OPINION Review of pulse oximetry screening for critical
congenital heart defects in newborn infants
Andrew K. Ewer
Purpose of review
The concept of using pulse oximetry as a screening method to detect undiagnosed critical congenital
heart defects (CCHD) in asymptomatic newborns was first explored over 10 years ago. A number of
studies were subsequently reported, which initially involved relatively small numbers of patients, low
prevalence of CCHD and heterogeneous methodology. As a consequence, the majority of clinicians felt
the case for routine pulse oximetry screening had not been proven.
In the last 3 years, four European studies reporting the test accuracy of routine pulse oximetry screening, and
involving over 150000 babies, have strengthened the argument. A systematic review and meta-analysis of
almost 230000 screened babies has also recently been published which reported high specificity, moderate
sensitivity and a low false-positive rate. In addition, acceptability to parents and staff, cost-effectiveness and
feasibility of implementing screening outside the research context have also been reported.
Pulse oximetry screening is a highly specific, moderately sensitive test, which is acceptable to parents and
staff, likely to be cost-effective and fulfils the criteria for universal screening. Routine screening for CCHD
using pulse oximetry is being increasingly supported and was added to the recommended uniform
screening panel in the USA in 2011.
critical congenital heart defects, detection, newborn infant, pulse oximetry, screening
critical congenital heart defects (CCHDs) (defects
leading to death or requiring invasive intervention
within 28 days of life) is between two and three per
1000 livebirths [1,2]. Early detection of CCHDs
reduces the risk of acute cardiovascular collapse,
acidosis and death and improves outcome [2,3
Screening for CCHDs has previously relied on
antenatal ultrasound and postnatal examination
but both have a relatively low detection rate
may be discharged with undiagnosed critical defects
Pulse oximetry is a well-established, accurate,
of hypoxaemia. On the basis of the rationale
that clinically undetectable hypoxaemia is present,
to some degree, in most CCHDs, the use of this
technique as a screening method for early detection
was first reported over 10 years ago [5,6] with a
In 2007, a systematic review drew attention to the
&&,4]. It is estimated that up to a third of babies
difficulties inprecise assessmentofthetrueaccuracy
of pulse oximetry screening because of small
numbers of patients recruited, the low prevalence
of CCHDs and methodological variations reported
in the studies . In 2009, a statement on behalf of
the American Heart Association and The American
Academy of Paediatrics , which included two
further screening studies [8,9], also concluded that
‘further studies in large populations and across
a broad range of newborn delivery systems were
needed to determine if (pulse oximetry screening)
should become a standard of care’ .
More recently, additional large studies have
added significant weight to the argument for
University of Birmingham and Birmingham Women’s Hospital, Birming-
Correspondence to Dr Andrew K. Ewer, MB, ChB, MD, MRCP,
FRCPCH, Neonatal Unit, Birmingham Women’s Hospital, Edgbaston,
Birmingham B15 2TG, UK. Tel: +44 122 472 1377; e-mail: a.k.ewer@
Curr Opin Cardiol 2013, 28:92–96
Volume 28 ? Number 2 ? March 2013
universal screening [10,11,12
vided essential data on important issues regarding
screening strategies [10,11,12
parents and clinical staff [3
This article will focus on the latest evidence and
give an update on the progress of the implementa-
tion of universal pulse oximetry screening.
&,13] and have pro-
&], acceptability to
&] and cost-effective-
RECENT TEST ACCURACY STUDIES
In 2009, de-Wahl Granelli et al.  from Sweden
published the results of a test accuracy study
involving over 39000 babies. Screening included
ation (as opposed to postductal testing alone, which
had been the strategy in the majority of previous
studies) and reported a sensitivity of 62% and speci-
postnatal age of 38h with a low false positive rate
of 0.17%. Pulse oximetry demonstrated additional
detection rate of 92% of CCHD when both methods
wereusedwasmuchimprovedwhen compared with
hospitals in other regions of Sweden where pulse
oximetry screening was not performed.
In 2010, a German study by Riede et al. 
screened over 40000 babies using only postductal
saturations and reported a sensitivity of 78%,
specificity of 99.9% and a false positive rate of
0.1%. Screening was performed between 24 and
72h, but the average screening time is not reported.
In 2011, the PulseOx study from the United
Kingdom reported the test accuracy, acceptability
to parents and clinical staff and cost-effectiveness of
pulse oximetry screening in over 20000 babies
an earlier time of screening (mean postnatal age
at screening – 12h) was adopted. Sensitivity for
detecting CCHD was 75%, specificity 99.1% and
the false positive rate was relatively high at 0.8%.
