Author + information
- Received September 30, 2014
- Revision received February 23, 2015
- Accepted February 24, 2015
- Published online May 12, 2015.
- Myrthe E. Menting, MD∗,
- Judith A.A.E. Cuypers, MD∗,
- Petra Opić, MSc∗,
- Elisabeth M.W.J. Utens, PhD†,
- Maarten Witsenburg, MD, PhD∗,
- Annemien E. van den Bosch, MD, PhD∗,
- Ron T. van Domburg, PhD∗,
- Folkert J. Meijboom, MD, PhD‡,
- Eric Boersma, PhD∗,
- Ad J.J.C. Bogers, MD, PhD§ and
- Jolien W. Roos-Hesselink, MD, PhD∗∗ ()
- ∗Department of Cardiology, Erasmus Medical Center, Rotterdam, the Netherlands
- †Department of Child and Adolescent Psychiatry and Psychology, Sophia Children’s Hospital, Rotterdam, the Netherlands
- ‡Department of Cardiology, University Medical Center Utrecht, Utrecht, the Netherlands
- §Department of Cardiothoracic Surgery, Erasmus Medical Center, Rotterdam, the Netherlands
- ↵∗Reprint requests and correspondence:
Dr. Jolien W. Roos–Hesselink, Erasmus University Medical Center, Department of Cardiology, Room Ba-583a, P.O. Box 2040, 3000 CA Rotterdam, the Netherlands.
Background Few prospective data are available regarding long-term outcomes after surgical closure of a ventricular septal defect (VSD).
Objectives The objective of this study was to investigate clinical outcomes >30 years after surgical VSD closure.
Methods Patients who underwent surgical VSD closure during childhood between 1968 and 1980 were reexamined every 10 years. In 2012, we invited eligible patients to undergo another examination, which included electrocardiography, Holter monitoring, echocardiography, bicycle ergometry, measurement of N-terminal pro–B-type natriuretic peptide, and subjective health assessment.
Results Cumulative survival was 86% at 40 years. Causes of mortality were arrhythmia, heart failure, endocarditis, during valvular surgery, pulmonary hypertension, noncardiac causes, and unknown causes. Cumulative event-free survival after surgery was 72% at 40 years. Symptomatic arrhythmias occurred in 13% of patients and surgical or catheter-based reinterventions in 12%. Prevalence of impaired right ventricular systolic function increased from 1% in 2001 to 17% in 2012 (p = 0.001). Left ventricular systolic function was impaired but stable in 21% of patients. Aortic regurgitation occurred more often in the last 20 years (p = 0.039), and mean exercise capacity decreased (p = 0.003). N-terminal pro–B-type natriuretic peptide (median: 11.6 pmol/l [interquartile range: 7.0 to 19.8 pmol/l]) was elevated (>14 pmol/l) in 38% of patients. A concomitant cardiac lesion, for example, patent ductus arteriosus, and aortic cross-clamp time were determinants of late events (hazard ratio: 2.84 [95% confidence interval: 1.23 to 6.53] and hazard ratio: 1.47 per 10 min [95% confidence interval: 1.22 to 1.99], respectively). Patients rated their subjective health status significantly better than a reference population.
Conclusions Survival up to 40 years after successful surgical VSD closure is slightly lower than in the general Dutch population. Morbidity is not negligible, especially in patients with a concomitant cardiac lesion.
Ventricular septal defect (VSD) is by far the most common congenital heart defect, with a birth prevalence of 2.62 per 1,000 live births (1,2). Small defects may not have hemodynamic consequences, but the presence of a significant left-to-right shunt can cause left ventricular (LV) overload, pulmonary arterial hypertension, ventricular dysfunction, arrhythmias, and aortic regurgitation (3,4). Surgical closure at a young age is still the treatment of choice, and mid- to long-term results are good with regard to survival, morbidity, and quality of life (5–7). Therefore, most such patients have been discharged from routine cardiological follow-up. For both patients and their treating physicians, it is essential to know whether the pre-operative left-to-right shunt and the VSD patch affect biventricular function or the conduction system in the long term. However, information on mortality and morbidity beyond 30 years after surgical VSD closure is scarce, and almost all such data were collected retrospectively, which introduces the possible bias of including only patients with residual morbidity who are still seen at outpatient clinics.
