The Journal of Maternal-Fetal & Neonatal Medicine

ISSN: 1476-7058 (Print) 1476-4954 (Online) Journal homepage: https://www.tandfonline.com/loi/ijmf20

Diagnosis and management of placenta previa and

low placental implantation

Souhail Alouini, Pascal Megier, Arnaud Fauconnier, Cyrille Huchon, Adele

Fievet, Anna Ramos, Charles Megier & Antoine Valéry

To cite this article: Souhail Alouini, Pascal Megier, Arnaud Fauconnier, Cyrille Huchon, Adele
Fievet, Anna Ramos, Charles Megier & Antoine Valéry (2019): Diagnosis and management of
placenta previa and low placental implantation, The Journal of Maternal-Fetal & Neonatal Medicine,
DOI: 10.1080/14767058.2019.1570118

To link to this article: https://doi.org/10.1080/14767058.2019.1570118

Published online: 27 Jan 2019.

Submit your article to this journal

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ijmf20
THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE
https://doi.org/10.1080/14767058.2019.1570118

ORIGINAL ARTICLE

Diagnosis and management of placenta previa and low placental

implantation
Souhail Alouinia, Pascal Megiera, Arnaud Fauconnierb, Cyrille Huchonb, Adele Fieveta, Anna Ramosa,
Charles Megiera and Antoine Val
eryc
a
Department of Gynaecology and Obstetrics, Centre Hospitalier R
egional d’Orl
eans, Orleans, France; bDepartment of Gynecology and
Obstetrics, University of Versailles Saint-Quentin in Yvelines, Versailles, France; cDepartment of Medical Informatics and Statistics,
Centre Hospitalier R
egional d’Orl
eans, Orleans, France

ABSTRACT ARTICLE HISTORY

Objective: To evaluate the migration of low-placental implantation (LPI) during the third trimes- Received 17 July 2018
ter of pregnancy and its effect on delivery and post-partum hemorrhage. Accepted 12 December 2018
Methods: We conducted a retrospective study at a level 3 maternity center including all cases
KEYWORDS
of placenta previa (PP) and LPI between 1998 and 2014. The distance (d) between cervical
Caesarean section; low-lying
internal os (CIO) and placental edge (PE) were measured by vaginal ultrasonography in the third placenta; placenta previa;
trimester of pregnancy at 32 and 3 weeks after. We analyzed CIO-PE distance, volume of post- placental edge; placental
partum hemorrhage, delivery decision, and mode of delivery using Kruskall–Wallis test. migration; post-
Results: In total, 319 patients presented with PP or LPI. All complete PP (121) and 90.6% (58 of partum hemorrhage
64) of the placentas less than 1 cm from the CIO did not migrate. Among the 138 placentas
with an initial CIO-PE d greater than 1 cm, only 17 (12.3%) did not migrate above 2 cm. The
patients for whom the decision to perform a cesarean section (C-section) was retained and real-
ized had a CIO-PE d significantly lower than those who delivered vaginally (p < .001). The
patients who delivered by C-section had a lower CIO-PE d when an emergency C-section was
performed, specifically for hemorrhage (p < .001). The mean volume of hemorrhage was signifi-
cantly higher for patients with a CIO-PE d less than 2 cm.
Conclusion: Complete PP and the majority of the placentas less than 1 cm from the CIO did not
migrate. Above 1 cm, the majority of the placentas migrated three to four weeks later. For the
placentas less than 1 cm from the CIO, a significant risk of hemorrhage at delivery was observed.
Thus, prophylactic cesarean section is required for CIO-PE distances <1 cm. For distances
between 1 and 2 cm, the volume of blood loss tends to be more important than for distances
>3 cm without statistical significance. A vaginal delivery could be tried after information
of patients.

