[Abstract]
Background Preterm birth is the main cause of child
death under 5years of age. The incidence of twin pregnancies is less
than 2%, but the incidence of preterm delivery is 50% and the risk of
neonatal death is 5 times higher in twin pregnancies than in singleton
pregnancies. However, there is still no consensus on the effect of
cervical pessary on preventing preterm delivery, prolonging the
pregnancy cycle, and improving maternal and infant outcomes in patients
with twin pregnancies.
Objectives To explore the effect of cervical pessary on
the pregnancy outcome of unselected twin pregnancy patients.
Search Strategy Up to Jan 2022, researchers searched
PubMed, EMBASE, COCHRANE, Web of Science, Wanfang, Weipu, and CNKI
databases for research.
Study eligibility criteria Randomized controlled trials
that compared cervical pessary with standard care (no pessary) or
alternative interventions (conventional and standard treatment (e.g.,
Atoxiban therapy) or vaginal progesterone) in patients with twin
pregnancies.
Study appraisal and synthesis methods.
Data Collection and Analysis Two authors independently
extracted information related to the study characteristics and test
results from each of the included literature, and used Revman 5.3 to
analyze the data. Pooled relative risks with 95% confidence intervals
were calculated. Cohrane collaborative tools were used to assess the
risk of bias in individual studies.The main results were premature
delivery at <34 weeks, preterm delivery <37 weeks,
and abortion <28 weeks. Secondary results included spontaneous
preterm delivery <34 weeks, spontaneous preterm delivery
<34 weeks, spontaneous abortion <28 weeks, and
preterm prelabour rupture of membranes <34 and preterm
prelabour rupture of membranes, vaginal bleeding, chorioamnionitis,
delivery week, vaginal infection, vaginal discharge, cesarean section,
intrauterine death or stillbirth, neonatal death, low weight birth, very
low weight birth, neonatal respiratory distress syndrome, neonatal
intraventricular hemorrhage, necrotizing enterocolitis, retinopathy,
sepsis.
Results The researchers included a total of 7 documents
with a total of 3120 patients. Among them, 4 studies included pregnancy
outcomes and neonatal outcomes for patients with cervix length
<25mm, and 7 studies included pregnancy outcomes and neonatal
outcomes for patients with cervix length <38mm. The results
showed that cervical pessary increase the incidence of delivery week,
vaginal discharge, and vaginal bleeding, which was statistically
significant. For neonates, cervical pessary decreases the incidence of
low-weight children, necrotizing enterocolitis and neonatal septicemia,
which were statistically significant. Subgroup analysis results based on
cervical length <38mm showed that cervical pessary could
reduce the preterm birth rate before 34 weeks, the spontaneous preterm
birth rate before 34 weeks, prolong the gestational week of delivery,
reduce neonatal mortality, occurrence of neonatal necrotizing
enterocolitis and neonatal sepsis. However, the incidence of events such
as increased vaginal discharge and vaginal bleeding in the experimental
group was significantly higher than that in the control group, and the
results were statistically significant. The results of subgroup analysis
based on the cervical length < 25mm showed that cervical
pessary was better than the control group in reducing the preterm birth
rate before 34 weeks, the spontaneous preterm birth rate before
<34 weeks, and the incidence of low-birth-weight infants, and
the results were statistically significant. Otherwise, the subgroup
analysis based on merely Arabin cervical pessary prove merely Arabin
cervical pessary has similar results.
Conclusion The cervical pessary can extend the
gestational week of short-cervix twin pregnancy without clinical
symptoms, reduce the premature birth rate before 34 weeks of gestation,
improve pregnancy outcome, reduce neonatal mortality, reduce necrotizing
enterocolitis incidence, neonatal sepsis incidence, and improve neonatal
outcome. For patients with a cervical length less than 38mm, a cervical
pessary can be performed to extend the gestational week. For patients
with a cervical length less than 25mm, a cervical pessary can
effectively prolong the gestational age and improve the maternal and
fetal outcomes. The cervical pessary is safe for patients with twin
pregnancies. Suggest that twin-preganct-patients with CL<38mm
should consider take cervical pessary in advance. In terms of long-term
efficacy, there is no evidence of cervical support placement on the
long-term maternal prognosis. In terms of economic benefits, cervical
support is better than vaginal progesterone, but this conclusion still
needs more research to prove.