Our study is part of a unique, ongoing longitudinal follow-up of patients with congenital heart defects who underwent surgery at a young age at our institution between 1968 and 1980. The current study had 3 objectives: first, to evaluate survival 30 to 40 years after surgical VSD closure in an unselected cohort; second, to investigate the current clinical condition of survivors by extensive in-hospital examination and to detect determinants of outcome; and third, to evaluate the present subjective health status of survivors.
All consecutive patients who underwent surgical VSD closure at our institution between 1968 and 1980 at <15 years of age formed the original study cohort. This cohort was first studied in 1990, with a second follow-up performed in 2001 (5,6). In 2010 to 2012, survival status was obtained from the Dutch National Population Registry, and all surviving patients who had participated in 1 or both of the previous studies were actively invited to participate in a third study for clinical examination at the outpatient clinic of Erasmus Medical Center. The study protocol was approved by the institutional Medical Ethics Committee. Written informed consent was obtained from all participants.
Survival rates were compared with the expected survival rates of an age-matched Dutch population. Adverse events included all-cause mortality, surgical or catheter-based cardiac reinterventions, symptomatic arrhythmias (requiring medication, cardioversion, ablation, or insertion of a pacemaker/implantable cardioverter-defibrillator [ICD]), endocarditis, and heart failure (requiring medication or hospital admission). Mortality and events were defined as “early” when they occurred within 30 days post-operatively and “late” when they occurred beyond 30 days. All events were assessed by 2 independent investigators (M.E.M., J.A.A.E.C.).
Examinations included history, physical examination, standard 12-lead electrocardiography (ECG), 24-h Holter monitoring, echocardiography, cardiopulmonary exercise testing, and N-terminal pro–B-type natriuretic peptide (NT-proBNP) measurement. If a patient was unwilling or unable to visit the outpatient clinic, questionnaires were sent to obtain information on morbidity and subjective health status and to receive permission to use the patient’s medical records.
Electrocardiography and 24-h holter monitoring
Standard 12-lead surface ECGs were analyzed for rhythm, PR interval, and QRS duration. A 24-h Holter monitoring was performed with a CardioPerfect Holter DR180+ 3-channel recorder (Welch Allyn Cardio Control, NorthEast Monitoring, Maynard, Massachusetts). Sinus node disease (SND) was defined according to the Kugler criteria: nodal escape rhythm, sinus arrest >3 s, or severe sinus bradycardia (<30 beats/min at night or <40 beats/min during daytime) (8).
Cardiopulmonary exercise testing
Maximal workload, heart rate, and peak oxygen consumption (peak o2) were assessed by bicycle ergometry with a gradual workload increment of 20 W/min (ramp protocol) and compared with normative values corrected for age, sex, height, and weight. Performance was considered maximal when a respiratory exchange ratio >1 was reached.
A complete 2-dimensional transthoracic echocardiogram was performed with the commercially available IE33 system (Philips Medical Systems, Best, the Netherlands). Cardiac dimensions, ventricular function, and valvular function were measured according to published guidelines (9–12). LV and right ventricular (RV) systolic function were assessed visually to enable comparison with the 2 previous studies. Systolic function was graded as normal or mildly, moderately, or severely impaired. Additionally, more objective measurements, including LV ejection fraction (Simpson’s method), RV fractional area change, and tricuspid annulus plane systolic excursion, were used to quantify systolic ventricular function. Measurements were obtained by 2 independent observers (M.E.M., J.A.A.E.C.).
Peripheral venous blood samples were collected after 30 min of rest. Plasma NT-proBNP levels were determined with use of the commercially available electrochemiluminescence immunoassay Elecsys (Roche Diagnostics, Basel, Switzerland). The normal value in our hospital is <14 pmol/l.
