DOI: 10.1111/1471-0528.

15566 Systematic review

www.bjog.org

Vaginal progesterone, oral progesterone,

17-OHPC, cerclage, and pessary for preventing
preterm birth in at-risk singleton pregnancies: an
updated systematic review and network
meta-analysis
A Jarde,a O Lutsiv,b J Beyene,c SD McDonalda
a
Department of Obstetrics and Gynecology, McMaster University, Hamilton, ON, Canada b Provincial Council for Maternal and Child
Health, Toronto, ON, Canada c Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada
Correspondence: A Jarde, Department of Obstetrics and Gynecology, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4K1,
Canada. Email: jarde@mcmaster.ca

Accepted 12 November 2018. Published Online 29 December 2018.

Background Recent progesterone trials call for an update of
previous syntheses of interventions to prevent preterm birth.
Objectives To compare the relative effects of different types and
routes of administration of progesterone, cerclage, and pessary at
preventing preterm birth in at-risk women overall and in specific
populations.
Search strategy We searched Medline, EMBASE, CINAHL,
Cochrane CENTRAL, and Web of Science up to 1 January 2018.
Selection criteria We included randomised trials of progesterone,
cerclage or pessary for preventing preterm birth in at-risk
singleton pregnancies.
Data collection and analysis We used a piloted data extraction
form and performed Bayesian random-effects network meta-
analyses with 95% credibility intervals (CrI), as well as pairwise
meta-analyses, rating the quality of the evidence using GRADE.
Main results We included 40 trials (11 311 women). In at-risk
women overall, vaginal progesterone reduced preterm birth <34
(OR 0.43, 95% CrI 0.20–0.81) and <37 weeks (OR 0.51, 95% CrI
0.34–0.74), and neonatal death (OR 0.41, 95% CrI 0.20–0.83). In
women with a previous preterm birth, vaginal progesterone
reduced preterm birth <34 (OR 0.29, 95% CI 0.12–0.68) and
<37 weeks (OR 0.43, 95% CrI 0.23–0.74), and 17a-
hydroxyprogesterone caproate reduced preterm birth <37 weeks
(OR 0.53, 95% CrI 0.27–0.95) and neonatal death (OR 0.39, 95%
CI 0.16–0.95). In women with a short cervix (≤25 mm), vaginal
progesterone reduced preterm birth <34 weeks (OR 0.45, 95% CI
0.24–0.84).
Conclusions Vaginal progesterone was the only intervention with
consistent effectiveness for preventing preterm birth in singleton
at-risk pregnancies overall and in those with a previous preterm
birth.
Keywords Cervical cerclage, cervical pessary, network meta-
analysis, preterm birth, progesterone, systematic review.
Tweetable abstract In updated NMA, vaginal progesterone
consistently reduced PTB in overall at-risk pregnancies and in
women with previous PTB.

Please cite this paper as: Jarde A, Lutsiv O, Beyene J, McDonald SD. Vaginal progesterone, oral progesterone, 17-OHPC, cerclage, and pessary for
preventing preterm birth in at-risk singleton pregnancies: an updated systematic review and network meta-analysis. BJOG 2018; https://doi.org/10.1111/
1471-0528.15566

risk of preterm birth in women at increased risk, including

Introduction
Globally, about 15 million pregnancies each year end in
preterm birth,1 i.e. before the 37th week of gestation, which
is a major cause of morbidity and mortality in children.1,2
Various interventions have been attempted to reduce the
progesterone, cervical cerclage, and cervical pessary. Proges-
terone can be natural, administered vaginally [per vagina
(PV)] or orally [per os (PO)], versus 17a-hydroxyprogester-
one caproate (17-OHPC), administered intramuscularly
(IM).

