OUP user menu

Outcome of patients after surgical closure of ventricular septal defect at young age: longitudinal follow-up of 22–34 years

J.W Roos-Hesselink, F.J Meijboom, S.E.C Spitaels, R van Domburg, E.H.M van Rijen, E.M.W.J Utens, A.J.J.C Bogers, M.L Simoons
DOI: http://dx.doi.org/10.1016/j.ehj.2004.04.012 1057-1062 First published online: 2 June 2004


Background Long-term survival and clinical outcome after surgical closure of a VSD is poorly documented. Such data are important for the future perspectives, medical care, employability, and insurability of these patients.

Methods 176 consecutive patients underwent surgical closure of an isolated VSD between 1968 and 1980 in our hospital. A systematic follow-up study was performed in 1990 and again in 2001.

Findings Late survival was poorer than in the general population. Pulmonary hypertension and right ventricular hypertrophy were present in the 4 patients who died suddenly, late after operation. During follow-up no new pulmonary hypertension became manifest. Re-operations were necessary in 6%. Some patients (4%) developed sinus node disease late after repair, requiring pacemaker implantation. At last follow-up (91 survivors) 92% of the patients were in NYHA class I. Pulmonary hypertension was found in 4%, and aorta insufficiency in 16%. Patients experienced difficulties when applying for insurance.

Conclusion Among patients with surgically repaired VSDs, late results were good, although some late sudden deaths occurred in the patients with pulmonary hypertension. Furthermore, some patients developed sinus node disease late after repair, requiring pacemaker implantation. Employability is good, but pregnancy and insurance matters need further attention.

  • VSD
  • Heart defect
  • Congenital
  • Survival
  • Follow-up study


Isolated VSD (VSD) is by far the most common congenital heart defect, and surgical closure of a VSD is the most common open-heart procedure performed in paediatric cardiac surgery.1 Apart from a high peri-operative mortality, patients operated on in the 1950s and 1960s showed a higher-than-normal prevalence of sudden death and serious arrhythmia during follow-up.2 Since then, significant changes have taken place such as early correction of a large VSD to avoid pulmonary vascular disease and broader indications for closure in patients with moderate shunts. Furthermore, many improvements in operative and peri-operative care were achieved. It has been postulated that patients with surgically closed VSDs in the1970s do well, with a normal life expectancy. Most adult patients have been discharged from routine cardiological follow-up.3,4 Partly as a result of this policy, data on late survival, sequelae such as arrhythmias, pulmonary hypertension and aortic insufficiency and employability and insurability is sparse.5–10

This report describes the long-term survival and clinical course in a single centre cohort of 176 consecutive patients operated upon in our institution between 1968 and 1980 at young age, with 22–34 years follow-up. Changes in ECG, exercise capacity and echocardiographic parameters were examined. Employability and insurability issues were addressed.


Patient population

All 176 patients with surgical repair of an isolated VSD at our institution between 1968 and 1980 at young age (Math15 years) were included in this study. The year 1968 was chosen as the starting point because, in this year, cardiac surgery was started in our centre. The details of pre-operative clinical findings, including cardiac catheterisation, localisation and type of VSD, data on surgical technique and post-operative course have been described previously and are summarised in Table 1.4

View this table:
Table 1

Baseline characteristics of 176 consecutive patients who underwent VSD closure between 1968 and 1980

Total group19902001P-value
Number of patients17610995
Age at operation (years)4 (0–13)4 (0–13)4 (0–13)0.9
Age Embedded Image1 year (%)3432320.9
Age at the time of follow-up (years)19 (10–33)30 (21–44)
Time from OK to follow-up (years)15 (11–23)26 (22–34)
Pre-op RV syst pressure (mmHg)67 (40–112)66 (40–110)66 (40–110)0.8
Pre-op QP/QS2.2 (0.7–6.0)2.2 (0.7–6.0)2.2 (0.8–6.0)1.0
Pulm resistance (dyne/scm−5)233 (50–980)246 (62–825)249 (62–825)0.9
Aortic cross clamp time (min)39 (10–85)35 (14–80)35 (14–80)0.9
Type of VSD: peri-memb (%)7785820.5
Previous PA banding (%)9650.8
RV-incision (%)4551470.3
  • Numbers between brackets give the range.

After a median follow-up of 15 years (range 11–23 years), 109 patients (79% of those eligible for follow-up) participated in the first follow-up study performed in 1991.4 The target population of the second follow-up (2001) consisted of the 109 patients of the first follow-up. The follow-up status was determined by examination at our institution and the cardiac examination included medical history, physical examination, standard 12-lead electrocardiography, 24-h ambulatory electrocardiography (Holter), echocardiography and bicycle ergometry. Furthermore, a psychological interview was performed. The Medical Ethical Committee approved this study and all patients gave their written consent.


