ventricular septal defect' incidence, morbidity, and...

British Heart Journal, 197I, 33, Supplement, 8I-87.

Ventricular septal defect'Incidence, morbidity, and mortality in variousage groups

J. D. Keith, V. Rose, G. Collins, and B. S. L. KiddFrom The Hospital for Sick Children, 555 University Avenue,Toronto 2, Ontario, Canada

The true incidence of isolated ventricular sep-tal defect and its subdivisions is difficult toestimate accurately, since the data availablein the past have been obtained in a variety ofincomplete ways. Necropsy studies usuallyincluded the complicated as well as the iso-lated cases. The clinical diagnostic criteriawere less accurate than those used currently.Some studies dealt only with adults; otherreports omitted infants or took patients intothe investigation series at different ages.

It is reasonable to conclude that no finalfigures will ever be arrived at, since the factorsaffecting prevalence data are changing con-tinuously and no two studies are identical orcomplete. However, it is worth while con-sidering some of the information obtained inthe past from certain specific points of view.

Pathological dataTable i presents four references on the inci-dence of ventricular septal defect at necropsy.

TABLE I Ventricular septal defect: incidenceat necropsy

Authors Rate per IooOnecropsies

Gelfman and Levine (I942) I-0Selzer (I949) 0-40

o-65Kaplan et al. (I96I) 0-57Bloomfield (I964) 0-25

Since the majority of infants with ventricu-lar septal defect survive, and many have thedefect closed spontaneously, it is obvious thatthe necropsy data do not give a true pictureof the prevalence of this anomaly in the popu-lation as a whole. It is clearly important toestimate incidence very early in life to arriveat a satisfactory figure. Two referencesattempt to achieve this (see Table 2).1 These investigations were supported by a grant fromthe Ontario Heart Foundation.

TABLE 2 Ventricular septal defect: incidencein live births

Authors Rate per IOOO

Hoffman and Rudolph (I965) iRose and Keith (1970) I.37

There are a few factors that might operatefrom time to time to increase the incidence ofventricular septal defect, such as rubella. Asfar as we can tell from the statistics in theToronto Heart Registry, the prevalence ofventricular septal defects remains about thesame from year to year and rubella or otherinfections have a minimal or minor role. Thetrue incidence, therefore, appears to be some-where between I/700 and i/Iooo live birthsas shown in Table 2.

Ventricular septal defect is the most com-mon heart anomaly in childhood, comprisingapproximately 25 per cent of all congenitalheart disease as an isolated defect, but it occursin 50 per cent of congenital heart disease if allcomplex forms are included. Among infantsdying with congenital heart disease in the firstmonth of life, ventricular septal defect issecond only to transposition of the greatarteries as a cause of death in this age group(Keith, Rowe, and Vlad, I967; Keith et al.,I970).

Present studyThe present study is an attempt to arrive ata clearer recognition of the true prevalence ofventricular septal defect and the various re-sponses to it of the human body.Our records now include approximately

I500 cases of infants, children, and adultswith isolated ventricular septal defect. Thegroup that offers the best appraisal of thecourse followed by these children is that whichstarts in the first year of life and in whichfollow-up has occurred over many years. Wehave 630in this category. Some 3i8 were diag-

on 8 Septem

ber 2018 by guest. Protected by copyright.

http://heart.bmj.com

/B

r Heart J: first published as 10.1136/hrt.33.S

uppl.81 on 1 January 1971. Dow

nloaded from

http://heart.bmj.com/

82 Keith, Rose, Collins, and Kidd

nosed clinically on the basis of a well-heardblowing systolic murmur over the lower prae-cordium without a thrill, and with the passageof time the presence of aortic stenosis, pul-monary stenosis, or patent ductus arteriosuswas ruled out. Those in whom there appearedto be any doubt about the diagnosis werecatheterized. Thus we have 3I8 diagnosedclinically and an additional 3I2 diagnosed bycatheterization, giving a total of 630 seen firstin the first year of life.