ª 2018 Royal College of Obstetricians and Gynaecologists 1

 Jarde et al.
A previous network meta-analysis combining direct and
indirect evidence for these interventions concluded that
progesterone, particularly natural progesterone, was the
best intervention for the primary prevention of preterm
birth and neonatal death in women at risk overall and in
women with a previous preterm birth.3 However, since the
last literature search, a number of studies have been pub-
lished, including the PROGRESS Study, one of the biggest
studies to date.4
Therefore, our objectives were to provide an up-to-date
network meta-analysis comparing the effectiveness of pro-
gesterone, cerclage, and pessary for preventing preterm
birth in women at risk overall, and for women with a pre-
vious preterm birth and women with a short cervix, sepa-
rating progesterone by route of administration.
We followed a similar protocol for this network meta-
analysis as in our prior publication (CRD42015016166),3
with the modifications specified below established before
beginning. No patients were involved in the development
of this study.
Information sources and search strategy
We searched Medline, EMBASE, CINAHL, Cochrane Cen-
tral Register of Controlled Trials, and ISI Web of Science
up to 1 January 2018 (Supporting Information
Appendix S1) and screened the reference lists of systematic
reviews and included studies.
Eligibility criteria
We included randomised controlled trials comparing pro-
gesterone, cerclage or pessary with a control group or
another intervention for the prevention of preterm birth
and/or associated adverse outcomes in at-risk singleton
pregnancies. Women were considered at risk based on their
history of preterm birth, cervical length or other factors
defined by the authors. We included studies with any type
of progesterone (natural or 17-OHPC) and route (PV, PO
or IM), and any type of cerclage (McDonald or Shirodkar).
The comparison group could have received placebo, bed
rest, treatment as usual or a different type/route of the
interventions listed above. We had no language exclusions.
We excluded studies in which a subgroup of women was
already receiving an intervention before randomisation to
our interventions of interest, unless this was based on
objective risk factors (e.g. cervical length ≤20 mm) and
unless results were reported separately for the groups of
women with and without a pre-randomisation intervention.
We included studies in which women received an addi-
tional intervention after randomisation, following the
intention-to-treat principle, but excluded them in sensitiv-
ity analyses. We also excluded studies assessing interven-
tions in the context of artificial reproductive therapy,
assessing combinations of interventions or interventions for
2 ª 2018 Royal College of Obstetricians and Gynaecologists
the secondary prevention of preterm birth (e.g. tocolytics
in women with contractions), as well as non-peer-reviewed
literature, studies published only as abstracts, and other
study designs.
Given that the PROGRESS Study4 included twin preg-
nancies (~2%, eight receiving progesterone and four receiv-
ing placebo) without stratifying results by twins, and as we
did not receive a response from the authors to a request
for singletons-only data, we subtracted, for the main analy-
ses, the number of twin infants from the total sample size
in each group, while maintaining the reported number of
cases (as if none of the twins had the outcome). In sensi-
tivity analyses, we considered the opposite scenario, sub-
tracting the number of twin infants not only from the total
sample size, but also from the number of cases (as if all
twins had the outcome). Other studies included only sin-
gleton pregnancies, reported stratified data by pregnancy
type, or the authors provided data.
Our primary outcomes were preterm birth <34 weeks
(PTB <34 weeks) and <37 weeks (PTB <37 weeks), overall
or specified as spontaneous. Our main infant secondary
outcome was neonatal death (NND) and other outcomes
included important infant and maternal adverse outcomes
(see Supporting Information Tables S1 and S2). Although
we had initially included medically indicated preterm birth
in the protocol, after further consideration it was removed
as it was not considered relevant here.
Data extraction and assessment of risk of bias
Two reviewers (AJ and OL) independently screened titles
and abstracts, assessed the full text of potentially relevant
papers and extracted the data from included studies using
a piloted form. A third reviewer (SDM) resolved disagree-
ments that could not be solved by discussion. We con-
tacted authors of the original studies to confirm inclusion
criteria if necessary.
We used the Cochrane Risk of Bias tool to assess each
study’s risk of bias,5 considering studies to have a low risk
of bias if none of the domains had a high risk and at least
four domains (not counting ‘Other’) were low risk (with at
least one being ‘Random sequence generation’ or ‘Alloca-
tion concealment’).
Data synthesis and statistical analyses
In the first set of analyses, we executed all the analyses and
subgroup analyses from the previously published protocol,
starting with overall network meta-analyses (any risk factor,
any intervention type), and then focusing on women with
a previous preterm birth and women with a cervical length
≤25 mm, and separating progesterone and cerclage by type,
independent of the risk factors.
The second set of analyses focused on the comparison
with the results of the PROGRESS Study,4 in which vaginal

