To examine long-term mortality in Danish patients with congenital heart defect (CHD), we performed a population-based follow-up study using nationwide registries. We identified all children born in Denmark from January 1, 1977 to January 1, 2006 from the Danish Civil Registration System. Children with a primary diagnosis of CHD, diagnosed before 1 year of age, were then identified in the National Registry of Patients. We computed cumulative mortality of patients and the background population according to birth period (1977 to 1986, 1987 to 1996, and 1997 to 2005). We identified 6,646 patients with CHDs. Overall cumulative mortality estimates in patients with CHDs at 1 year and 10 and 25 years of age were 20% (95% confidence interval [CI] 19 to 21), 25% (95% CI 24 to 26), and 28% (95% CI 27 to 30). In Danes born in the same period equivalent mortality estimates were 0.6% (95% CI 0.6 to 0.6), 0.8% (95% CI 0.7 to 0.8), and 1.3% (95% CI 1.3 to 1.3). Mortality differed substantially according to heart defect type and mortality at 10 years of age ranged from 9% (95% CI 6 to 12) in patients with atrial septal defects (n = 361) to 55% (95% CI 45 to 66) in patients with common arterial trunk (n = 78). Mortality decreased during the study period; 1-year mortality was 28% (95% CI 26 to 31) for patients born from 1977 to 1986 (n = 2,907) compared to 13% (95% CI 12 to 15) for patients born from 1997 to 2005 (n = 2,741). Mortality decreased in all heart defect type categories. In conclusion, mortality in patients with CHD was high compared to the general population, especially in infancy, but also after 10 years of age, emphasizing the need for long-term medical follow up. Mortality at 1 year of age has decreased substantially during recent decades.
More knowledge on mortality from birth to adulthood of the growing population of patients with congential heart defect (CHD) is important to evaluate treatment results beyond in-hospital mortality, to evaluate the need for additional long-term medical follow-up, and to provide information to patients and their parents. We therefore examined the mortality of Danish patients diagnosed with CHD before the age of 1 year and compared it to the mortality of the general population using nationwide databases with up to 30 years of follow-up from birth.
Methods
We used the civil registration system to identify the 1,795,298 subjects born in Denmark from January 1, 1977 to January 1, 2006 and followed them until January 1, 2007. The study period started on the date of initial availability of data from the Danish National Registry of Patients. The 10-digit unique civil registration number assigned to every Danish resident since 1968 allows for valid individual-level linkage between Danish public registries. The Danish National Health Service provides tax-supported health care for all inhabitants, guarantees free access to general practitioners and hospitals, and refunds a variable proportion of the prescription medication costs. Patients with CHD are exclusively treated at public hospitals ( http://www.sum.dk/publikationer/healthcare_in_dk_2008/index.htm ).
Using computerized data from the Danish National Registry of Patients we identified all patients in the study population with a primary discharge diagnosis of CHD before 1 year of age from January 1, 1977 to January 1, 2007. The registry is updated daily and contains information on patient civil registration number, dates of admission and discharge, and surgical procedures coded exclusively by doctors according to the International Classification of Diseases (ICD). The eighth edition of the ICD (ICD-8) was used until the end of 1993 and the 10th edition (ICD-10) thereafter. ICD-8 codes used to identify patients with CHD were 746 to 747 (except for 746.7 and 747.5 to 747.9, which were not specific to CHD) and ICD-10 codes were Q20 to Q26 (except for Q26.5 to Q26.6, which were not specific to CHD). Diagnoses of patent ductus arteriosus were considered only for mature infants (gestational age ≥37 weeks). We assigned each patient 1 CHD diagnostic code, namely the primary diagnosis from the first admission with CHD as the primary diagnosis. We grouped ICD-10 diagnostic codes according to the corresponding ICD-8 diagnostic codes to provide a uniform categorization of the study cohort during the study period ( Table 1 ).