Introduction characterize the relation between the placenta and

Placenta previa (PP) and low-placental implantation the cervix [8]. According to Bhide et al. [9], the dis-
(LPI) are obstetrical situations that pose a high risk of tance to be measured is between the CIO and the PE.
antepartum and post-partum hemorrhage as well as Cases in which the placenta completely covers the cer-
perinatal mortality [1–5]. Therefore, ultrasound (US) vical os are defined as complete PP. The placenta is
screening of the placental implantation is essential. considered low lying at a distance of up to 3–4 cm
Advances in ultrasonography performance support the from the CIO [10]. Some authors characterized the pla-
precise measurement of the distance between the pla- centa as low lying when the distance between the
cental edge (PE) and the cervical internal os (CIO). This CIO and the PE is less than 2 cm [11]; according to
measurement was previously performed abdominally others, this distance could be extended to 3 or 3.5 cm.
and the classifications of PP were based on abdominal Some authors reported the possible migration of LLP
measures [5,6]. Abdominal measures can still be used above 1 or 2 cm [12]. The relation between CIO-PE
for screening purposes [7]. distance (d), mode of delivery and the occurrence
The developments in vaginal probes make vaginal of post-partum hemorrhage (PPH) is inad-
measurements more precise and seem to better equately determined.

CONTACT Souhail Alouini alouini.s@orange.fr Department of Gynecology and Obstetrics, Centre Hospitalier R
egional d’Orl
eans, 14 avenue de
l’H^opital, CS 86709, Orleans CEDEX 2, France
ß 2019 Informa UK Limited, trading as Taylor & Francis Group
2 S. ALOUINI ET AL.

Table 1. General characteristics of the patients with placental placenta. The echographs used were Voluson 730
edge to cervical internal os distance less than or equal to (RIC5-9H transvaginal probe), followed by Voluson E8
4 cm (n ¼ 319).
(GE Ric 5–9D vaginal transducer probe).
% Standard deviation
The measures were reported by intervals in centi-
N patients 319
Mean age (years) 33 6 meters. We defined five categories of CIO-PE distances:
Previous 84 26 0, 0–1, 1–2, 2–3, and >3 cm. Complete previa was
C-section
(n patients) defined as 0 cm between the CIO and the PE.
Previous curettage (n patients) 87 27 Cases of CIO-PE d greater than 4 cm were excluded
Mean term of delivery (WG) 37.41 2.53
Birth weight of newborn (g) 2942 583 in the first US measure. Cases of placenta accreta or
Mean volume of post-partum 694 873 percreta diagnosed by histopathologic exams were
hemorrhage (ml)
excluded from the study.
The volume of blood loss and mode of delivery
were recorded. PPH was quantified during delivery
According to some authors, vaginal delivery should
using a graduated bag placed under the patient’s but-
be attempted when the CIO-PE d is greater than 1 cm
tocks. In case of a C-section, blood loss was quantified
[13]; others propose vaginal delivery when the CIO-PE
by collecting the blood sucked into a graduated jar
d is greater than 2 cm [9].
The present study evaluated the evolution of the and weighing the compresses soaked in blood.
CIO-PE d during the third trimester in patients who According to the French guidelines of the College of
presented with PP or LPI and assessed its effect on Obstetricians and Gynecologists, a planned C-section
mode of delivery and PPH. should be proposed for patients presenting with com-
plete PP. A trial of vaginal delivery could be proposed
for CIO-PE distance >2 cm or less than 2 cm if blood
Methods loss is not important and controlled [14].
We conducted a retrospective study in the maternity Data were acquired from medical charts and each
department of the Regional Hospital Center of Orleans case was anonymized by numbering. The study was
between 1998 and 2014 including all cases of PP or approved by the French committee (Commission of
LLP. All patients were managed in the protocol frame- Informatics and Liberty, number 2017–005).
work of the department. According to the framework,
in case of a suspected diagnosis of a PP or LPI in the Statistical analysis
third trimester of pregnancy (31–32 weeks) by abdom-
inal US exam or of bleeding, another ultrasound exam Following a descriptive analysis (mean, 95% confi-
would be performed by an endovaginal probe to dence interval, standard deviation, and percentage),
measure the CIO-PE distance. All US examinations the means were compared using the Student’s t-test
were performed by a senior specialist in obstetrics or the Wilcoxon, test for not normal distributions. For
ultrasonography. The measure was realized when the multiple means comparison, we used the
image of the whole length of the cervix and the lower Kruskal–Wallis test. When the results were statistically
PE was obtained. A complete previa was diagnosed if significant, comparison by pairs was performed to cal-
the placenta covered the internal cervical os. When culate the p values, which were adjusted using
the placenta did not cross the CIO, we measured the Holm’s method.
smallest distance between the PE and the CIO. For binary variables, the proportions were com-
In France, three US examinations are routinely per- pared using the chi-squared test of independence or
formed to screen fetal or placental abnormalities at Fisher’s exact test (if a number was at least theoretic-
12, 22, and 32 weeks of gestation according to the ally less than five). Nominal requirements were used
French guidelines on obstetrics and gynecology. These for more than two classes: Pearson’s chi-squared or
exams are usually performed through the abdom- Fisher’s exact test. In case of significance, Holm’s
inal way. method was used for a two-to-two post hoc compari-
A second US examination was performed at 35 or son. The software application used was R version 3.1.2.
3 weeks after the first one to evaluate the potential
migration of the placenta and to plan a delivery route.
Results
All patients with a CIO-PE d less than or equal to 4 cm
were included in the subsequent analysis. Calipers In total, 319 patients presented with PP or LPI in the
were placed at the CIO and the lower extremity of the third trimester of pregnancy. The mean age of
 THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE 3