Keywords: cervical pessary; pregnancy outcome; fetal
outcome; preterm birth;preterm delivery;
Introduction
Studies have shown that the mortality rate for children under 5 years of
age in China is 37%, with the main cause of death being complications
from premature birth, accounting for about 17% of all
deaths[1].Surviving preterm infants are at greater risk for
short-term complications, with higher rates of respiratory distress
syndrome, bronchopulmonary dysplasia, necrotizing enterocolitis, sepsis,
intraventricular hemorrhage, paraventricular leukodystrophy, and
retinopathy than in term-born neonates[2,3].The incidence of twin
pregnancies is less than 2%, but the incidence of preterm delivery is
50% and the risk of neonatal death is 5 times higher in twin
pregnancies than in singleton pregnancies[4,5].Cervical
insufficiency as a cause of spontaneous preterm delivery in patients
with twin pregnancies[6].
Currently, the main therapeutic measures regarding the prevention of
preterm delivery in patients with twin pregnancies are the vaginal
progesterone, cervical cerclage and cervical pessary. In the Clinical
Guidelines for the Management of Twin Pregnancies published in China in
2020, it is clearly stated that in asymptomatic patients with twin
pregnancies with a short cervix, the use of progestins can effectively
reduce the risk of preterm delivery before 35 weeks of
gestation[7].As for cervical cerclage, for singleton patients, this
treatment is currently considered effective in preventing preterm
birth[8],but for patients with twin pregnancies, the efficacy of the
treatment remains controversial[7,9].
Originally used to treat pelvic organ prolapse, cervical pessary
placement has been used to prevent preterm birth since 1990. Currently,
the Arabin uterine support is widely used in the treatment of
spontaneous preterm labor. It is designed with the intention not only to
support and compress, but also to tilt the cervix and possibly rotate it
towards the sacrum, mainly by supporting the inner cervical opening and
preventing it from being overburdened with gravity[10].Currently
studies show that cervical pessary placement is effective in reducing
the rate of preterm birth in patients with singleton
pregnancies[11,12],while cervical pessary placement remains
controversial in patients with twin pregnancies. A retrospective
analysis in 2016 showed that cervical pessary placement reduced the rate
of preterm delivery before 37 and 34 weeks of gestation in patients with
twin pregnancies[13].In 2019, an RCT trial suggested that cervical
pessary placement reduces preterm birth rates and improves pregnancy
outcomes in patients with twin pregnancies[14],but There were RCT
experiments and meta-analyses that came to the opposite
conclusion[15-17].
No specific and effective treatment exists to prevent preterm birth in
the 2014 ACOG guidelines for twin women[18]. In 2019, the Canadian
Association of Obstetricians and Gynecologists (SOGC) still believed
that, even in twin patients with a short cervix, there was still no
evidence that cervical pessary could effectively prevent premature
birth[9]. China updated its guidelines in 2020 and did not give
clear recommendations on the use of cervical pessary to prevent
premature birth in twin patients. However, Chinese guidelines clearly
support that vaginal progesterone can reduce the preterm birth and
neonatal prevalence before 35 weeks in pregnant women with asymptomatic
ultrasound showing a short cervix[7]. As for cervical cerclage, as
an invasive treatment, its efficacy varies due to its different timing
and indications. In 2014, ACOG noted that existing data demonstrated
that ring ligation actually significantly increased preterm birth rates
in asymptomatic twin pregnancies with CL <25mm[18]. The
SOGC guidelines also indicate that cerclage increases asymptomatic
preterm rates in twin pregnancies with CL <25mm, arguing that
without physical evidence of physical examination, taking cerclage
merely indicated by ultrasound hints of cervical shortening or previous
second trimester abortion history increases the risk of preterm
pregnancy.
In conclusion, there is still no consensus on the effect of cervical
pessary on preventing preterm delivery, prolonging the pregnancy cycle,
and improving maternal and infant outcomes in patients with twin
pregnancies. In this article, we present a meta-analysis of the effects
of cervical pessary placement on pregnancy outcomes in patients with
twin pregnancies and discuss the effectiveness of cervical pessary
placement in improving adverse pregnancy outcomes and neonatal outcomes.
Materials and Methods
1.1 Literature Search
This study was conducted and reported in accordance with the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)
statement[19],and was registered with PROSPERO, number
CRD42021275530.Two authors independently extracted all study data into a
canonical form. When there is a difference of opinion, the two authors
reach a consensus through negotiation.