In an attempt to evaluate our experiencemost profitably we have divided our patientsinto 4 groups:

(A) A group of 200 infants catheterized in thefirst year of life and in the majority of casesrecatheterized some years later.(B) A group of II2 that were catheterized atvarious ages for diagnostic purposes but allseen initially in the first year of life.(C) A group of 3I8 infants seen first in thefirst year of life whose findings satisfied thecriteria for diagnosis ofventricular septaldefectas listed above but on whom no catheterstudies were performed. This group con-sisted chiefly of small defects for which de-tailed studies did not appear to be necessary.From our experience over the years withmanysimilar small defects investigated in the pastin complete detail, we have arrived at a con-clusion that the diagnosis can be substantiatedclinically with at least 95 per cent accuracy.We feel that it is important to include thisbenign group, since without it a very lop-sided picture of the severity and incidence ofthe complications would be obtained. Thus,when groups A, B, and C are put togetherwe have 630 cases of isolated ventricular sep-tal defect seen first in the first year of life. Todate they have now been followed an averageof 7-5 years.(D) A group of295 cases of isolated ventricu-lar septal defect, now in the teenage or adultcategory, is being followed. Initially they wereall seen in infancy or childhood at The TorontoHospital for Sick Children. All in this groupwere born before I955, and have a follow-upof I6 to 42 years.

Group AThe 200 infants catheterized in the first yearof life were divided into 6 haemodynamicgroups according to the classification of Kiddet al. (I965). The groups are as follows: (I)pulmonary to systemic flow ratio of less than2: I and a normal pulmonary vascular resist-ance; (II) pulmonary to systemic flow ratioof greater than 2: I with a normal pulmonaryvascular resistance; (III) pulmonary to sys-

temic flow ratio greater than 2: I with slightincreased resistance; (IV) pulmonary to sys-temic flow ratio greater than 2: I with amoderate increase in resistance; (V) pulmon-ary to systemic flow ratio of less than 2: I withhigh pulmonary vascular resistance but noreversal of flow; and (VI) pulmonary to sys-temic flow ratio of less than 2: I with highpulmonary vascular resistance and reversalof flow through the defect (Eisenmengersyndrome).The haemodynamic groups at the start of

observation in the first year of life and atfollow-up are shown in Table 3 (for group Aonly).

TABLE 3 Group A: infant catheter studiesin ventricular septal defect

Haemo- Total Haemodynamic group Otherdynamic cases at follow-upgroup atstart I II III IV V VI

I 43 39 4 0 0 0 0 0II 64 33 I6 1 2 2 0 10III 46 I4 4 7 6 0 3 12IV 36 8 5 6 5 I 2 9V II 7 I I 0 0 0 2

Total 200 IOI 30 15 13 3 5 33

Initial catheter first year of life.

Haemodynamic group A I Ninety percent of those that started in group A I re-mained in group A I - none developed pul-monary infundibular stenosis, none died,none required surgery, and there was no in-crease in pulmonary vascular resistance. Thisis obviously a very benign group comparedwith the others. At the end of this follow-upin the total group of 200 cases there were ioithat eventually fell into group A I, indicatingthe recognized tendency of a ventricular sep-tal defect to decrease in size and decrease theflow through it (see Table 3).

Haemodynamic group AII There were64 cases initially in this category, but on fol-low-up only 25 per cent persisted in group II;5i per cent became group I, and the rest werescattered over the other categories as shownin Table 3. Five, or 9 per cent, developed pul-monary infundibular stenosis, 17 per centdied, 23 per cent required surgery, and 5 percent had a conspicuous increase in pulmonaryvascular resistance. It is of special interestthat about half of these cases ended up ingroup A I after starting in group A II.The relatively high incidence of deaths and

surgery indicates the presence of intractableheart failure that one associates with a large

on 8 Septem



/B



nloaded from


Ventricular septal defect 83

defect. There is always the problem of howlong one should treat these infants medicallyin the hope that the failure will be controlledor the defect will get smaller, or both. Therelatively high incidence of deaths and surgeryin groups II, III, and IV indicated the lackof conspicuous success in medical or surgicaltreatment in many instances.

