Inclusion criteria were as follows: (1) women aged 18 years and older, (2) GA of 14–40 weeks with normal anomaly scan, (3) no known medical or obstetric complications, (4) reliable GA based on regular menstrual cycle and certain last menstrual period consistent with sonographic fetal biometry in the first half of pregnancy and (5) normal fetal heart rate with no arrhythmia. Women at 14–15 weeks of pregnancy could be enrolled if first trimester scan was normal, but were then withdrawn from the study if the second trimester complete anomaly scan revealed any abnormalities.

Exclusion criteria were (1) multi-fetal pregnancies; (2) abnormal chromosomes in the fetus; (3) abnormal fetal growth (either restriction or macrosomia); (4) women with immune system disorders, including systemic lupus erythematosus, anti-phospholipid syndrome, rheumatoid arthritis, Sjogren syndrome, hyperthyroidism and undifferentiated autoimmune diseases; (5) women with a positive blood test for anti-SSA/Ro or anti-SSB/La auto-antibodies and (6) women taking medications (i.e., beta-adrenergic agonists) that affect fetal heart rate.

All fetal echocardiogram were performed by specialists in maternal-fetal medicine (MFM) with qualified diploma of the Thai Subspecialty Board of Maternal and Fetal Medicine issued by the Medical Council of Thailand. At least 150 cases per month of fetal echocardiogram were performed in our MFM division. All pulse wave Doppler investigations were performed during fetal quiescence and apnea on Samsung HS60 abdominal 1–5 MHz curvilinear transducer (Samsung Medison, Korea). The setting of pulse wave Doppler included a wall motion filter of 120 Hz, a sweep speed of 117 mm/s to obtain 4–5 waveform images, pulse repetition frequencies of 5.5–6 kHz and an angle between the ultrasound beam and blood flow of less than 20${}^{\circ }$. In each woman, FAVTIs were measured using three different assessment (ALDAF-AO, ALDAF-PA and LV In/Out), repeated three times, all of which were averaged.

In ALDAF-AO, we sought an apical view of the left ventricular outflow tract (LVOT). Transducer orientation was then adjusted to ensure an insonation angle 20 degrees along the direction of the aortic blood flow. A Doppler gate of 1–3 mm was selected and placed within the aorta and distal to the aortic valves (Fig. 1A) [13]. FAVTI measurement was set at onset of atrial systole (A-wave; red line) to onset of ventricular systole (V-wave; yellow line) during the same cardiac cycle (Fig. 2A).

Fig. 1.

Schematic drawing of appropriate positions of Doppler gate. (A) LVOT; (B) RVOT with appropriate positions of Doppler gate that can create waveform of outflow tract to identify ALDAF-AO/PA. (C) drawing of LV In/Out with appropriate positions of Doppler gate that create waveform of mitral valve inflow and aortic valve outflow to identify LV In/Out waveform.
LVOT, left ventricular outflow tract; RVOT, right ventricular outflow tract; ALDAF, antegrade late diastolic arterial blood flow; AO, aorta; PA, pulmonary artery; LV In/Out, left ventricular inflow/outflow.

Fig. 2.

Position of sample volume and pulse wave Doppler waveform patterns in each method, red line represent onset of A wave and yellow line represent onset of V wave. (A) ALDAF-AO; (B) ALDAF-PA; (C) LV In/Out.
ALDAF, antegrade late diastolic arterial blood flow; AO, aorta; PA, pulmonary artery; LV In/Out, left ventricular inflow/outflow; A, Atrial contraction peak velocity; V, ventricular systole; E, Early diastolic peak velocity.

In ALDAF-PA, we sought a five-chambered view, with rotation or tilting of the transducer cephalad in order to view the PA arising from the RV. Transducer orientation was then adjusted to ensure an insonation angle 20 degrees along the direction of the pulmonary artery blood flow. A Doppler gate of 1–3 mm was selected and placed within the pulmonary artery and distal to the pulmonic valves (Fig. 1B) [13]. FAVTI measurement was set at onset of atrial systole (A-wave; red line) to onset of ventricular systole (V-wave; yellow line) during the same cardiac cycle (Fig. 2B).

