Author + information
- Received November 2, 2016
- Revision received March 21, 2017
- Accepted March 21, 2017
- Published online May 29, 2017.
- Signe H. Larsen, MD, PhDa,∗ (, )
- Morten Olsen, MD, PhDb,
- Kristian Emmertsen, MD, DMSca and
- Vibeke E. Hjortdal, MD, DMScc
- aDepartment of Cardiology, Aarhus University Hospital, Aarhus, Denmark
- bDepartment of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark
- cDepartment of Cardiothoracic Surgery, Aarhus University Hospital, Aarhus, Denmark
- ↵∗Address for correspondence:
Dr. Signe Holm Larsen, Department of Cardiology, Aarhus University Hospital, Palle Juul-Jensens Boulevard 99, 8200 Aarhus N, Denmark.
Background The treatment of congenital heart (CHD) has changed rapidly.
Objectives The authors reviewed CHD treatment through a 39-year nationwide population-based study on congenital heart surgery and catheter-based interventions, unbiased by referral patterns.
Methods Using medical registries, the authors identified children (<18 years of age) treated for CHD in Denmark from 1977 to 2015, their need for reinterventions, and their long-term survival. Ten controls per patient, matched by sex and year of birth, allowed comparison with the background population. Survival was described using Kaplan-Meier curves.
Results A total of 9,372 patients underwent 11,968 cardiac surgeries and 1,912 catheter-based interventions. Median age at first procedure decreased from 3.4 years (5th and 95th percentiles: 0.01 to 15.4 years) in 1977 to 1989 (period 1), 0.8 years (5th and 95th percentiles: 0.003 to 13.8 years) in 1990 to 2002 (period 2), and to 0.6 years (5th and 95th percentiles: 0.0 to 14.9 years) in 2003 to 2015 (period 3). More patients were born preterm (<37 weeks) in period 3 compared with those in period 1 (18.5% vs. 6.7%). Catheter-based interventions, not recorded before 1990, were increasingly used as the initial procedure in 5.8% of patients in period 2 and 25.9% of patients in period 3. An increasing part of the population did not undergo surgery at all (4.8% in period 2; 24.0% in period 3). Thirty-day survival increased from 97% (period 1) to 98% (period 2) to 100% (period 3). Ten-year survival increased from 80% (period 1) to 87% (period 2) to 93% (period 3). Compared with the background population, CHD was associated with lower survival in all 3 time periods.
Conclusions Interventional treatment of CHD has evolved from 1977 to 2015 and is now performed on younger and more preterm patients, often with catheter-based interventions. However, compared with the background population, survival remains significantly reduced.
The treatment of congenital heart disease (CHD) has changed rapidly. In 1938, Robert Gross was the first to ligate a patent arterial duct (1). In 1944, the Blalock-Taussig shunt (2) introduced palliation of ductal-dependent congenital heart defects, whereas the development of the heart-lung machine in 1953 (3) made it possible to operate directly on the heart. The introduction of prostaglandin in 1974 (4) allowed clinicians to stabilize children with complex CHD before surgery. Echocardiography, with the development of 2-dimensional views and color Doppler (5), provided cardiologists with improved diagnostics. New surgical techniques, including the Fontan (6) and Norwood (7) procedures, enabled staged reconstruction of infants with single ventricles. Catheter-based interventions started in 1966 with balloon atrial septostomy (8), and today these interventions include balloon dilation, stenting, device closures, and transcatheter valve implantation (9). Overall, treatment is now available for most CHDs, and early survival has increased to nearly 100%; thus, the focus today is on long-term outcomes.
In Denmark, 5 cardiothoracic surgery departments initially conducted treatment for CHD. In recent years, the treatment has been performed in just 2 centers: Aarhus University Hospital (Aarhus, Denmark) and Rigshospitalet (Copenhagen, Denmark). Every Danish citizen has a unique social security number that is used in the health care system from birth until death. This information is collected in nationwide health registries (10,11) and provides a unique opportunity to report long-term outcomes.
