Four-limb blood pressure screening for congenital malformations of aorta in high-risk infants

doi:10.21203/rs.3.rs-1857382/v1

Objectives

Four-limb blood pressure (BP) had been used in congenital malformations of aorta (CMoA) studies with conflict results. To determine the accuracy of four-limb BPs measurement as an approach for CMoA detection in high-risk infants.

Methods

We conducted a retrospective cohort study among infants with suspected congenital heart disease between 2013 and 2019 at a regional tertiary cardiac transferring center. The golden standard of CMoA was echocardiography or cardiac computed tomography results within 2 days after admission. The optimal cutoff of BP gradient for CMoA was determined by Youden Index in the discovery group (admitted between 2013 and 2017) and was verified in the validation group (admitted between 2018 and 2019). These infants were stratified by age to determine the test accuracy of the optimal BP gradient at different ages.

Results

Of 1604 infants (median age, 4 days (1-16 days)) included in the final analysis, 254 CMoA were identified. In infants with pre-test probability of 15.9%, the accuracy test results for CMoA in both groups were similar with the optimal cutoff of BP gradient of 10 mm Hg, area under the ROC curve of 74.8% (95% CI, 68.0%-81.5%), sensitivity of 40.3% (95% CI, 29.3%-51.2%), specificity of 96.3% (95% CI, 94.3%-98.3%). The BP gradient in infants older than 2 days of age was more likely to predict CMoA.

Conclusions

As no suitable alternative tests, four-limb BP measurement with BP gradient of 10 mm Hg is an acceptable approach to screen out CMoA in high-risk infants, especially in infants older than 2 days of age.

Registration: No NCT05022836 (Four-limb Blood Pressures in Infants With High Risk of Congenital Malformation of Aorta)

congenital malformation of aorta

blood pressure

diagnostic test

infants

Congenital heart diseases (CHD) are the most common congenital anomalies and affect nearly 1.8 cases per 100 live births¹. About 30% of these newborns remain undiagnosed upon discharge from the nursery². More than a decade years ago, hospitals in developed countries initiated Pulse Oximetry Screening (POS) for CHD. More than 90% of infants with critical CHD (CCHD) were screened out. However, CHD screening with the saturation of peripheral capillary oxygen (SpO₂) is not designed to detect cardiac defects without hypoxemia. More than 60% of CHD is actually non-cyanotic³. At 2 days of age, more than 60% of those with a ductus-dependent circulation including congenital malformation of aorta (CMoA) had not been diagnosed⁴. Meanwhile, prenatal screening programs were also introduced to screen CCHD out. After upgrading the fetal echocardiography strategy, the prenatal detection rate of all CHD increased significantly from 35.8% to 59.7% and early infant deaths from CCHD decreased by 33.4%^5,6. However, given enhanced prenatal detection of outflow tract anomalies with fetal echocardiography, CMoA was still out of the spotlight^7-9. More than 60% of CMoA was late diagnosed, with 0.2% diagnosed at autopsy¹⁰.

CMoA was supposed to present with four-limb blood pressure (BP) gradient. Hypotension in lower limbs could be noticed by four-limb BP measurement before life-threatening shock developed. Moreover, the 2018 ACC/AHA guidelines for management of adults with CHD recommend to measuring resting blood pressure in upper and lower extremities in all adults with CMoA¹¹. However, a four-limb BP screening study showed negative results in infants with low pre-test probability. The incidence of CMoA in that study was as low as 0.01%. We hypothesized that four-limb BPs measurement could be used to screen CMoA out in infants with high risk of CHD.

Participants

This was a retrospective, cohort study for CMoA. Infant admitted at the Guangdong Provincial People’s Hospital, a tertiary cardiac transferring center, between January 1st, 2013 and December 31st, 2019 were enrolled in the study. Four-limb BPs in infants of confirmed or suspected CHD admitted at our hospital were prospectively collected at admission following the Pediatric Department protocol for CHD infants in which all infants with high risk of CHD, defined as confirmed or suspected CHD by clinical teams, were screened by measuring four-limb BPs by certificated nurses. Suspected CHD was defined as unreasonable low SpO₂, cyanosis, murmur, suspected by prenatal ultrasound or postnatal echocardiography at other hospitals, and unexplained poor cardiac output¹². Three groups of infants were eligible for this study. First, infants confirmed or suspected with CHD by other hospitals and transferred to our hospital. Second, infants suspected of CHD by fetal ultrasound or echocardiography and born in our hospital. Third, others suspected by clinicians by auscultation of heart murmurs, SpO₂<90%, fainter femoral pulse than radial pulse, or other signs of poor cardiac output. When infants were admitted for respiratory disease, it was decided by clinician team whether to measure four-limb BPs or not.