Up to Jan 2022, researchers searched PubMed, EMBASE, COCHRANE, Web of
Science, Wan fang, Wi Pu and CNKI databases, for example, in the PubMed
database, they searched for ((((((((cervical) OR pessary) OR cervical
pessary)) OR pessary[MeSH Terms])) AND ((((((prematurity) OR
prematurity[MeSH Terms]) OR premature birth) OR premature
birth[MeSH Terms]) OR Preterm delivery) OR Preterm delivery[MeSH
Terms])) AND (((((Twin) OR pregnancy) OR Twin pregnancy[MeSH
Terms]) OR Multiple pregnancy) OR Multiple pregnancy[MeSH Terms]))
AND ((cervical length) OR short cervix). Investigators selected studies
for inclusion that met the inclusion standards. Relevant publications
were searched. Researchers also systematically reviewed the references
of the literature included in the study. Clinical pregnancy outcomes and
neonatal outcomes were collected in the cervical pessary group
(experimental group) and the non-cervical pessary group (control group).
Cervical cerclage was excluded at an early stage of the study for the
following reasons: 1. Meta-analysis has been conducted in relevant
studies[20] ; Cervical cerclage is more irritating to patients than
cervical pessary, cervical pessary as a non-invasive means, combined
with non-invasive conventional fetal preservation methods such as bed
rest, atosiban, vaginal progesterone, etc., compared with conventional
fetal care methods, lower heterogeneity, more reliable results.
Two reviewers reviewed each potential eligibility article separately,
analyzed the quality of the studies according to the Corhrane scale, and
extracted data. Two authors (W-Y and M-D)independently performed the
original screening of all study titles and abstracts, excluding
literature that was deemed irrelevant by both observers. The PRISMA
flowchart provides more detailed information about the article selection
process (picture 1: flowchart). The researchers recorded in detail the
year of publication of the records, country, study type, number of
participants, week of gestation and range of cervical length (CL) at the
time of placement of the cervical pessary, and mode of treatment for
both groups of patients (table 1: characteristics of the included
studies).
1.2 Eligibility criteria
The purpose of the included study was to investigate the effect of
cervical pessary placement on pregnancy outcomes in patients with twin
pregnancies. The inclusion criteria were1. Patients with twin
pregnancies who underwent gynecological trans-vaginal ultrasound after
16 weeks of gestation and underwent cervical pessary because of the
short cervix. There were no special requirements for the mode of
pregnancy (ART and non-ART), and medical history of patients with twin
pregnancy), 2.Must be done for RCT studies, 3.The experimental group
must be patients undergoing cervical pessary, which can be performed in
combination with conventional and standard treatment (e.g.Atoxiban
therapy) or vaginal progesterone, 4.The control group must be routine
noninvasive treatment such as conventional therapy (such as atoxeban) or
vaginal progesterone. The exclusion criteria were: 1. The patient
underwent cervical cerclage at this pregnancy, 2. patients who needed to
receive fetoscopy, 3. Exclusion abnormal fetal development, pregnancies
of three and more fetuses, medically indicated preterm birth of medical
origin (Twin-twin transfusion syndrome, severe preeclampsia, placenta
abruption, placenta previa, prenatal bleeding); 4. non-RCT experiments.
1.3 Outcome measures and data item extraction
Two researchers extracted information related to the study
characteristics and test results from each of the included literature.
The main outcome was preterm delivery (ptd) before 34 weeks gestation,
preterm delivery before 37 weeks, miscarriage a before 28 weeks(China
officially defines miscarriage as occurring before 28 weeks’ gestation).
Secondary outcomes included spontaneous preterm delivery(sptd) before 34
weeks, spontaneous preterm delivery before 37 weeks, spontaneous
miscarriage before 28 weeks(SM), preterm prelabor rupture of
membranes(PPROM),preterm prelabor rupture of membranes before 34 week,
deliver week(DW),vaginal bleeding(VB),chorioamnionitis(C),vaginal
infection(VI),vaginal discharge(VD),C-section, intrauterine death or
stillbirth, neonatal death , low-birth weight (<2500g), very
low birth weight (<1500g), Respiratory distress syndrome
(RDS),Intraventricular hemorrhage(IVH),Necrotizing Enterocolitis
(NE),Retinopathy (RE),Sepsis (S).
