Fetal Measurements: Normal and Abnormal Fetal Growth and Assessment of Fetal Well-Being

First Trimester

Sonographic findings and measurements allow highly accurate dating from 5 weeks' gestation until the end of the first trimester. The earliest sign of an intrauterine pregnancy is identification of a gestational sac in the uterine cavity. This appears as a round or oval fluid collection within the uterine cavity first seen on transvaginal sonography at approximately 5 weeks' gestation. In some cases, the early gestational sac is surrounded by one or two echogenic rings, formed by the proliferating chorionic villi and the deeper layer of the decidua vera. These rings are not consistently present, however, and any round or oval fluid collection in the mid uterus of a woman with a positive pregnancy test is highly likely to be a gestational sac ( Fig. 42.1 ).

From 5 to 6 weeks' gestation, two methods can be used to assign gestational age by ultrasound: (1) measurement of mean sac diameter (MSD) or (2) sonographic identification of gestational sac contents. The MSD is calculated as the average of the anteroposterior diameter, the transverse diameter, and the longitudinal diameter. It increases from 2 mm at 5 weeks to 10 mm at 6 weeks, a growth pattern that can be used to assign gestational age during this period ( Table 42.1 ).

The second method, based on the sonographic findings within the gestational sac, is best done by transvaginal sonography and relies on the observation that, on average, the gestational sac is first identifiable at 5.0 weeks, the yolk sac at 5.5 weeks ( Fig. 42.2 ), and the embryo and embryonic heartbeat at 6.0 weeks ( Fig. 42.3 , Video 42.1 ). The timing of these milestones is subject to minimal variability, with a 95% confidence range of approximately 0.5 weeks. Gestational age can be assigned based on these milestones ( Table 42.2 ).

From 6 weeks until the end of the first trimester, gestational age correlates closely with the crown-rump length (CRL) of the embryo or fetus. The term embryo is commonly used up to 10 weeks' gestation, and the term fetus applies thereafter. The CRL is the length of the embryo or fetus from the top of its head to the bottom of its torso. It is measured as the longest dimension of the embryo, excluding the yolk sac and extremities. From 9 or 10 weeks onward, the CRL measurements are most accurate if taken with the fetus in neutral position ( Fig. 42.4 ). The CRL can be used to assign gestational age accurately up to 14 weeks because there is little biologic variability in fetal length up until that age ( Table 42.3 ). After that point, the CRL of the longer, more developed fetus becomes less reliable. At this later stage, the CRL is affected by the fetal position, measuring shorter in a fetus whose spine is flexed and longer in a fetus whose spine is extended.

Second and Third Trimesters

Many sonographic parameters have been proposed for estimating gestational age in the second and third trimesters. These include several fetal measurements: biparietal diameter (BPD), head circumference (HC), abdominal circumference (AC), femur length (FL), as well as combinations of two or more fetal measurements: the corrected-BPD and composite age formulas. Measurements of structurally abnormal fetal body parts should not be used in the assignment of gestational age.

Fetal Head Measurements

Three measurements or parameters involve the fetal head: BPD, occipitofrontal diameter (OFD), and HC. All three measurements are taken from transaxial sonograms of the fetal head at the level of the paired thalami and cavum septi pellucidi ( Fig. 42.5 ). These head measurements are most accurate if the image is a true axial plane demonstrating the head as oval with symmetry of the thalami and cerebral hemispheres and without visualization of the posterior fossa. The BPD is measured from the outer edge of the cranium nearest the transducer to the inner edge of the cranium farthest from the transducer (“leading edge to leading edge”) ( Fig. 42.5 ). The OFD is obtained from the same transaxial image as the BPD and is measured from mid-skull to mid-skull along the long axis of the fetal head ( Fig. 42.5 ). This latter measurement is used in conjunction with the BPD to calculate the corrected-BPD using the following formula :

$\text{Corrected-BPD}=\sqrt{(\text{BPD}\times \text{OFD})/1.265}$

The rationale for the corrected-BPD is that it represents the BPD of the standard-shaped head (one with an OFD/BPD ratio of 1.265) of the same cross-sectional area. The same tables or formulas used to determine gestational age from the BPD are used to determine gestational age from the corrected-BPD ( Table 42.4 ).

