Ultrasound Obstet Gynecol 2001; 17: 496 – 501
The relevance of placental location at 20–23 gestational
BlackwellArticle
Original Science, Ltd
weeks for prediction of placenta previa at delivery:
evaluation of 8650 cases
R. H. BECKER, R. VONK, B. C. MENDE, V. RAGOSCH and M. ENTEZAMI
Free University of Berlin, University Hospital Benjamin Franklin, Department of Obstetrics and Gynecology, Berlin, Germany
K E Y W O R D S: Placenta previa, Prenatal diagnosis, Second trimester, Ultrasound
ABSTRACT
Objective To determine the correlation between placental
position at 20–23 weeks and incidence of birth complications
caused by placental position.
Subjects and methods In an ongoing prospective study,
placental position was determined by transabdominal sono-
graphy as part of anomaly scanning at 20 –23 gestational
weeks, followed by transvaginal sonography in uncertain or
suspicious situations. Examination was performed in 9532
cases; feedback was obtained from 8650 patients (90.7%).
Results Transabdominal sonography was followed by
transvaginal scan in 363 of 8650 cases (4.2%). In 8551 cases
(98.9%), we found normal placental position, with the
placenta not reaching the internal os and a Cesarean section
rate of 17.1% (1458/8551). The incidence of ‘low placental
position’, with the placenta reaching the internal os was
0.66% (57/8650), with a Cesarean section rate of 21% (12/
57). In 0.49% (42/8650) of cases, the placenta overlapped
the internal os at 20 –23 weeks; Cesarean section because of
placenta previa or bleeding was performed in 28 of 8650
cases (0.32%). Vaginal delivery was possible in 43% of cases
(13/30), when the overlap did not exceed 25 mm. If the
overlap exceeded 25 mm (12 cases), no vaginal delivery was
reported. There was no reported case of placenta previa
missed at the 20–23-week scan.
Conclusion At 20 –23 weeks, a combination of routine
transabdominal and indication-based transvaginal loca-
tion of placental position is a powerful tool in predicting
placenta previa at delivery. The advantage of determining
placental position at this stage of pregnancy is a low false-
positive rate compared to at earlier stages of pregnancy. We
conclude that an overlapping placenta at 20 –23 weeks has
the consequence of a high probability of placenta previa at
delivery. An overlap of 25 mm or more at 20 –23 weeks
seems to be incompatible with later vaginal delivery.
Correspondence: Dr R. Becker, Free University of Berlin, University Klinikum Benjamin Franklin, Kurfürstendamm 199, 10719 Berlin, Germany
(e-mail: bedaktari@t-online.de)
Received 8-3-00, Revised 14-12-00, Accepted 9-3-01
496
INTRODUCTION
Placenta previa is a rare but dangerous, sometimes fatal, con-
dition: it increases maternal, fetal and neonatal morbidity
and mortality1–3. During pregnancy, placenta previa is one of
the leading causes of third-trimester bleeding4. At delivery,
women with placenta previa have an increased risk of abruptio
placentae, Cesarean delivery, fetal malpresentation and post-
partum hemorrhage5. In the United States between 1979 and
1986, 44 maternal deaths were related to placenta previa as an
underlying disease, mainly because of bleeding and dissem-
inated intravascular coagulation5.
To prevent severe complications, the accurate and timely
diagnosis of placenta previa is indispensable. Unfortunately,
location of placental position produces a diagnostic dilemma:
on one hand it would be advantageous to be informed about
a dangerous placental position as early as possible; on the
other hand, it is well known that the incidence of so called
placenta previa decreases with advancing gestational age,
especially during the second trimester. Many cases of placenta
previa diagnosed during the early second trimester have an
outcome of normal delivery6. Therefore too early assessment
of placental position results in unnecessary patient anxiety7,8.
Our aim was to determine the accuracy of placental location
at a 20–23-week anomaly scan in predicting complications
related to placental position at delivery.
SUBJECTS AND METHODS
In an ongoing prospective study beginning in September
1992, the position of the placenta was registered in 9532 non-
selected consecutive patients undergoing anomaly scanning
at 20–23 gestational weeks in our institution, which serves
as one of the referral centers of Berlin. We received feed-
back concerning the outcome of pregnancies in 8650 cases
(91.9%); evaluation was done for these 8650 patients only.
The study group included 255 sets of twins, 19 sets of triplets
O RIGIN AL PAPER
Placenta previa Becker et al.

