Cardiovascular malformations account for up to 6–10% of infant deaths and 20–40% of deaths caused by congenital malformations.1 The most deadly malformations are those where a persistent ductus arteriosus is necessary for survival, which are called duct-dependent malformations and are present in 1–1.8 babies per 1000 live births.1 Across the world the average inpatient stay following a normal delivery is steadily shortening, putting these babies at particular risk since the timing of ductal constriction causing clinical signs of circulatory insufficiency is increasingly occurring after discharge from hospital.
In parallel to the shortening of maternity stay there has been a significant decrease in the proportion of critical congenital heart disease detected before hospital discharge.2 Often these babies have no murmurs or other clinical signs that are picked up on the routine neonatal physical examination, and among those that leave hospital with their critical heart disease undetected 18–38% die at home, constituting around 5% of all infants with critical congenital heart disease.1,3 Those who survive collapse at home often have severe morbidity with signs of hypoxic/ischaemic brain injury and periventricular leukomalacia is reported in up to 39% of neonates with critical congenital heart disease leading to concern about long-term neurological handicap.1,4
Can an experienced paediatrician or midwife recognise cyanotic heart disease in the absence of a murmur?
Unfortunately, the detection of severe cyanotic disease in neonates is particularly difficult because of the high oxygen affinity of foetal haemoglobin. A human eye needs about 50g/litre of reduced haemoglobin in arterial blood in order to detect visible cyanosis and with foetal haemoglobin you need to have an arterial pO2 of <4.1kPa in order to have arterial oxygen saturation of <72%. Systematic studies have documented the fallibility of visual detection of cyanosis,5 and with the shortening maternity stays there has been a significant fall in the detection of critical cyanotic heart disease to the extent that 44% of babies with transposition of the great arteries left hospital undetected from hospitals not using pulse oximetry screening.1 The logic of this observation is to try to improve detection of significant arterial desaturation by routinely using a pulse oximeter, the feasibility of which has been shown in several pilot studies.
The benefits of pulse oximetry screening
The optimal cut-offs for a pulse oximetry screening protocol were defined in a systematic study comparing pulse oximetry readings in 66 consecutive infants with duct-dependent congenital heart disease to 200 normal newborns with ultrasound confirmed normal cardiac anatomy.6 This study showed that the use of two rather than one criteria optimised detection: <95% saturation and/or a difference of >3% between right hand and foot saturation. It is noteworthy that some infants with duct-dependent arch obstruction had a foot saturation >95% but still had a >3% differential between right hand and foot. Furthermore some children with the combination of tricuspid atresia and/or transposition with either a coarctation or an interrupted arch had foot saturations >95% but lower hand saturations, and in total six additional diagnoses of critical duct-dependent heart disease was obtained by including the >3% differential criterion, as compared with only relying on post-ductal oxygen saturations for screening.6
Repeating pulse oximetry measurement three times with short intervals before declaring a definite positive screening reduces the false positive rate.1 These criteria were subsequently applied in a large prospective regional neonatal screening study on 39,821 newborns, where the results could be compared with 108,604 children born in the same time period in regions that were not using pulse oximetry screening, but referred their children with neonatal heart disease to the same tertiary hospital. This study showed that blind physical examination alone was significantly inferior to physical examination plus pulse oximetry in the detection of critical heart disease and that total pre-discharge detection of duct-dependent heart disease in the screening region was 92% as compared with 72% in the non-screening regions (p=0.0025).
The false positive rate of pulse oximetry screening was substantially lower than that of neonatal physical examination, 0.17% versus 1.90%, and the increase in echocardiography referrals was negligible. In fact most of the “false positives” had other pathology causing desaturation and benefitting from early medical interventions such as sepsis, pneumonia and pneumothorax, or less severe cyanotic heart disease. Furthermore children who had been discharged home with undetected critical heart disease had significantly higher incidence of severe preoperative acidosis (p=0.0025) and higher surgical mortality than those who had been diagnosed before discharge (p=0.0054).1 Cost–benefit analysis showed that pulse oximetry screening only triggered 2.3 cardiac ultrasounds with normal cardiac findings for each true case with critical duct-dependent heart disease found, and thus the estimated cost is offset by costs for resuscitation and ambulance transport for each infant that collapses at home with duct-dependent heart disease. Consequently the introduction of generalised screening was estimated to be cost neutral.1 The same size of statistically significant benefit from pulse oximetry screening has been demonstrated in large prospective pulse oximetry screening studies, both in Britain (20,055 infants)7 and in China (122,738 infants)8 showing that the benefit is equally great, and the false positive rate low, both in high-income and low-income countries, and in total there are five large prospective population screening studies that have demonstrated benefit from pulse oximetry screening.1,7–10 Two reviews in 2011 already suggested that it was time to implement routine neonatal pulse oximetry screening.11,12
Implementation of generalised pulse oximetry screening
In 2011 the Secretary of Health in the USA, Kathleen Sebelius, recommended that pulse oximetry screening for critical heart disease should be included in the uniform screening panel: a recommendation that was endorsed by the American Academy of Pediatrics.13 A rolling implementation in the United States has spawned several smaller studies that have shown that the same screening criteria can be used on premature babies14 and at 2700 feet altitude.15 They confirm a low false positive rate, and that the procedure imposes a low burden on nursing staff and has a low cost.16 In Scandinavia there have been no recommendations from central government, only from associations of paediatricians; however, demand from the medical profession has led to very high coverage with pulse oximetry screening. In June 2013 implementation of screening in Finland was 97%, in Sweden 91% (planned to be 100% in 2014), and in Norway 90%17 and uniform Nordic guidelines using both right hand and foot, that is, pre- and post-ductal screening, before 24 hours of age have been proposed.17 Importantly, new studies have also confirmed that babies who fail pulse oximetry screening without having congenital heart disease mostly have other conditions that benefit from early medical detection such as sepsis, pneumonia and undiagnosed pneumothorax.1,16,18
Do you need pulse oximetry screening if you perform prenatal cardiac ultrasound?