TABLE 42.4

Gestational Age Estimation by Biparietal Diameter (BPD)

Values from Doubilet PM, Benson CB. Improved prediction of gestational age in the late third trimester. J Ultrasound Med. 1993;12(11):647-653.

BPD OR BPDc (mm)	Gestational Age (Weeks)	BPD OR BPDc (mm)	Gestational Age (Weeks)
20	13.2	60	24.2
21	13.4	61	24.5
22	13.6	62	24.9
23	13.8	63	25.3
24	14.0	64	25.7
25	14.3	65	26.1
26	14.5	66	26.5
27	14.7	67	26.9
28	14.9	68	27.3
29	15.1	69	27.7
30	15.4	70	28.1
31	15.6	71	28.5
32	15.8	72	29.0
33	16.1	73	29.4
34	16.3	74	29.9
35	16.6	75	30.3
36	16.8	76	30.8
37	17.1	77	31.2
38	17.3	78	31.7
39	17.6	79	32.2
40	17.9	80	32.7
41	18.1	81	33.2
42	18.4	82	33.7
43	18.7	83	34.2
44	19.0	84	34.7
45	19.3	85	35.2
46	19.6	86	35.8
47	19.9	87	36.3
48	20.2	88	36.9
49	20.5	89	37.4
50	20.8	90	38.0
51	21.1	91	38.6
52	21.4	92	39.2
53	21.7	93	39.8
54	22.1	94	40.4
55	22.4	95	41.0
56	22.8	96	41.6
57	23.1	≥97	42.0
58	23.5
59	23.8

BPDc, Corrected-BPD.

The HC is the length of the outer perimeter of the cranium, made on the same transaxial image of the fetal head. It can be measured by using an electronic ellipse ( Fig. 42.6 , Table 42.5 ) or calculated from the outer-edge-to-outer-edge analogs of the BPD and OFD:

$\text{HC}=1.57\times \left[\right(\text{Outer-to-outer BPD})+(\text{Outer-to-outer OFD}\left)\right]$

TABLE 42.5

Gestational Age Estimation by Head Circumference (HC)

Values derived from formula in Law RG, MacRae KD. Head circumference as an index of fetal age. J Ultrasound Med. 1982;1(7):281-288.

HC (mm)	Gestational Age (Weeks)	HC (mm)	Gestational Age (Weeks)
80	13.4	225	24.5
85	13.7	230	25.0
90	14.0	235	25.5
95	14.3	240	26.1
100	14.7	245	26.6
105	15.0	250	27.1
110	15.3	255	27.7
115	15.6	260	28.3
120	16.0	265	28.9
125	16.3	270	29.4
130	16.6	275	30.0
135	17.0	280	30.7
140	17.3	285	31.3
145	17.7	290	31.9
150	18.1	295	32.6
155	18.4	300	33.3
160	18.8	305	33.9
165	19.2	310	34.6
170	19.6	315	35.3
175	20.0	320	36.1
180	20.4	325	36.8
185	20.8	330	37.6
190	21.3	335	38.3
195	21.7	340	39.1
200	22.2	345	39.9
205	22.6	350	40.7
210	23.1	355	41.6
215	23.6	360	42.4
220	24.0

Although the BPD is simpler to measure than the corrected-BPD or HC, it has the disadvantage of being the only one of the three measurements that disregards head shape. This means that two heads of equal widths but different lengths will have the same BPD, but the longer head will have a greater corrected-BPD and HC than the shorter head ( Fig. 42.7 ). The fetus with the longer head will therefore be assigned a greater gestational age based on the corrected-BPD or HC; however, both fetuses will be assigned the same gestational age if the BPD is used as the basis for age assignment.