Subjective health assessment
The 36-item short-form health survey (SF-36) was completed to assess subjective health status. Results for the patients were compared with their results from 10 years earlier and with normative data from the general Dutch population (13).
Continuous data are presented as mean ± SD or median with interquartile range (IQR) depending on the data distribution. Categorical data are presented as frequencies and percentages. Changes in patient characteristics since the follow-up study in 1990 and 2001 were evaluated by estimating a trend by use of mixed models, which take missing values into account. Changes in characteristics between 2001 and 2012 were analyzed by paired Student t tests. Differences between independent subgroups were evaluated by unpaired Student t tests or Mann-Whitney U tests (continuous data) and by chi-square test or Fisher exact test (categorical data). To quantify correlations between 2 variables, the Spearman correlation test was used.
Cumulative survival and event-free survival for all patients and for patients with successful VSD surgery (i.e., excluding early post-operative mortality) were determined by the Kaplan-Meier method. We compared the cumulative survival of isolated and nonisolated VSD patients by the log-rank test. Cumulative event incidences were computed with the use of a nonparametric estimator of cumulative incidence functions, with death as a competing risk. Univariable and multivariable Cox regression analyses were used to identify determinants of pre-defined adverse events: all-cause mortality, arrhythmias, reinterventions, heart failure, and endocarditis.
The mixed models were estimated by use of SAS software (version 9.3, SAS Institute, Inc., Cary, North Carolina). The cumulative incidence functions were estimated with R (version 3.1.1, R Foundation for Statistical Computing, Wien, Austria). All other statistical analyses were performed with the Statistical Package for Social Sciences (version 21, SPSS Inc., Chicago, Illinois). The statistical tests were 2-sided, and p < 0.05 was considered statistically significant. Further information on statistical analysis can be found in the Online Appendix.
The original study cohort consisted of 174 consecutive patients who underwent surgical VSD closure between 1968 and 1980. Figure 1 presents an overview of patient participation for the current study. Baseline characteristics, including surgical details and follow-up duration, are presented in Table 1. Further baseline and surgical details have been reported previously (5,6).
Median follow-up duration of the actively included patients was 35.8 years (IQR: 34.0 to 37.3 years; range 30.4 to 40.3 years). Of 91 patients, 70 participated in-hospital, and 21 completed the questionnaires and gave permission to use their hospital records. Twenty-seven of the 91 patients (30%) had 1 or more concomitant cardiac lesions (nonisolated VSD), including patent foramen ovale (n = 11), pulmonary stenosis (n = 10), patent arterial duct (n = 6), aortic coarctation (n = 5), atrial septal defect (n = 4, of whom 2 had partial abnormal pulmonary venous return), mitral stenosis (n = 2), and aortic stenosis (n = 1). A patent foramen ovale was included as a concomitant lesion when an intervention was performed to close it. There were no discrepancies between assessments of the 2 evaluators regarding post-surgical events.
Information on survival was available for 156 patients (90%). Cumulative survival after surgical closure, including early post-operative mortality, was 89% at 10 years, 87% at 20 years, 85% at 30 years, and 78% at 40 years, which was significantly lower than in the general Dutch population of comparable age (Figure 2A). Thirty-one patients died, with 17 deaths occurring within 30 days of surgery. Cumulative survival after successful surgery, excluding early post-operative mortality, was 99% at 10 years, 96% at 20 years, 95% at 30 years, and 86% at 40 years, which was slightly lower than the general Dutch population (Figure 2B). Successful surgery was defined as the group without early post-operative mortality.
In the last 10 years, 8 patients died. One patient died of ventricular fibrillation 34 years after surgery at 34 years of age. He had received a pacemaker for SND 20 years earlier. One patient who had developed moderate aortic stenosis had sudden death, presumably due to an arrhythmia, 39 years after surgery at the age of 49 years. One patient died of heart failure as a result of severely dilated cardiomyopathy, 28 years after surgery, at the age of 28. There were 3 noncardiac deaths, attributable to breast cancer, lung cancer, and alcohol abuse 29, 33, and 31 years after surgery, respectively. The cause of death was unknown for 2 patients, who died 34 and 42 years after surgery, respectively.