&]. Once again, both preductal and postductal
Data from these recent large studies were added
to those from existing studies and in 2012 a further
systematic review and meta-analysis was performed
examining test accuracy of pulse oximetry which
involved almost 230000 babies – an addition of
over 100000 since the previous reviews. With such
a large cohort, and the subsequent improvement
in the prevalence of CCHD, the overall test accuracy
could be reported with more precision. The system-
atic review reported that pulse oximetry was
a highly specific test for CCHD (99.9%) with
moderate sensitivity (76.5%) and a low false positive
rate (0.14%). This was particularly low if the test was
performed after 24h of age (0.05%). The authors
concluded that pulse oximetry met the criteria for
universal screening. They also concluded that, in
view of the number of babies who had now been
studied, it was unlikely that any further research
would demonstrate substantially different findings.
Since the publication of the systematic review, a
further study of over 50000 babies in Poland has
been reported . Testing postductal saturations
within the first 24h (mean age at testing 7h),
a sensitivity and specificity of 78.9 and 99.9%
respectively and a very low false positive rate of
0.026% were reported, which is broadly consistent
with the findings of the systematic review.
was variable between studies and only two studies
those dying from unidentified CCHD, but this is
unlikely to affect the conclusions significantly.
Importantly, the authors of all four studies
emphasized that pulse oximetry should not replace
the existing screening strategies but should be
employed as an adjunct investigation. Riede et al.
 describes a ‘diagnostic gap’ for CCHD, that is,
babies who are missed by routine screening alone.
When pulse oximetry is included in addition
to antenatal ultrasound and physical examination
the percentage of babies with CCHD who are
identified prior to discharge is increased to over
90% (range 92–96%) [10,11,12
CHDs and significant non-cardiac conditions such
as respiratory problems or early-onset sepsis is
reported as an additional benefit. It is note-
worthy that these clinically important conditions
contributed between 37 and 70% of the false
positive groups (average 50%) [10,11,12
&&]. This review included 13 good quality studies
&] included mortality databases to identify
There is a degree of heterogeneity between the
screening protocols used in the reported studies.
? Screening newborn infants for critical congenital heart
defects using pulse oximetry is feasible and adds value
to existing screening techniques.
? The addition of pulse oximetry screening to antenatal
ultrasound and physical examination may increase
detection rates for CCHD to over 90%.
? Pulse oximetry screening has been endorsed by an
increasing number of professional bodies and universal
screening is being considered by health-care systems
across the world.
Review of pulse oximetry screening for congenital heart defects Ewer
0268-4705 ? 2013 Wolters Kluwer Health | Lippincott Williams & Wilkins
The main differences are as follows: timing of
screening; postductal or preductal and postductal
saturation measurement; saturation threshold for a
Timing of screening
There is no doubt that later screening (>24h) has
a lower rate of false positive results than early screen-
ing (<24h); this has been a consistent finding in
published reviews [2,7,16
Polish study (which has not been included in a sys-
tematic review) screened at an average age of 7h and
reported one of the lowest false positive rates .
A high false positive rate could potentially
have significant impact on clinical services and
is, therefore, an important consideration. How-
ever, a number of countries (including the United
Kingdom) have an increasing trend towards earlier
discharge, and discharge before 24h is often the
norm for uncomplicated deliveries. In this situation
will place an unmanageable pressure on resources
and will, therefore, be untenable. In the later screen-
ing studies up to half of the babies with CCHDs who
prior to the onset of screening, and in de-Wahl
in hospital prior to diagnosis . In Riede ’s study
18/36 babies presented with symptoms prior to
screening, although details of the severity of symp-
toms are not described . In the PulseOx study
which screened at an average time of 12h no babies
are reported as presenting prior to screening taking
place and the only case of cardiovascular collapse
was in a baby who was discharged home after
passing all screening tests [12
Polish study, which screened at an average of 7h,
reported that 68% of babies with undiagnosed
CCHD presented with symptoms prior to screening,
which suggests an exceptionally high degree of
close observation of babies in the postnatal nursery,
although the number of babies presenting with
collapse prior to diagnosis is not reported .
It is also worth considering that babies with
important non-cardiac conditions such as con-
genital infection and pulmonary hypertension are
more likely to present in the first 24h and the early
identification of these problems may also reduce
morbidity and mortality.