In LV In/Out, we sought an apical view of LVOT, then used pulsed Doppler with a sample volume wide enough to cover both diastolic inflow via the mitral valve and systolic outflow via the aortic valve. With a sample gate adjustment of 5–10 mm [14], the angle of insonation was fitted to get as close as possible to zero degrees (up to 20 degrees) (Fig. 1C). FAVTI measurement was set at onset of the mitral A-wave (red line) between the E-peak and A-peak to the beginning of the aortic V-wave (yellow line) in the aortic outflow tract (Fig. 2C).

We used SPSS version 26.0 (IBM Corp., Armonk, NY, USA) to analyze the data. FAVTI data in ALDAF-AO and ALDAF-PA of each GA were created at reference range percentile values of 2.5, 5, 10, 25, 50, 75, 90, 95 and 99. Descriptive values of the three Doppler measurements were expressed as means $\pm$ standard deviation. Data distribution from our large sample size (N = 528) tended towards normal [15, 16]. Categorical data were expressed in terms of number or percentage (%). Statistical significance was set at a two-tailed P value of 0.05.

Correlations between FAVTI with GA in ALDAF-AO, ALDAF-PA and LV In/Out methods were analyzed with linear regression. Bland-Altman analysis and Pearson correlation coefficient were used to compare the fetal AV time interval between measurement techniques. Assessing the reliability of FAVTI measurements from all three methods involved intraobserver and interobserver approaches with r values $>$0.80 indicating high accuracy.

A total of 528 pregnant women participated in the study. Table 1 quantifies the maternal demographic characteristics. This study found no neonatal structural heart anomalies or cardiac arrhythmias. The longest measured time was in LV In/Out: (118.83 milliseconds (ms) $\pm$ 13.27), shortest was in ALDAF-PA: (112.51 $\pm$ 15.03) and in between for ALDAF-AO (115.02 $\pm$ 14.98). The normal reference values for FAVTI measured by ALDAFAO and ALDAF-PA divided into GA from 14 to 40 complete weeks are shown in Table 2.

Linear regression analysis shows highly significant (P 0.001) positive correlations between FAVTI outcome and GA in all three methods. The coefficients of determination (R${}^{\mathrm{2}}$) were 0.272, 0.177 and 0.201 with ALDAF-AO, ALDAF-PA and LV In/Out measurement, respectively (Fig. 3). The strongest correlation was identified in ALDAF-AO, in which the predicted estimation was 1.02 $\times$ GA (weeks) $\pm$ 87.82. Fig. 4A–C show outcomes of the Bland-Altman analysis with a FAVTI correlation strongest in ALDAF-AO versus ALDAF-PA (r = 0.57, P 0.001) (Fig. 4C). This magnitude of correlation was confirmed in the mean FAVTIs difference values (95% CI) of 2.5 (–24.7, 29.7) between ALDAF-AO and ALDAF-PA, ALDAF-AO versus LV In/Out at –3.8 (–30.3, 22.7) and ALDAF-PA versus LV In/Out at –6.3 (–33.9, 21.3) (Table 3).

Fig. 3.

Linear regression analyses between FAVTI and gestational age (weeks). (A) ALDAF-AO; (B) ALDAF-PA; (C) LV In/Out.
FAVTI, fetal atrioventricular time intervals; ALDAF, antegrade late diastolic arterial blood flow; AO, aorta; PA, pulmonary artery; LV In/Out, left ventricular inflow/outflow; R${}^{\mathrm{2}}$, square of the correlation. Lines denote regressions and 95% confidence limits for individual observations.

Fig. 4.

Bland-Altman analysis of FAVTI. (A) ALDAF-AO versus LV In/Out; (B) ALDAF-PA versus. LV In/Out; (C) ALDAF-AO versus ALDAF-PA.
FAVTI, fetal atrioventricular time intervals; ALDAF, antegrade late diastolic arterial blood flow; AO, aorta; PA, pulmonary artery; LV in/out, left ventricular inflow/outflow.

Intraobserver reliability coefficients of FAVTI for ALDAF-AO, ALDAF-PA and LV In/Out were 0.967, 0.979 and 0.978, respectively (95% CI for the three values ranged from 0.96 to 0.98). Interobserver reliability coefficients of FAVTI for ALDAF-AO, ALDAF-PA and LV In/Out were 0.991, 0.959 and 0.989, respectively (95% CI for the three values ranged from 0.923 to 0.996).