Standard 12-lead electrocardiograms were analysed for cardiac rhythm, the height of the P wave (measured in lead II), duration of the P wave and the PR interval. A first-degree atrioventricular block was defined by a PR interval Math200 ms. Furthermore the Median frontal plane P wave axis and QRS axis were determined, as was the widest QRS duration. A QRS duration Math120 ms was defined to be a complete bundle branch block: a positive QRS complex in lead V1 was categorised as right bundle branch block and a negative QRS complex as left bundle branch block. A single observer made all ECG measurements.

Holter monitoring

Sinus node dysfunction was assessed during 24-h Holter monitoring using the modified Kugler criteria: nodal escape rhythm, sinus arrest Math3 s or severe sinus bradycardia (Math30 beats/min at night or Math40 beats/min during daytime).11 Ventricular tachycardia was defined as 3 or more consecutive ventricular beats with a heart rate of Math100 beats/min.

Two-dimensional echocardiography

Echocardiography was performed using a Hewlett-Packard Sonos 5500 echocardiograph. M-Mode measurements of left atrial and left ventricular end-diastolic and end-systolic dimensions were made in the parasternal view. A left atrium Math45 mm and left ventricle end-diastolic dimension Math58 mm were considered enlarged. A fractional shortening of the left ventricle of Math30% was considered abnormal. Right ventricular dimensions and function were judged by visual estimate by two experienced cardiologists. Dimensions were scored as normal, mildly, moderately or severely dilated. Right ventricular function was graded as normal, mildly, moderately or severely impaired. Of each patient the study of 1990 was compared with that of 2001. The degree of tricuspid regurgitation (minimal, moderate, or severe) was estimated with colour-Doppler by the width and length of the regurgitant jet. Pulmonary hypertension was defined by an early diastolic pulmonary regurgitation flow velocity of Math2.5 m/s or, in the absence of right ventricular outflow obstruction, a tricuspid regurgitation flow velocity Math3.0 m/s. Multiple echocardiographic views were examined using color flow to identify residual shunts.

Bicycle ergometry

Maximal exercise capacity was assessed by bicycle ergometry with stepwise increments of the workload by 20 W/min. Exercise capacity was compared to that of normal individuals corrected for age, sex and body height.

Cardiovascular events were defined as either re-operation, pace-maker implantation, tachy-arrhythmia requiring treatment (medication, electrical cardioversion or ablation), endocarditis or congestive heart failure.

Statistical analysis

Data are presented as median and range, unless indicated otherwise. The Math and Fisher's exact test were used for the comparison of discrete variables. The Student t-test was used to compare continuous variables. All tests used were two-tailed. The Mc Nemar test of symmetry was used to compare the 15 and 26 years outcome. Cumulative survival curves were constructed using the Kaplan–Meier method. Among patient sub-groups the log rank test was used to compare survival curves. The level of significance was chosen at Math. Multivariate Cox-regression analysis was performed for survival.

The variables tested were birth weight, type of palliation, main pulmonary artery systolic pressure, pulmonary artery/aorta ratio, age at operation, year of operation, use of deep hypothermia, use of cold cardioplegia, aortic cross-clamp time, right ventricular incision, peri-operative complications, residual lesions. Continuous variables were not categorised in the model. The proportional hazards assumptions were tested by constructing interaction terms between the variables and time to each end-point. Cox regression analyses showed no statistically significant interactions with time (each Math). The model selection is based on the step-wise principle, where the limit to enter and to remove a variable was both 0.05.



Median age at operation was 4 years (range 0–13). The baseline characteristics of all 176 patients are described in Table 1. Information on survival is complete for the total cohort of 176 patients. Twenty-three (12%) patients died before 1991. Of the 109 patients who participated in the first follow-up study, 2 patients died, 7 were lost to follow-up and 5 participated only by written questionnaire. The remaining 95 patients (95% of those eligible for follow up) fully participated in the second follow-up study with a median follow-up of 26 years (range 22–34 years) after surgery, and with a median age at the time of study of 30 years (range 21–44). No significant differences were found between the baseline characteristics of the patients who participated in the follow-up studies and the patients who did not (Table 1). There were 57 males (60%) and 38 females (40%).