Population-based long-term outcomes following congenital heart surgery have been reported in Norway (12,13) and Finland (14,15). However, this is the first nationwide study that describes the treatment of CHD including both congenital heart surgery and catheter-based interventions over a 39-year period without bias from referral patterns.
The Danish National Registry of Patients (10) was used to identify children (<18 years of age) treated for CHD in Denmark from January 1, 1977, to December 31, 2015. The International Classification of Diseases-Eighth Revision was used from 1977 to 1993, and the International Classification of Diseases-Tenth Revision was used from 1994 to 2015. Congenital heart surgery was defined as procedures with codes 30 to 33 and codes KF, except for codes specific for catheter-based interventions. Diagnostic catheterizations were not included in the analyses. Specific codes for catheter-based interventions were first introduced to the classifications with aortic balloon valvotomy in 1987, dilation of coarctation and mitral and pulmonary balloon valvotomy in 1990, and atrial septostomy in 1994. Catheter-based interventions before 1990 were not included in the analyses. Finally, reoperations before discharge of initial procedure, extracorporeal membrane oxygenation, use of aortic balloon pump, myocardial biopsies, pacemaker implantations, and ablation procedures were excluded.
The specific CHD of each patient was identified from the primary diagnosis of each hospital contact. When a patient had more than 1 CHD, the most severe defect was chosen according to a hierarchy published previously (16,17). Furthermore, the diagnosis of each patient was subdivided into simple and complex defects (12). Gestational age was identified in the Danish Medical Birth Registry. Prematurity was defined as a gestational age <37 weeks. New operations or catheter-based interventions were defined as new procedures performed during a new admission. All patients were followed from date of surgery or intervention until death, emigration, or end of follow-up (February 24, 2016) according to the Danish Civil Registration System (11).
To compare survival in the CHD population with the Danish background population, the Danish Civil Registration System (11) was used to identify 10 controls per patient, matched by sex and year of birth. The controls were chosen so they were alive at the date of first procedure for each patient. Statistics Denmark was used to identify the number of live births in Denmark from 1977 to 2015 (18).
The study was approved by the Danish Data Protection Agency (record number 2015-41-4034). Because no patients were directly involved in the study, informed consent was not indicated per Danish law.
Data are presented as mean ± SD or median (5th and 95th percentiles), as appropriate. To describe changes over time, the study period was divided into 3 periods: period 1 was defined as 1977 to 1989, period 2 was defined as 1990 to 2002, and period 3 was defined as 2003 to 2015. We described survival and freedom from reinterventions and death using Kaplan Meier curves with 95% confidence intervals (CIs). The 95% CIs were calculated by using the asymptotic variance of ln[−ln S (t)], where sums were calculated over j | tj ≤ t (19). Short-term mortality was defined as death within 30 days following the first procedure. Long-term survival was defined as overall survival including short-term mortality at the end of follow-up. One patient was excluded from the survival analysis due to incomplete follow-up data. Stata SE version 14.0 (StataCorp, College Station, Texas) was used for the analyses.
A total of 9,372 patients underwent 11,968 cardiac surgeries and 1,912 catheter-based interventions in Denmark from 1977 to 2015. During the study period, 216 emigrated, and 1,331 died. Median follow-up was 14.2 years (5th and 95th percentiles: 0.03 to 35.8 years) with a maximum follow-up of 39.1 years.
The study population represented 3.9‰ of live births in Denmark (Table 1). Over time, median age at first procedure decreased, and more patients were born preterm in the latter period compared to those in the first period (Table 1). The pattern of surgery versus catheter intervention changed over time to include more catheter-based interventions (Table 2). Summaries of the type of surgery (Table 3) and type of catheter-based interventions (Table 4) revealed changing patterns as well.