Haemodynamic group A m The casesseen initially in group A III have a relativelyserious prognosis. Only 30 per cent decreasedin size to end up in group A I. Three cases,or 6 per cent, developed pulmonary infun-dibular stenosis, 20 per cent died, 43 per centrequired surgery, and 20 per cent ultimatelydeveloped a conspicuous increase inpulmonaryvascular resistance.

Haemodynamic group A IV This groupalso has a serious prognosis. Only 22 per cent'decreased to group I, 25 per cent died, 50 percent had surgery related to the heart, and 22per cent developed a high pulmonary vascularresistance.

Haemodynamic group A V This groupappears to be relatively benign. Seven of theii cases ended up in group I. One case diedand one required surgery. All had a declineof the pulmonary vascular resistance to nor-mal levels. This group therefore appears tohave persistence of the foetal pulmonary vas-cular response after birth which lasts forseveral months but declines in the latter halfof the first year of life. We think that one canreasonably conclude that these are eithersmall defects to begin with, or become so inthe first year or two of life, and thereforeusually have a favourable response in spiteof the initial high resistance of the lung fields.

Haemodynamic group A VI (see Table 3)These cases were obviously not in group VIin the first year of life, since it takes severalyears before fully developed pulmonary vas-cular disease occurs in isolated ventricularseptal defect. It may occur more rapidly intransposition of the great vessels than in ven-tricular septal defect. However, at the end of7 years, 5 cases could legitimately be put inthe category of Eisenmenger syndrome withreversal of flow through the defect.

Other The cases listed under the headingof 'Other' in Table 3 are those in which itwas not possible to do a second catheteriza-tion either because they were operated on ortreated medically and died or, in a few cases,were lost to follow-up.7

Deaths in group A These occurred almostentirely in haemodynamic groups II, III, andIV, and were due to heart failure, or failureplus respiratory infection, orunsuccessful sur-gery whether this was of direct closure type orbanding of the pulmonary artery to controlexcessive pulmonary blood flow. Group Acases had more severe or larger defects thanthe other groups, hence the higher mortality(see Table 4, and Table 8 which shows theoverall mortality of the whole group of 630children).

TABLE 4 Infant catheter study: deaths byhaemodynamic group in group A (200 infantswith repeat catheter studies)

Group No. in group Deaths

I 43 0II 64 II (I7%)III 46 9 (20%)IV 36 9 (25%)V II I (9%)

Total 200 30 (I5%)

Development ofinfundibular pulmonarystenosisSince Gasul, Dillon, and Vrla in I957 firstdescribed the development of infundibularstenosis in a child who had previously had anisolated ventricular septal defect, a number ofinvestigators have attempted to determine thetrue incidence of this developing complica-tion. As we have shown, it appears to varywith the haemodynamic group concerned. Itdid not occur in haemodynamic groups I, V,or VI. It did occur in 6 to ii per cent of thecases in groups A II, A III, and A IV (thosethat had a flow ratio greater than 2: I withor without some increase in resistance). Thegradient at the time of first catheter was mini-mal or absent in the I3 cases recorded here.At the end of the follow-up period thegradient varied from 15 to 8o mm. Hg. Twocases by that time were classified as typicaltetrad of Fallot and were operated on. Theinitial gradient in each had been less than I0mm. Hg.