Nineteen patients died within 30 days after operation (Fig. 1). Late death occurred in 6 patients (4%). Of these 6 patients who died late post-operatively 4 had documented residual pulmonary hypertension and all died suddenly 3, 4, 16 and 18 years after the operation, respectively. These 4 patients all showed right ventricular hypertrophy on their ECG. One patient died during re-operation for aortic valve surgery and 1 patient died of a non-cardiac cause (fire accident).

Fig. 1

(Event-free) survival after VSD closure.

Cardiovascular events

The 25 years event-free survival for the hospital survivors was 80% (Fig. 1). Re-operations were performed in 6 patients: 2 residual VSD (10 and 21 years after the initial operation), 1 discrete sub-aortic stenosis and 3 RV-outflow tract obstruction (at a median interval of 6 years after the initial operation). Additional cardiac surgery was performed in 4 patients: 1 persisting arterial duct, 1 aortic valve replacement after endocarditis, and 2 aortic coarctation. Pacemaker implantation was performed in 6 patients; 2 for surgical atrioventricular block shortly after operation and 4 because of sick sinus syndrome more than 15 years after surgery. One patient underwent radio frequency catheter ablation for intra-atrial re-entry tachycardia. Two patients with co-morbidity, suffered from endocarditis late after VSD closure leading to aortic valve replacement in one, and pacemaker replacement in the other patient.

Clinical evaluation

Of the patients, 92% were in NYHA class 1, and 8% in class 2. Five patients (5%) were taking medication: oral anti-coagulation in 1 (artificial aortic valve), β-blockers in 2, and ace-inhibitors in 2 patients. Oxygen saturation (measured with Nellcor) was 98% (range 94–100%). No patient showed signs of heart failure.


The ECG findings are described in Table 2. We found no difference between the trans-atrial and trans-ventricular approach (with right ventricular incision) regarding the incidence of a right bundle branch block on the ECG during follow-up. A prolongation of the QRS-complex and P-wave duration occurred between 1990 and 2001.

View this table:
Table 2

Standard 12-lead ECG long term after surgical VSD repair

Number of patients10995
Rhythm (%)
Sinus106 (97)88 (93)0.3
Nodal1 (1)2 (2)
Atrial1 (1)1 (1)
Pacemaker1 (1)4 (4)
PR interval±SD (ms)147±30152±280.05
PR Embedded Image200 ms (%)250.2
QRS duration±SD (ms)101±20113±260.01
RBBB (%)23290.06
QTc segment±SD (ms)398±40398±270.9
QRS axis±SD (°)52±3043±330.01
P duration±SD (ms)84±1591±170.01
P-wave height±SD (cm)0.17±0.070.17±0.070.3
P-wave axis±SD (°)39±2635 ±270.09
LVH (%)7110.05
RVH (%)870.5

Twenty-four hour ambulatory monitoring

None of the patients had atrial flutter or fibrillation on the 24-h ambulatory monitoring in 1990 or 2001. Junctional escape rhythm was found in 2001 in 23% of the patients, signs of sinus node disease were found in 9% of the patients, but no ventricular pauses longer than 3 s occurred. Ventricular tachycardia of more than 10 complexes was not found, and 8% showed ventricular tachycardia of 3–10 complexes.


Echocardiography was performed in 95 patients (Table 3). Left ventricular dimensions were normal in 96% of the patients. Pulmonary hypertension was found in 4%. These 4% were not different from the total group concerning the age at operation. We found no change in median pulmonary artery pressure between 1990 and 2001. Aortic regurgitation was present at the last follow-up in 15 patients (16%); this was mild in 13 and moderate in 2 patients. Over the last 10 years, 2 patients showed progression from mild to moderate aortic regurgitation. No severe aortic valve regurgitation was encountered and no significant change in mitral, tricuspid or pulmonary regurgitation was noted (Table 3). Using Cox multivariate regression we found no relationship between the presence of moderate aortic or mitral insufficiency and the duration of follow-up, age at surgery, right ventricular incision during surgery, type of VSD or left ventricular function (fractional shortening on echo).

