Send letter to journal:
Re: Pulse Oximetry – not a sensitive screening tool for CHD

Dear Editor,

Thangaratinam et al(1) should be complimented on the well conducted systematic review on the accuracy of pulse oximetry in screening for congenital heart disease (CHD) in asymptomatic neonates. It needs to be stressed that more than 60% of CHD is actually non-cyanotic and pulse oximetry would offer little if any assistance in their detection. Clinical detection of cyanosis is highly clinician dependent and mild cyanosis is likely to be missed by the relatively inexperienced. Routine pulse oximetry would certainly avoid this risk. Indeed low oxygen saturations in a well neonate are likely to indicate an underlying cyanotic heart disease/duct dependent condition. The sensitivity however will depend on the timing of pulse oximetry. If done too early pulse oximetry is likely to reveal normal oxygen saturations even in the presence of cyanotic heart disease because of a patent ductal circulation.

In our hospital pulse oximetry is not a part of routine pre-discharge neonatal check. It is however an important component of the assessment of any neonate with a heart murmur. The detection of oxygen saturations of 95% or less in such a neonate is an indication for urgent echocardiography(2). A review of our practice revealed that in the last one year the only babies who were noted to have low oxygen saturations(with or without a murmur) were clinically unwell and symptomatic with signs of respiratory distress. Echocardiography confirmed a diagnosis of persistent pulmonary hypertension in these cases. None of the babies picked up to have a heart murmur on the post-natal wards were found to have low oxygen saturations. As per our departmental protocol all neonates with heart murmurs who are clinically well and have normal pulses, blood pressure, oxygen saturations (pre and post-ductal) and ECG are discharged home to be followed up in our cardiac clinic in 8-12 weeks. In the cardiac clinic repeat pulse oximetry is carried out along with echocardiography. Two of the babies who had oxygen saturations of 100% at the time of neonatal assessment were noted to have oxygen saturations below 95% when seen in this clinic at about 8-10 weeks of age. Both were confirmed to have cyanotic heart disease (Truncus arteriosus and Tetralogy of Fallot). Clearly pulse oximetry in the neonatal period had failed to detect these cases. Thus the sensitivity of pulse oximetry in the early neonatal period appears to be low even in the presence of a heart murmur. This raises doubts regarding feasibility of pulse oximetry as a screening tool for cyanotic CHD. Perhaps as the authors suggest a large well designed study is required to answer this important question.

References:

1. Thangaratinam S, Daniels J, Ewer AK, Zamora J, Khan KS. Accuracy of pulse oximetry in screening for congenital heart disease: a systematic review. Arch Dis Child Fetal Neonatal Ed 2007;92:F176-F180.