Recently Tyrrell et al. have presented dataon 22 cases at The Toronto Hospital for SickChildren which initially had gradients acrossthe pulmonary valve of less than 20 mm. Hg,and all increased significantly with follow-up(Tyrrell et al., I970). Two were identified asdue to anomalous right ventricular musclebundle and the others had the usual type ofinfundibular stenosis in the outflow tract.

on 8 Septem



/B



nloaded from



Pulmonary vascular resistanceOne of the long-term complications of theventricular septal defect is the developmentof progressive increase in pulmonary vascularresistance until there is reversal of flowthrough the defect and a resultant drop in thearterial oxygen saturation. In time cyanosisappears and the full-blown picture of theEisenmenger syndrome is complete. Thenatural history of such cases leads to death,usually in the teens or twenties, but somesurvive to the thirties or forties.The data indicated above reveal that such

cases do not come from haemodynamic groupI but evolve from group II, III, or IV, even-tually presenting with the flow and resistancepatterns seen in group A VI. Infants ingroup A V in the first year of life with isolatedventricular septal defect appear to resolvetheir raised pulmonary resistance and end upwith a benign prognosis. Any child with a flowgreater than 2: 1 in early life should be seenby a physician at regular intervals, repeatingthe radiograph and electrocardiogram ateach visit. If there is increasing right ventricu-lar hypertrophy in the electrocardiogram anddecreasing size of heart in the radiograph thecardiac catheter study should be repeated.Early operation is imperative in many suchcases before the pulmonary vascular changesbecome irreversible. Most children with 2: Iflow ratio and rising resistance will respondfavourably if operated on by 2 years, eitherby direct closure or banding of the pulmonaryartery. If it is left until 5 years there will besome who are certain to progress to fataloutcome in spite of adequate surgery.For those who have a low pulmonary resist-

ance and increased flow in childhood, opera-tion can be postponed until it becomes clearwhether the defect is getting smaller, but if arelatively high flow continues such cases mayultimately develop pulmonary vascular dis-ease at a later age, as happens in some casesof atrial septal defect in the second, third, orfourth decade. Direct closure is thereforeindicated between 5 and I0 years of age ifflow ratio is persistently greater than 2: 1.

Combined infant groups A, B, and C atfollow-upAs indicated earlier, we have 630 infants seenin the first year of life and followed up foran average of 7.5 years or more in categoriesA, B, and C.A summary of the haemodynamic groups

in the 630 cases at the most recent follow-upis given in Table 5.There are 3 main divisions, those that had

TABLE 5 Haemodynamic groups: categoriesA, B, and C at latest follow-up

Haemo- Catheter Catheter Clinical Totalsdynamic and alone evalua-group surgery tions

I I I43 292 437 (71%)II 21 42 25 88 (I4%)III 64 II I 75 (I2%)IV 15 9 0 24 (4%)V 2 2 0 4 (o-6%)VI I I 0 2 (0o3%)

Total I04 208 318 630

both catheter studies and surgery (the mostserious cases), those that had catheter studiesbut surgery was not carried out (an inter-mediate group), and finally the most benigncategory - those in whom no catheter studyor operation was carried out.As might be expected, those operated on

were chiefly in haemodynamic groups II toIV (high flow), whereas those who had no sur-gery were chiefly in haemodynamic group I(low flow), although a few were found to bein group II (high flow, low resistance).

Another striking difference between thethree categories was the incidence of spon-taneous closure of the defect. There was ofcourse no spontaneous closure in the surgicalgroup. In the non-surgical catheter studygroup ii per cent, and in the clinical group26 per cent closed spontaneously.

Spontaneous closureThere are a variety of figures given in theliterature of the last 2 to I2 years on the inci-dence of spontaneous closure ofthe ventricularseptal defect. Four of these are shown inTable 6.

TABLE 6 Spontaneous closure of ventricularseptal defect

Authors Percentage closing

Keith, Rowe, and Viad (I958) I0Hoffman and Rudolph (I965) 24Ash (1964) I5Li et al. (I969) 22

The present report summarizes 2 groups:(i) The 630 cases followed from infancy for

an average of 7.5 years with an overall closurerate of I7 per cent (Table 7). The smalldefects identified clinically had the highestclosure rate (26%).