Data collection

All the charts of these infants were reviewed. The diagnosis was retrieved from the discharge note. The diagnosis of CMoA, including Coarctation of aorta (CoA) (Q25.1), Atresia of aorta or interrupted aortic arch (IAA) (Q25.2), Aplasia of aorta or Hypoplasia of aorta (HA) (Q25.4), were identified. Four-limb BPs at admission and echocardiography or cardiac computed tomography (CT) results were also retrieved. The gold standard for CMoA is anomalies confirmed by echocardiography or cardiac CT within 2 days after admission. The echocardiography results obtained just before admission by the same group of examiners were also acceptable. The post-surgery diagnosis was retrieved if there was an inconsistency between echocardiography and cardiac computed tomography results. If neither echocardiography nor cardiac CT had been performed, the baby’s records were excluded from the final analysis. Infants admitted directly to Cardiac Surgery Department without four-limb BPs before surgery were also excluded. As to multiple admission, only the first admission data was included. Echocardiography was routinely performed for all these infants due to CHD within 2 days after admission. Cardiac CT was ordered if more details of the arteries out of the heart or the coronary arteries were requested by cardiac surgeons. Echocardiograph examiners and cardiac CT radiologists had no information about the four-limb BPs results. General characteristics included gender, delivery, age at admission, weight at admission, gestational age, and birth weight. The status at which four-limb BP was collected, were also retrieved, such as sepsis including clinical or confirmed sepsis, inotrope (dopamine, dobutamine, adrenaline, milrinone, and other alpha or beta-energetic agents), prostaglandin E, mechanical ventilation, status of patent ductus arteriosus.

BP and BP gradient

BP measurements were taken in both arms and legs by automated oscillometric devices (Draeger Infinity Kappa or Infinity Delta) when infants were quiet and calm. All these devices were regularly quality-controlled by the Guangzhou Bureau of Quality and Technical Supervision. The BP measurements were performed by well-trained nurses who don’t necessarily know the specific diagnosis. The cuff size was chosen at the discretion of the bedside nurse based on the length of arm or leg. Repeat BP measurements were performed at clinician team’s discretion. As our previous research showed no significant difference of BPs between right and left arm, the BP gradient was calculated as the difference of BP between right arm and both legs¹². BP gradient=right arm systolic BP – (right leg systolic BP + left leg systolic BP)/2.

Study design and statistical analysis

The object was to determine the optimal cutoff of BP gradient for CMoA screening. Infants between 2013 and 2017 were grouped into the discovery group and infants between 2018 and 2019 into the validation group. In the discovery process, receiver operating characteristics (ROC) curve was used to analyze the performance of the BP gradient as a diagnostic test for CMoA and Youden Index (=sensitivity + specificity - 1) was used to determine the optimal BP gradient cutoff point to identify CMoA in this population. Evidence of good validation was defined as an area under the curve of at least 0.700, the accepted level required as evidence of robust validation in other studies^13,14. Demographic and hemodynamic variables were compared between CMoA group and non-CMoA group. In the validation process, the optimal BP gradient cutoff was used to identify CMoA. For defining accuracy, sensitivity, specificity, Youden index, area under the ROC curve, likelihood ratio for positive test results, likelihood ratio for negative test results, positive predictive value, and negative predictive value were calculated in the whole population, CoA infants, IAA infants, and HA infants respectively. The accuracy tests of a range of BP gradient from 5 mm Hg to 25 mm Hg were also performed.

Categorical variables were compared with Pearson χ2 tests or Fisher exact test and presented as number and percentage. Continuous variables were compared using the t test or Mann-Whitney test and presented as mean (SD) values or median values and interquartile ranges (IQRs) as appropriate. A 2-sided P value less than .05 was considered statistically significant. Statistical analyses were performed using IBM SPSS Statistics, version 21.0 (IBM).

This diagnostic test was reported according to the Standards for Reporting of Diagnostic Accuracy (STARD) reporting guideline. During the study period, 11401 infants were admitted to our hospital, with 9570 excluded with no suspect of CHD, 175 without four-limb BP, 2 withdrawals of care without echocardiography and cardiac CT tests, and 50 for multiple admissions (Figure 1). Therefore, 1604 infants (median age, 4 days (1-16), 61.5% male) were recruited in the study, 1172 in the discovery group and 432 in the validation group. The clinical and demographic features of infants in CMoA and non-CMoA groups are summarized (Table 1). The index test and echocardiography were done within 72 h of each other. In total, 1527 was confirmed of CHD by gold standard test. Among them, 160 (10.5%) were confirmed of CoA, 53 (3.5%) of IAA, and 41 (2.7%) of HA. The other 77 infants included 32 isolated atrial septal defect, 21 isolated patent foramen ovale, 9 persistent pulmonary hypertension of the newborn, 1 idiopathic pulmonary arterial hypertension, 1 pulmonary sequestration, 1 pulmonary artery embolism, and 12 others.