TABLE 2. Evolution of the placental edge to cervical internal For cesarean section, the volume of blood loss was
os distance (n ¼ 319). significantly higher when the CIO-PE d was <1 cm
[0] [0–1] [1–2] [2–3] >3 Total than when it was >3 cm (adjusted p ¼ .0267).
N initial measure of CIO–PE d 123 62 60 43 31 319 When the CIO-PE d was between 1 and 2 cm, there
% 38.56 19.43 18.8 13.48 9.72 100
N final measure of CIO–PE d 123 58 17 26 95 319 was a tendency to more important bleeding but it
% 38.56 18.18 5.33 8.15 29.78 100 was not statistically significant.
In 62% of the cases, the decided mode of delivery
patients was 33 years ± 6. Among the patients, 84 was prophylactic C-section, but the effective rate of C-
(26%) previously had one or more C-sections and 87 sections was significantly higher at 69%. Delivery deci-
(27%) previously had one or more curettages. The sion seemed more closely correlated with the final US
mean term of delivery was 37 weeks and the mean measure, which is nearer the delivery. Most of the dis-
newborn birth weight was 2941 g. The mean volume crepancies between the delivery decision and the
of hemorrhage was 693 ml ± 872. The patients’ gen- effective mode of delivery were related to the per-
eral characteristics are summarized in Table 1. formance of a C-section. Patients for whom the deci-
sion to perform a C-section was retained and realized
had a significantly lower PE-CIO d than those who
Migration of the placenta delivered vaginally (p < .001) (Table 4 and Figure 2).
Among the 123 cases of complete PP, none had
moved at the time of the second US and remained PPH and CIO-PE d
complete. Among the patients in group 0–1, the pla-
centa did not move significantly. The most notable The mean volume of PPH in cases of vaginal delivery
migration of the placenta occurred in cases of a CIO- was 321 ml ± 389 and that in cases of C-section was
852 ml ± 969 (p < .001). The mean volume of hemor-
PE d between 1 and 4 cm (Table 2). In the majority of
rhage was significantly higher for C-section. Thirty per-
the cases, placentas less than 1 cm from the CIO at
cent of the patients had an emergency C-section,
32 weeks did not migrate.
including 24% for hemorrhage. For patients who deliv-
ered by C-section, the PE-CIO d was lower for emer-
Mode of delivery and PPH (Table 3) gency C-sections, especially for those involving
hemorrhage (p < .001). The mean volume of hemor-
In the complete PP group, 123 patients (100%) had a
rhage differed significantly according to the PE-CIO d
C-section and the mean volume of hemorrhage was (Kruskal–Wallis test, p < .001).
918 ml ± 1034. In the group [0–1] (n ¼ 58), only one As shown in Figure 3, the mean volume of hemor-
patient delivered vaginally (700 ml of blood loss) and rhage was significantly higher for patients with a
57 of 58 (98%) had a C-section. The mean volume of CIO–PE d less than 2 cm (p < .001). In total, the mean
hemorrhage was 680 ml ± 471. volume of blood loss during delivery was higher for
In the group [1–2] (n ¼ 17), 9 patients (52.9%) patients who had a CIO–PE d less than 2 cm and for
underwent a C-section (793 ml ± 482 of blood loss) women who underwent a C-section (Figure 4). The
and 8 delivered vaginally with a mean blood loss of threshold of 2 cm distinguishes deliveries with a high
522 ml ± 433. In these three groups, 189 of the 198 volume of blood loss from those with a low volume of
patients (89.5%) underwent a C-section and 9 deliv- blood loss.
ered vaginally.
The volume of blood loss tends to be more import-
ant in vaginal delivery (VD) when the CIO-PE d is less Discussion
than 2 cm (Figure 1), however, the number of patients The initial mean of the CIO-PE d at 32 weeks corre-
is insufficient to show statistically significant difference lated with the placenta migratory prognosis. Indeed,
(p ¼ .11, Kruskal–Wallis test). the CIO-PE d at 32 weeks did not change for measures
In total, 219 of the 319 patients (69%) had a C-sec- between 0 and 1 cm; the distances only increased sig-
tion and 100 patients underwent vaginal delivery. nificantly above 1 cm. According to Haino et al. [12],
Only one vaginal delivery occurred below 1 cm 51% of the patients (23 cases) with LLP normalized
between the CIO and the PE. Only one tentative vagi- their placental position between 30 and 33 weeks and
nal delivery was successful but the patient lost 700 ml 34–38 WG. Oppenheimer et al. [15] reported that
of blood. migration of LLP occurred in all cases (n ¼ 29) at a
4 S. ALOUINI ET AL.