To our knowledge, the present study provides the first normal reference values for each week of gestation between 14 and 40 weeks of FAVTI derived from ALDAF-AO and ALDAF-PA. Here, we examined FAVTI in a large population (N = 528) between 14 and 40 weeks of gestation according to previous studies [9, 12]. We demonstrated GA specific normal reference values of FAVTI, which is different from previous investigations presenting with groups of GA [9, 11, 12, 17, 18, 19]. Detailed reference values of FAVTI for each week of gestation should be more accurate in diagnosing CHB than group GA. FAVTI is longest in LV In/Out and ALDAF-PA is shorter than ALDAF-AO, which was similar to findings observed by Howley and colleagues [13]. As a result of fetal cardiac cycle, ALDAF-AO/PA were obtained beyond aortic and pulmonic valve in proximal of vessels and the distance of RV apex to pulmonic valve is shorter than distance of LV apex to aortic valve [20]. Therefore, Doppler signal time in ALDAF-PA was shorter than ALDAF-AO.

We succeeded in obtaining FAVTI for ALDAF-AO/PA in 100% of examinations, as compared to 80% in a previous study [13]. Due to technical limitations, clear identification of the interrogated great artery is difficult in fetuses with gestational ages 11 weeks [13], which may have led to lower success rate of ALDAF AO/PA in this prior work. Therefore, in this study, pregnant women were selected with gestational age greater than or equal to 14 weeks to improve the success rate. Furthermore, ALDAF-AO/PA uses a single valve or vessel, in which pulse wave Doppler measurements were easier to obtain than other techniques (LV In/Out, SVC/AO or PV/PA) requiring two valves or vessels for measurement. This led to high reliability coefficients ($>$0.9) by intraobserver and interobserver ratings in our study, indicating ALDAF shows high reliability and reproducibility method. The ability to obtain FAVTI is based on well-trained physician, high resolution ultrasound, and most importantly the ability to accurately position the sample volume and the angle of insonation is parallel to the signal of AO or PA outflow. Overall, an average time of 30 min was necessary to complete the echocardiographic assessment.

The present study showed FAVTI obtained from ALDAF-AO, ALDAF-PA and LV In/Out FAVTI was significant correlated with advancing GA. This finding corroborates majority of previously published data [9, 12, 13, 17, 18, 19, 21, 22], although a few studies have not identified a correlation between FAVTI and GA [11, 23]. This correlation is likely attributable to enhancement of fetal cardiac size and chamber with progressive GA, which results in prolonged time of myocardium depolarization and repolarization leading to increasing FAVTI [19].

Regarding thresholds for FAVTI measurements, we established cut-off values for ALDAF-AO/PA for CHB diagnosis at $>$99th percentile for each specific GA between GA 14 to 40 weeks (Table 2) which was not similar to those with previous studies [9, 19, 21, 24, 25]. In this study, cut-off FAVTI value obtained from ALDAF-AO and ALDAF-PA are 115–153.7 milliseconds and 111–155 milliseconds respectively as more detail show in Table 2.

ALDAF-AO/PA has more advantages than LV In/Out in the aspect of fetal heart rate. At high heart rate, with no fusion of mitral E and A wave in ALDAF, FAVTI can be determined but by the LVI In/Out method [10, 26, 27]. It was reported that in 39% of moderate to severely prolonged FAVTI cases [28], the LV In/Out method could not identify A wave. However, ALDAF AO/PA can obtain FAVTI in all conditions.

Future studies are warranted to validate the ability of the ALDAF technique to diagnose congenital heart block in pregnancy with positive anti Ro/La autoantibody, and other pregnancies with risk factors for congenital heart block. Our study provides convincing evidence that ALDAF-AO and/or ALDAF-PA is a good technique, with minimal bias, high reliability, and high reproducibility for evaluation of prenatal congenital heart block. Both ALDAF-AO/PA can be used instead LV In/Out. Of note, we did not identify any differences in effectiveness of ALDAF-AO and ALDAF-PA. Additional studies should be done to compare FAVTI measured by ALDAF-AO/PA with other techniques such as SVC/AO or PV/PA. Larger sample sizes for each week of gestation may provide more accuracy in values of FAVTI with ALDAF-AO/PA.

We identified limitations in our study: (1) we did not compare our findings with electrical PR neonatal time intervals. However, the accuracy of FAVTI from ALDAF-AO/PA (mechanical PR interval) showed good correlation with neonatal (EKG) in a previous study [13]. (2) We were likewise unable to compare FAVTI obtained from Doppler measurement to other techniques such as RV tissue Doppler image, fetal EKG and fetal magnetocardiography because of supplier limitation.