Table 1. Characteristics of infants with high risk of congenital heart disease

	Total			Discovery group			Validation group
	CMoA (n=254)	Non-CMoA (n=1350)	P value	CMoA (n=177)	Non-CMoA (n=995)	P value	CMoA (n=77)	Non-CMoA (n=355)	P value
Age, mean (SD), d	13.9 (17.3)	12.1 (22.2)	0.246	15.8 (19.5)	14.1 (24.5)	0.398	9.5 (9.2)	6.5 (12.3)	0.054
Age, median (IQR), d	9 (1,20)	3 (0,15)	0.246	10 (1,22)	4 (0,18)	0.398	8 (0,17)	2 (0,10)	0.054
Male, No. (%)	170 (66.9)	817 (60.5)	0.058	114 (64.4)	591 (59.4)	0.243	56 (72.7)	226 (63.7)	0.147
Vaginal delivery, No. (%)	146 (57.5)	800 (59.3)	0.387	73 (41.2)	617 (62.0)	<0.001	42 (54.5)	183 (51.5)	0.006
Gestational age, wk	38.5 (1.6	37.8 (4.1)	0.015	38.4 (1.7)	37.8 (4.6)	0.077	38.6 (1.6)	37.9 (2.1)	0.010
Birth weight, g	3039 (574)	2960 (576)	0.047	3062 (579)	2965 (576)	0.038	2984 (562)	2948 (576)	0.618
Weight, g	3228 (615)	3106 (677)	0.008	3261 (617)	3143 (705)	0.038	3152 (606)	3001 (579)	0.040
Sepsis, No. (%)	6 (2.4)	11 (0.8)	0.040	6 (3.4)	7 (0.7)	0.007	0 (0)	4 (1.1)	1.000
Inotrope, No. (%)	41 (16.1)	185 (13.7)	0.160	29 (16.4)	141 (14.2)	0.420	12 (15.6)	44 (12.4)	0.456
Prostaglandin, No. (%)	12 (4.7)	107 (7.9)	0.089	7 (4.0)	62 (6.2)	0.299	5 (6.5)	45 (12.7)	0.168
Patent ductus arteriosus, No. (%)	166 (65.4)	719 (53.3)	<0.001	107 (60.5)	488 (49.0)	0.006	59 (76.6)	231 (65.1)	0.061
Ventilation, No. (%)	60 (23.6)	160 (11.9)	<0.001	42 (23.7)	102 (10.3)	<0.001	18 (23.4)	58 (16.3)	0.141
BP gradient, mm Hg	-9.1 (17.4)	4.0 (9.1)	<0.001	-8.9 (17.4)	3.9 (9.3)	<0.001	-9.6 (17.7)	4.3 (8.3)	<0.001
Significant BP gradient, No. (%)	103 (40.6%)	72 (5.3%)	<0.001	72 (40.7%)	59 (5.9%)	<0.001	31 (40.3%)	13 (3.7%)	<0.001
Right arm SBP, mm Hg	77.8 (16.4)	69.9 (11.3)	<0.001	78.3 (16.8)	71.0 (11.5)	<0.001	76.6 (15.5)	66.9 (10.3)	<0.001
Right arm DBP, mm Hg	45.3 (12.3)	39.6 (9.3)	<0.001	45.5 (12.6)	40.4 (9.4)	<0.001	44.9 (11.7)	37.2 (8.8)	<0.001
Left arm SBP, mm Hg	77.3 (16.1)	69.7 (11.3)	<0.001	78.2 (16.7)	71.0 (11.4)	<0.001	75.3 (14.5)	66.3 (10.2)	<0.001
Left arm DBP, mm Hg	44.7 (12.4)	39.7 (9.4)	<0.001	45.3 (13.0)	40.7 (9.5)	<0.001	43.5 (10.9)	36.8 (8.6)	<0.001
Right leg SBP, mm Hg	68.7 (12.8)	73.7 (12.6)	<0.001	69.6 (13.3)	74.5 (12.9)	<0.001	66.8 (11.6)	71.2 (11.2)	0.002
Right leg DBP, mm Hg	36.7 (9.9)	38.3 (13.4)	0.073	37.5 (10.7)	38.7 (9.6)	0.132	34.9 (7.6)	37.2 (20.6)	0.343
Left leg SBP, mm Hg	68.7 (12.1)	74.1 (12.7)	<0.001	69.3 (12.4)	75.1 (12.9)	<0.001	67.3 (11.4)	71.4 (11.7)	0.005
Left leg DBP, mm Hg	36.6 (9.9)	41.8 (111.7)	0.475	37.4 (10.2)	43.9 (130)	0.506	35.3 (8.0)	35.8 (9.0)	0.660

Abbreviation: BP, blood pressure; CMoA, congenital malformation of aorta; DBP, Diastolic blood pressure; IQR, interquartile range; SBP, systolic blood pressure; SD, standard deviation