progesterone was assessed in women with a previous pre-
term birth. Therefore, these additional, planned subgroup
analyses involved splitting progesterone by route and sepa-
rating studies by their participants’ risk factors.
When there were at least 10 included studies for a given
comparison, we performed Bayesian random-effects net-
work meta-analyses.6 We used the Markov chain Monte
Carlo method, specifying vague prior distributions for
intervention effects and assuming common heterogeneity
parameters across each network to obtain pooled estimates.
We used 100 000 iterations with a burn-in of 4000 itera-
tions, using four chains and a thinning interval of 10. We
assessed convergence using Gelman and Rubin’s conver-
gence diagnostic criteria.7 When there were fewer than 10
included studies per comparison, we performed standard
pairwise inverse variance random effects meta-analyses
(DerSimonian and Laird).8
We reported the posterior median odds ratios (OR)
and median mean differences for binary and continuous
outcomes, respectively. For each measure we reported its
95% credibility interval (CrI, in network meta-analyses)
or confidence interval (CI, in pairwise meta-analyses), the
number needed to treat (NNT, calculated in statistically
significant results using the estimated effect size and the
pooled prevalence in the control group),5 and its quality
rating [assessed using the Grading of Recommendations
Assessment, Development and Evaluation (GRADE)
Working Group approach].9 The CrI is the equivalent of
the CI in the Bayesian framework, the NNT indicates
how many additional women at risk would need to be
treated to prevent one outcome, and the quality rating
reflects the certainty that the results are close to the true
effect.
Three basic assumptions underlie a network meta-analy-
sis: first, that studies comparing interventions directly (di-
rect estimates) are similar (absence of heterogeneity);
second, that study characteristics and baseline characteris-
tics are similar in the studies involved in the direct com-
parisons used to calculate indirect estimates (transitivity);
and third, that the direct and indirect estimates of the same
comparison are similar (coherence). If these assumptions
are not met, the results of the network meta-analysis are
challenged.
We used the I-squared statistic (I2) to quantify the
amount of heterogeneity in the direct comparisons.10 To
assess intransitivity in the indirect estimates we compared
the inclusion criteria (risk factors) and baseline characteris-
tics of the group of studies in each of the direct compar-
isons underlying the indirect comparison. We assessed
incoherence between the direct and indirect effect estimates
using the node-splitting method.11 We then used the
GRADE approach for network meta-analysis to rate the
quality of the evidence, downgrading its rating if there was
ª 2018 Royal College of Obstetricians and Gynaecologists
Progesterone, cerclage, and pessary to prevent preterm birth
heterogeneity, intransitivity or incoherence, among other
aspects such as high risk of bias in the included studies,
small pooled sample size or publication bias in the direct
comparisons, which we assessed using Duval and Tweedie’s
trim-and-fill method when there were at least 10 studies in
a comparison.9,12,13
Whenever we performed a network meta-analysis we
ranked the interventions according to their Surface Under
the Cumulative RAnking curve values (SUCRA).14 An
intervention consistently being among the most effective
(in the multiple iterations of the Markov chain Monte
Carlo) would have a higher SUCRA value.
We used R for all network meta-analyses (using the
‘gemtc’ package).15 For all pairwise meta-analyses, we used
REVIEW MANAGER (v.5.3).
Subgroup and sensitivity analyses
In addition to the subgroup analyses based on the partici-
pants’ common risk factors (previous preterm birth or cer-
vical length ≤25 mm) or intervention type (natural
progesterone versus 17-OHPC, or McDonald versus Shi-
rodkar cerclage), we separated progesterone by route (PO,
PV, and IM), overall and in subgroups defined by the com-
mon risk factor of the participating women. We undertook
these additional subgroup analyses for PTB <34 and
PTB <37 weeks, and for NND, our key secondary outcome.
To explore the possibility that the dose of PV progesterone
(≤200 versus >200 mg/day) could explain the differences
(heterogeneity) in PTB <37 weeks between studies in
women with a previous preterm birth, we performed a
post-hoc subgroup analysis.
We performed sensitivity analyses including only studies
with low risk of bias and excluding studies with additional
post-randomisation interventions. We repeated all the anal-
yses and comparisons assuming that all the twins in the
PROGRESS Study4 had the outcome.
Results
Our search strategy identified 15 009 citations, of which 40
met our inclusion criteria, comprising 11 311 women (Fig-
ure 1).
Study characteristics
Eighteen studies compared progesterone with a control
group (ten studies used PV progesterone, including the
PROGRESS Study,4,16–24 six used IM 17-OHPC,25–30 and
two used PO progesterone31,32), 11 studies compared
cerclage with a control group (eight used McDonald cer-
clage,33–40 two used Shirodkar cerclage,41,42 and one used
both43) and four studies compared an inert pessary (Ara-
bin) with a control group.44–47 One study compared cer-
clage (McDonald) with progesterone (IM 17-OHPC)48 and
3
 Jarde et al.

Studies identified through database searching (14,886) and additional sources (reference screening: 122, contact
with experts: 1): n = 15 009

Duplicate publications: n = 5101

Screening of titles and abstracts: n = 9908

Citations excluded based on titles and abstracts: n = 9530

Full-text screening: n = 378

Full-text articles excluded (n = 338)

• Publication not full report (e.g. conference abstracts,
editorials, etc.) (112)
• Study design not randomized controlled trial (86)
• Population not women at risk of preterm birth and eligible
for primary prevention (44)
• Twin population (27)
• Intervention group not meeting inclusion criteria (11)
• Comparison group not meeting inclusion criteria (10)
• No relevant outcome data (9)
• Overlap of samples (11)
• Duplicate studies (21)
• Ongoing trials (3)
• Not possible to locate (2)
• Artificial reproduction therapy studies (2)

Publications included in systematic synthesis: n = 40 (11 311 pregnancies)

• Progesterone vs control (18):
o Vaginal progesterone vs control: 10
o 17-OHPC vs control: 6
o Oral progesterone vs control: 2
• Cerclage vs control (11)
o McDonald vs control: 8
o Shirodkar vs control: 2;
o Shirodkar & McDonald vs control: 1
• Pessary vs control: 4
• Cerclage vs progesterone : 1
• Vaginal progesterone vs 17-OHPC: 6

Figure 1. Flow diagram of the systematic review.

six studies compared PV progesterone with IM 17-
OHPC.49–54
Only four studies included women with a previous pre-
term birth excluding those with a short cervix (defined as
<15, <20, <25, and <28 mm), and three studies included
women with a short cervix excluding those with a history
of preterm birth (including one study with women with a
cervical length of <30 mm). Twelve studies included
women with a previous preterm birth, regardless of cervical
length, and ten studies included women with a short cervix
(usually defined as ≤25 mm, although one study included
women with a cervical length of <30 mm), regardless of
their preterm birth history. The remaining 11 studies
included women with a variety of risk factors (e.g. preterm
birth in a sister30 or using a scoring system of several risk
factors,38 Tables 1 and 2).
Overall analyses (women with any risk factor)
Progesterone (any of natural progesterone PO or PV or IM
17-OHPC) significantly reduced both PTB <34 (OR 0.45,
95% CrI 0.23–0.81; NNT 9; Moderate quality of evidence)
and <37 weeks (OR 0.52, 95% CrI 0.36–0.73; NNT 8; Low
quality of evidence), compared with control, ranking first
and second, respectively, according to the SUCRA values.
Pessary reduced PTB <37 weeks. Among studies labelling
preterm birth specifically as ‘spontaneous’, pairwise meta-
analyses resulted in progesterone significantly reducing PTB
<34 weeks and pessary significantly reducing PTB
<37 weeks, relative to control (Table S1).
Among our secondary outcomes, progesterone (any type
and route) significantly reduced the odds of NND com-
pared with control, as well as showing significant benefits
regarding admission to the neonatal intensive or special

