFEUNCES

I. Swan HJC. Ganz W. Forrester JS, et al: Catheterization ofthe heart inman with the use of a flow-directed balloon-tipped catheter. N Engl JMed 283:447-451, 1970

2. Forrester'JS. Ganz W. Diamond G, et al: Thermodilution cardiac out-put determination with a single now-directed catheter. Am Heart J83:306-311, 1972

3. Ganz W, Donoso R, Marcus HS, et al: A new technique for measure-ment of cardiac output by thermodilution in man. Am J Cardiol 27:392-396. 1971

4. Rahimtoola SH, Loeb HS, Ehsani A. et al: Relationship of pulmonaryartery to left ventricular diastolic pressures in acute myocardial infarc-tion. Circulation 46:283-290. 1972

5. Falicov RE, Resnekov L: Relationship of the pulmonary artery end-diastolic presaure to the left ventricular end-diastolic and mean tillingpressures in patients with and without left ventricular dysfunction. Cir-culation 42:65-73, 1970

6. Bouchard RJ, Gault JH, Ross J Jr: Evaluation of pulmonary arterialend-diastolic pressure as an estimate of left ventricular end-diastolicpressure in patients with normal and abnormal left ventricular perfor-mance. Circulation 44:1072-1079, 1971Jenkins BS, Brandley RD, Branthwaite MA: Evaluation of pulmonaryarterial end-diastolic pressure as an indirect estimate of left atrial meanpresaure. Circulation 42:75-78, 1970Lipp H, Gambetta M, Schwartz J, et al: Intermittent pansystolic mur-mur and presumed mitral regurgitation after acute myocardial infarc-tion. Am J Cardiol 30:690-694, 1972Heikkili J: Mitral incompetence complicating acute myocardial infarc-lion. 8r Heart J 29: 162-169. 1967Forrester JS, Diamond G. Freedman S. et al: Silent mitral insufficiencyin acute myocardial infarction. Circulation 44:877-883, 1971Swan HJC, Forrester JS, Diamond G, et al: Hemodynamic spectrum ofmyocardial infarction and cardiogenic shock: a conceptual model. Cir-culation 45: 1097-1 110. 1972Diamond G. Forrester JS: Effect of coronary artery disease and acutemyocardial infarction on left ventricular compliance in man. Circula-tion 45:11-19, 1972Smith M, Ratshin RA, Russell RO Jr, et al: Early consecutive leftventricular compliance changes after acute myocardial infarction. Am JCardiol 31:158. 1973Hood WB Jr, Bianco JA, Kumar P, et al: Experimental myocardial in-farction. IV. Reduction of left ventricular compliance in the healingphase. J Clin Invest 49:1316-1323, 1970Regan TJ, Passannante AJ, Khan MI, et al: Influence of scar on leftventricular performance at the onset of myocardial ischemia: shockversus heart failure. J Clin Invest 50:534-542, 1971Forrester JS. Diamond G. McHugh TJ, et 81: Fillin, pressures in theright and left sides of the heart in acute myocardial infarction: a reap-praisal of central-venous pressure monilorin,. N Eng! J Med 285: 190-193. 1971Diamond G, Marcus H, McHugh T, et al: Catheterization of the leftventricle in acutely ill patients. 8r Heart J 33:489-493, 1971Forrester JS, Diamond G. Swan HJC: Bedside diagnosis of latent car-diac complications in acutely ill patients. JAMA 30:338-342, 1972Pale DL, Caultield J8, Kastor JA. et al: Myocardial chanles associatedwith cardiogenic shock. N Eng! J Med 285: 133-137, 1971

1.

8.

U.

13.

,4.

IS.

16.

fvol. 295 ]IDec. 9, 1916ENGLAND jOURNAL OF MED.CINE

Proa Car-ZO.

21.

24.

15.

CADETER]

26.

r7.AS the

~ dosagent enthe wodlthe agen

In Ire!inf1uenz~per year1918 hacIy. The 1was the

Althouglyears of

epidemi,was 151900.3 1\

. high inf world c,l status 0i mining j

I that ofnyears of

impress'i nourish.I studies,

f showed~ strain;

A/swin.Epid~

cheobro1918 pntients ...early; iIafter hedilatedwas an

quentlybrane 1

21.

'..

30.

34.

..40.

41.

From ItNY, andStony 8rcvice, YetC