The distributions of BP gradient values of infants with CMoA and without CMoA in the discovery group and the validation group are shown (Figure 2). The ROC curves were generated (Figure 2). In the discovery group, Youden Index showed that the optimal cutoff of BP gradient for CMoA screening was 10 mm Hg, with an AUC of 0.723 (95% CI, 0.678-0.768), sensitivity of 40.7% (95% CI, 33.4%-47.9%), specificity of 94.1% (95% CI, 92.6%-95.5%), likelihood ratio for positive test results of 6.86 (95% CI, 5.06-9.30), likelihood ratio for negative test results of 0.631 (95% CI, 0.558-0.713), positive predictive value of 55.0% (95% CI, 46.4%-63.5%) and negative predictive value of 89.9% (95% CI, 88.1%-91.7%). In the validation group, for distinguishing CMoA from non-CMoA, this cutoff of 10 mm Hg generated an AUC of 0.748 (95% CI, 0.680-0.815) (Figure 2), with sensitivity of 40.3% (95% CI, 29.3%-51.2%), specificity of 96.3% (95% CI, 94.4%-98.3%), likelihood ratio for positive test results of 11.0 (95% CI, 6.0-20.0), likelihood ratio for negative test results of 0.602 (95% CI, 0.516-0.746), positive predictive value of 70.5% (95% CI, 57.0%-83.9%) and negative predictive value of 88.1% (95% CI, 84.9%-91.4%).

When these infants were divided into five groups by age, the sensitivity of BP gradient of 10 mm Hg increased dramatically after 48 h after birth (Figure 3 and Table 2). After incorporating specificity with sensitivity, the likelihood ratio of positive test results increased dramatically after 24 h after birth (Table 2). With a pre-test probability of about 16%, the positive post-test probability increased from 26.7% at the first day of age to 75.0% after 1 month of age.

Table 2. Accuracy performance of 10 mm Hg BP gradient in infants of different age

	BP gradient, mm Hg	CMoA	Non-CMoA	Sensitivity, %	Specificity, %	PPV, %	NPV, %	PTP, %	LR+	PPTP, %	LR-	NPTP, %	FPR, %	FNR, %
Total	>10	103	72	40.4	94.7	58.9	89.4	15.9	7.57	58.9	0.630	10.6	5.3	59.6
	<10	152	1277
Day 1	>10	8	22	13.3	95.0	26.7	88.9	12.1	2.65	26.7	0.913	11.1	5.0	86.7
	<10	52	416
Day 2	>10	4	5	26.7	94.9	44.4	89.4	13.3	5.23	44.4	0.773	10.6	5.1	73.3
	<10	11	93
Day 3-7	>10	18	11	50.0	96.0	62.1	93.7	11.5	12.59	62.1	0.521	6.3	4.0	50.0
	<10	18	266
Week 2-4	>10	55	28	50.5	92.4	66.3	86.3	22.8	6.65	66.3	0.536	13.7	7.6	49.5
	<10	54	341
Month 1 more	>10	18	6	52.9	96.4	75.0	91.0	16.8	14.82	75.0	0.488	9.0	3.6	47.1
	<10	16	162

Abbreviation: BP, blood pressure; CMoA, congenital malformation of aorta; PPV, positive predictive value; NPV, negative predictive value; LR+, Likelihood ratio for positive test results; LR-, Likelihood ratio for negative test results; PTP, Pre-test probability; PPTP, Positive post-test probability; NPTP, Negative post-test probability; FPR, False positive rate; FNR, False negative rate;

The summary of diagnostic test accuracy of 5 to 25 mm Hg BP gradients in the whole population was shown (Figure 4 and Table 3). As expected, the sensitivity decreased with BP gradient increasing, while the specificity, the positive predictive value and the likelihood ratio for positive test results increased with BP gradient. The optimal cutoff of BP gradient was 10 mm Hg.

Table 3. Accuracy performance of blood pressure gradient ranging from 5 mm Hg to 25 mm Hg