4 ª 2018 Royal College of Obstetricians and Gynaecologists

 Progesterone, cerclage, and pessary to prevent preterm birth

Table 1. Main characteristics of studies included in network meta-analysis (assessing progesterone)

First author, year Indication for preterm Intervention group: Comparison group: Gest. RoB
(Country) birth prevention sample size (details) sample size (PTB rate) age

Progesterone versus control

Fonseca, 200720 Cervix ≤15 mm Vaginal progesterone: 114 Placebo: 112 20–25 +
(Multinational*) (history of PTB not considered) (200 mg/day suppositories) (PTB <34 weeks: 32%)
O’Brien, 200717 History of PTB Vaginal progesterone: 309 Placebo: 302 (41%) 18–23 +
(Multinational*) (cervical length not considered) (90 mg/day gel)
Majhi, 200923 History of PTB Vaginal progesterone: 50 NI/TAU: 50 (38%) 20–24 +
(India) (cervical length not considered) (100 mg/day suppositories)
Akbari, 200916 History of PTB or cervical cerclage or uterus Vaginal progesterone: 69 NI/TAU: 72 (32%) NR
(Iran**) anomaly (cervical length not considered) (100 mg/day suppositories)
Cetingoz, 201124 History of PTB Vaginal progesterone: 37 Placebo: 34 (50%) 24 +
(Turkey) (cervical length not considered) (100 mg/day suppositories)
Hassan, 201118 Cervix = 10–20 mm Vaginal progesterone: 235 Placebo: 223 (34%) 20–24 +
(Multinational*) (history of PTB not considered) (90 mg/day gel)
Van Os, 201522 Cervix <30 mm and no history of PTB Vaginal progesterone: 41 Placebo: 39 (18%) 22 +
(Netherlands) (200 mg/day capsules)
Norman, 201621 History of PTB or cervix <25 mm; or positive Vaginal progesterone: 600 Placebo: 597 18–24 +
(Multinational*) fetal fibronectin test and history of PTB, 2nd (200 mg/day gel) (PTB <34 weeks: 17%)
trimester loss, PPROM or cervical
procedure to treat abnormal smears
Azargoon, 201619 History of PTB and cervix >28 mm (only Vaginal progesterone: 50 Placebo: 50 (76%) 16–22
(Iran) data from this subgroup included) (400 mg/day suppositories)
Crowther, 20174 History of PTB Vaginal progesterone: 390 Placebo: 385 (38%) 20–24 +
(Multinational*) (cervical length not considered) (100 mg/day pessary)
Rai, 200932 History of PTB Oral progesterone: 74 Placebo: 74 (59%) 18–24 +
(India) (cervical length not considered) (100 mg twice daily
capsules)
Glover, 201131 History of PTB Oral progesterone: 19 Placebo: 14 (57%) 16–20 +
(USA) (cervical length not considered) (400 mg/day capsules)
Johnson, 197528 History of two PTB and/or abortions 17-OHPC: 18 Placebo: 24 (NA) <24
(USA) (cervical length not considered) (250 mg/week)
Meis, 200326 History of PTB 17-OHPC: 306 Placebo: 153 (55%) 16–20 +
(USA*) (cervical length not considered) (250 mg/week)
Ibrahim, 201027 History of PTB 17-OHPC: 25 Placebo: 25 (52%) 2nd trim.
(Egypt) (cervical length not considered) (250 mg/week)
Saghafi, 201125 History of PTB 17-OHPC: 50 NI/TAU: 50 (60%) 16
(Iran) (cervical length not considered) (250 mg/week)
Grobman, 201229 Cervix <30 mm 17-OHPC: 327 Placebo: 330 (24%) 16–22.5 +
(USA*) (history of PTB not considered) (250 mg/week)
Shahgheibi, 201630 History of preterm labor <37 weeks; history 17-OHPC: 50 Placebo: 50 (58%) 24 +
(Iran) of PTB; history of PTB in sister; acquired (250 mg/week)
and/or congenital uterine abnormalities

Comparison group:
sample size (details)

Vaginal progesterone versus 17a-hydroxyprogesterone caproate

Maher, 201352 History of PTB or history of cervical cerclage Vaginal progesterone: 253 17-OHPC: 249 14–18 +
(Saudi Arabia) (women with cervix ≤25 mm were excluded) (90 mg/day gel) (260 mg/week)
El-Gharib, 201353 History of PTB and cervix ≥15 mm Vaginal progesterone: 80 17-OHPC: 80 20–24
(Egypt) (200 mg/day gel) (100 mg every 3rd day)
Bafghi, 201554 History of PTB or cervix <25 mm Vaginal progesterone: 39 17-OHPC: 39 16–20 +
(Iran) (200 mg/day suppositories) (250 mg/week)