Cumulative survival rates of isolated versus nonisolated VSD patients up to 39 years after surgery were comparable (93% vs. 88%) (Figure 3A).
Cumulative survival free of adverse events was 63% at 40 years (Figure 2A). When early post-operative mortality was excluded, event-free survival was 72% (Figure 2B). The event-free survival at 40 years was significantly better in isolated versus nonisolated VSD patients (79% vs. 57%, respectively) (Figure 3B).
Cumulative incidence of late symptomatic arrhythmias including pacemaker implantation at 40 years was 13% (Figure 4C). In the last decade studied, 7 patients developed new symptomatic arrhythmias: 1 patient had recurrent atrial flutter and fibrillation, treated with electrical cardioversion and catheter ablation; 1 patient who had a pacemaker for 19 years had recurrent atrial fibrillation treated with electrical cardioversion; 1 patient had atrial fibrillation treated with verapamil; and 1 patient had paroxysmal atrioventricular reentrant tachycardia. The cumulative incidence of late pacemaker implantation, including ICD in 2 patients, was 7% at 40 years. In the last decade studied, 3 pacemakers were implanted: 1 for SND, 1 for severe bradycardia, and 1 for complete heart block. The last patient received a biventricular ICD because of ventricular tachycardias and diminished ejection fraction due to dilated cardiomyopathy. During total follow-up, 3 patients developed complete heart block.
Cumulative incidence of late endocarditis at 40 years was 4%. During the last 10 years, 1 patient was diagnosed with pacemaker endocarditis caused by Staphylococcus aureus, which required treatment with antibiotic drugs and replacement of the pacemaker system.
Cumulative incidence of late heart failure at 40 years was 4%. During the last decade studied, 3 patients developed heart failure. One of them died during hospitalization, 1 was admitted to the hospital several times, and 1 was hospitalized for heart failure triggered by a pneumonia. One of these patients developed heart failure 9 years after epicardial pacemaker implantation, and 1 had diminished LV function before pacemaker implantation.
Electrocardiography and holter monitoring
The ECG and Holter findings are summarized in Table 3. None of the patients had ventricular pauses longer than 3 s.
Cardiopulmonary exercise testing
Table 3 shows the results of bicycle ergometry. Of the 33 patients with a diminished exercise capacity (i.e., workload <85% of expected), 6 (18%) had 1 or more reinterventions compared with 1 (3%) in the group with normal exercise capacity (p = 0.049). No differences were found with regard to current age, surgical characteristics, isolated versus nonisolated VSD, QRS duration, or ventricular function.
Echocardiographic findings are summarized in Tables 3 and 4⇓. Systolic LV function was mildly impaired in 11 patients (14%), moderately impaired in 3 (4%), and severely impaired in 2 (3%). The percentage of patients with impaired systolic LV function was 14% in 2001 and 21% in 2012 (p = 0.180). Normal diastolic LV function was observed in 63 patients (89%), pseudonormal diastolic function in 6 (8%), and restrictive relaxation pattern in 2 (3%). Systolic RV function was mildly impaired in 10 patients (13%) and moderately impaired in 3 (4%). The percentage of patients with impaired systolic RV function increased significantly over the last decade of study from 1% to 17% (p = 0.001). Patients with impaired systolic RV function more often had an elevated estimated RV systolic pressure than patients with normal systolic RV function (42% vs. 12%, p = 0.036), more often had a pacemaker (23% vs. 3%, p = 0.033), and more often had impaired systolic LV function (69% vs. 11%, p < 0.001). Diastolic LV function did not differ significantly between patients with impaired or normal systolic RV function. There were no significant differences in systolic biventricular function between patients with isolated and nonisolated VSD, with and without RV incision, or with and without VSD patch.
The percentage of patients with mild or moderate aortic regurgitation increased significantly over the last 20 years from 11% in 1990 to 21% in 2012 (p = 0.039).