and 1 set of quadruplets. Each multiple pregnancy was
counted as one case. We included only pregnancies with a
duration of more than 24 weeks; pregnancies terminated
because of malformations were excluded from evaluation.
Most of the patients were delivered in one of the several
hospitals of the city or the surrounding area. Feedback of
information concerning details of delivery including com-
plications was achieved by return of a form handed out to the
patient after examination; if omitted, information was
obtained by telephone contact in most cases with the patient.
In 1998 and 1999, this was performed in 958 of 2333 cases
(41%); we therefore estimate the rate of spontaneous return
of information in our whole study group to be around 50%.
The mean age of the patients was 31.1 (15 –46) years. The
mean gestational age was 21.6 weeks. Examinations were
performed with an Acuson 128 XP10 (Mountain View, CA,
USA) ultrasound machine. Transabdominal scans were per-
formed with 4- and 5-MHz probes and transvaginal scans
with 7- or 8.5-MHz probes. Experience of the participating
physicians at the beginning of the study was more than
2 years of exclusive sonographic work. The placenta was
located by abdominal scan with a normally filled bladder,
followed by transvaginal scan in uncertain or suspicious sit-
uations (n = 363, 4.2%). For evaluation, observations were
divided into three groups:
Group I: ‘normal’ position of placenta: the lower margin
of the placenta did not reach the internal os;
Group II: ‘low position’ of placenta: the lower margin of
the placenta reached the internal cervical os but did not
overlap (Figure 1);
Group III: ‘overlapping placenta’: the lower margin of the
placenta overlapped the internal os (Figure 2).
When an overlapping placenta was suspected, the patient
and her physician were informed, and follow up was recom-
mended at 28 weeks. Calculations were performed using SAS
(SAS, Heidelberg, Germany). Exact Clopper–Pearson confidence
intervals were used for the proportions.
RESULTS
The frequency of different classes of placental position is
listed in Table 1. The correlation of gestational age and fre-
quency of ‘low’ or ‘overlapping’ placenta is presented in
Table 2. An overview of the frequency of Cesarean sections
in correlation to position of placenta at 20–23 weeks is
shown in Table 3. The frequency of Cesarean section in the
whole study group was 17.3% (1499/8650). Patients with
‘normal placental position’ (Group I) were delivered by
Cesarean section with a frequency of 17.1% (1458/8551). In
‘low position’ (Group II), the frequency was 21% (12/57);
this frequency was not statistically different from that of Group
I (P = 0.38, Fisher’s exact test). In 42 patients we found a
placenta overlapping the internal os containing one set of
twins (Group III). We had occasion to perform follow-up
examinations in 22 of these cases. In six of 22 cases, placental
position had normalized in later phases of pregnancy, all of

Table 1 Placental position at 20–23 weeks

Figure 1 Low position of placenta of uterine anterior wall, visualized by
transvaginal sonography with an 8.5-MHz probe. P, placenta; FH, fetal Placental position n (%) 95% confidence interval
head, large arrow, internal cervical os; small arrows, cervical channel.
‘Normal’ (Group I) 8551 (98.90) 98.61–99.07%
‘Low’ (Group II) 57 (0.66) 0.50–0.85%
‘Overlapping’ (Group III) 42 (0.49) 0.35–0.66%
> 0–10 mm 14 (0.16) 0.09–0.27%
11–24 mm 16 (0.19) 0.11–0.30%
≥ 25 mm 12 (0.14) 0.07–0.24%

Percentages are based upon the total population (n = 8650).

Table 2 Gestational age and placental extension

Placenta extended to or over

the internal cervical os
Gestational age
(weeks) n n Proportion (%)

20 + 0 to 20 + 6 661 11 1.66
21 + 0 to 21 + 6 3613 44 1.22
Figure 2 Placenta overlapping the internal cervical os by 16.6 mm, 22 + 0 to 22 + 6 3381 35 1.04
visualized by transvaginal sonography at 21 + 1 weeks with a 7-MHz 23 + 0 to 23 + 6 995 9 0.90
probe. P, placenta; large arrow, internal cervical os; small arrows, Overall 8650 99 1.14
cervical channel.