Survival, new cardiac surgeries, and catheter-based interventions
Overall survival for simple and complex CHD in the 3 time periods compared with the background population varied between patients with simple versus complex CHD (Central Illustration). Short- and long-term survivals in different time periods are listed in Table 5. Compared with the background population, both simple and complex CHD was associated with lower survival in all 3 time periods.
Freedom from death or reinterventions, including cardiac surgery or catheter-based intervention, is presented in Figure 1. Thirty-day freedom from death or reintervention increased from 82% (95% CI: 81% to 84%) to 90% (95% CI: 89% to 91%) and finally to 99% (95% CI: 99% to 99%) in periods 1 through 3, respectively. One-year freedom from death or reintervention slowly increased from 71% (95% CI: 69% to 73%) in period 1 and 74% (95% CI: 72% to 75%) in period 2 to 85% (95% CI: 84% to 86%) in period 3. A similar pattern was seen for 10-year freedom from death or reintervention, with little change in the first 2 periods (61% [95% CI: 59% to 63%] in 1977 to 1989 and 63% [95% CI: 62% to 65%] in 1990 to 2002), then increasing to 76% (95% CI: 75% to 78%) in 2003 to 2015.
Trends were similar when only considering freedom from new cardiac surgeries or death. Thirty-day freedom from new cardiac surgeries or death was 82% (95% CI: 81% to 84%) in the early period, improving to 90% (95% CI: 89% to 91%) in the middle period, and finally to 99% in the most recent period (95% CI: 99% to 99%). One-year freedom from new cardiac surgeries or death was 71% in period 1 (95% CI: 69% to 73%) and 74% in period 2 (95% CI: 73% to 76%), with an increase to 87% (95% CI: 86% to 88%) in 2003 to 2015. Ten-year freedom from new cardiac surgeries or death was 61% in period 1 (95% CI: 59% to 63%), 65% in period 2 (95% CI: 63% to 66%), improving to 80% (95% CI: 78% to 81%) in period 3.
This was the first study to report on nationwide short- and long-term outcomes of surgery and catheter-based treatment of CHD in children. The lack of similar studies makes direct comparisons difficult. Studies from Norway (12,13) and Finland (14,15) reported long-term outcomes following congenital heart surgery in large population-based cohorts. We previously reported long-term outcomes following catheter-based interventions at our institution (20), which serves a little more than one-half of the Danish population for the treatment of CHD.
During the 3 time periods, the age at first operation decreased and the frequency of prematurity increased similar to other reports (12,14). This probably reflected advances in surgical techniques and improved neonatal care, enabling treatment of still more complex lesions in younger children. Consequently, the proportion of the birth cohort treated for CHD increased between periods 1 and 2. However, no further increase was observed between periods 2 and 3, and there were fewer interventions for complex disease in the latter period. This was probably the result of the introduction of second trimester screening for CHD in the third period, enabling a parental decision to terminate pregnancy. Fewer complex patients might also explain why we found a higher freedom from new interventions in the last period. A study from the southern part of Denmark (21) showed that during 1995 to 2008, an increased number of severe congenital anomalies were diagnosed prenatally, leading to frequent terminations of pregnancy. A Danish nationwide study (22) demonstrated a decrease in the birth incidence of univentricular hearts from 1977 to 2009 and an increased termination rate. The Danish Health Authorities have recommended second-trimester screening for all women since 2004. The fact that many Danes choose to terminate pregnancy when expecting a child with severe CHD must be considered when comparing this population with others because it may limit generalizability to other settings.
Survival, new cardiac surgeries, and catheter-based interventions
Survival improved overall, as well as for most diagnoses, within the 3 time periods; findings that were consistent with Finnish (14) and Norwegian (12) studies. However, some diagnoses (such as univentricular heart disease) had better survival in 1977 to 1989 than in 1990 to 2002. This finding might reflect a selection bias because patients with very complex heart disease might not have been offered surgical treatment in the first period. Finally, we found that both simple and complex heart defects had decreased survival compared with the background population, a finding that has been maintained in the current era.