NT-proBNP was measured in 68 patients. The median level was 11.6 pmol/l (IQR: 7.0 to 19.8 pmol/l), with an elevated level (>14.0 pmol/l) measured in 26 patients (38%). The highest value was 56.5 pmol/l, measured in a woman with an isolated VSD, in sinus rhythm, and without clinical signs of heart failure. Median NT-proBNP was comparable between patients with an isolated VSD (n = 50) and those with a nonisolated VSD (n = 18; 11.2 pmol/l [IQR: 6.9 to 18.4 pmol/l] vs. 14.3 pmol/l [IQR: 6.8 to 28.1 pmol/l], p = 0.436). Patients with impaired systolic RV function (n = 8) tended to have a higher NT-proBNP than patients with normal RV function (n = 59; 16.2 pmol/l [IQR: 12.9 to 26.6 pmol/l] vs. 11.1 pmol/l [IQR: 6.4 to 19.5 pmol/l], p = 0.063). No relationships were found with systolic LV function, ventricular dimensions, age at operation, current age, or exercise capacity.
Subjective health status assessment
Seventy-four patients completed the SF-36 health survey. The results are depicted in Figure 5. Patients obtained significantly better scores than the reference population on all scales except for general health perceptions, which were comparable (p = 0.089). The patients’ results were comparable with their own results 10 years earlier, except for mental health, which they now rated better (85 ± 13% vs. 80 ± 11%, p = 0.030).
Determinants of clinical outcome
Higher pre-operative systolic RV pressure tended to be a determinant of mortality (hazard ratio [HR]: 1.02; 95% confidence interval [CI]: 1.00 to 1.04).
Univariable regression analyses identified nonisolated VSD, early post-operative arrhythmias, and aortic cross-clamp time as determinants of late events, including mortality. In multivariable regression analysis with these determinants, patients with nonisolated VSD or patients with longer aortic cross-clamp time had higher risk for events (HR: 2.84 [95% CI: 1.23 to 6.53]; HR: 1.47 per 10 min [95% CI: 1.22 to 1.99 per 10 min], respectively). Early post-operative arrhythmias showed a trend toward predicting later events (HR: 2.59 [95% CI: 0.92 to 7.29]).
Patients with a complete heart block in the early post-operative period more often developed symptomatic arrhythmias during follow-up (HR: 9.7 [95% CI: 2.1 to 44.6]). No other baseline characteristics were significant determinants of outcome.
In this unique longitudinal cohort study of VSD patients who underwent surgical repair at a young age, survival up to 40 years was relatively good but lower than in the general Dutch population. Although morbidity was substantial, especially among patients with nonisolated VSD, the reported subjective health status was even better than normative data.
Mortality and adverse events
Cumulative survival 40 years after surgical VSD closure was 78% in our cohort. One-half of these deaths occurred within 30 days after surgery. Currently, pulmonary artery banding is performed only rarely, and advances in surgical and anesthesiology techniques and improvements in peri-operative care have greatly reduced peri-operative mortality (14). In addition, the care during follow-up is better organized today. Therefore, these early results are not applicable to patients who have undergone surgery in more recent years. With the exclusion of in-hospital operative mortality, however, cumulative survival in our cohort was 86% at 40 years, which was still slightly lower than in the general population. In more than one-half of the cases, late mortality was cardiac related: cardiac arrest/sudden death, reoperation, and heart failure (Central Illustration).
Cumulative survival was similar between patients with isolated and nonisolated VSD. Morbidity was higher in patients with nonisolated VSD and was dominated by reinterventions for concomitant lesions, for example, aorta-related problems or pulmonary stenosis.
Health status and ventricular function
Most of our patients rated their physical functioning better than the reference population; however, their maximal workload at exercise testing was clearly lower than the reference population, with almost one-half of the patients having a diminished maximal workload. Furthermore, the maximal workload decreased over the last decades studied. The median peak o2 of 87% in our study population was a bit higher than the 73% in the study by Kempny et al. (15). This difference could probably be explained by the fact that they included patients who underwent exercise testing as part of their ongoing outpatient care, whereas our study included many patients who had been discharged from routine clinical follow-up.