Ultrasound in Obstetrics and Gynecology 497

Placenta previa
them with an overlap less than 25 mm; 5 of these were delivered
vaginally; one was by Cesarean section because of malpres-
entation. The patient with the greatest amount of overlap and
normal delivery had an overlap of 20 mm. The frequency of
Cesarean section was 69% (29/42) including the one case of
Cesarean section because of malpresentation (overlap 10 mm
at 22 + 6 weeks; no overlap at 26 + 6 weeks). In all other
cases, Cesarean section was performed because of placenta
previa according to the decision of the responsible physician,
nine of them because of vaginal bleeding between 31 and 39
gestational weeks. All patients with an overlap of 25 mm or
more delivered by Cesarean section. In patients with pla-
centa overlapping less than 25 mm (n = 30), the frequency
of Cesarean section because of placenta previa in the sub-
group of 1–10 mm overlap was 43% (6/14) compared to
62.5% (10/16) in the subgroup of 11–24 mm (P = 0.46,
Fisher’s exact test). The correlation between degree of
placental overlap and sensitivity for Cesarean section
because of placenta previa in Group III (28/42, 67%) is
shown in Figure 3. An overview of the duration of pregnancies
in the three groups of different placental position is given in
Table 4.
Table 3 Cesarean section in relation to placental position at 20–
23 weeks
95% confidence11
Placental position n (%) interval (%)
‘Normal’ (Group I) 1458/8551 (17.10) (16.26 –17.87)
‘Low’ (Group II) 12/57 (21.05) (11.38 –33.89)
‘Overlapping’ (Group III) *29/42 (69.05) (52.91–82.39)
> 0 –10 mm *7/14 (50.00) (23.04 –76.96)
11–24 mm 10/16 (62.50) (35.44 –84.80)
≥ 25 mm 12/12 (100.00) (73.54 –100.00)
Overall 1499/8650 (17.33) (16.54 –18.14)
Percentages are based on the total number of patients in the group.
*Includes one case of Cesarean section because of fetal malpresentation
with overlapping placenta at 22 weeks and normal placental position at
control sonography.
Figure 3 Receiver operating characteristic (ROC) curve for Cesarean
section due to persistence of placenta previa depending upon the degree
of overlap over the internal os in group III. (Group III consisted of 42
patients; 28 of these had a Cesarean section because of placenta previa.)
Area under curve (overlap) is 0.786 (standard error, 0.069). The dashed
line represents: Sensitivity = 1 – Specificity.
498
Becker et al.
There was neither fetal nor maternal death in Group III.
One patient with a history of Cesarean section and an overlap
of posterior wall placenta of more than 50 mm experienced
massive hemorrhage during Cesarean section at 39 weeks
and underwent hysterectomy because of placenta increta. In
another patient, using color Doppler sonography we detected
overlapping placenta in combination with velamentous inser-
tion with an arterial umbilical vessel on the surface of the
overlapping part of the placenta (Figures 4 and 5); she under-
went emergency Cesarean section following vaginal bleeding
at home at 32 + 4 weeks. One patient presented with a vas
previum detected at 23 + 0 gestational weeks (Figure 6); she
was admitted to hospital at 30 weeks and delivered by emer-
gency Cesarean section following vaginal bleeding at 32 + 5
gestational weeks. In all cases, outcome for mother and child
was favorable and all of the pregnancies of this group had a
duration of at least 31 weeks (Table 4).
DISCUSSION
Placenta previa at delivery is a rare condition. Frequencies
reported at birth range from 0.14%9 to 0.75%10 with a mean
of 0.38%1,5,8–16. An overview of frequencies of placenta pre-
via is given in Table 5. In our study group, the incidence of
placenta previa at delivery with the consequence of Cesarean
section was 0.32% (28/8650; 95% confidence interval
Table 4 Duration of pregnancy in relation to placental position at 20–
23 weeks
Duration of pregnancy (weeks)
Placental position n Median Range
‘Normal’ (Group I) 8551 39.0 24–43
‘Low’ (Group II) 57 39.0 32–42
‘Overlapping’ (Group III) 42 38.0 31–41
Overall 8650 39.0 24–43
Figure 4 Placenta overlapping the internal os at 25 + 6 gestational weeks
visualized by transabdominal scan, with a 4-MHz probe. Overlap was
detected at 20 + 6 gestational weeks, with an arterial vas previum
(umbilical artery) on the surface of the overlapping part of the placenta
continuing as insertio velamentosa. X, insertion of the umbilical cord at
the ventral uterine wall; large arrow, internal cervical os; small arrows,
cervical channel.
Ultrasound in Obstetrics and Gynecology
Placenta previa Becker et al.