This study also showed the increased use of catheter-based interventions since 1990, as either the only treatment modality or an intervention combined with cardiac surgery. This partly explained why we found a higher freedom from new surgeries in the last period. However, the overall freedom from death and new interventions, including cardiac surgery and catheter-based interventions, was also significantly higher in the last period. Despite this progress, it is evident that the improvement lies within the first few years of follow-up. Thereafter, the CHD population continues to have a need for new cardiac surgeries or catheter-based interventions as well as a decreased survival. Overall, the population of patients surviving with CHD is increasing. Many of these patients have neurodevelopmental disorders (23,24) and special needs and require support in school (23,25). This should be acknowledged when planning future care for these patients.
There is always a risk for misclassification of diagnoses and procedures. Diagnoses of congenital heart defects from the Danish National Registry of Patients have been shown to have a positive predictive value of 90% when a patient has been seen at a university hospital (26). In the present study, the expected positive predicted value for the procedures was higher because the surgeon or catheter interventionalist was coding the procedures. However, compared with a previous study (20) from our institution, based on a local database, the frequency of the atrial septostomy and pulmonary valve dilation procedures in the Danish National Registry of Patients was low. The reason might be that atrial septostomies performed during the same admission as surgery have not been properly registered, thereby leading to an underestimation of these catheter-based interventions. Furthermore, coding for a large range of different catheter-based interventions was not introduced in the Danish National Registry of Patients until 1994, which is a limitation. Finally, no specific surgical code existed for an interrupted aortic arch in the International Classification of Diseases-Eighth Revision classification until 1994. Therefore, these operations were most likely misclassified as coarctation operations.
The treatment of CHD has changed over the last 39 years. Cardiac operations and catheter-based interventions are now performed on a larger portion of the birth cohort and on younger and more preterm children. The use of catheter-based interventions has increased since 1990, and some defects can now be treated with catheter-based intervention alone. The survival time for both simple and complex CHD has increased. However, survival remains significantly reduced compared with the background population and a large part of the CHD population still needs new cardiac surgeries or catheter-based interventions.
COMPETENCY IN MEDICAL KNOWLEDGE: Treatment of CHD has evolved over the past 40 years to include cardiac surgery or catheter-based interventions with nearly 100% early survival. Despite improved survival for both simple and complex CHD, many patients need subsequent interventions, and their long-term survival remains significantly reduced compared with the overall population.
TRANSLATIONAL OUTLOOK: Further studies are needed to identify patients with CHD at increased risk of adverse outcomes to improve long-term morbidity and survival.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- congenital heart disease
- confidence interval
- Received November 2, 2016.
- Revision received March 21, 2017.
- Accepted March 21, 2017.
- 2017 American College of Cardiology Foundation
- Olley P.M.,
- Coceani F.,
- Bodach E.
- Fontan F.,
- Baudet E.
- Arzamendi D.,
- Miro J.
- Erikssen G.,
- Liestol K.,
- Seem E.,
- et al.
- Raissadati A.,
- Nieminen H.,
- Jokinen E.,
- Sairanen H.
- Nieminen H.P.,
- Jokinen E.V.,
- Sairanen H.I.
- ↵Statistics Denmark. Live births by age of mother and sex of child. Available at: http://www.statistikbanken.dk/statbank55/SelectVarVal/Define.asp?MainTable=FOD&PLanguage=1&PXSId=0&wsid=cftree. Accessed March 15, 2016.
- Kalbfleisch J.D.,
- Prentice R.L.
- Larsen S.H.,
- Emmertsen K.,
- Hjortdal V.E.,
- Nielsen-Kudsk J.E.
- Wernovsky G.,
- Stiles K.M.,
- Gauvreau K.,
- et al.
- Agergaard P.,
- Hebert A.,
- Bjerre J.,
- Sorensen K.M.,
- Olesen C.,
- Ostergaard J.R.