ª 2018 Royal College of Obstetricians and Gynaecologists 5

 Jarde et al.
Table 1. (Continued)
Elimian, 201650 History of PTB
(USA) (cervical length not considered)
Wajid, 201649 History of PTB
(Pakistan) (cervical length not considered)
Pirjani 201751 Cervix ≤25 mm
(Iran) (history of PTB not considered)
17-OHPC, 17a-hydroxyprogesterone caproate; Gest. Age, gestational age (weeks) at entry into trial; NI/TAU, no intervention/treatment as usual;
NR, not reported; PTB, preterm birth; PPROM, preterm premature rupture of membranes; Gest; RoB, risk of bias: studies considered to have low
risk of bias are indicated with a positive (+) sign, otherwise with a negative ( ) sign.
*Multi-centre study.
**Data not extracted by duplicate but with the aid of a translator.
care unit, gestational age at birth and, in pairwise meta-
analyses, PTB <33 and <35 weeks, birthweight <1500 g,
respiratory distress syndrome, and early onset of sepsis.
Cerclage significantly increased the gestational age at birth
and, in pairwise meta-analyses, reduced PTB <33 weeks,
while increasing the length of antepartum hospital stay.
Pessary significantly reduced the odds of preterm prema-
ture rupture of membranes and, in pairwise meta-analyses,
sepsis (without specified time of onset, Tables S1 and S2).
Overall analyses (women with any risk factor)
separating interventions by type and route
When analysing progesterone by route, PV progesterone
significantly reduced the odds of PTB <34 weeks (OR 0.43,
95% CrI 0.21–0.78; NNT 9; Low quality of evidence),
PTB <37 weeks (OR 0.51, 95% CrI 0.34–0.74; NNT 7; Low
quality of evidence), and NND (OR 0.41, 95% CrI 0.20–
0.83; NNT 30; Moderate quality of evidence), ranking
highest, according to SUCRA values, for PTB <34 weeks
and NND, and third highest for prevention of PTB
<37 weeks. PO progesterone did not significantly reduce
PTB <34 weeks, PTB <37 weeks or NND, and IM 17-
OHPC significantly reduced only PTB <37 weeks (OR 0.61,
95% CrI 0.39–0.92; NNT 9; Moderate quality of evidence;
Supporting Information Figure S1 and Table 3). See Sup-
porting Information Figures S2–S4 for individual study
data.
Similar results by type of progesterone (natural proges-
terone PO or PV versus IM 17-OHPC) are presented in
Table S3. When separating cerclage into McDonald and
Shirodkar cerclage, neither type had a significant effect on
PTB <34 weeks, PTB <37 weeks or NND (Table S3). We
did detect significant incoherence in the comparison of
McDonald versus Shirodkar for NND. However, no
6 ª 2018 Royal College of Obstetricians and Gynaecologists
Comparison group:
sample size (details)
Vaginal progesterone: 66 17-OHPC: 79 16–20 + 6 +
(100 mg/day suppositories) (250 mg/week)
Vaginal progesterone: 400 17-OHPC: 400 20–24
(200 mg/day pessary) (250 mg/week)
Vaginal progesterone: 147 17-OHPC: 150 16–24
(400 mg/day suppositories) (250 mg/week)
incoherence was found in any of the other network meta-
analyses of this study, according to the node-splitting test.
Subgroup analyses by common risk factors
Women with a previous preterm birth
In the subgroup of women with a ‘previous preterm birth’
(regardless of cervical length), progesterone (any type and
route) significantly reduced the odds of PTB <34 weeks,
PTB <37 weeks, and NND, compared with control. No
study assessed pessary in this subpopulation and only two
studies assessed cerclage, resulting in no significant benefit
(Supporting Information Table S4).
When separating progesterone by its route (Table 3), PV
progesterone reduced PTB <34 (OR 0.29, 95% CI 0.12–
0.68; NNT 8; Moderate quality of evidence) and <37 weeks
(OR 0.43, 95% CrI 0.23–0.74; NNT 6; Moderate quality of
evidence), but not NND. In addition, we found statistically
significant differences in PTB <37 weeks between the stud-
ies using PV progesterone with a dose of ≤200 mg/day (OR
0.67, 95% CI 0.40–1.13) and those using a higher dose
(OR 0.18, 95% CI 0.05–0.58). PO progesterone significantly
reduced PTB <34 weeks (OR 0.42, 95% CI 0.22–0.83; NNT
5; Low quality of evidence) but not PTB <37 weeks or
NND. IM 17-OHPC reduced PTB <37 weeks (OR 0.53,
95% CrI 0.27–0.95; NNT 7; Moderate quality of evidence),
as well as NND (OR 0.39, 95% CI 0.16–0.95; NNT 24; Low
quality of evidence, Tables 3 and S4). None of the studies
including women with a previous preterm birth assessed
the effect of IM 17-OHPC on PTB <34 weeks. However,
PV progesterone reduced PTB <34 compared with IM 17-
OHPC (OR 0.63, 95% CI 0.43–0.93; NNT 14). See Sup-
porting Information Figures S5–S7 for individual study
data.
 Progesterone, cerclage, and pessary to prevent preterm birth