BP gradient, mm Hg	Sensitivity, %	Specificity, %	Youden index	PPV, %	NPV, %	PTP, %	PPTP, %	NPTP, %	LR+	LR-
5	46.9	86.4	0.333	39.3	89.6	15.8	39.3	10.4	3.44	0.615
6	44.5	89.0	0.335	43.3	89.5	15.8	43.3	10.5	4.06	0.623
7	43.3	91.3	0.346	48.3	89.5	15.8	48.3	10.5	4.96	0.621
8	42.1	92.4	0.345	51.2	89.5	15.8	51.2	10.5	5.58	0.626
9	41.3	93.9	0.352	55.9	89.5	15.8	55.9	10.5	6.72	0.625
10	40.6	94.7	0.353	58.9	89.4	15.8	58.9	10.6	7.60	0.628
11	38.2	95.6	0.338	61.8	89.2	15.8	61.8	10.9	8.59	0.647
12	36.2	96.2	0.324	64.3	88.9	15.8	64.3	11.1	9.59	0.663
13	34.7	97.2	0.319	69.8	88.8	15.8	69.8	11.2	12.31	0.672
14	32.2	97.6	0.298	71.3	88.5	15.8	71.3	11.6	13.21	0.694
15	32.3	97.8	0.301	73.2	88.5	15.8	73.2	11.5	14.53	0.693
16	31.5	98.2	0.297	76.9	88.4	15.8	76.9	11.6	17.72	0.697
17	29.5	98.3	0.278	76.5	88.1	15.8	76.5	11.9	17.33	0.717
18	27.6	98.7	0.263	79.6	87.9	15.8	79.6	12.1	20.67	0.734
19	25.2	99.0	0.242	82.1	87.6	15.8	82.1	12.5	24.30	0.756
20	23.6	99.4	0.230	88.2	87.4	15.8	88.2	12.6	39.86	0.768
21	22.4	99.6	0.220	90.5	87.2	15.8	90.5	12.8	50.49	0.779
22	22.1	99.6	0.217	91.8	87.2	15.8	91.8	12.8	59.53	0.782
23	21.3	99.7	0.210	93.1	87.1	15.8	93.1	12.9	71.75	0.790
24	20.9	99.7	0.206	93.0	87.0	15.8	93.0	13.0	70.42	0.794
25	19.3	99.8	0.191	94.2	86.8	15.8	94.2	13.2	86.81	0.809

Abbreviation: BP, blood pressure; PPV, positive predictive value, NPV, negative predictive value, PTP, Pre-test probability; PPTP, Positive post-test probability; NPTP, Negative post-test probability; LR+, Likelihood ratio for positive test results; LR-, Likelihood ratio for negative test results

The test accuracy performances for CoA, IAA and HA were also shown (Table 4). With the optimal cutoff, the BP gradient of 10 mm Hg was less sensitive in infants with HA than those in CoA or IAA infants (24.4% vs. 43.5% and 43.4% respectively). The positive post-test probability was higher in CoA infants (49.3%). The Likelihood ratio for positive test results was 8.15 in CoA infants. As for all, the positive post-test probability was 58.9%. The likelihood ratio of positive test result was 7.57. Five percent of non-CMoA infants presented with significant BP gradient at admission.

Table 4. Test accuracy performance of BP gradient in infants with different congenital malformation of aorta

	Gradient, No.	no gradient, No.	Sensitivity, %	Specificity, %	PPV, %	NPV, %	PTP, %	LR+	PPTP, %	LR-	NPTP, %	FPR, %	FNR, %
CMoA	103	152	40.4	94.7	58.9	89.4	15.9	7.57	58.9	0.63	10.6	5.3	59.6
CoA	70	91	43.5	94.7	49.3	93.4	10.7	8.15	49.3	0.60	6.7	5.3	56.5
IAA	23	30	43.4	94.7	24.2	97.7	3.8	8.13	24.2	0.60	2.3	5.3	56.6
HA	10	31	24.4	94.7	12.2	97.6	3.0	4.57	12.2	0.80	2.4	5.3	75.6
Healthy	72	1277

Abbreviation: BP, blood pressure; CMoA, congenital malformation of aorta; CoA, coarctation of aorta; HA, hypoplasia of aorta; IAA, interrupted aortic arch; PPV, positive predictive value, NPV, negative predictive value, LR+, Likelihood ratio for positive test results; LR-, Likelihood ratio for negative test results; PTP, Pre-test probability; PPTP, Positive post-test probability; NPTP, Negative post-test probability; FPR, False positive rate; FNR, False negative rate;

In this respective cohort study assessing 1604 infants with high risk of CMoA, four-limb BP measurement is an acceptable approach to screen CMoA out in infants with high risk of CHD and the BP gradient test with the optimal cutoff of 10 mm Hg performed better, especially in infants older than 2 days of age and in infants with CoA or IAA.

POS has been added into the Recommended Uniform Screening Panel for ten years. The addition of POS increases the proportion of CCHD identified from 50%–70% to over 90%^15,16. A reduction of early cardiac deaths due to CCHD by 30% had also been reached.7 This reduction in mortality resulted in ∼120 lives saved per year in the US and an additional 21% reduction in other or unspecified CHD deaths. However, POS had a low sensitivity for detection of CMoA, due to its non-cyanotic profile¹⁷. Tests targeted at the low cardiac output profile of CMoA had been performed with peripheral perfusion index (PI) or femoral pulse palpation^18-23. Some researchers suggested that PI test was better used as an addition of POS for the same pulse oximeter being used at the same time²⁴. Unfortunately, the optimal cutoff of PI to screen CMoA out was not determined so far. And its sensitivity varies greatly from 25% to 75%^24,25. Femoral pulse palpation test was also used to screen CMoA with a very low sensitivity of 19.2%²⁰. Meanwhile, BP gradient between four limbs has been supposed to be a profile of CMoA. And this BP gradient seems stable with gestational age²⁶. But there are overlaps of BP gradients between CMoA and healthy infants. Moreover, Boelke et al retrospectively reviewed 10012 infants screened by right arm and leg BP in an advanced fetal diagnostic center with a very low rate of positive result²⁷. After that, no large sample size study on CMoA screening has been published.

