Is it worth screening for congenital structural heart disease in newborn infants?

Cohort studies are used to discover findings with regard to specific diseases and are one of the foundations of everyday clinical research. Some researchers have extended their cohort study by using statistical samples from the general population and in most cases these investigate adult age groups. Cohort data can cover such areas as anthropometry, biobanking of blood or other tissues, organ imaging and measurements of physical activity. Extensive questionnaires explore, for example mental well-being and food intake. Examples of such studies in the field of cardiology include the Copenhagen City Heart Study,1 the Dallas Heart Study2 and the Multi-Ethnic Study of Atherosclerosis.3 Another example is the large Swedish Cardiopulmonary Bioimage Study of 30 000 people aged 50–64 years.4 Examples of population-based paediatric studies include examining the epidemiology and outcomes of out-of-hospital paediatric cardiopulmonary arrest5 and the use of magnetic resonance imaging by the Generation R study.6 Birth cohorts have included The Danish National Birth Cohort in 20017 and the Norwegian Mother and Child Cohort study in 2006.8 In this issue of Acta Paediatrica, Molin et al.9 report data on structural heart disease in twins. Their research formed part of a very large postnatal screening study, the Copenhagen Baby Heart Study, which used echocardiography and electrocardiograms. Pregnant mothers from three large Copenhagen maternity wards who were undergoing prenatal ultrasound at 18–20 weeks of gestation were given the opportunity to take part in the study, and 55% were accepted. The authors had access to echocardiography and electrocardiograms from 17 480 newborn infants who were originally included in the study from 2016 to 2018. Data on the mothers and their newborn infants were retrieved from the existing databases and electronic medical records. Newborn twins were characterised based on chorionicity. In total, 412 twins were included in the sub-study and matched 1:3 to 1236 singletons from the cohort. Pregnancy risk factors and International Classification of Diseases, Tenth Revision diagnoses were recorded. Echocardiography was recorded according to the guidelines by specially trained physicians or sonographers. The studies were reviewed by paediatric cardiologists when the findings were abnormal. Because of the varying gestational ages at birth, open ducts and interatrial communications were not characterised as abnormal. Most standard measurements were similar in the two study groups. Only two measurements remained significantly different after the multivariable analysis. The left ventricular posterior wall was thinner in the twins than in the singletons (1.82 mm vs. 1.87 mm), and the left atrial systolic diameter was lower in the twins (10.6 mm vs. 11.1 mm). The main question posed by the study was whether the prevalence of structural heart disease was higher in twins than in singletons, in particular, ventricular septal defects, mitral valve prolapses, bicuspid aortic valves and aortic valve regurgitation. In total, structural heart disease was found in 6.6% of the twins ventricular septal defect (VSD in 3.9%) and in 4.0% of the singletons (p = 0.04) (VSD in 2.3%). Since the prevalence of congenital heart disease often is given as eight per 1000 births, this study reported much higher numbers. This finding was discussed by the authors, who argued that the general application of echocardiography and electrocardiograms would detect all structural heart diseases, whether they were symptomatic or not. Thus, most mitral valve prolapses, bicuspid aortic valves and even ventricular septal defects were not symptomatic. An earlier study from Japan showed 2% prevalence of trabecular VSD with spontaneous closure of 76% within the first year, most often within the first 6 months.10 This finding of an elevated prevalence of structural heart disease was made, despite the likely influence from the healthy participant effect. This expression is based on the assumption that the parents of healthy babies are more likely to participate in this type of health screening than the parents of babies who have already been diagnosed with more or less severe disease that might require surgery. Children with severe coronary heart disease and those undergoing heart surgery in the neonatal period were underrepresented in the cohort. This means that the results may not apply to hospitalised or newborn infants with severe congenital heart disease. Findings from electrocardiograms have been published elsewhere and only minor differences between twins and singletons were noted. The authors concluded that their results did not support the need for additional systematic evaluations of cardiac parameters in newborn twins. However, the general finding of structural heart disease in newborn infants is thought-provoking and was obtained with standard echocardiography. Newer techniques, such as deformation imaging using speckle tracking, or volume measurements using biplane Simpson or 4D acquisition were not applied. We look forward to a possible later follow-up study of this population-based cohort! None.