(2) Among the 295 cases followed throughthe teens into the twenties 4 per centclosed spontaneously (see Table io). It will

on 8 Septem



/B



nloaded from



TABLE 7 Spontaneous closure

Group No. of No. of spontaneouscases closures

Catheter and surgery I04 0Catheter study only 208 22 (II %)Clinical 318 82 (26%)

Total 630 I04 (I7%)

be necessary to follow the infant groupthrough the teens and re-evaluate the adultgroup from time to time to arrive at a more

final figure. However, it is quite suitable forthe moment to add the 17 per cent closure inthe first 7 years of life to the 4 per cent of thepost-childhood group, giving a total of 2I percent. This would appear to be a minimumfigure which will rise with the passage oftime the longer a group of patients is followed.

Perhaps equally important is to recognizeearly in life which cases are likely to close oftheir own accord, particularly in the first yearor two of life. It is obvious that this ismost likely to occur in the smaller defects,since approximately 30 per cent of the cases

identified as haemodynamic group I by cathe-ter study or clinical examinations have sub-sequently closed and the murmur disappeared.There is a good anatomical experimental

basis for expecting 25 to 30 per cent to closespontaneously, since of our cases (95) comingto necropsy with ventricular septal defect 27per cent had a defect 5 mm. or less. Further-more, it has been shown experimentally indogs that a hole 5 mm. or less created in theseptum will gradually become obliterated ofits own accord over a few months.

DeathsIn 630 infants followed from the first year oflife for an average of 7.5 years the deathsamounted to 6-9 per cent (Table 8). Therewere no deaths in the very small ventricularseptal defects diagnosed by clinical evaluadonand 7 per cent mortality in those that had a

catheter study only and did not require sur-

gery. However, among those that did requiresurgery the mortality was 26 per cent. Themajority of these were in intractable failureand required either a banding operation or an

attempt at direct closure. In either case therisk is high in the first 6 months of life.

Apart from the cases that died during or

immediately after surgery we have not hadany deaths in recent years in the children over

2 years of age with isolated ventricular septaldefect due to the defect itself until they gotinto the adolescent age group. In the infantage group the deaths were almost entirely

TABLE 8 Deaths: average duration offollow-up 7 years

Group No. of Deathscases

Clinical evaluation 318 (o)Catheter study only 208 I5 (7%)Catheter and surgery I04 27 (26%)

Total 630 42 (6 9%)

limited to those with a large pulmonary bloodflow ( > 2: i) and normal or moderately raisedpulmonary vascular resistance. Deaths fromthe Eisenmenger syndrome rarely occur beforethe age of I0 years.

In the 295 cases born before I955 andfollowed until the present time i0, or 3.4 percent, have died in the late teens or twenties.Most of these deaths have been due to pro-gressive pulmonary vascular disease, but twoof them were associated with attempts atcorrective surgery.

Bacterial endocarditisIt is very difficult to work out the true in-cidence of bacterial endocarditis in ventricularseptal defect. If one turns to earlier data, inI942 Gelfman and Levine found that of 3Icases of ventricular septal defect at necropsy,43 per cent had bacterial endocarditis. Theseobviously occurred in the preantibiotic era.More recent data are available which can

be related to the patient years of the groupfollowed over a period of time (Table 9).

TABLE 9 Bacterial endocarditis in ventricularseptal defect

Authors Patient years Subacutebacterialendocarditis

Wood (Bloomfield, I964) 638 0Griffiths et al. (I964) 837 4Bloomfield (I964) 222 IWalker et al. (I965) I407 6

Total 3I04 II

These groups taken together indicate I case ofbacterial endocarditis in 282 patient years.

Shah et al. in i966 reviewed the patientmaterial at The Hospital for Sick Children andpresented further data on the incidence ofbacterial endocarditis in ventricular septal de-fect. In the 3 to I2 year age group coveringIo years the frequency of bacterial infectionwas 2 per cent during that interval (or i in470 patient years).