As reported²⁸, high antenatal detection rate leads to a low POS rate and diagnostic test performance could be enhanced in patients with high pre-test probability, our study implemented four-limb BP screening in infants with high risk of CHD. By calculating the Youden index, 10 mm Hg of BP gradient was determined as the optimal cutoff for CMoA²⁶, with sensitivity of 40.6% and specificity of 94.7%. The probability of CMoA increased from 15.8% before test to 58.9% after test. The positive predictive value was much higher than that in PI screening study and femoral pulse palpation test (58.9% vs. 3.5% and 1.5% respectively). But the negative predictive value was lower (89.4% vs. 99.9% and 99.9% respectively). The likelihood ratio for positive test results was 7.6. As for different types of CMoA, the likelihood ratios for positive test results for CoA, IAA and HA were 16.98, 15.75 and 6.59 respectively. As different BP gradients presented with different likelihood ratios, the higher the BP gradient was, the higher the likelihood ratio for positive test results and for negative test results. The positive post-test probability increased from 39.3% at 5 mm Hg BP gradient to 94.2% at 25 mm Hg BP gradient, while the sensitivity decreased progressively from 46.9% to 19.3%. Therefore, the accuracy performances in Figure 4 and Table 3 could work as a guide for clinical decision-making for infants with different BP gradients.

Although BP gradient was with low sensitivity, as our pilot case series study showed that only a small portion of infants with CoA presented with both cyanosis and BP gradient, BP gradient could also be an addition to POS screening, especially in infants with high risk of CMoA in resource-limited areas where the blood pressure measurement might be the only way accessible or in areas with high prevalence^29,30. Geggel’s study from Harvard Medical School also confirmed the importance of physical examination⁸. Although the sensitivity of four-limb BP test was as low of about 40% as that of the PI screening study²⁴, the specificity of both tests was higher than 95%. In population of low-risk of CMoA, the positive predictive value of PI study was 1.5%~19%, while the positive predictive value of the femoral pulse palpation study was 3.5%. And the negative predictive values of both tests were higher than 99%. But in the population of high risk of CMoA, the positive predictive values of BP gradients from 5 to 25 mm Hg ranged from 40% to 94%, while the negative predictive values ranged from 85% to 90%.

As for the progress of BP gradient, different from Patankar’s study²⁶, in this study, BP gradient changed with aging and became significant after 2 days after birth. The main difference in Patankar’s study was that BP gradient came from a cross-section study with gestational age between 22 weeks and 42 weeks, not a consecutive BP trend, and that four-limb BPs were taken shortly after birth²⁶. As our data presented, infants with CMoA might develop BP gradient later than 2 days of age.

The limitations

Firstly, this is not a population-based cohort study and a large proportion of infants were transferred from other hospitals. The prevalence of CMoA couldn’t be calculated. Missing asymptomatic CMoA infants who needn’t attend hospital might exaggerate the test performance. Secondly, the false positive rate of 5.3% in this study is much higher than that of 0.06% in POS screening. This rate was still as high as 3.6% after 1 month of age. The false-negative rate decreased from 86.7% on the first day after birth to 47.1% after 1 month of age. Thirdly, non-invasive BP measurement is not as accurate as invasive BP measurement. The power of one measurement of four-limb BP at admission to screen CMoA out seems low. Repeat for one more measurement might enhance the test accuracy. Fourthly, poor cardiac output might alleviate the BP gradient. But in this study, we failed to retrieve this data to determine the relationship between cardiac output and BP gradient in CMoA infants. Fifthly, in China, CMoA has not been included in routine fetal cardiac anomaly screening category. The comprehensive fetal echocardiography could only be performed in a small part of tertiary hospitals. A large proportion of infants with CMoA were discharged without being diagnosed. This might be an explanation of high incidence of CMoA in this cohort.

As no suitable alternative tests, four-limb BP measurement with BP gradient of 10 mm Hg is an acceptable approach to screen out CMoA in infants with high risk of CHD, especially in infants older than 2 days of age. Attention should be paid to the low sensitivity and specificity of the test. Further study should focus on improvement of test accuracy of the diagnostic workup of CMoA by repeated measurement of BP, a combination of physical examination, BP gradient, PI, and other approaches of cardiac output evaluation. Long-term follow-up is also necessary to further determine the high-risk population of CMoA who need surgical intervention in their early life.