ranges, 0.22–0.47%) which lies close below the mean of pub-
lished data. The differences of incidence in different study
groups may be due to the higher amount of high-risk preg-
nancies in patients of referral hospitals as well as racial and
genetic differences. Also, factors like confusing abruption with
placenta previa, quality of outcome information or size of
study group may be of importance.
A crucial point in evaluation is the verification of placenta
previa at delivery. In our study group, this was performed by
the responsible physician in the different hospitals of the city.
In nine cases, patients were delivered by Cesarean section in
acute clinical situations of bleeding following the diagnosis of
placenta previa by previous sonograms. In one case, Cesar-
ean section was performed because of malpresentation of the
fetus in a situation where ‘overlapping’ placenta had normal-
ized; this case was counted as non-persistence. In 19 cases of
Cesarean section following overlapping placenta at the 20–
23-week scan, the information of a Cesarean section having
been carried out because of placenta previa was given to us
Figure 5 Transvaginal sonography (7-MHz probe) of the same situation
as in Figure 4 (In Figure 4, fetal head moved to lateral part of amniotic
cavity.) with pulsed Doppler of the vas praevium. P, placenta; FH, fetal
head; large arrow, internal cervical os; small arrows, cervical channel.
by the patients according to the information they had received
from their physicians in charge. In our opinion this seemed
to be sufficiently reliable but may bear a small risk of incorrect
information. With an incidence of 0.32% in our study group
and 0.38% as the mean of incidence of more than 500 000
patients (Table 5) our assessment of placenta previa most prob-
ably does not include a large amount of incorrect information.
During pregnancy, the frequency of so-called ‘placenta
previa’ decreases with increasing gestational age due to an
effect described as ‘placental migration’17, caused by the fact
that the placenta-free uterine wall grows faster than the placenta-
covered area11. This effect was observed in our study group
with the frequency of ‘low’ or ‘overlapping’ placenta decreasing
from 1.66% at week 20 to 0.90% at week 23 (Table 2).
The consequence of this effect is a higher amount of false-
positive cases of diagnosis of ‘placenta previa’ in earlier
phases of pregnancy.
An overview of the reported incidence and regression rate
of so-called ‘placenta previa’ is given in Table 6. Incidence
at the time of scan decreased from 2.4% at 12–16 weeks11
to 0.48% at 20–23 weeks, depending on the definition of
‘placenta previa’. Parallel to this effect, the persistence rate
of ‘placenta previa’ until delivery increased from 6.4% at
12–16 weeks11 to 67% at 20–23 weeks.
Before introduction of transvaginal sonography, diagnosis
of placenta previa was difficult and had to rely on indirect
signs16. Since the introduction of transvaginal sonography,
this method has become the gold standard for the diagnosis
of placenta previa in all phases of pregnancy18,19. Performed
accurately, it seems to pose no threat to the pregnancy18–21.
In our study group, one patient with an overlap of 16 mm
reported bleeding in the evening after vaginal examination.
This was possibly due to abruption of the overlapping part
of the placenta following contractions induced by trans-
vaginal sonography; after 8 days in hospital, her pregnancy
went on to 40 weeks and ended in a spontaneous delivery.
It is still unclear whether accurate diagnosis of placenta
previa is possible only by routine application of transvaginal
sonography9,11. Our experience with the use of transvaginal
sonography only in uncertain or suspicious situations, which

Table 5 Incidence of placenta previa at delivery

Incidence
Time span Study group (n) n %

Taipale et al. 19989 1995–96 3696 5 0.14

Taipale et al. 199711 1993–94 6428 10 0.16
Lauria et al. 199612 > 2910 5 0.17
Taylor et al. 199413 1984–87 278 933 917 0.33
Crane et al. 19991 1988–95 92 983 305 0.33
McClure and 0.47
Dornan 19908 1986 1490 7
Iyasu et al. 19935 1979–86 0.48
Frederiksen et al. 199914 1976–97 93 384 514 0.55
Figure 6 Vasa praevia at 23 + 0 gestational weeks visualized by Williams et al. 199115 1977–80 12 420 69 0.56
transvaginal scan with an 8.5-MHz probe: low position of placenta of Rizos et al. 197916 1972–77 1098 6 0.58
uterine anterior wall with the umbilical cord overlapping the internal os Sherman et al. 199210 1990 8410 63 0.75
and continuing on the posterior wall as insertio velamentosa. P, placenta; All 501 752 1901 0.38
FH, fetal head; arrows, cervical channel.