TABLE 3. Volume of blood loss according to CIO-PE distances and mode of delivery.
Cesarean section Caesarean section Vaginal delivery Vaginal delivery
CIO-PE d (cm) N Mean ± Sd blood loss (ml) N Mean ± Sd blood loss (ml)
[0] 123 1022.13 ± 1186.04 0 0
[0;1] 57 679.09 ± 471.28 1 700 ± 0
[1;2] 9 793.75 ± 482.88 8 522.5 ± 433.65
[2;3] 12 495 ± 409.91 14 381.82 ± 454.57
[3;n] 18 452.78 ± 509.46 77 283.8 ± 371.37
Total 219 852.11 ± 9 69.11 100 321.21 ± 389.36
Sd: standard deviation.

The repetition of measurements 3–4 weeks later (at

34–35 weeks) is essential either to confirm the first
measurement in the case of the placenta covering the
OI and up to 1 cm or to observe placental migration
away from the CIO when the initial measure is greater
than 1 cm. The latest US measure correlated with the
mode of delivery. The patients for whom the C-section
decision was chosen and performed had a lower CIO-
PE d than those who delivered vaginally. Similarly, the
patients who underwent emergency C-section for
bleeding had a lower CIO-PE d than those who deliv-
ered vaginally. Both PP and LLP were high-risk situa-
tions for bleeding and must be controlled upstream.
According to Ohira et al. [17], the weak migration
of LLP was associated with a high C-section rate
Figure 1. Volume of blood loss in vaginal delivery according
to CIO-PE distance. because of antepartum vaginal bleeding. In our study,
the mean volume of PPH was significantly higher for
patients with a CIO-PE d less than 2 cm. The threshold
TABLE 4. Blood loss and mode of delivery according to the value of 2 cm facilitated the identification of signifi-
CIO–PE d. cantly more hemorrhagic deliveries, below 2 cm, deliv-
Placental edge to ery was significantly more hemorrhagic than above
cervical internal os Mean volume of
distance (cm) Rate of C-section hemorrhage 2 cm. In this situation, delivery must be controlled and
<2 95.79% 898.57 ± 1005 the patient should be informed of the increased hem-
2 28.13% 355.65 ± 415 orrhagic risks.
p<.001 p < .001
A planned C-section at 36–37 weeks is imperative in
cases of a CIO-PE d less than 1 cm. According to
Dashe [18], a C-section was performed for bleeding in
CIO-PE d greater than 20 mm. The US exam was one-third of the cases in which the placenta was
repeated after 4 weeks (third trimester). within 2 cm of the os. Vergani et al. [13] reported a C-
The initial US measurement of the CIO-PE d should section rate of 75% for a distance of 1–10 mm (24
be repeated at 35–36 weeks before the delivery patients) and 31% for a distance of 11–20 mm (29
because of the possible migration of the placenta patients). Bhide et al. [9] reported a C-section rate of
from the CIO and to plan the delivery route. The pre- 90% for a CIO-PE d less than 2 cm (64 patients). A trial
sent study found no migration of the placenta cover- of vaginal delivery was considered appropriate when
ing the internal orifice of the cervix or that up to 1 cm the CIO-PE d was greater than 2 cm.
from the cervical os in the first ultrasound. Significant Patients should be managed in a maternity center
migrations were observed only for initial distances with a blood bank and informed preoperatively [19].
greater than 1 cm. The intervention should be performed during the day
Haino et al. [12] revealed that US measurement at by operators who are familiar with vascular ligations,
30–33 weeks should be repeated 3 weeks later, as the hemostatic sutures, and other hemostatic techniques
diagnosis of placental position changed for 49.5% of [20]. In case of placental migration above 2 cm in the
the patients. Durst et al. [16] reported the migration last US examination, the risk of hemorrhage remains
of the majority of CIO-PE d greater than 1 cm. greater than that of placental fundal insertion but
 THE JOURNAL OF MATERNAL-FETAL & NEONATAL MEDICINE 5