The study was performed in accordance with the Declaration of Helsinki. This study was approved by Research Ethics Committee at the Guangdong Provincial People’s Hospital (application numbers: GDREC2019338H(R2)). The need for informed consent was waived by the Research Ethics Committee of the Guangdong Provincial People’s Hospital, because of the retrospective nature of the study.

All coauthors agreed to submit this manuscript to the BMC Cardiovascular Disorders journal.

Availability of Data and Materials

The datasets generated and/or analysed during the current study are not publicly available due to a deep exploration project on these data but are available from the corresponding author on reasonable request.

All coauthors declare no conflicts of interest.

Funding

Natural Science Foundation of Guangdong Province (2020A1515010904); Science and Technology Planning Project of Guangdong Province (2019B020230003)

Author contributions

Dr. Yifei Wang had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Yifei Wang, Shifang Yang, Xiaobing Liu, Chengcheng Pang, Jian Zhuang

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Ruikun Zou, Yuan Ren.

Critical revision of the manuscript for important intellectual content: Yifei Wang, Shifang Yang, Xiaobing Liu, Chengcheng Pang, Yunxia Sun, Jian Zhuang

Statistical analysis: Yifei Wang, Shifang Yang

Obtained funding: Jian Zhuang, Yifei Wang

Supervision: Yifei Wang

Acknowledgements

Disclosure of grants or other funding: Natural Science Foundation of Guangdong Province (2020A1515010904)[Yifei Wang]; Science and Technology Planning Project of Guangdong Province (2019B020230003)[Jian Zhuang]