It has been argued from foetal sonographers that pulse oximetry screening would be unnecessary with increasing expertise in the prenatal ultrasound detection of severe congenital heart disease. However, although individual tertiary centres of excellence in large cities claim very high prenatal detection rates (99% tertiary centre versus 35% in outlying hospitals) among babies operated at the tertiary centre,19 this study included no ascertainment of children with critical heart disease dying at home, or not making it to the referral centre, so both figures are likely overestimates.
The reality from geographically based cohort studies, including all hospitals in a region, suggest an average prenatal detection rate of 42–43%,20 which with maximal use of several cardiac views and colour-Doppler can be brought to an average of 66–67%.7,21 Notably, there are some important duct-dependent lesions such as transposition of the great arteries, pulmonary atresia with VSD, coarctation of the aorta and interrupted aortic arch that have much lower prenatal detection rates than the duct-dependent lesions with single ventricle physiology.3 The one study that has systematically compared prenatal detection with pulse oximetry screening has concluded that pulse oximetry conferred significant additional benefit.7 A formal Health Technology Assessment confirmed earlier results about cost effectiveness and concluded, “pulse oximetry is a simple, safe, feasible test that is acceptable to parents and staff and adds value to existing screening”.22
Furthermore, for religious or cultural reasons some women do not participate in prenatal ultrasound scanning,23 whereas uptake of pulse oximetry screening is 99.8%.1 There is clearly benefit in a prenatal cardiac ultrasound diagnosis of a potentially duct-dependent cardiac malformation, since it allows for planning delivery at a suitable medical facility, with prompt postnatal cardiac assessment.
However, there is no doubt that in terms of cost effectiveness it should be a higher priority to implement universal pulse oximetry screening in the first 24 hours of life, before discharge from the birthing facility, or at home if a home delivery is planned.23 This is because the detection rate of duct-dependent heart disease is higher with pulse oximetry screening and routine neonatal examination than that with prenatal cardiac ultrasound, it is faster taking only a few minutes as compared to around 30 minutes for a foetal cardiac scan, it uses inexpensive equipment as compared to very expensive ultrasound machines and it can be satisfactorily carried out by nurses or nursery nurses after only brief training1,6 rather than needing highly skilled ultrasonographers.
Benefit in remote medical facilities
The fact that it can be accurately carried out by nurses or nurse-helpers is an additional benefit in remote medical facilities where a physician may not be easily accessible and is a great help in deciding whether a newborn needs to be transported to a specialist centre. It can be combined with both high-tech and low-tech telemedicine.
For example in Brazil a programme introducing pulse oximetry screening in remote maternity hospitals which had no access to paediatric cardiology expertise and then transmitting simple ultrasound pictures via ordinary mobile phones to the specialist centre, has been very successful (Sandra Mattos, personal communication).
Possible further improvements of pulse oximetry screening
The types of critical duct-dependent congenital heart disease that may be missed by pulse oximetry screening are not surprisingly the non-cyanotic lesions with aortic arch obstruction: coarctation of the aorta, and interrupted aortic arch with VSD, where approximately 50% are detected with current pulse oximetry screening. A possible tool to increase the detection of these lesions would be to utilise the other measurement available on many, but not all, pulse oximeters, namely a digital readout of the peripheral perfusion index (PPI). The PPI reflects stroke volume in pulsatile arterial flow and is decreased in conditions with low systemic cardiac output. A survey of PPI in 10,000 normal newborn babies established that the fifth centile value was 0.70, and the first centile value 0.50.24 Among 11 infants with duct-dependent systemic circulation, the lowest limb PPI was <0.7 in six (56%) and all were below the interquartile range of 1.18–2.5. A PPI of <0.7 gave an odds ratio of 23.75 for the presence of left heart severe obstructive disease.24 Thus this measurement has the potential to increase the sensitivity for the detection of left heart obstructive disease, but how this measurement could be incorporated in pulse oximetry screening without causing an undue increase in false positive rate awaits further studies.