View this table:
Table 3

Echocardiographic results in 95 patients after surgical VSD repair

Number of patients10995
LA dimension±SD (mm)32±537±60.2
LVED dimension±SD (mm)50±452±50.1
LA dilatation (%)4 (4)9 (9)0.02
LV dilatation (%)11 (10)4 (4)0.01
RA dilatation (%)19 (20)19 (20)0.9
RV dilatation (%)13 (14)13 (14)0.7
LVH (%)5 (5)9 (10)0.1
RVH (%)4 (4)7 (7)0.3
Mean FractShort.LV (%)34340.8
Good RV syst function (%)109 (100)94 (99)0.8
Valve insuff (%)
AoI16 (15)15 (16)0.6
Mild16 (15)13 (14)
Moderate02 (2)
MI13 (12)11 (12)0.9
Mild13 (12)11 (12)
PI26 (24)27 (28)0.3
Mild26 (24)25 (26)
Moderate02 (2)
TI45 (41)40 (42)0.7
Mild43 (39)36 (38)
Moderate2 (2)4 (4)
Embedded Image TI±SD2.4±0.42.4±0.40.8
Embedded Image PI±SD1.5±0.31.6±0.30.6
Pulmonary hypertension (%)7 (6)4 (4)0.4
Small residual VSD (%)9 (8)8 (8)0.8

Exercise capacity

Ninety-three patients exercised to maximal effort; the median exercise capacity in these patients was 91% (range 43–226) of the predicted values in 2001, while it was 100% in 1990. During the test 5 patients had an increase of ventricular extra systoles and 1 patient developed supra-ventricular tachycardia. No ventricular tachycardia occurred. We found no relation between exercise capacity in 2001 and Median pulmonary artery pressure before surgery, duration of aortic cross clamp time, presence or absence of a right bundle branch block on the ECG, young age (Math1 year) at operation or left ventricular function (fractional shortening on echo).

Risk factors for late death

Multivariate Cox-regression revealed two predictors for late mortality (see Table 4). A median pulmonary artery pressure of more than 70 mmHg before operation and peri-operative complications (re-operation, arrhythmia, infection) were predictors for late mortality.

View this table:
Table 4

Multivariate Cox regression

End point: late death (Embedded Image1 year after surgery) independent predictorsHR95% CI
Median pulmonary artery pressure before surgery per 10 mmHg increase1.29[1.05–1.58]
Post-operative complications7.8[2.6–24.1]

Social life

Of the 95 patients seen at the last follow-up, 70% had some sort of relationship (stable relationship 16%, co-habiting 19%, married 35%). A total of 72% had no children, 12% had 1 child and 16% had 2 or more children. Regarding the daily activities, 82% of the patients had a paid job. On the question if the patient had ever encountered difficulties with respect to insurance policies, 75% of the patients answered yes. With respect to which insurance policy, 33% had problems with health insurance, 42% with mortgage, and 55% with life insurance.


This series is unique in that it comprises a cohort of consecutive patients operated on at young age at a single institution with longitudinal follow-up of 22–34 years. Apart from a considerable historical early mortality (13%), we experienced only a 4% (6 patients) late mortality. This study also provides new data documenting sinus node disease developing late after surgery and the insurability of the patients.


The operative mortality of 13% found in this study represents the patients with large VSDs operated upon in the time period of 1968–1980. At present these figures have improved dramatically. The focus of this study was on late events in hospital survivors, and statistical analyses were confined to this patient population.

There was a 4% (6 patients) late mortality among the 153 patients who survived the operation, and 5 of the 6 deaths were attributed to cardiovascular causes. This is higher than in the general population. For comparison, in patients undergoing surgical closure of an atrial septal defect who were studied in the same manner, no cardiac deaths occurred during the 26 years follow-up.12 The most likely explanation for this late mortality is a right ventricular hypertrophy due to long-standing right ventricular pressure overload, causing ventricular arrhythmias. Right ventricular hypertrophy was found in all 4 patients who died suddenly late after the operation. Cardiac hypertrophy is the common denominator in all cases reported in detail of VSD-associated sudden death.13 It is known that the incidence of ventricular arrhythmias and sudden death in non-operated VSD-patients is high (20–90%), suggesting that damage to the ventricle is particularly due to pressure overload.2,13 Ventricular arrhythmias were found in 42% of our patients on ambulatory ECG monitoring in 1990, none of these patients experienced clinical significant arrhythmias or sudden death in the following 11 years. Therefore, we conclude that there is no clear predictive value of ventricular ectopic activity found on 24-h Holter monitoring. Others reported ventricular arrhythmias varying from 18% to 39%.8,10 We did not find patients developing pulmonary hypertension late after surgery. Pulmonary pressure neither diminished, nor progressed during follow-up. Our study reports the late outcome of patients operated on between the late 1960s and 1970s. Since then, changes have occurred and for some time, VSDs with large left to right shunts are treated in infancy and therefore at the present time pulmonary hypertension late after surgery will largely be prevented. The 4% of patients in our study with pulmonary hypertension at present are at risk for sudden death in the future and for these high-risk patients intensified surveillance is justified and possible prophylactic implantation of an internal defibrillator should be considered. Since no follow-up data are available for patients who died prior to the follow-up, we only included the follow-up data of the surviving patients. This may lead to biased results.