Femur Length

The length of the ossified diaphysis of the fetal femur is often used for gestational age estimation. Careful measurement technique is necessary to obtain an accurate estimate of gestational age by FL ( Fig. 42.8 , Table 42.6 ). The transducer must be aligned to the long axis of the diaphysis, perpendicular to the beam, with both ends of the ossified diaphysis appearing squared off. The cursors should be positioned at the junction of the echogenic bone with the hypoechoic cartilage, and the thin, bright reflection of the cartilaginous epiphysis should not be included in the measurement.

TABLE 42.6

Gestational Age Estimation by Femur Length (FL)

Values from Doubilet PM, Benson CB. Improved prediction of gestational age in the late third trimester. J Ultrasound Med. 1993;12(11):647-653.

FL (mm)	Gestational Age (Weeks)	FL (mm)	Gestational Age (Weeks)
10	13.7	45	24.5
11	13.9	46	24.9
12	14.2	47	25.3
13	14.4	48	25.7
14	14.6	49	26.2
15	14.9	50	26.6
16	15.1	51	27.0
17	15.4	52	27.5
18	15.6	53	28.0
19	15.9	54	28.4
20	16.2	55	28.9
21	16.4	56	29.4
22	16.7	57	29.9
23	17.0	58	30.4
24	17.3	59	30.9
25	17.6	60	31.4
26	17.9	61	31.9
27	18.2	62	32.5
28	18.5	63	33.0
29	18.8	64	33.6
30	19.1	65	34.1
31	19.4	66	34.7
32	19.7	67	35.3
33	20.1	68	35.9
34	20.4	69	36.5
35	20.7	70	37.1
36	21.1	71	37.7
37	21.4	72	38.3
38	21.8	73	39.0
39	22.2	74	39.6
40	22.5	75	40.3
41	22.9	76	40.9
42	23.3	77	41.6
43	23.7	≥78	42.0
44	24.1

Abdominal Circumference

The fetal AC is the length of the outer perimeter of the fetal abdomen, measured on transverse view at the level of the intrahepatic portion of the umbilical vein. Ideally the stomach will also be visible ( Fig. 42.9 ). Ideally, the abdomen should appear round. The level chosen should demonstrate only a short segment of the umbilical vein in the anterior third of the liver. Measurement of the AC involves placing the ellipse along the outer skin surface ( Fig. 42.9A ). Alternatively, the AC may be calculated from two orthogonal abdominal diameters (AD ₁ , AD ₂ ), one anteroposterior and the other transverse ( Fig. 42.9B ), measured on the same image, as follows :

$\text{AC}=1.57\times ({\text{AD}}_1+{\text{AD}}_2)$

Composite Formulas

Gestational age can be estimated from measurements of the head, abdomen, or femur by means of tables or formulas that present the mean value of each measurement for a given gestational age ( Tables 42.4 to 42.6 ). Composite age formulas that combine several fetal measurements can also be used to estimate gestational age.

Because composite age formulas use a combination of two or more measurements in conjunction to estimate gestational age, a potential disadvantage of using such formulas is that an abnormal measurement or anomaly might be obscured. For example, in a fetus with a skeletal dysplasia manifested by shortened long bones and a normal head size, the gestational age based on the composite formula will be an underestimation, falling between that predicted by the HC and that predicted by the short FL. As a result, the FL might not appear to be abnormally small when compared to this improperly calculated gestational age.

Gestational Age Estimation by Ultrasound: Most Accurate Approach at Each Stage of Pregnancy

When sonographic, as opposed to clinical, dating is used to assign gestational age, it is important to use the most accurate method or measurement available ( Table 42.7 ). This is most relevant at the time of the initial scan in a woman's pregnancy, since, as noted previously, it is inappropriate to reassign gestational age at the time of the second or subsequent scan during the pregnancy.