Ultrasound in Obstetrics and Gynecology 499

Placenta previa Becker et al.

Table 6 Incidence of placenta previa at ultrasound, persistence until delivery and incidence at birth

GA at US Incidence at US Persistence Incidence at birth

n (weeks) P.p. def US type (n (%)) (n (%)) (n (%))

Taipale et al. 199711 6428 12 –16 > 14 mm TV 156/6428 (2.4) 10/156 (6.4) 10/6428 (0.16)
Lauria et al. 199612 > 2910 15–20 ≥ 0 mm TA + TV 31/2910 (1.1) 5/36 (14) 5/ > 2910 (< 0.17)
Rizos et al. 197916 1098 16 –18 > 0 mm TA 58/1098 (5.3) 5/52 (10) 5/1098 (0.46)
Taipale et al. 19989 3696 18–23 > 0 mm TV 57/3969 (1.5) 5/57 (8.8) 5/3696 (0.14)
> 15 mm 27/3969 (0.68) 5/27 (18.5) 5/3696 (0.14)
≥ 25 mm 10/3696 (0.27) 4/10 (40) 4/3696 (0.11)
Present study 8650 20 –23 > 0 mm TA + TV 42/8650 (0.48) *28/42 (67.0) *28/8650 (0.32)
≥ 25 mm 12/8650 (0.14) 12/12 (100) 12/8650 (0.14)

*Excluding one case of Cesarean section because of fetal malpresentation with overlapping placenta at 22 weeks and normal placental position at
control sonography. GA, gestational age; US, ultrasound; P.p. def, definition of placenta previa (degree of overlap over internal os); TA,
transabdominal scan; TV, transvaginal scan.