| Diagnostic Category | ICD-8 Codes | ICD-10 Codes |
|---|---|---|
| Common arterial trunk | 7460 | Q200 |
| Transposition of great vessels | 7461 | Q203, Q205 |
| Tetralogy of Fallot | 7462 | Q213 |
| Ventricular septal defect | 7463 | Q210 |
| Atrial septal defect | 7464 | Q211 |
| Atrioventricular septal defect | 7465, 74641 | Q212, Q218B |
| Anomalies of heart valve | 7466 | Q220-Q229, Q230–Q239 |
| Other heart anomalies | 7468, 7469 | Q201, Q202, Q204, Q206, Q208, Q209, Q214, Q218, Q219, Q240–Q248, Q249 |
| Patent ductus arteriosus | 7470 | Q250 |
| Coarctation of aorta | 7471 | Q251 |
| Other anomalies of great arteries | 7472, 7473 | Q252–Q254, Q255–Q256, Q257–Q259 |
| Malformations of great veins | 7474 | Q260–Q264, Q268–Q269 |
We categorized patients into 2 main groups according to treatment (surgery or catheter-based treatment) or no treatment. We used the following codes to identify diagnoses of extracardiac defect (ECD) and chromosomal abnormalities in the Danish National Hospital Registry: ICD-8 codes 740.99 to 759.99 and ICD-10 codes Q0.00 to Q99.9, except for CHD-related codes. According to a guideline from the European Surveillance of Congenital Anomalies (EUROCAT), we disregarded isolated minor defects such as torticollis (Q68.0) or protuberant ears (Q17.3). We obtained data on gestational age from the Danish Medical Birth Registry that contains computerized data on all newborns in Denmark from 1973 onward and defined preterm birth as a gestational age <37 weeks. We retrieved data on vital status for the entire cohort through linkage with the civil registration system that has kept daily updated electronic records on date of birth, date of emigration, and exact date of death for all Danish residents since 1968.
We computed the proportion of patients who had ECDs or chromosomal abnormalities, the proportion who were born before term, and the proportion who underwent therapeutic interventions at any time during the study period, before 30 days of age, or before 1 year of age. Based on the Kaplan-Meier estimator, we computed cumulative mortality 1 year and 10 and 25 years from birth for overall patients with CHD and in subgroups according to defect type and presence of ECDs or chromosomal abnormalities and for the remaining Danish population. We computed cumulative mortality curves for patients with CHD by period of birth (1977 to 1986, 1987 to 1996, and 1997 to 2005).
Using Cox regression analysis we computed the hazard ratio as an estimate of the relative risk in patients with CHD of dying during the first year of life in the most recent birth period compared to the earliest birth period, adjusting for gender. We computed overall and type-specific hazard ratios. In patients undergoing therapeutic procedures we computed the cumulative mortality at 30 days, 1 year, and 5 years from discharge after the first therapeutic procedure. We computed hazard ratios to compare mortality during 1 year from discharge in the recent and early birth periods in those undergoing surgery before 1 year of age. We adjusted for gender and age at surgery (<30 or ≥30 days of age). Analyses were performed using STATA 10.1 (STATA Corp. LP, College Station, Texas).
Results
We identified 6,646 patients with CHD diagnosed before 1 year of age, corresponding to a prevalence at birth of 3.7 per 1,000 live births. Overall, approximately 1 of 5 had ECDs or chromosomal anomalies and 11% were born before term ( Table 2 ).
| Patients | ECD or Chromosomal Abnormality (%) | Preterm Birth (%) | First Therapeutic Procedure Before 1 Year of Age (%) | First Therapeutic Procedure After 1 Year of Age (%) (median age at first intervention) | Patients Undergoing Therapeutic Intervention According to Birth Period (%) | |||
|---|---|---|---|---|---|---|---|---|
| 1977–1986 | 1987–1996 | 1997–2005 | ||||||
| All patients with congenital heart defect | 6,646 | 22% | 11% | 40% | 12% (2.0 years) | 751 (43%) | 1,259 (49%) | 1,443 (63%) |
| Diagnostic category | ||||||||
| Common arterial trunk | 78 | 19% | 22% | 58% | 3% (1.4 years) | 6 (38%) | 17 (53%) | 24 (80%) |