If the initial scan is performed in the early first trimester, when ultrasound demonstrates no embryo or an embryo less than 5 mm in length, two approaches to gestational age assignment are equal in accuracy, both with 95% confidence ranges of ±0.5 weeks: date the pregnancy based on MSD (see Table 42.1 ) or on sac contents (see Table 42.2 ). For the rest of the first trimester, when the CRL measures 5 to 80 mm, dating based on CRL (see Table 42.3 ) is very accurate, with a 95% confidence range of ±0.7 weeks.

In the second and third trimesters, there are several options for sonographic dating, including measurements of the head, femur, or abdomen, as well as composite formulas based on measurements of two or more of these. When both the width and length of the head (BPD and OFD) are visible, the optimal approach to sonographic age assignment in the second trimester is to use either the corrected-BPD or the HC, and, in the third trimester, use the corrected-BPD, the HC, or the FL. When either the anterior or posterior aspects of the cranium are not well seen on ultrasound, so that the OFD and HC cannot be measured accurately, it is best to date based on the BPD or FL in the second trimester and on the FL in the third trimester. These approaches are the most accurate single-body-part sonographic dating methods and are preferable to composite age formulas for reasons presented earlier.

Estimation of Fetal Weight

Numerous formulas have been published for estimating fetal weight based on sonographic measurements, most of which use one or more of these fetal body parts: head (BPD or HC), abdomen (AD or AC), and femur (FL). Other measurements, such as thigh circumference, have been used as well. Formulas that estimate fetal weight using three-dimensional (3D) sonography and 3D magnetic resonance imaging (MRI) have also been published.

The accuracy of a weight prediction formula is determined by assessing how well the formula works in a group of fetuses scanned close to delivery. An important measure of a formula's performance is its 95% confidence range. If the 95% confidence range is ±18%, for example, the estimated weight will fall within 18% of the actual weight in 95% of cases, and the error will be greater than 18% in only 5% of cases. The narrower the confidence range is, the more reliable is the formula.

Many published studies provide information on this measure of a formula's accuracy ( Table 42.8 ). The following points are noteworthy:

• The accuracy of weight prediction formulas improves as the number of measured body parts increases up to three, achieving greatest accuracy when measurements of the head, abdomen, and femur are used. There is no apparent improvement by adding the thigh circumference as a fourth measurement and no proven benefit from using 3D sonography or MRI.

• Even when based on measurements of the head, abdomen, and femur, sonographic weight prediction has a rather wide 95% confidence range of at least ±15%. Based on the abdomen and either the head or femur, the range is at least ±16% to 18%. Precision is considerably worse when only the abdomen is used.

• A number of factors have been studied to determine their effect on accuracy of weight prediction. Accuracy appears to be worse in fetuses that weigh less than 1000 g than in larger fetuses. Over the rest of the birth weight range, however, accuracy is fairly constant. Weight prediction is less accurate in diabetic than in nondiabetic mothers: in diabetic mothers, formulas that use measurements of the head, abdomen, and femur have a 95% confidence range of ±24%, wider than the range of ±15% in the general population. The presence of oligohydramnios or polyhydramnios has no impact on accuracy. Scan quality may have an effect on accuracy. Studies have shown a trend toward greater accuracy in scans that were rated “good” compared with those rated “poor” based on ability to visualize anatomic landmarks.

Weight Assessment in Relation to Gestational Age

When an ultrasound is performed in the third trimester, best estimates of gestational age and fetal weight should be established. The gestational age may be based on a prior ultrasound, clinical dating criteria, or current measurements; fetal weight is always calculated from current measurements. The two values should be assessed in relation to one another to determine whether the fetus is appropriate in size for dates. This can be accomplished by using a table that provides norms for fetal weight as a function of gestational age ( Table 42.10 ), several of which appear in the literature.