occurred in 4.2% of cases and to our knowledge missed no

diagnosis of placenta previa at delivery, confirms reports12,22
which demonstrate that transabdominal ultrasound is satis-
factory for screening for placenta previa.
Another topic of discussion is the question: what is the cut-
off point of placental overlapping of the internal cervical os
for predicting placenta previa with a specificity of 100%?
Regarding the natural history of the condition, it seems likely
that the dimension at which spontaneous regression is pos-
sible varies with gestational age. An overlap of 25 mm or
more had a persistence rate of placenta previa to term of 40%
in a study group performing ultrasound at 18 –23 weeks9. In
our group with a further advanced period of assessment of
20–23 weeks, all 12 patients with an overlap of 25 mm or
more delivered by Cesarean section, indicating that an over-
lap of 25 mm or more might be incompatible with the chance
of regression if assessed at this time of pregnancy.
As a consequence of these data, it seems justified to inform
patients in these situations about their prognosis as follows:
• In the case of ‘low’ placental position, there is a probability
of Cesarean section similar to that of ‘normal’ placental
position.
• In the case of ‘overlapping’ placenta,
with an overlap of up to 20 mm there is a high probability of
placental-position-originated complications at delivery, with
a chance of spontaneous regression; follow-up sonography
should be performed;
with an overlap of 25 mm or more there is a very high probabil-
ity of the necessity of Cesarean section, with a very low chance
of regression.
Our experience with placental localization at 20 –23 weeks
shows that this phase of pregnancy is late enough to reduce
to a minimum the frequency of false-positive cases of predic-
tion of placenta previa at birth. With an incidence in our
group of 0.32% of placenta previa at birth, it seems unlikely
that we lacked reports of considerable numbers of placenta
previa because of incomplete feedback from the patients;
thus the sensitivity at this stage of pregnancy seems to be high
using a combination of routine transabdominal and indic-
ated transvaginal sonography. Furthermore, in Group III,
there was neither fetal nor maternal death: at this phase of
pregnancy it is not yet too late to prevent severe complications
of pregnancy by adequately counseling the pregnant women.
REFERENCES
1 Crane JMG, Van Den Hof MC, Dodds L, Armson A, Liston R.
Neonatal outcomes with placenta previa. Obstet Gynecol 1999; 93:
541–4
2 Gorodeski IG, Bahari CM. The effect of placenta previa localization
upon maternal and fetal-neonatal outcome. J Perinat Med 1987; 15:
167–77
3 Jouppila P. Vaginal bleeding in the last two trimesters of pregnancy:
a clinical and ultrasonic study. Acta Obstet Gynecol Scand 1979; 58:
461–7
4 McShane PM, Heyl PS, Epstein MF. Maternal and perinatal
morbidity resulting from placenta previa. Obstet Gynecol 1985;
65: 176–82
5 Iyasu S, Saftlas AK, Rowley DL, Koonin LM, Lawson HW, Atrash HK.
The epidemiology of placenta previa in the United States, 1979 through
1987. Am J Obstet Gynecol 1993; 168: 1424–9
6 Wexler P, Gottesfeld KR. Second trimester placenta previa. An
apparently normal placentation. Obstet Gynecol 1977; 50: 706–9
7 Andersen ES, Steinke NMS. The clinical significance of asymptomatic
mid-trimester low placentation diagnosed by ultrasound. Acta
Obstet Gynecol Scand 1988; 67: 339–41
8 McClure N, Dornan JC. Early identification of placenta previa.
Br J Obstet Gynaecol 1990; 97: 959–61
9 Taipale P, Hiilesmaa V, Ylöstalo P. Transvaginal ultrasonography at
18–23 weeks in predicting placenta previa at delivery. Ultrasound
Obstet Gynecol 1998; 12: 422–5
10 Sherman SJ, Carlson DE, Platt LD, Medearis AL. Transvaginal ultra-
sound: does it help in the diagnosis of placenta previa? Ultrasound
Obstet Gynecol 1992; 2: 256–60
11 Taipale P, Hiilesmaa V, Ylöstalo P. Diagnosis of placenta previa by
transvaginal sonographic screening at 12–16 weeks in a nonselected
population. Obstet Gynecol 1997; 89: 364–7
12 Lauria MR, Smith RS, Treadwell MC, Comstock CH, Kirk JS, Lee W,
Bottoms SF. The use of second-trimester transvaginal sonography
to predict placenta previa. Ultrasound Obstet Gynecol 1996; 8:
337–40
13 Taylor VM, Kramer MD, Vaughan TL, Peacock S. Placenta previa
and prior Cesarean delivery: how strong is the association? Obstet
Gynecol 1994; 84: 55–7
14 Frederiksen MC, Glassenberg R, Stika CS. Placenta previa: a 22-year
analysis. Am J Obstet Gynecol 1999; 180: 1432–7
15 Williams MA, Mittendorf R, Lieberman E, Monson RR, Schoenbaum SC,
Genest DR. Cigarette smoking during pregnancy in relation to placenta
previa. Am J Obstet Gynecol 1991; 165: 28–32
16 Rizos N, Doran TA, Miskin M, Benzie RJ, Ford JA. Natural history
of placenta previa ascertained by diagnostic ultrasound. Am J Obstet
Gynecol 1979; 133: 287–91
17 Townsend RR, Laing FC, Nyberg DA, Brooke Jeffrey R, Wing VW.
Technical factors responsible for ‘placental migration’: sonographic
assessment. Radiology 1986; 160: 105–8

500 Ultrasound in Obstetrics and Gynecology

Placenta previa
18 Farine D, Fox HE, Jakobson S, Timor-Tritsch IE. Vaginal ultrasound for
diagnosis of placenta previa. Am J Obstet Gynecol 1988; 159: 566–9
19 Farine D, Fox HE, Jakobson S, Timor-Tritsch IE. Is it really a
placenta previa? Eur J Obstet Gynecol Reprod Biol 1989; 31: 103–8
20 Timor-Tritsch IE, Yunis RA. Confirming the safety of transvaginal
sonography in patients suspected of placenta previa. Obstet Gynecol
1993; 81: 742 – 4
Ultrasound in Obstetrics and Gynecology
Becker et al.
21 Leerentveld RA, Gilberts ECAM, Arnold MJCW, Wladimiroff JW.
Accuracy and safety of transvaginal sonographic placental localization.
Obstet Gynecol 1990; 76: 759–62
22 Smith RS, Lauria MR, Comstock CH, Treadwell MC, Kirk JS, Lee W,
Bottoms SF. Transvaginal ultrasonography for all placentas that
appear to be low-lying or over the internal cervical os. Ultrasound
Obstet Gynecol 1997; 9: 22–4
501