A study of congenital cardiac disease in a neonatal population--the validity of echocardiography undertaken by a neonatologist.

OBJECTIVES

To estimate the incidence of, and profile the spread of, congenital cardiac defects, and to assess the accuracy of the echocardiographic diagnosis as performed by a neonatologist.

DESIGN

Hospital-based study.

METHODS

All neonates meeting our criterions, specifically those with a persistent murmur 48 hours after birth, underwent an echocardiographic examination to exclude an underlying congenital cardiac defect. All scans performed by the neonatologist were videotaped and reviewed by a paediatric cardiologist. We assessed concordance according to inter-rater agreement.

RESULTS

Out of 11,085 live births, there were 83 infants with a congenital cardiac defect, giving an incidence of 7.49 per 1000 live births, with 95% confidence interval from 5.88 to 9.09. Of the infants with a murmur persisting at or greater than 48 hours after birth, who had a median age of 4.5 days, 75% had a congenital heart defect, with ventricular septal defect being the commonest, encountered in 56.7% of cases. Concordance between the neonatologist and the cardiologist was good, with Cohen's Kappa coefficient being calculated at 0.68, and 95% confidence interval from 0.51 to 0.85.

CONCLUSION

The incidence of congenital cardiac malformations as determined in our hospital-based study in the United Arab Emirates is similar to that described in the Gulf region and worldwide. A persistent murmur at or greater than 48 hours after birth is strongly suggestive of an underlying congenital cardiac malformation. Our experience shows that a neonatologist appropriately trained in echocardiography can perform as well as a paediatric cardiologist. Where specialist resources are limited, this allows for early diagnosis, earlier referral if necessary, and early institution of appropriate therapy.