Table 2. Main characteristics of studies included in network meta-analysis (assessing cerclage, pessary, and cerclage versus progesterone)

Cerclage versus control

Lazar, 198438 Scoring system based on several risk factors McDonald: 268 NI/TAU: 238 (5%) <28
(France*) (including cervical length and history of PTB)
Rush, 198436 History of PTB (≥2 previous PTB; women with McDonald: 96 NI/TAU: 98 (32%) 15–21
(South Africa) 2 cm long or dilated cervix excluded)
MacNaughton, 199337 Obstetrician’s uncertainty whether to McDonald or similar: 635 NI/TAU: 629 (30%) ≤29*** +
(Multinational*) advise cerclage (a minority received a
Shirodkar cerclage
or similar)
Rust, 200035 **** Prolapse of membranes >25% or cervix McDonald: 28 NI/TAU: 26 (38%) 16–24 +
(USA) <25 mm (history of PTB not considered)
Althuisius, 200140 Cervix ≤25 mm McDonald: 19 NI/TAU: 16 <27 +
(Netherlands) (history of PTB not considered) (and bed rest) (PTB <34
weeks: 44%)
Rust, 200134 Dilation of the internal os and either prolapse McDonald: 47 NI/TAU: 52 16–24 +
(USA) of membranes >25% or cervix <25 mm (PTB <34
(history of PTB not considered) weeks: 33%)
Berghella, 200439 Cervix ≤25 mm or significant funneling (>25%) McDonald: 31 NI/TAU: 30 14–24 +
(USA) in women with history of PTB, ≥2 curettage (and bed rest (PTB <34
procedures, diethylstilbestrol exposure, and education about weeks: 41%)
cone biopsy or Mullerian anomaly. preterm labour)
Ezechi, 200433 History of PTB McDonald: 38 NI/TAU: 43 (37%) 14
(Nigeria) (cervical length not considered)
Otsuki, 201643 Cervix ≤25 mm McDonald: 34; NI/TAU: 33 (30%) 16–26 +
(Japan*) (history of PTB not considered) Shirodkar: 34
Szeverenyi, 199242 History of PTB or of miscarriage Shirodkar: 108 NI/TAU: 108 (25%) NR
(Hungary**)
To, 200441 Cervix ≤15 mm Shirodkar: 127 NI/TAU: 125 22–24 +
(Multinational*) (history of PTB not considered) (PTB <33
weeks: 26%)
Pessary versus control
Goya, 201247 Cervix ≤25 mm Arabin: 190 NI/TAU: 190 (59%) 18–22 +
(Spain) (history of PTB not considered)
Hui, 201346 (Hong Cervix ≤25 mm (women with a known history Arabin: 53 NI/TAU: 55 (18%) 20–24 +
Kong, China) of cervical incompetence were excluded)
Nicolaides 201645 Cervix ≤25 mm Arabin: 261 NI/TAU: 248 20–24 +
(Multinational*) (history of PTB not considered) (PTB <34
weeks: 4%)
Saccone 201744 Cervix 20–25 mm and not receiving Arabin: 17 NI/TAU: 25 18–23 + 6 +
(Italy) progesterone (only data from this (PTB <34
subgroup were included) weeks: 8%)

Comparison group:
sample size (details)

Cerclage versus progesterone

Keeler, 200948 Cervix ≤25 mm McDonald: 42 17-OHPC: 37 16–24 +
(USA) (history of PTB not considered) (250 mg/week)

17-OHPC, 17a-hydroxyprogesterone caproate; Gest. Age, gestational age (weeks) at entry into trial; NI/TAU, no intervention/treatment as usual;
NR, not reported; PTB, preterm birth; Gest; RoB, risk of bias: studies considered to have low risk of bias are indicated with a positive (+) sign,
otherwise with a negative ( ) sign.
*Multi-centre study.
**Data not extracted by duplicate but with the aid of a translator.
***87% of women were recruited before 20 completed weeks of pregnancy.
****Subsample of Rust34; only outcomes not reported in Rust34 were used.

ª 2018 Royal College of Obstetricians and Gynaecologists 7

 Jarde et al.

Women with a short cervix (≤25 mm)
In the subgroup of studies with women with a cervical
length ≤25 mm (regardless of history of preterm birth), the
only statistically significant result was the reduction of PTB
<34 weeks in women receiving progesterone (specifically
PV progesterone) compared with control (Table 3,
Supporting Information Table S5). See Supporting Infor-
mation Figures S8–S10 for individual study data.
Sensitivity analyses
Of the 40 included studies, 28 were considered to have low
risk of bias. The network including only low risk of bias

Table 3. Effectiveness of progesterone (PV, PO, IM 17-OHPC), cerclage, and pessary compared with control in women with a singleton
pregnancy at overall risk of preterm birth and in subgroups defined by the common risk factor (previous preterm birth or cervical length ≤25 mm)