Figure 4. Volume of blood loss in cesarean section according

to CIO-PE distance.

unquantified, especially in cases of vaginal delivery

[21]. Only hemorrhage during delivery was quantified
using a graduated bag placed under the patient’s but-
Figure 2. Mode of delivery by CIO–PE distance. tocks. Our statistical analysis showed that the CIO-PE d
was the major factor in bleeding. In a study on 73
patients, Matsubara et al. [11] reported that total
blood loss was greater in women with a CIO-PE d less
than or equal to 4 cm and especially those with a CIO-
PE d less than 2 cm. According to Taga et al. [22], the
LLP group bled more in a planned C-section (seven
patients) than in a vaginal delivery (11 patients)
(Figure 4).

Limitations of the study

As this work was a retrospective study, some biases
were likely to occur. Nevertheless, the large sample of
patients with PP and LPI (319) means that the results
have significant implications for decision making in
high-risk obstetrical situations, such as post-par-
tum hemorrhage.
Figure 3. Hemorrhagic volume according to final measure of
PE-CIO d (n ¼ 319)
Conclusion
vaginal delivery may be proposed and discussed with
the patient. In our study, the vaginal US measure of the distance
In our study, hemorrhage volume was greater dur- between CIO and PE in the third trimester of preg-
ing a C-section than during a vaginal delivery. Even in nancy at around 32 weeks enabled the precise diagno-
cases of normally positioned placentas, a C-section is sis and prognosis of PP or LPI. When the CIO-PE d in
usually more hemorrhagic than a vaginal delivery (9). the first US examination was less than 1 cm, the pla-
A quarter of the patients had a previous C-section, centas did not migrate 3 weeks after. For distances
which was identified as a risk factor of PPH. greater than 1 cm, the majority of the placentas
Approximately, a quarter of the C-sections was per- migrated within three to four weeks and a repetition
formed for emergency antepartum vaginal bleeding. of the US examination was necessary in deciding the
Moreover, repeated antepartum hemorrhage remains mode of delivery.
6 S. ALOUINI ET AL.