GBD 2017 Congenital Heart Disease Collaborators. Global, regional, and national burden of congenital heart disease, 1990-2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet Child Adolesc Health. 2020;4(3):185-200. doi: 10.1016/S2352-4642(19)30402-X
Peterson C, Ailes E, Riehle-Colarusso T, et al. Late detection of critical congenital heart disease among US infants: estimation of the potential impact of proposed universal screening using pulse oximetry. JAMA Pediatr. 2014;168(4):361-370. doi:10.1001/jamapediatrics.2013.4779
Mawson IE, Babu PL, Simpson JM, Fox GF. Pulse oximetry findings in newborns with antenatally diagnosed congenital heart disease. Eur J Pediatr. 2018;177(5):683–689. doi: 10.1007/s00431-018-3093-2
Mellander M, Sunnegardh J. Failure to diagnose critical heart malformations in newborns before discharge – an increasing problem? Acta Paediatr 2006;95(4):407–413. doi: 10.1080/08035250500541910
van Velzen CL, Clur SA, Rijlaarsdam ME, et al. Prenatal detection of congenital heart disease—results of a national screening programme. BJOG. 2016;123(3):400–407. doi: 10.1111/1471-0528.13274
Abouk R, Grosse SD, Ailes EC, Oster ME. Association of US State Implementation of Newborn Screening Policies for Critical Congenital Heart Disease With Early Infant Cardiac Deaths. JAMA. 2017;318(21):2111–2118. doi: 10.1001/jama.2017.17627
Mouledoux JH, Walsh WF. Evaluating the diagnostic gap: statewide incidence of undiagnosed critical congenital heart disease before newborn screening with pulse oximetry. Pediatr Cardiol. 2013;34(7):1680–1686. doi: 10.1007/s00246-013-0697-1
Geggel RL. Coarctation of the Aorta- Delay in Diagnosis and Referral Basis from Infancy to Adult Age. J Pediatr. 2021 Dec 1:S0022-3476(21)01164-1. doi: 10.1016/j.jpeds.2021.11.066
Janicki MB, Fernandez CG, Wakefield D, Shepherd JP, Figueroa R. Improving fetal congenital heart disease screening using a checklist-based approach. Prenat Diagn. 2020;40(2):223-231. doi: 10.1002/pd.5581
Ailes EC, Gilboa SM, Honein MA, Oster ME. Estimated number of infants detected and missed by critical congenital heart defect screening. Pediatrics. 2015;135(6):1000–1008. doi: 10.1542/peds.2014-3662
Stout KK, Daniels CJ, Aboulhosn JA, et al. 2018 AHA/ACC Guideline for the Management of Adults With Congenital Heart Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019 Apr 2;139(14):e698-e800. doi: 10.1161/CIR.0000000000000603
Ruikun Zou, Yifei Wang, Chengcheng Pang, Yunxia Sun, Chen Chen, Jian Zhuang. Screening Coarctation of Aorta with Clinical and Echocardiographic Profiles in Infants: A Pilot Study. Congenit Heart Dis. 2021;16(2): 159–170. doi: 10.32604/CHD.2021.015050
Schisterman EF, Perkins NJ, Liu A, Bondell H. Optimal cut-point and its corresponding Youden Index to discriminate individuals using pooled blood samples. Epidemiology. 2005;16(1):73-81. doi: 10.1097/01.ede.0000147512.81966.ba
Goobie SM, Faraoni D, Zurakowski D, DiNardo JA. Association of Preoperative Anemia With Postoperative Mortality in Neonates. JAMA Pediatr. 2016;170(9):855-862. doi: 10.1001/jamapediatrics.2016.1032
Martin GR, Ewer AK, Gaviglio A, et al. Updated Strategies for Pulse Oximetry Screening for Critical Congenital Heart Disease. Pediatrics. 2020;146(1):e20191650. doi: 10.1542/peds.2019-1650
Badawi D, Watson J, Maschke S, Reid L. First-Year Outcomes of Critical Congenital Heart Disease Screening in Maryland. Glob Pediatr Health. 2019;6:2333794X19868226. doi: 10.1177/2333794X19868226
Cloete E, Bloomfield FH, Cassells SA, de Laat MWM, Sadler L, Gentles TL. Newborn pulse oximetry screening in the context of a high antenatal detection rate of critical congenital heart disease. Acta Paediatr. 2020;109(1):93-99. doi: 10.1111/apa.14946
Siefkes H, Kair L, Tancredi DJ, et al. Oxygen Saturation and Perfusion Index-Based Enhanced Critical Congenital Heart Disease Screening. Am J Perinatol. 2020;37(2):158-165. doi: 10.1055/s-0039-1685445
Searle J, Thakkar DD, Banerjee J. Does pulsatility index add value to newborn pulse oximetry screening for critical congenital heart disease? Arch Dis Child. 2019;104(5):504-506. doi: 10.1136/archdischild-2018-315891
Khammari Nystrom F, Petersson G, Stephansson O, Johansson S, Altman M. Diagnostic values of the femoral pulse palpation test. Arch Dis Child Fetal Neonatal Ed. 2020;105(4):375-379. doi: 10.1136/archdischild-2019-317066
Crossland DS, Furness JC, Abu-Harb M, Sadagopan SN, Wren C. Variability of four limb blood pressure in normal neonates. Arch Dis Child Fetal Neonatal Ed. 2004;89(4):F325-F327. doi: 10.1136/adc.2003.034322
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 39,821 newborns. BMJ. 2009;338:a3037. doi: 10.1136/bmj.a3037
Zuppa AA, Riccardi R, Catenazzi P, et al. Clinical examination and pulse oximetry as screening for congenital heart disease in low-risk newborn. J Matern Fetal Neonatal Med. 2015;28(1):7-11. doi: 10.3109/14767058.2014.899573
Schena F, Picciolli I, Agosti M, et al. Perfusion Index and Pulse Oximetry Screening for Congenital Heart Defects. J Pediatr. 2017;183:74-79.e1. doi: 10.1016/j.jpeds.2016.12.076
Granelli Ad, Ostman-Smith I. Noninvasive peripheral perfusion index as a possible tool for screening for critical left heart obstruction. Acta Paediatr. 2007;96(10):1455-1459. doi: 10.1111/j.1651-2227.2007.00439.x
Patankar N, Fernandes N, Kumar K, Manja V, Lakshminrusimha S. Does measurement of four-limb blood pressures at birth improve detection of aortic arch anomalies? J Perinatol. 2016;36(5):376-380. doi: 10.1038/jp.2015.203
Boelke KL, Hokanson JS. Blood pressure screening for critical congenital heart disease in neonates. Pediatr Cardiol. 2014;35(8):1349-1355. doi: 10.1007/s00246-014-0935-1
Banait N, Ward-Platt M, Abu-Harb M, Wyllie J, Miller N, Harigopal S. Pulse oximetry screening for critical congenital heart disease: a comparative study of cohorts over 11 years. J Matern Fetal Neonatal Med. 2020;33(12):2064-2068. doi: 10.1080/14767058.2018.1538348
Karande S, Kumar S, Vaideeswar P. How often is coarctation of aorta correctly diagnosed antemortem in children with fatal illnesses? A retrospective review of medical and autopsy records. J Postgrad Med. 2020;66(3):169-171. doi: 10.4103/jpgm.JPGM_125_20
Klein J, Dwyer G, Donofrio MT, Krishnan A. Geographic distribution of congenital heart disease: a single surgical center experience. J Pediatr. 2021:S0022-3476(21)00865-9. doi: 10.1016/j.jpeds.2021.08.080. Epub ahead of print. PMID: 34480916

No competing interests reported.

STARDChecklistof4BPforCMoA.docx

Four-limb blood pressure screening for congenital malformations of aorta in high-risk infants

Status:

Version 1

Abstract

Figures

1 Introduction

2 Methods

3 Results

4 Discussion

5 Conclusions

DECLARATIONS

REFERENCES

Additional Declarations

Supplementary Files

Status:

Version 1