Outcome Intervention k n SUCRA OR NNT Quality of

(studies in the NMA) (95% CrI/CI) (95% CI) evidence

Subgroup analyses by route of administration of progesterone

Preterm birth <34 Progesterone (PV) 7 1853 75% 0.43 (0.21–0.78) 9 (6–25) Low***,****
weeks (k = 21)* Progesterone (PO) 1 148 67% 0.42 (0.08–2.07) Very Low
Cerclage 5 508 56% 0.54 (0.22–1.20) Moderate
Pessary 4 1036 48% 0.60 (0.25–1.68) Very Low***
17-OHPC 1 657 44% 0.64 (0.24–1.54) Low
Preterm birth <37 Pessary 2 488 85% 0.31 (0.13–0.82) 5 (4–21) Very Low***
weeks (k = 31)*,** Progesterone (PO) 2 181 77% 0.37 (0.13–1.03) Low
Progesterone (PV) 8 2289 64% 0.51 (0.34–0.74) 7 (5–15) Low***,****
17-OHPC 5 1366 45% 0.61 (0.39–0.92) 9 (6–49) Moderate***
Cerclage 7 2416 25% 0.78 (0.48–1.25) Very Low****
Neonatal death Progesterone (PV) 5 3121 76% 0.41 (0.20–0.83) 30 (22–104) Moderate
(k = 18)* Progesterone (PO) 1 148 71% 0.38 (0.06–1.88) Very Low
17-OHPC 3 1166 53% 0.58 (0.28–1.16) Moderate****
Cerclage 5 640 44% 0.68 (0.29–1.46) Very Low
Pessary 2 488 39% 0.91 (0.02–36.81) Very Low ***

Subgroup analyses by route of administration of progesterone in women with a previous preterm birth
Preterm birth <34 Progesterone (PV) 3 224 n.a. (MA) 0.29 (0.12–0.68) 8 (6–18) Moderate
weeks (no NMA) Progesterone (PO) 1 148 n.a. 0.42 (0.22–0.83) 5 (4–22) Low
17-OHPC 0 0 – – –
Cerclage 0 0 – – –
Pessary 0 0 – – –
Preterm birth <37 Progesterone (PO) 2 181 75% 0.37 (0.11–1.18) Moderate
weeks (k = 17)* Progesterone (PV) 5 1610 72% 0.43 (0.23–0.74) 6 (4–14) Moderate***
17-OHPC 4 616 52% 0.53 (0.27–0.95) 7 (4–80) Moderate
Cerclage 2 275 46% 0.60 (0.19–1.74) Very low***
Pessary 0 0 – – –
Neonatal death 17-OHPC 2 509 n.a. (MA) 0.39 (0.16–0.95) 24 (17–295) Low
(no NMA) Progesterone (PO) 1 148 n.a. 0.40 (0.10–1.63) Very low
Progesterone (PV) 2 1386 n.a. (MA) 0.76 (0.28–2.07) Low
Cerclage 1 194 n.a. 1.02 (0.39–2.70) Very low
Pessary 0 0 – – –

Subgroup analyses by route of administration of progesterone in women with a cervical length ≤25 mm
Preterm birth <34 Cerclage 2 136 n.a. (MA) 0.22 (0.01–4.99) Very low
weeks (no NMA) Progesterone (PV) 1 226 n.a. 0.45 (0.24–0.84) 7 (5–28) Low
Pessary 4 1036 n.a. (MA) 0.68 (0.20–2.29) Very low
Progesterone (PO) 0 0 – – –
17-OHPC 0 0 – – –
Preterm birth <37 Pessary 2 488 n.a. (MA) 0.36 (0.09–1.48) Very low
weeks (no NMA) Cerclage 1 101 n.a. 0.83 (0.33–2.07) Very low
Progesterone (PV) 1 458 n.a. 0.84 (0.57–1.24) Low
Progesterone (PO) 0 0 – – –
17-OHPC 0 0 – – –

8 ª 2018 Royal College of Obstetricians and Gynaecologists

 Progesterone, cerclage, and pessary to prevent preterm birth

Table 3. (Continued)

Outcome Intervention k n SUCRA OR NNT Quality of

(studies in the NMA) (95% CrI/CI) (95% CI) evidence

Neonatal death
(no NMA) Cerclage 3 389 n.a. (MA) 0.55 (0.18–1.68) Low
Progesterone (PV) 1 458 n.a. 0.56 (0.13–2.39) Very low
Pessary 2 488 n.a. (MA) 1.02 (0.11–9.90) Low
Progesterone (PO) 0 0 – – –
17-OHPC 0 0 – – –

17-OHPC, 17a-hydroxyprogesterone caproate (intramuscular); CI, confidence interval (MA); CrI, credibility interval (NMA); k, number of studies;
MA, pairwise meta-analysis; MD, mean difference; n, sample size; n.a., not applicable; NMA, network meta-analysis; NNT, number needed to
treat; OR, odds ratio; PO, per os; PV, per vagina; SUCRA, Surface Under the Cumulative Ranking curve.
Statistically significant results are given in bold. Interventions sorted by SUCRA values (network meta-analyses) or effect size (pairwise meta-
analyses). Quality of evidence assessed using the GRADE approach. No significant incoherence between direct and indirect estimates was
detected.
*Additional studies assessed PV progesterone versus 17-OHPC.
**One additional study compared progesterone with McDonald cerclage.
***More than moderate heterogeneity was detected (I2 > 60%).
****We estimated that there was intransitivity between the studies of the direct comparisons that underlie the indirect comparison.