1.4 Risk assessment of bias
Cochrane collaborative tools were used to assess the risk of bias in
individual studies,including: 1.selection bias; 2. implement bias; 3.
detection bias; 4. attribution bias; 5. reporting bias; 6. Other bias
(including measuring bias; sampling bias and follow-up bias) (See
details in the table2:bias of included literature assessment table).The
quality of evidence for primary and
secondary outcomes was assessed by using the GRADE approach, which takes
into account 5 domains: risk of bias, inconsistency, indirectness,
imprecision, and publication bias.The GRADE
approach categorizes the certainty of the evidence into 4 levels: (1)
high: we are very confident that the true effect lies close to that of
the estimate of the effect,and further research is unlikely to change
our confidence in the estimate other effect; (2) moderate: we are
moderately confident in the effect estimate, and the true effect is
likely to be close to the estimate of the effect, but there is
apossibility that it is substantially different; (3) low: our confidence
in the effect estimate is limited, and the true effect may be
substantially different from the estimate of the effect; and (4) very
low: we have very little confidence in the effect estimate, and the true
effect is likely to be substantially different from the estimate of
effect.
1.5 Data processing and analysis
Researchers used Revman 5.3 to analyze the data. It belonged to two
categorical outcome variables and relative rate (RR) was used as the
effect indicator. For the outcome index belonging to the continuous
variables, the standards mean difference (SMD) was used as the effect
indicator.95% confidence interval was calculated to evaluate the
strength of the association between cervical pessary and the risk of
adverse pregnancy-related outcomes. The RR values were calculated by the
Z test. The P-value of <0.05 was defined as meaningful.
Random-effects and fixed-effects models were applied in this
meta-analysis. To assess inter-study heterogeneity, the Q test was
applied to calculate I2. The I2values were defined as 25%, 50%, and 75%, representing low, moderate,
and high heterogeneity, respectively. When high heterogeneity was
observed, random-effects models were used to pool results, and a
fixed-effect model. When the heterogeneity is too high, there should be
further excluded clinical and methodological heterogeneity, and a random
effect model is used for analysis. If there was evidence of statistical
heterogeneity (I2>50%), it is necessary
to explore the possible sources by using sensitivity and subgroup
analyses to search for evidence of bias or methodological differences
among trials. Researchers used the exclusion method article by piece
exclusion literature method for sensitivity analysis. Differences in the
elimination results and the original merger results were also assessed.
Publication bias was visually judged by drawing funnel plots.
Results
2.1 Results of literature search
A total of 7 literature studies have included 3120 patients, and 4
studies containing <25mm for CL.For patient pregnancy and
neonatal outcomes, seven studies included pregnancy and neonatal
outcomes for patients with CL <38mm.
2.2 Primary and secondary outcome summary
2.2.1.A meta analysis was performed for all of the literature
First, we performed meta analyses on
all included literature to compare the efficacy of cervical support in
improving pregnancy outcomes with neonatal outcomes (see table3: Summary
Results 1 ),It was found that the patients in the experimental group
(cervical pessary group) had a longer deliver week (RR 6.82,
P<0.00001), higher incidence of vaginal discharge (RR12.96 ,
P<0.00001) and vaginal bleeding (RR 5.34, P<0.0001)
than in the control group, which was statistically significant. As
neonates, in experimental groups, the incidence of low birth weight (RR
4.14, P<0.00001), NE (RR2.45, P=0.01) and neonatal septicemia
(RR2.21, P=0.03) were significantly lower than control group, which were
statistically significant. Compared DW, vaginal discharge, vaginal
bleeding, low birth weight, S, remarkable heterogeneity could be seen.
The sensitivity analysis was performed article by article, and the
exclusion literature name and the p values after analysis are shown in
the figure below. The funnel map suggests publication bias in part of
the study.
2.2.2 Subgroup analysis was performed based on the CL
During the process of literature inclusion, the researchers found
differences in the CL of the included patients, and to avoid the outcome
bias caused by different CL, the subgroup analysis was
performed.
Results of the subgroup analysis based on CL <38mm rows are
shown in table
4.
It can be seen that cervical pessary was better than the control group
of preventing spontaneous premature delivery before 34 weeks(RR2.18,
P=0.03) and premature delivery before 34 weeks (RR2.9, P=0.004),
prolonged delivery week (RR6.31, P<0.00001), reducing neonatal
mortality(RR2.47, P0.01), neonatal necrotic enterocolitis (RR3.52,
P=0.0004) and neonatal sepsis(RR2.21, P=0.03).All the results were
statistically significant. However, the incidence of increased vaginal
discharge and vaginal bleeding in the cervical pessary group was
significantly higher than in the control group. Some studies funnel
plots suggest publication bias.