There is some debate about a number of aspects of fetal weight assessment in relation to gestational age. Should a weight table or chart based on neonatal weights or on estimated fetal weights be used? Should the table or chart be derived solely from pregnancies of low-risk mothers? Should “population norms” or “customized norms” be used?

Most weight norms (tables or charts) are derived from large data sets of neonatal birth weights from babies born at a known gestational age. At least one chart, on the other hand, was produced using estimated fetal weights instead of neonatal birth weights. A rationale for the latter is that several studies have shown that fetuses that deliver preterm are smaller, on average, than those of the same gestational age that remain in utero. Babies born preterm thus represent an abnormal group with a negatively skewed weight distribution. This supports the argument that fetuses should be compared to fetuses, not to babies.

A large international study, INTERGROWTH-21st, used another approach to produce weight norms derived from a healthy population. Although it is based on birth weights rather than fetal weights, the study population used to generate these norms consists of babies born at or beyond 33 weeks to mothers with no known pregnancy-related risk factors. The INTERGROWTH-21st norms, however, are of limited use for assessing fetal weight in relation to gestational age, because they do not cover the period prior to 33 weeks' gestation.

Perhaps the biggest controversy in weight assessment is whether the estimated weight of a fetus should be compared to norms from the overall population or to customized norms derived from a subgroup of fetuses similar to that fetus. For example, if an African-American fetus has an estimated weight of 1200 g at 30 weeks' gestation, should the weight percentile be determined from overall population norms or from African-American norms? In that example, the population-based percentile would be lower than the customized percentile, because African-American fetuses and neonates are smaller, on average, than those in the general population. Although there are proponents of customized norms, a major concern with this approach is that it can do harm by inadvertently normalizing the weight of a fetus who is small on a pathologic basis. This viewpoint is supported by the fact that at least some population groups with small babies have elevated rates of postnatal complications and by the INTERGROWTH-21st finding that neonates of healthy mothers do not differ in size based on ethnic background.

Overall, our recommended approach is to use population norms. Although it would theoretically be preferable to use norms derived from estimated fetal weights, norms derived from birth weights are better established and are based on larger study populations.

The weight gain between two ultrasound examinations can be estimated as the difference between the two estimated weights. Adequacy of weight gain can be assessed by comparing this difference to established normal fetal growth rate as a function of gestational age. INTERGROWTH-21st data indicate that median fetal weight gain per week decreases progressively from 33 weeks of gestation onward, with a maximum rate of 270 g per week at 33 weeks, down to 100 g per week at 41 weeks. Other older growth tables suggest that weight gain may increase until 36 weeks, then decline steadily thereafter. The longer the time is between scans, the more accurate the sonographic estimate of interval weight gain will be. When two scans are performed within 1 week of each other, weight gain cannot be determined reliably, because weight prediction is too imprecise to detect small changes in growth. Furthermore, estimating weight too close to the time of the prior scan could raise unnecessary concern by a spurious finding that the fetus has grown subnormally or even lost weight. Thus there is little or no value in computing an estimated weight at the time of the second scan 1 week after the first. Instead, it is recommended that fetal weight gain only be assessed after an interval of at least 2 weeks' duration.

When several examinations have been performed, fetal growth can be depicted graphically by means of a trend plot or growth curve. One form of growth curve plots the estimated fetal weight versus gestational age, with the curve for the fetus being examined superimposed on lines depicting the 1st, 10th, 50th, 90th, and 99th percentiles ( Fig. 42.10A ). An alternative mode of display plots the estimated fetal weight percentile versus gestational age ( Fig. 42.10B ). In this latter format, the graph for a normally growing fetus will be a horizontal line, indicating maintenance of a particular weight percentile throughout gestation. A down sloping line suggests subnormal growth rate.

Calculation of weight percentiles and plotting of growth curves are most easily accomplished by computer, using an obstetric ultrasound software package that performs these tasks. Alternatively, similar results can be achieved by means of a calculator and manual plotting of data.