The threshold CIO-PE d of 2 cm enabled the differ- [9] Bhide A, Prefumo F, Moore J, et al. Placental edge to
entiation of high-risk hemorrhagic deliveries from internal os distance in the late third trimester and
mode of delivery in placenta praevia. BJOG. 2003;110:
those at moderate risk. Below 1 cm, the delivery is at a
860–864.
significantly high risk of post-partum hemorrhage and [10] D’Antonio F, Bhide A. Ultrasound in placental disor-
a planned C-section under optimal conditions is indi- ders. Best Pract Res Clin Obstet Gynaecol. 2014;28:
cated. For a CIO-PE d greater than 2 cm, a trial of vagi- 429–442.
nal delivery could be proposed to patients. For [11] Matsubara S, Ohkuchi A, Kikkawa M, et al. Blood loss
in low-lying placenta: placental edge to cervical
distances between 1 and 2 cm, the volume of blood
internal os distance of less vs. more than 2 cm.
loss tends to be more important than for distances J Perinat Med. 2008;36:507–512.
superior to 3 cm; however, the tendency is not statis- [12] Haino K, Ishii K, Kanda M, et al. Variations of placental
tically significant. A vaginal delivery could be tried migration in patients with early third trimester malpo-
after information of patients concerning its risks sition. J Med Ultrason. 2018;45:99–102.
and benefits. [13] Vergani P, Ornaghi S, Pozzi I, et al. Placenta previa:
distance to internal os and mode of delivery. Am J
Obstet Gynecol. 2009; 201:266.e1–266.e5.
Disclosure statement [14] Sentilhes L, Vayssiere C, Deneux-Tharaux C, et al.
Postpartum hemorrhage: guidelines for clinical prac-
No potential conflict of interest was reported by the authors. tice from the French College of Gynaecologists and
Obstetricians (CNGOF): in collaboration with the
French Society of Anesthesiology and Intensive Care
References (SFAR). Eur J Obstet Gynecol Reprod Biol. 2016;198:
12–21.
[1] Kramer MS, Berg C, Abenhaim H, et al. Incidence, risk
[15] Oppenheimer L, Holmes P, Simpson N, et al.
factors, and temporal trends in severe postpartum
Diagnosis of low-lying placenta: can migration in the
hemorrhage. Am J Obstet Gynecol. 2013;209:
third trimester predict outcome? Ultrasound Obstet
449.e1–449.e7.
Gynecol. 2001;18:100–102.
[2] Kollmann M, Gaulhofer J, Lang U, et al. Placenta prae-
[16] Durst JK, Tuuli MG, Temming LA, et al. Resolution of
via: incidence, risk factors and outcome. J Matern
a low-lying placenta and placenta previa diagnosed
Fetal Neonatal Med. 2016;29:1395–1398. at the midtrimester anatomy scan. J Ultrasound Med.
[3] Wortman AC, Twickler DM, McIntire DD, et al. 2018;37:2011–2019.
Bleeding complications in pregnancies with low-lying [17] Ohira S, Kikuchi N, Kobara H, et al. Predicting the
placenta. J Matern Fetal Neonatal Med. 2016;29: route of delivery in women with low-lying placenta
1367–1371. using transvaginal ultrasonography: significance of
[4] Vahanian SA, Lavery JA, Ananth CV, et al. Placental placental migration and marginal sinus. Gynecol
implantation abnormalities and risk of preterm deliv- Obstet Invest. 2012;73:217–222.
ery: a systematic review and meta-analysis. Am J [18] Dashe JS. Toward consistent terminology of placental
Obstet Gynecol. 2015;213:S78–S90. location. Semin Perinatol. 2013;37:375–379.
[5] Fan D, Xia Q, Liu L, et al. The incidence of postpartum [19] Pande B, Shetty A. An audit to review the characteris-
hemorrhage in pregnant women with placenta previa: tics and management of placenta praevia at
a systematic review and meta-analysis. Plos One. Aberdeen Maternity Hospital, 2009–2011. J Obstet
2017;12:e0170194. Gynaecol. 2014;34:403–406.
[6] Heer IM, M€ uller-Egloff S, Strauss A. Placenta praevia – [20] Huchon C, Dumont A, Traor
e M, et al. A prediction
comparison of four sonographic modalities. score for maternal mortality in Senegal and Mali.
Ultraschall Med. 2005;27:355–359. Obstet Gynecol. 2013;121:1049–1056.
[7] Quant HS, Friedman AM, Wang E, et al. [21] Chama CM, Wanonyi IK, Usman JD. From low-lying
Transabdominal ultrasonography as a screening test implantation to placenta praevia: a longitudinal
for second-trimester placenta previa. Obstet Gynecol. ultrasonic assessment. J Obstet Gynaecol. 2004;24:
2014;123:628–633. 516–518.
[8] Vintzileos AM, Ananth CV, Smulian JC. Using ultra- [22] Taga A, Sato Y, Sakae C, et al. Planned vaginal deliv-
sound in the clinical management of placental ery versus planned cesarean delivery in cases of low-
implantation abnormalities. Am J Obstet Gynecol. lying placenta. J Matern Fetal Neonatal Med. 2017;30:
2015;213:S70–S77. 618–622.