Results of the subgroup analysis based on the <25mm line of CL
are shown in table 5.Due to insufficient data, comparisons were made
only between premature delivery before 34 weeks (RR3.82, P=0.0001) ,
spontaneous premature delivery before 34 weeks (RR2.7, P=0.04), low
weight birth (RR6.32 ,P<0.00001) and vary low weight birth
(RR0.16, P0.87). Results showed that cervical pessary was better in
reducing spontaneous preterm birth <34 weeks and preterm birth
<34, decreased the born of lower weight infants, and the
results were statistically significant. Some study funnel plots
suggested publication bias.
2.2.3 Subgroup analysis was performed based on the type of cervical
pessary.
Because of Bioteque cervical pessary in Berghella2017, which could cause
performance bias, the result of subgroup analysis after excluding
Berghella2017 show in table 6 that the patients in the experimental
group (merely Arabin cervical pessary group) had a longer deliver week
(RR6.82 P<0.00001), higher incidence of vaginal discharge
(RR12.23, P<0.00001) , C-section (RR2.05, P=0.04), vaginal
bleeding(RR 5.43, P<0.04),which was statistically significant.
For neonates, In experimental groups, the incidences of low birth
weight(RR 4.14, P<0.0001), NE(RR2.59, P=0.01) and S (RR2.61,
P=0.009) were lower ,which was statistically significant. Remarkable
heterogeneity could be seen. The sensitivity analysis was performed
article by article, and the exclusion literature name and the p values
after analysis are shown in the figure below. After sensitivity
analysis, the ptd <34 weeks, DW, VD, PPROM, DE, low birth
weight were statistically significant, and the P value of sptd
<34 weeks and S changed, but were not statistically
significant. The funnel map suggests publication bias in part of the
study.
Results of the subgroup analysis based on CL <38mm rows are
shown in table 7. It can be seen that Arabin cervical pessary was better
than the control group of preventing ptd<34 (RR3.05, P=0.002)
and sptd<34 (RR2.24, P=0.03),prolonged delivery week(RR2.24,
P<0.03), increased VD (RR9, P<0.00001)and VB(RR5.34,
P<0.00001), decreased incidence of PPROM (RR2.31,
P<0.02),reducing neonatal mortality(RR6.02,
P<0.006),the incidences of low birth weight(RR 4.14,
P<0.00001), NE (RR3.68, P=0.0002), S(RR 2.61,
P<0.009),which was statistically significant. Remarkable
heterogeneity could be seen. The sensitivity analysis was performed
article by article.Funnel plots show the publication bias in part of the
studies.
Results of the subgroup analysis based on the CL<25mm are
shown in table 8. Due to insufficient data, comparisons were made only
between premature delivery before 34 weeks(RR3.82 ,P=0.001) ,
spontaneous premature delivery before 34 weeks(RR2.07, P=0.04), low
weight birth (RR6.32, P<0.00001) and vary low weight birth (RR
0.16, P=0.87). Results showed that merely Arabin cervical pessary was
better in reducing spontaneous preterm birth <34 weeks and
preterm birth <34, decreased the born of lower weight infants,
and the results were statistically significant. Some study funnel plots
suggested publication bias. Remarkable heterogeneity could be seen. The
sensitivity analysis was performed article by article, and the exclusion
literature name and the p values after analysis are shown in the figure
below. After sensitivity analysis, sptd<34 keep his
statistically significant, but low birth weight was not.
2.3 Quality evaluation and bias evaluation
These literatures were assessed for the risk of bias (as table 2 and
picture2). Considering the particularity of the study, that the patients
in the experiment knew all about their intervention after participating
in the experiment, the researchers default that patient’s blindness are
low-risk, and only evaluate the blind method of the researchers and the
results assessors. The high uncertain risk of Merced 2019,Norman 2021and
Berghella 2017 are exposed as missing visits and bias burned when
selected patients.Dang’s high uncertain risk is from missing visits,and
it has uncertain selection bias and reporting bias. Berghella 2017 has
high preform bias.The results of GRADE approach categorizing the
certainty of the evidence are in table 3-8.During analysis,we found that
the quality of evidences could be influenced by Berghella 2017 because
of his bias from patients selection.besides,the bias from Merced
2019,Norman 2021 and Dang 2019 could influence the certificate evidence
quality.All in all,the quality of evidences in table 3-5 is low.But
after subgroup analysis, it is obvious that the quality of evidences has
improved.