Clinical condition

The functional outcome in late survivors was good with the vast majority of patients being in NYHA class I, a median exercise capacity of 91% and only five patients (5%) using medication. Although a second cardiac operation was necessary in 10% of the patients, in only 2% it concerned a residual VSD. A small, haemodynamically insignificant residual VSD was found in 8% of the patients, which is in line with, or slightly better than in other reports.2,14

Bacterial endocarditis rarely occurred (2 patients), and was not even directly related to the VSD; in 1 patient the aortic valve was involved and the other patient suffered from pacemaker-endocarditis. Also others report a very low incidence of endocarditis after VSD closure, so endocarditis prophylaxis should only be given to patients who have concomitant pathology such as pacemaker leads, valve disease or have a residual VSD. Routine-use of prophylaxis does not seem to be indicated after successful closure of an isolated VSD.

The occurrence of aortic insufficiency is one of the issues after VSD surgery. The incidence in this cohort of patients was 16%, which is higher than in other reports.2 In most studies, however, no extensive cardiovascular examination was performed. The degree of aortic insufficiency was mild and did not change significantly in 10 years time. Aortic valve surgery was necessary in 1 patient only, and that was because of endocarditis. Left ventricular dilatation was found in a minority of patients only. No significant heart failure was found and no signs of left ventricular dysfunction, which is very important for the future perspectives of these patients.

Conduction disturbances

Pacemaker implantation was necessary in 6 patients (4%). The 2 patients with surgical block had a pacemaker implanted shortly after surgery. The fact that 4 patients developed sinus node disease with the need for pacemaker implantation more than 15 years after surgery is striking. A possible cause of this late sinus node dysfunction is cannulation of the right atrium for cardiopulmonary bypass, but it is only after a surprisingly long period of time that this sinus node disease becomes manifest.15 The progressive nature of the disease is confirmed by 20% of the patients having minor indicators of sinus node dysfunction in 1990, and 4 of these then needing subsequent pacemaker implantation.4 Furthermore, the PR-interval increased along with the incidence of first degree AV block. The prolongation of the QRS-complex may indicate late myocardial reaction to the earlier ventricular overload. The late occurrence of clinically significant sinus node disease necessitating pacemaker implantation has not been reported before in this patient group, perhaps because most of these patients do not have regular cardiological follow-up.

Social life

Employability is excellent in our patient group and it appears that finding a paid job is not a major issue. However, starting a family seems more problematic. Only 28% of the patients had children, although the fast majority had a stable relationship. This is low compared with the general population.16 The explanation may be fear that pregnancy may cause damage to the female patient or fear for congenital heart disease in the offspring. Another explanation may be that fertility is negatively influenced by heart surgery at a young age. The numbers are relatively small in our study and more research is necessary in this field. Probably, more attention should be paid to give proper information to these patients about the different aspects of pregnancy. Finally, because the long-term prognosis is excellent in the absence of pulmonary hypertension, application for insurance should be able without problems in the large majority of patients.17 However, 75% of the patients in our study experienced obstacles to obtain insurance policies, especially with life insurance. Since the prognosis after surgical correction of VSD has further improved over the last 30 years, the ease of gaining insurance should improve in these patients.


Adults who underwent surgical VSD closure in childhood generally do well, but survival is lower than that of the general population. Late mortality was found in 4%, particularly in the patients with pulmonary hypertension and right ventricular hypertrophy. In our study no new pulmonary hypertension developed long after surgery. The overall clinical condition of the survivors at last follow-up was satisfying and complications such as endocarditis, arrhythmias and clinical significant aorta insufficiency rarely occurred. Sinus node disease requiring pacemaker implantation was found in 4% of the patients, more than 15 years after surgery. This emphasises the need to follow patients, who have been subjected to open heart surgery before 1980, indefinitely with special attention for the development of sinus node disease. Employability is good, but pregnancy needs further attention. The results of this study can be used as a guideline for insurance policies.


We thank Mrs. J. Mc Ghie, Mrs. V.E. Kleyburg-Linkers (echocardiography laboratory), and Mrs. E.M. Peterse-Dekkers (Holter laboratory) for the excellent technical support and Mrs. W. van der Bent for her help in preparing this manuscript. The report was written as part of a project funded by the Netherlands Heart Foundation (No. 99.033).


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
View Abstract