The occurrence of aortic regurgitation appears to be an important issue after VSD surgery. In our cohort, 1 patient developed severe aortic regurgitation that necessitated aortic valve replacement, and in the other patients, the prevalence of aortic regurgitation nearly doubled over the last 20 years of the study, from 11% in 1990 to 21% in 2012.
Systolic LV function remained stable over the last decade of study, but impaired systolic RV function increased. In our cohort, systolic RV dysfunction was not related to VSD patch or RV incision, but the vast majority of patients with systolic RV dysfunction had systolic LV dysfunction and more often had elevated RV systolic pressure. This relationship between RV and LV dysfunction could be explained in part by the presence of a pacemaker in a quarter of patients with RV dysfunction or by systolic ventricular interaction. In the case of systolic LV dysfunction, pulmonary pressure increases, which results in reduction of RV contractility. Moreover, ventricular interaction is mediated by forces at the interventricular septum and mechanical coupling through shared myocardial fibers: When the LV becomes more spherical, the fibers of the interventricular septum become less oblique, which reduces and impairs RV contractile function (16,17). Systolic ventricular interaction is also observed in other congenital heart defects (18,19). Another explanation for the higher prevalence of RV dysfunction over time may be the use of a more sophisticated echocardiographic assessment to facilitate the detection of dysfunction. Using tissue Doppler imaging, Klitsie et al. (20) found a systolic RV impairment up to 20 months after surgical VSD closure in children. Although subclinical RV impairment may not be of direct clinical relevance, it could be a first sign of systolic RV dysfunction and does stress the importance of detailed evaluation of RV function after VSD surgery.
NT-proBNP levels were elevated in more than one-third of the patients. This is striking, because the majority of patients were asymptomatic and had normal systolic biventricular function. No studies were found in the published data on BNP levels in adults after surgical VSD closure, and only a few in children (21), which reported mildly increased levels. Prospective studies are necessary to elucidate the prognostic value of BNP in these patients.
In the last 10 years of the study, 4 patients developed supraventricular arrhythmias, and 3 required pacemaker implantation. Cumulative incidence of late symptomatic arrhythmias at 40 years was 13%. This was much lower than the incidence after surgical VSD closure at adult age (22) but higher than in patients who underwent surgery after 1980 (14,23). The incidence of arrhythmias after surgical closure of an atrial septal defect at young ages is higher (24).
Recently, interest has been generated in development of percutaneous techniques to close VSDs; however, this technique is not implemented in routine clinical practice and has even ceased in most clinics because of the high rate of post-procedural heart block (up to 8%) (4,25). This is higher than the incidence in our cohort after surgical closure at a young age (3%), as well as at an adult age (22,23).
Subjective health status assessment
A remarkable finding of the SF-36 results was that patients reported a more favorable functioning than the reference Dutch population on 7 of the 8 scales and comparable functioning on 1 scale (general health perceptions). These propitious results for VSD patients may be attributable to different frames of reference than the normal population, more adequate coping with the disadvantageous consequences of the congenital heart defect at advanced age, overcompensation, or social desirability (26).
Determinants of clinical outcome
Although numbers in our study were relatively small, patients with early post-operative arrhythmia tended to develop more late events. In another cohort of patients with surgically repaired VSD, transient and complete heart blocks were a risk factor for late mortality (27). In that study, a substantial percentage of deceased patients had not received a pacemaker because either pacemakers had not been developed at that time or they were only newly available. The relation between early post-operative arrhythmias and arrhythmias during follow-up was also described after surgery at young age for other congenital heart defects (24,28). The relation may be explained by surgical damage to the conduction system or by post-operative scar tissue and fibrosis. Early arrhythmias might therefore reveal patients at risk for late complications.