In the end, a balance needs to be struck between
timely diagnosis of life-threatening conditions and
an excess of false positives. It should be remembered
that all false positives are babies with low oxygen
saturations and in principle no baby should have
persistent unexplained hypoxemia.
&&]. Interestingly, the recent
&]. In contrast, the
Preductal or postductal saturation
The majority of pulse oximetry studies (60%)
used postductal measurements only, and in meta-
analysis there was no difference in sensitivity
between postductal testing and preductal testing
included almost twice as many patients and the
sensitivity estimates in the two sub-groups were
too imprecise to make any inference [16
of the recent studies using preductal and postductal
testing reported that, if postductal testing alone
had been used, babies with CCHD would have been
missed; three babies in Ewer’s study [12
de-Wahl Granelli’s .
&] and one in
Saturation threshold for a positive test
The lower limit for a positive test varies between
92 and 95% with the majority of studies using
less than 95%. The numbers of patients in the
different subgroups were too small to identify sen-
The studies employing preductal and postductal
saturations used two different thresholds – both less
than 95% or a difference of more than 3% and either
less than 95% or a difference of more than 2%
former threshold (although the patients were
screened much later and there was an additional
retest) and sensitivity of the latter was higher, but
again numbers were too small to identify differences
in sensitivity with precision [16
&]. The false positive rate was lower with the
When new neonatal screening procedures are
introduced it is important to consider their accept-
ability to both parents and clinical staff and also the
psychological impact that screening may have on
parents, particularly raising anxiety that the child
has a serious health condition. Screening accept-
ability may have an effect on uptake, and the
psychological effects of an inaccurate result may
extend over a long period of time. Only two studies
have addressed the issue of acceptability. In the
recent Polish study all the parents of screened babies
were asked two questions – one pertaining to the
easeof deciding to participate and the secondasking
if the parent felt that screening should be adopted as
a routine for all babies in Poland. In both instances
91% of respondents gave positive responses .
In the United Kingdom PulseOx study, a
detailed psychological questionnaire was given to
mothers of all babies with false positive and true
positive results and to a sample of those with a true
Volume 28 ? Number 2 ? March 2013
negative result [3
rigorous methodology to assess satisfaction with
the test, anxiety and depression, general feelings
about the test and illness perception. The results
showed that parents were predominantly satisfied
with the test and those whose babies had a false
positive result were no more anxious than those
with true negative. Parents generally perceived it
as an important and valued test to detect ill babies
in testing were also sought, either at focus group
meetings or by e-mail questionnaires. All staff
groups (healthcare assistants, midwives, nurses
and doctors) were predominantly positive about
the testing procedure and perceived the test as
&]. The questionnaire used
&]. In addition, the views of clinical staff involved
In 2005, an evaluation of newborn screening for
CHD published by Knowles et al.  used decision
analytic modelling to estimate the cost of screening.
The model suggested an additional cost per timely
diagnosis of £4500 for pulse oximetry compared
with £4.5million for routine screening echocardio-
graphy. de-Wahl Granelli et al. , using data
from an older model, estimated cost at £3430 per
In the PulseOx study, a cost-effectiveness
analysis was also performed employing the model
used by Knowles et al. and using data from the test
accuracy study [3
clinical examination alone with clinical examin-
ation and pulse oximetry and found that the
addition of pulse oximetry was twice as costly, but
identified an additional 30 cases of CHD per 100000
livebirths compared with examination alone. The
incremental cost-effectiveness ratio for this strategy
was £24000 per timely diagnosis, and probabilistic
sensitivity analysis suggested that, at a willingness-
to-pay threshold of £100000, the probability of
pulse oximetry screening being cost-effective is
more than 90%. The report concluded that, at
current thresholds, the addition of pulse oximetry
was likely to be cost-effective [15
remember that this analysis was performed on data
from a study with a relatively high false positive rate
and an antenatal detection rate for CCHD of 50%.
It was also assumed that a diagnostic echocardio-
gram would be performed in all test positive cases.
In situations with a lower false positive rate and/or
a lower antenatal detection, the cost-effectiveness is
likely to improve. The diagnosis of a non-cardiac
cause for hypoxemia would obviate the need
for an echocardiogram, potentially reducing costs
&]. The analysis compared
&]. It is important to
It is important to assess the ability to perform
routine pulse oximetry screening in all newborns
outside a research study and in all clinical settings
In Switzerland, 76% of maternity units screen all
newborns prior to discharge, although screening
is performed significantly less in birthing centres
than in hospitals . In 2010, it was reported that
7% of UK maternity units undertook routine screen-
ing  and effective screening was established in
a community hospital in the USA [20
particularly relevant, as smaller hospitals may not
always benefit from a robust antenatal screening
programme. These reports demonstrate that screen-
ing can be achieved without additional staff and
without overburdening clinical services, suggesting
that implementation of universal pulse oximetry is
feasible and achievable.