Since then we have reviewed 295 cases of

on 8 Septem



/B



nloaded from



isolated ventricular septal defect followed dur-ing adolescence and adult life, covering anaverage period of i0 years from I3 to 22 years.The incidence of bacterial endocarditis was Iper cent (3 cases) or i in 787 patient years.Thus it is clear that bacterial endocarditis

is an infrequent occurrence in early adult life,and furthermore when children or youngadults do contract this infection the prognosisis usually good. All the cases with ventricularseptal defect and bacterial endocarditis wehave seen in the past IO years have recoveredwith medical treatment.

Teenage and adult group: group DTable IO shows the haemodynamic categoriesof the teenage and adult group of 295 casesall born before 1955 and seen at some time inchildhood at The Hospital for Sick Children.

TABLE Io Teenage and adult group (bornbefore I955)

Group Haemodynamic group Totals

I II III IV V VI

CHnical I55 I 0 0 0 2 I58Catheter 72 2 I I 0 23 99Catheterandsurgery 8 I2 4 I0 I 3 38

Total 235 I5 5 II I 28 295Percentage 8o 5 I-7 3.7 0-2 9-4 IOO

4 per cent closed spontaneously. 3.4 per cent died(all in teens or twenties).

Incidence of bacterial endocarditis: age 3-I3 years,2 per cent; I4-22 years, I per cent.

The decade or two in which these indi-viduals came as children to the Cardiac Clinicat The Hospital for Sick Children before I955produced a natural selection by our latefollow-up that has eliminated those who diedin the first two or three years of life, and italso eliminated those whose defects closedspontaneously in the first five or ten years oflife, but it has included those whose defectshad closed spontaneously since the age of I2.We were therefore left with the data shown

in Table I0, and I58 were evaluated on clinicalexamination alone, 99 had catheterization, and38 had catheterization plus surgery, for a totalof 295. If one looks at the total one sees that80 per cent are now found in category I andhave small defects with low flow and low pul-monary vascular resistance. Five per cent arein group II, I-7 per cent in group III, 3.7 percent in group IV, 0o2 per cent in group V,and 9 4 per cent in group VI. This latter group

is composed of those who developed a highpulmonary vascular resistance with reversalof flow through the defect and they may there-fore be classified as having the Eisenmengersyndrome. The outlook for them is a verylimited one, and according to the data ofClarkson et al. (I968), one-third to one-halfwill die in the next i0 years.Thus in the teenage or adult group 80 per

cent have small defects and are leading normallives, fully active, with no functional dis-ability. It is worth re-emphasizing that 4 percent have closed spontaneously and only i percent have developed bacterial endocarditis.

Comparison of all groupsTable ii compares the various groups dis-cussed in this presentation.

TABLE II All ventricular septal defectgroups combined

Haemo- Groups A, Groups A, Group Ddynamic B, and C at B, and C at teen-adultgroup iyear (630) 75years (630) (295)

I 50°0% 7I-o% 8o*o%II 26.5% I4'0% 5-°%III i6-o% I2-0% 1P7%IV 6-o% 4 0% 3-7%V I.5% o-6% 0o2%VI oo% 0°3% 9'4%

Comparing children or infants seen in thefirst year of life with the same group an aver-age of 7-5 years later, there is a moderatelyhigh incidence ofhaemodynamic group I withsmall defects in the first year of life, but thepercentage increases at 7.5 years from 50 percent to 7I per cent. Haemodynamic group IIdeclines from 26 per cent to 14 per cent,group III declines from i6 per cent to 12 percent, group IV from 6 per cent to 4 per cent,and group V from I-5 per cent to o-6 per cent.There were no cases, of course, with devel-oped pulmonary vascular disease in the firstyear of life, but there is a small percentage,0o3 per cent, by 7-5 years and nearly I0 percent in the adult group. With present investi-gative and therapeutic measures, medical andsurgical, this latter group will usually beprevented in the future by early banding oftl-e pulmonary artery or direct closure of thedefect. There will always be a few who escapemedical attention and will be found to havethis complication at a later date when thechanges are irreversible.Group D cannot be compared with

groups A, B, and C adequately, but it can beconsidered as a fair representation of a long

on 8 Septem



/B



nloaded from



follow-up beginning I5-20 years ago. Thereis obviously a large number of young adultswith small ventricular septal defects who areleading a normal life and have very little riskof developing bacterial endocarditis. Theystill have murmurs, but not as well heard asin childhood. Further study is needed todetermine their ultimate course, but it seemslikely that a significant number will continueto close spontaneously with the passage oftime or remain patent but lead a normal life.