A simple concomitant cardiac lesion and longer aortic cross-clamp time during surgery were determinants of late events. The fact that patients with a concomitant cardiac lesion have a greater risk of events appears directly related to the relatively higher number of reinterventions in this group.
Although the number of patients was relatively small, we report results of a longitudinal follow-up of consecutive patients without selection bias related to disease severity. After a median follow-up of 36 years, we gathered medical information on 84% of the eligible patients. We found no significant differences in baseline characteristics between participating and nonparticipating patients. Therefore, we believe that we have minimized selection bias.
Diagnostic methods have changed over time. For comparisons of echocardiographic data with previous studies, we had to use the same methods as were used in the past, which may not be considered current state of the art. However, we also performed and reported innovative diagnostic methods available in the current era.
Survival up to 40 years after successful surgical VSD closure is good but slightly lower than in the general Dutch population. Although many patients have been discharged from routine follow-up at outpatient clinics, morbidity is substantial, especially in patients with nonisolated VSD. There is concern about systolic LV and RV dysfunction, which were observed in 21% and 17% of patients, respectively, and about occurrence of aortic regurgitation, which almost doubled over the last 20 years of the study. Early post-operative arrhythmia is a borderline determinant and aortic cross-clamp time a significant determinant of late events. Therefore, clinical follow-up with long intervals seems advisable. Despite the reported morbidity, the subjective health status is excellent.
COMPETENCY IN MEDICAL KNOWLEDGE: Surgical closure represents the standard approach to management of patients with congenital VSD, but the incidence of post-operative right ventricular systolic dysfunction has increased over time.
TRANSLATIONAL OUTLOOK: Longer-term follow-up studies of a larger number of patients could better define the risks of heart failure, arrhythmias, and death in the years after repair of congenital VSD.
The authors thank Elisabeth E.M. Konings and Maarten A. Slager for their contribution to data collection and Sara J. Baart for her contribution to statistical analysis.
For an expanded Methods section, please see the online version of this article.
This study was supported by a grant from the Dutch Heart Foundation (grant number 2009-B-073). The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- confidence interval
- hazard ratio
- implantable cardioverter-defibrillator
- interquartile range
- left ventricular
- N-terminal pro–B-type natriuretic peptide
- peak o2
- peak oxygen consumption
- right ventricular
- 36-item short-form health survey
- sinus node disease
- ventricular septal defect
- Received September 30, 2014.
- Revision received February 23, 2015.
- Accepted February 24, 2015.
- American College of Cardiology Foundation
- van der Linde D.,
- Konings E.E.,
- Slager M.A.,
- et al.
- Hoffman J.I.,
- Kaplan S.
- Minette M.S.,
- Sahn D.J.
- Roos-Hesselink J.W.,
- Meijboom F.J.,
- Spitaels S.E.,
- et al.
- Meijboom F.,
- Szatmari A.,
- Utens E.,
- et al.
- van Rijen E.H.,
- Utens E.M.,
- Roos-Hesselink J.W.,
- et al.
- Kugler J.
- Lang R.M.,
- Bierig M.,
- Devereux R.B.,
- et al.
- Rudski L.G.,
- Lai W.W.,
- Afilalo J.,
- et al.
- Kempny A.,
- Dimopoulos K.,
- Uebing A.,
- et al.
- Menting M.E.,
- Eindhoven J.A.,
- van den Bosch A.E.,
- et al.
- Klitsie L.M.,
- Kuipers I.M.,
- Roest A.A.,
- et al.
- Mongeon F.P.,
- Burkhart H.M.,
- Ammash N.M.,
- et al.
- Bol Raap G.,
- Meijboom F.J.,
- Kappetein A.P.,
- Galema T.W.,
- Yap S.C.,
- Bogers A.J.
- Cuypers J.A.,
- Opić P.,
- Menting M.E.,
- et al.
- Dumitrescu A.,
- Lane G.K.,
- Wilkinson J.L.,
- Goh T.H.,
- Penny D.J.,
- Davis A.M.
- Cuypers J.A.,
- Eindhoven J.A.,
- Slager M.A.,
- et al.