British Journal of Obstetrics and Gynaecology August 1979. Vol. 86. pp 623-632
ULTRASOUND ASSESSMENT OF FETAL GROWTH BY
THANKAM R. VARMA, Senior Lecturer and Consultant Gynaecologist Department of Obstetrics and Gynaecology H. TAYLOR, Chief Physicist AND
C . BRIDGES, Physicist Department of Medical Physics St George’s Hospital and Medical School, London
Summary Consecutive ultrasonic measurements of the fetal head area, thorax area, abdomen area, head/thorax (H/T) area ratio and head/abdomen (H/A) area ratio were obtained from 100 patients with a normal pregnancy and from 186 patients with suspected intrauterine fetal growth retardation. Of all the measurements used, H/A area ratio was found to be most accurate in identifying intrauterine fetal growth retardation predicting 82.9 per cent at 33 weeks and 85.7 per cent at 36 to 38 weeks of gestation. This technique was also found to be useful in distinguishing between the two types of growth retardation. The H/A area ratio was helpful in identifying the growth acceleration pattern of a large baby.
GOLDSTEINAND PECKHAM (1 975) have shown that birth weight is the principal variable affecting late fetal and neonatal mortality and that low birth weight babies, both pre-term and light-for-dates, are at greatest risk of anoxic stillbirth and neonatal death. In a series of 44256 consecutive births from 1958 to 1971, Usher and McLean (1974) found a two-fold increase in the perinatal mortality for babies born more than 2 SDs below mean weight over those whose birth weight was within 2SD of the mean. To improve the existing mortality and morbidity associated with a growth retarded fetus, appropriately timed delivery with intensive intrapartum and neonatal care is necessary. From ultrasound cephalometry it would appear that attempts to predict birth weight from a single measurement of the biparietal diameter are of
little value (Varma, 1974). It has been shown that a growth retarded baby has a compensatory ability to continue with normal brain growth, at the expense of body growth. Attention has therefore been directed recently to ultrasound measurements of circumference or area of fetal head, thorax and abdomen, and total uterine volume in relation to fetal growth retardation (Levi, 1972; Hansmann et a2, 1973; Campbell and Wilkin, 1975; Higginbottom et al, 1975; Kurjak and Breyer, 1976; Lunt and Chard, 1976; Wladimiroff et al, 1977). The object of this study was to compare the value of antenatal measurements of the fetal head area, the thorax area, the abdomen area, and the head to thorax and head to abdomen area ratios as indices of fetal growth in utero. 623
TAYLOR AND BRIDGES
METHODS All ultrasound measurements were made from B-scan echograms obtained with the diasonograph 4102 (Nuclear Enterprises Limited, Edinburgh) using a frequency of 2.5 MHz and a velocity setting of 1540 m/s. Measurements of the biparietal diameter (BPD) were made by the combined A and B scan method described by Campbell (1968). The head area was measured from the plane of the measured BPD. The thoracic area was measured from the plane at right angles to the fetal spine at the level of the heart when the maximum movement of one of the valves, either mitral or tricuspid, was identified (Schlensker and Bolte, 1973). Longitudinal scans were made to identify the fetal abdominal aorta or the fetal spine and scans were then made orthogonal to the long axis of the fetal body until the appearance of the umbilical vein as it passed under the fetal liver to measure the abdominal area. When the fetal spine was anterior, the fetal stomach was identified to take the cross section of the fetal upper abdominal area. All areas were measured using a planimeter from polaroid photographs after making the appropriate corrections. Patients The study was arranged in two parts. In the first part (normal series), serial measurements
of the fetal head area, the thorax area, the abdomen area, the head to thorax and the head to abdomen area ratios were obtained from 100 normal patients of known maturity with an uncomplicated pregnancy who were delivered of an infant whose weight was on the tenth centile or above, standardised for maternal parity, sex of the infant and the duration of gestation (Thomson et al, 1968). Maturity of the fetus was confirmed in all patients using cephalometry in the early second trimester. Similar measurements were made in pregnancies with suspected fetal growth retardation (‘at risk’ series). This group consisted of 186 patients, of whom 35 were delivered of a Iightfor-dates infant. An infant was considered lightfor-dates if its birth weight was below the tenth centile standardised for its maturity, maternal parity and sex of the infant (Thomson et al, 1968). Statistical differences between the two groups were termed by Student’s ‘t’ test.
RESULTS Table I shows the pregnancy complications and the number of normal weight and light-fordates infants in the ‘at risk’ series. Of the 35 infants, 26 (74.3 per cent) had a birth weight at or below the fifth centile, and 9 (25.7 per cent) had a birth weight above the fifth centile, but below the tenth centile.
TABLE I Complications of pregnancy in the ‘at risk‘ series Normal infant weight group
Pre-eclampsia Hypertension Static weight Subfertility Clinical growth retardation Antepartum haemorrhage Elderly multigravidae Elderly primigravidae Bad obstetric history Previous neonatal death Previous stillbirth Diabetes Premature labour Urinary tract infection Alcoholic
30 25 30 18
6 6 6 3 2 5 1 2 2 0
5 4 3
0 2 1 1
0 0 0 1
ULTRASOUND ASSESSMENT OF FETAL GROWTH
Head area Normal data. Figure 1 shows the mean, 5th and 95th per cent confidence limits for the fetal head area for each week of gestation from 26 to 40 weeks. The growth of the head area was fairly rapid up to 35 weeks after which there was a gradual reduction in rate up to 40 weeks of gestation.
Fetal growth retardation. Figure 1 shows a significant difference between the mean head areas for the normal series and light-for-dates group in the 'at risk' series from 35 weeks onwards (p