&]. This is
IMPLEMENTATION OF UNIVERSAL
In 2005, the Swiss Society of Neonatology and the
Swiss Society of Paediatric Cardiology recommen-
ded that all neonates in Switzerland should undergo
first day pulse oximetry screening  and in 2010,
this was also recommended by the Polish Ministry
for Health . In 2011, the US Secretary’s Advisory
Committee in Heritable Disorders in Newborns
reviewed the available evidence and recommended
a standard protocol for routine screening. The sub-
sequent statement by this group [21
by a number of professional bodies, including
the American Academy of Paediatrics, American
Heart Association and the US Health and Human
Services Secretary . States across the USA
are currently considering implementation of this
recommendation and, to date, four states are
currently screening all neonates, with the majority
of the other states making progress towards this
goal. In the United Kingdom, the National Screen-
ing Committee is currently undertaking a review of
screening for paediatric congenital heart disease,
which includes considering the value of the intro-
duction of pulse oximetry screening.
&&] was endorsed
In high altitude settings, the normal threshold for
saturations may not apply and may need to be
adjusted. Evidence to define this adjustment more
precisely is currently lacking.
Although the evidence strongly supports the
conclusion that the addition of pulse oximetry will
Review of pulse oximetry screening for congenital heart defects Ewer
0268-4705 ? 2013 Wolters Kluwer Health | Lippincott Williams & Wilkins
significantly reduce the number of babies leaving Download full-text
hospital with undiagnosed CCHD, it is vital to
remember that it will not detect all these babies.
The sensitivity of around 75% means that approxi-
mately a quarter of babies with CCHD will not be
detected by this method. Combining with other
screening methods will reduce this diagnostic gap
but some babies will still be missed. Most studies
report that the commonest lesions missed are those
causing obstruction to aortic outflow (e.g. coarcta-
necessarily be associated with hypoxemia. The use
of preductal and postductal saturation difference
may improve this, but both recent studies employ-
ing this technique reported a failure to detect
these conditions [10,12
which one may assume will always be associated
with cyanosis (such as transposition of the great
arteries), may occasionally be missed [12
It is essential that both clinical staff and parents
are made aware of the limitations of this technique
so that false expectations are not raised.
&]. Interestingly, even lesions
Recent studies, particularly when combined with
data from previous studies, provide compelling
evidence for the introduction of pulse oximetry
screening into routine clinical practice. A recent
Lancet editorial described the technique as ‘a new
milestone in the history of congenital heart disease’
. The test is well tolerated, simple and feasible;
it is highly specific and sufficiently sensitive to
qualify for screening. It is acceptable to parents
and clinical staff and cost-effective in the current
clinical setting and is endorsed by an increasing
number of professional and national institutions.
Conflicts of interest
There are no conflicts of interest.
REFERENCES AND RECOMMENDED
Papers of particular interest, published within the annual period of review, have
been highlighted as:
of special interest
&& of outstanding interest
Additional references related to this topic can also be found in the Current
World Literature section in this issue (p. 260).
1. Hoffman JIE, Kaplan S. The incidence of congenital heart disease. J Am Coll
Cardiol 2002; 39:1890–1900.
2. Mahle WT, Newburger JW, Matherne GP, et al. Role of pulse oximetry in
examining newborns for congenital heart disease: a scientific statement from
the AHA and AAP. Pediatrics 2009; 124:823–836.
Ewer AK, Furmston AT, Middleton LJ, 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
This monograph describes in detail the conclusions of the PulseOx study, includ-
ing test accuracy, acceptability to parents and clinical staff, and health economic
4. Hoffman JIE. It is time for routine neonatal screening by pulse oximetry.
Neonatology 2011; 99:1–9.
5. Richmond S, Reay G, Abu Harb M. Routine pulse oximetry in the asympto-
matic newborn. Arch Dis Child Fetal Neonatal Ed 2002; 87:F83–F88.