ReferencesAsh, R. (I964). Natural history of ventricular septal

defects in childhood lesions with predominantarteriovenous shunts. Journal of Pediatrics, 64, 45.

Bloomfield, D. K. (I964). The natural history ofventricular septal defect in patients surviving in-fancy. Circulation, 29, 914.

Clarkson, P. M., Frye, R. L., DuShane, J. W.,Burchell, H. B., Wood, E. H., and Weidman, W.H. (I968). Prognosis for patients with ventricularseptal defect and severe pulmonary vascular ob-structive disease. Circulation, 38, I29.

Gasul, B. M., Dillon, R. F., and Vrla, V. (1957). Fur-ther observations of the natural course of ventricu-lar septal defect: new clinical and physiologic data(Abstr.). Circulation, I6, 885.

Gelfman, R., and Levine, S. A. (I942). The incidenceof acute and subacute bacterial endocarditis in con-

genital heart disease. American Journal of theMedical Sciences, 204, 324.

Griffiths, S. P., Blumenthal, S., Jameson, A. G., Ellis,K., Morgan, B. C., and Malm, J. R. (I964). Ven-tricular septal defect: survival in adult life. Ameri-can Journal of Medicine, 37, 23.

Hoffman, J. I. E., and Rudolph, A. M. (I965). Thenatural history of ventricular septal defects in in-fancy. American journal of Cardiology, I6, 634.

Kaplan, S., Daoud, G. I., Glass, I. H., Shemtob, A.,and McGuire, J. (I96I). Natural course of ventricu-lar septal defect (Abstr.). Circulation, 24, 968.

Keith, J. D., Collins, G., Rose, V., and Kidd, L. (1970)Course and prognosis in ventricular septal defect.In Symposium on Natural History of CongenitalHeart Disease, Toronto, December z969. Charles C.Thomas, Springfield, Illinois. In Press.

, Rowe, R. D., and Vlad, P. (Eds.) (I958). HeartDisease in Infancy and Childhood, p. 233, Macmillan,New York.

, , and -- (I967). Heart Disease in Infancyand Childhood, 2nd ed., p. 4. Macmillan, New York.

Kidd, L., Rose, V., Collins, G., and Keith, J. (I965).The hemodynamics in ventricular septal defect inchildhood. American Heart J7ournal, 70, 732.

Li, M. D., Colfins, G., Disenhouse, R., and KeithJ. D. (I969). Spontaneous closure of ventricularseptal defect. Canadian Medical Association Jour-nal, 100, 737.

Rose, V., and Keith, J. D. (I970). Toronto HeartRegistry I964. To be published.

Selzer, A. (I949). Defect of the ventricular septum:summary of twelve cases and review of the litera-ture. Archives of Internal Medicine, 84, 798.

Shah, P., Singh, W. S. A., Rose, V., and Keith, J. D.(I966). Incidence of bacterial endocarditis in ven-tricular septal defects. Circulation, 34, I27.

Tyrrell, M. et al. (I970). To be published.Walker, W. J., Garcia-Gonzalez, E., Hall, R. J.,

Czarnecki, S. W., Franklin, R. B., Das, S. K., andCheitlin, M. D. (I965). Interventricular septaldefect: analysis of 4I5 catheterized cases, go withserial hemodynamic studies. Circulation, 31, 54.

Wood, P. Posthumous papers. (Quoted by Bloomfield,D. K. (I964). The natural history of ventricularseptal defect in patients surviving infancy. Circula-tion, 29, 914.)

on 8 Septem



/B



nloaded from


ventricular septal defect' incidence, morbidity, and...

Documents