| Transposition of great arteries ⁎ | 461 | 11% | 8% | 64% | 16% (1.2 years) | 87 (70%) | 131 (82%) | 154 (88%) |
| Tetralogy of Fallot | 381 | 27% | 15% | 65% | 23% (1.7 years) | 60 (79%) | 116 (87%) | 161 (94%) |
| Ventricular septal defect | 1,559 | 20% | 11% | 29% | 13% (2.1 years) | 164 (43%) | 244 (40%) | 252 (45%) |
| Atrial septal defect | 361 | 28% | 20% | 23% | 14% (3.3 years) | 25 (42%) | 51 (37%) | 58 (36%) |
| Atrioventricular septal defect | 354 | 69% | 13% | 70% | 11% (1.6 years) | 47 (68%) | 104 (79%) | 136 (88%) |
| Anomalies of heart valve | 510 | 13% | 14% | 53% | 7% (2.7 years) | 12 (46%) | 100 (52%) | 194 (67%) |
| Other anomalies of heart | 2,092 | 19% | 9% | 25% | 11% (2.6 years) | 265 (31%) | 283 (32%) | 196 (54%) |
| Patent ductus arteriosus | 300 | 21% | — | 45% | 5% (1.5 years) | 31 (62%) | 52 (44%) | 65 (49%) |
| Coarctation of aorta | 334 | 16% | 15% | 78% | 7% (3.2 years) | 35 (61%) | 117 (89%) | 132 (90%) |
| Other anomalies of great arteries | 177 | 25% | 15% | 57% | 6% (2.2 years) | 19 (44%) | 39 (66%) | 53 (71%) |
| Malformations of great veins | 39 | 26% | 13% | 56% | 3% (1.0 year) | 0 (0%) | 5 (56%) | 18 (67%) |
| Extracardiac defect or chromosomal anomalies † | ||||||||
| No | 5,191 | — | 10% | 37% | 11 (2.0 years) | 559 (40%) | 895 (44%) | 1,053 (60%) |
| Yes | 1,455 | — | 15% | 51% | 14 (1.8 years) | 192 (51%) | 364 (65%) | 390 (75%) |
⁎ Complete and congenitally corrected transposition.
One-half of patients with CHD received invasive treatment (surgery and/or catheter-based procedure; Table 2 ) and as expected the proportion of treated patients was closely associated with type of CHD. The highest treatment rates were found in patients diagnosed with tetralogy of Fallot, coarctation of the aorta, atrioventricular septal defect, and transposition of the great arteries. The proportion of patients undergoing treatment in infancy increased during the study period; approximately 20% of patients born from 1977 to 1986 received invasive treatment in infancy compared to 60% of patients born from 1997 to 2005 ( Table 3 ).
| Birth Period (Patients) | Proportion Who Underwent Therapeutic Cardiac Intervention Before 30 Days or 1 Year of Age (95% CI) ⁎ | CMPs at Specified Times From Discharge After First Intervention (95% CI) † | |||
|---|---|---|---|---|---|
| 30 Days | 1 Year | 30 Days | 1 Year | 5 Years | |
| 1977–1986 (1,767) | 4% (3–4) | 21% (19–23) | 25% (21–30) | 34% (29–39) | 42% (37–47) |
| 1987–1996 (2,591) | 13% (12–14) | 37% (35–38) | 15% (13–17) | 23% (20–25) | 27% (24–30) |
| 1997–2005 (2,288) | 22% (20–24) | 60% (58–62) | 7% (6–8) | 13% (11–15) | 14% (12–16) |
⁎ Denominators are all patients born in the same period.
† Mortality in patients who underwent a therapeutic intervention before 1 year of age.
In this cohort of patients diagnosed with CHD before 1 year of age, 1/5 died within the first year of life. Mortality remained high throughout the follow-up period; an additional 5% died before 10 years of age and another 3% died from 10 to 25 years of age. In the comparison cohort (the remaining Danish population) only 0.6% died as infants and an additional 0.7% had died by 25 years of age. Mortality varied by type of CHD and presence of ECD or chromosomal abnormalities ( Table 4 ). During the study period overall cumulative mortality at 1 year of age decreased from 28% (95% confidence interval [CI] 26 to 31) in the early period to 13% (95% CI 12 to 15) in the most recent period ( Figure 1 ). The corresponding overall mortality rate ratio was 0.42 (95% CI 37 to 49) when comparing patients from the most recent birth period to patients born from 1977 to 1986. Lower mortality rate ratios were also consistently found in the defect-specific analyses ( Table 5 ). Of patients born in the early period (1977 to 1986) who underwent a cardiac therapeutic intervention before 1 year of age, 34% (95% CI 29 to 39) died within 1 year after the first intervention. In comparison, only 13% (95% CI 11 to 15) died within 1 year after the first intervention in the late period (1997 to 2005; Table 3 ). This corresponded to an overall mortality rate ratio of 0.25 (95% CI 0.20 to 0.32).