6. Koppel RI, Druschel C, Carter T, et al. Effectiveness of pulse oximetry
screening for congenital heart disease in asymptomatic newborns. Pediatrics
7. Thangaratinam S, Daniels J, Ewer AK, et al. The accuracy of pulse oximetry in
screening for congenital heart disease in asymptomatic newborns: a sys-
tematic review. Arch Dis Child Fetal Neonatal Ed 2007; 92:F176–F180.
8. Sendelbach DM, Jackson GL, Lai SS, et al. Pulse oximetry screening at 4h of
age to detect critical congenital heart defects. Pediatrics 2008; 122:e815–
9. Meberg A, Brugmann-Pieper S, Reidar D Jr, et al. First day of life pulse
oximetry screening to detect congenital heart defects. J Pediatrics 2008;
10. de-Wahl Granelli A, Wennergren M, Sandberg K, et al. Impact of pulse
oximetry screening on the detection of duct dependent congenital heart
disease: a Swedish prospective screening study in 39821 newborns.
BMJ 2009; 338:a3037.
11. Riede FT, Worner C, Dahnert I, 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;
Ewer AK, Middleton LJ, Furmston AT, et al. Pulse oximetry as a screening test
for congenital heart defects in newborn infants (PulseOx): a test accuracy
study. Lancet 2011; 378:785–794.
This study evaluated the test accuracy of pulse oximetry screening using robust
methodology and follow-up to 1 year to identify all false negatives.
13. Turska Kmiec ´ A, Borszewska Kornacka MK, Błaz ˙ W, et al. Early screening for
critical congenital heart defects in asymptomatic newborns in Mazovia pro-
vince: experience of the POLKARD pulse oximetry programme 2006–2008
in Poland. Kardiologia Polska 2012; 70:370–376.
Powell R, Pattison HM, Bhoyar A, et al. Pulse oximetry as a screening test for
congenital heart defects in newborn infants: an evaluation of acceptability to
mothers. Arch Dis Child 2013; 98:F59–F63.
This is the first study to report the acceptability to mothers of pulse oximetry
screening. The results of screening (including false positives) did not increase
Roberts TE, Barton P, Auguste P, et al. Pulse oximetry as a screening test for
congenital heart disease in newborn infants: a cost effectiveness analysis.
Arch Dis Child 2012; 97:221–226.
This study reports the cost-effectiveness of screening compared with physical
examination alone using data from the PulseOx study. The conclusion is that pulse
oximetry screening is cost-effective in the light of currently accepted thresholds.
Thangaratinam S, Brown K, Zamora J, et al. Pulse oximetry screening
for critical congenital heart defects (CCHD) in asymptomatic newborns:
a systematic review and meta analysis. Lancet 2012; 379:2459–2464.
This meta-analysis identified 13 studies investigating the use of pulse oximetry
screening. The overall sensitivity for pulse oximetry screening was 76.5%, the
specificity 99.9% and false positive rate 0.14%. The review concludes that pulse
oximetry is highly specific for detection of critical congenital heart defects with
moderate sensitivity and meets criteria for universal screening.
17. Knowles R, Griebsch I, Dezateux C, et al. Newborn screening for congenital
heart defects: a systematic review and cost-effectiveness analysis. Health
Technol Assess 2005; 9:1–168.
for congenital heart defects in Switzerland: most but not all maternity units
screen their neonates. Swiss Med Wkly 2009; 139:699–704.
19. Kang SL, Tobin S, Kelsall W. Neonatal pulse oximetry screening: a national
survey. Arch Dis Child Fetal Neonatal Ed 2011; 96:F312.
Bradshaw EA, Cuzzi S, Kiernan SC, et al. Feasibility of implementing pulse
oximetry screening in a community hospital. J Perinatol 2012; 32:710–715.
This study reports the feasibility of implementing pulse oximetry screening in a
community hospital outside a research project. Screening was shown to be
feasible without additional staff.
Kemper AR, Mahle WT, Martin GR, et al. Strategies for implementing screening
for critical congenital heart disease. Pediatrics 2011; 128:e1259–e1267.
This article reports the deliberations of the Secretary’s advisory committee on
heritable disorders in newborns and children working group on pulse oximetry
screening and the recommended US national guidelines for screening.
22. Mahle WT, Martin GR, Beekman RH III, et al. Endorsement of Health and
Human Services recommendation for pulse oximetry screening for critical
congenital heart disease. Pediatrics 2012; 129:190–192.
23. A new milestone in the history of congenital heart disease [editorial].
Lancet 2012; 379:2401.
Volume 28 ? Number 2 ? March 2013