Generalised pulse oximetry screening is a significant advance for the detection of immediately life threatening congenital heart malformations. It easily justifies the modest initial costs by reducing costs for resuscitation and emergency transport with intensive care ambulances, shorter perioperative need for intensive care beds, better postoperative survival and quite likely a lower long-term morbidity in neurological sequelae after neonatal circulatory collapse with unrecognised duct-dependent heart disease.
- de Wahl Granelli A et al. Impact of pulse oximetry screening on the detection of duct dependent congenital heart disease: a Swedish prospective screening study in 39,821 newborns. BMJ 2009;338:a3037.
- Mellander M, Sunnegardh J. Failure to diagnose critical heart malformations in newborns before discharge–an increasing problem? Acta Paediatr 2006;95:407–13.
- Wren C, Reinhardt Z, Khawaja K. Twenty year trends in diagnosis of life-threatening neonatal cardiovascular malformations. Arch Dis Child Fetal Neonatal Ed 2008;93:F33–5.
- Mahle WT et al. Role of pulse oximetry in examining newborns for congenital heart disease: a scientific statement from the AHA and AAP. Pediatrics 2009;124:823–36.
- O’Donnell CP et al. Clinical assessment of infant colour at delivery. Arch Dis Child Fetal Neonatal Ed 2007;92:F465–7.
- de Wahl Granelli A et al. Screening for duct-dependant congenital heart disease with pulse oximetry: a critical evaluation of strategies to maximize sensitivity. Acta Paediatr 2005;94:1590–6.
- Ewer AK et al. Pulse oximetry screening for congenital heart defects in newborn infants (PulseOx): a test accuracy study. Lancet 2011;378:785–94.
- Zhao QM et al. Pulse oximetry with clinical assessment to screen for congenital heart disease in neonates in China: a prospective study. Lancet 2014;384:747–54.
- Meberg A et al. Pulse oximetry screening as a complementary strategy to detect critical congenital heart defects. Acta Paediatr 2009;98:682–6.
- Riede FT et al. Effectiveness of neonatal pulse oximetry screening for detection of critical congenital heart disease in daily clinical routine–results from a prospective multicenter study. Eur J Pediatr 2010;169:975–81.
- Hoffman JI. It is time for routine neonatal screening by pulse oximetry. Neonatology 2011;99:1–9.
- Kemper AR et al. Strategies for implementing screening for critical congenital heart disease. Pediatrics 2011;128:e1259–67.
- Mahle WT et al. Endorsement of Health and Human Services recommendation for pulse oximetry screening for critical congenital heart disease. Pediatrics 2012;129:190–2.
- Manja V et al. Critical congenital heart disease screening by pulse oximetry in a neonatal intensive care unit. J Perinatol 2015;35:67–71.
- Han LM et al. Feasibility of pulse oximetry screening for critical congenital heart disease at 2643-foot elevation. Pediatr Cardiol 2013;34:1803–7.
- Garg LF et al. Results from the New Jersey statewide critical congenital heart defects screening program. Pediatrics 2013;132:e314–23.
- de Wahl Granelli A et al. Nordic pulse oximetry screening – implementation status and proposal for uniform guidelines. Acta Paediatr 2014.
- Bhola K, Kluckow M, Evans N. Postimplementation review of pulse oximetry screening of well newborns in an Australian tertiary maternity hospital. J Paediatr Child Health 2014;50:920–5.
- Johnson LC et al. Prenatal and newborn screening for critical congenital heart disease: findings from a nursery. Pediatrics 2014;134:916–22.
- McBrien A et al. Impact of a regional training program in fetal echocardiography for sonographers on the antenatal detection of major congenital heart disease. Ultrasound Obstet Gynecol 2010;36:279–84.
- Eggebo TM et al. Routine use of color Doppler in fetal heart scanning in a low-risk population. ISRN Obstet Gynecol 2012;2012:496935.
- Ewer AK et al. Pulse oximetry as a screening test for congenital heart defects in newborn infants: a test accuracy study with evaluation of acceptability and cost-effectiveness. Health Technol Assess 2012;16:v–xiii,1–184.
- Lhost JJ et al. Pulse oximetry screening for critical congenital heart disease in planned out-of-hospital births. J Pediatr 2014;165:485–9.
- Granelli A, Ostman-Smith I. Noninvasive peripheral perfusion index as a possible tool for screening for critical left heart obstruction. Acta Paediatr 2007;96:1455–9.