studies found no significant effect of any of the interven-
tions on reducing PTB <34 weeks. The results for PTB
<37 weeks and NND remained similar to the original ones
(Supporting Information Table S6).
Given the lack of stratified data for singletons in the
PROGRESS Study, we assumed none of the twins had the
outcome and that, in sensitivity analysis, all had, and found
minimal differences between these two extreme approaches
(Supporting Information Tables S7 and S8).
Six studies assessing progesterone reported the possibility
or actual placement of a cerclage as an additional interven-
tion. When reported, the proportion of women receiving
such an additional intervention ranged from 2 to 17%,
although it was similarly distributed between the
groups.18,21,28,29,32,50 Removing these studies from the anal-
yses did not change the conclusions for the effect of pro-
gesterone on PTB <34 or <37 weeks, but the odds of NND
were no longer statistically significant.
As a minimum of ten studies is recommended to be able
to assess the potential for publication bias12, the trim-and-
fill method was applied in only four comparisons. In three
of these analyses the results changed from statistically sig-
nificant to borderline significant after imputing additional
studies, and the quality of evidence was lowered accord-
ingly.
Discussion
Main findings
Our network meta-analysis of 40 trials on preterm birth
prevention in singletons found that PV progesterone
reduced PTB <34 weeks, PTB <37 weeks, and NND,
making it the most consistently effective among the
interventions studied in women at overall risk of preterm
birth. In women with a short cervix (≤25 mm), PV pro-
gesterone reduced PTB <34 weeks, but data were sparse.
In women with a previous preterm birth, PV proges-
terone significantly reduced PTB <34 and <37 weeks,
although not NND. IM 17-OHPC significantly reduced
PTB <37 weeks and NND; PTB <34 weeks could not be
assessed due to lack of data. Pessary and PO proges-
terone showed significant results in only one comparison
each.
Strengths and limitations
The main strength of this study is its design with both an
overarching synthesis of the published trials on interven-
tions to prevent preterm birth (progesterone, cerclage, and
pessary), and a detailed analysis by types and routes of
administration, and by subgroups based on risk factors (i.e.
previous preterm birth or short cervix). Another strength
of our study was the combination of direct and indirect
estimates using network meta-analyses, and assessing the
underlying assumptions and the quality of evidence using
the GRADE approach.
Our study has several limitations, as we were limited by
the data and subgroups reported in the included studies,
yielding sparse data for some subgroups, such as in women
with a short cervix, and consequently wider CrI/CI and
uncertainty in the results. Similarly, there is a disparity in
the outcomes reported in studies using different interven-
tions. For instance, whereas half of the studies assessing the
effect of PV progesterone on PTB <34 and/or <37 weeks
reported results for both outcomes, only one study

ª 2018 Royal College of Obstetricians and Gynaecologists 9

 Jarde et al.
assessing the effects of 17-OHPC did so. This disparity in
the amount of data for each intervention and outcome puts
some restraints on our results. Finally, in some cases the
assumptions underlying a number of network meta-ana-
lyses were challenged, and the quality of evidence was ‘low’
or ‘very low’. However, in life-threatening situations and
no evidence of harm it is justified to make strong recom-
mendations in favour of an intervention even when there is
‘low’ or ‘very low’ quality of evidence.55
Interpretation in light of other evidence
We identified PV progesterone as an effective intervention
to prevent preterm birth. Including the PROGRESS Study,4
five studies assessed PV progesterone in women with previ-
ous preterm birth and reported PTB <37 weeks (the PRO-
GRESS Study did not report PTB <34 weeks), with a
pooled effect size of 0.43 (95% CrI 0.23–0.74, Moderate
quality of evidence), although with substantial heterogene-
ity. Three studies had positive, statistically significant effects
with progesterone,19,23,24 while the remaining two, which
account for 86% of the sample, centred on the null
effect.4,17 One possible explanation is publication bias.
Alternatively, when comparing the characteristics of the
two groups of studies we found that studies including
women with previous preterm births and using vaginal
suppositories reported significant, positive results, whereas
those using vaginal gel or pessary did not.
A subgroup analysis by dose of progesterone (≤200 ver-
sus >200 mg/day) found statistically significant differences
between the groups in favour of higher doses, although as
all but one of the studies (Azargoon19 used 400 mg/day)
used very similar low doses (90 or 100 mg/day), this does
not explain the differences between the large non-signifi-
cant studies and the smaller significant ones.
Recently Romero et al.56 pooled individual participant
data from five studies (all of which were included in our
overall analyses, although stratified data for women with a
short cervix were available for only two of them),18,20 and
found similarly significant results for PTB <34 weeks, and
non-significant ones for PTB <37 weeks and NND. Besides
these outcomes, however, they found a benefit with vaginal
progesterone on many outcomes.56
Areas for future research
Further research is required to explore PV progesterone’s
heterogeneity, examine interventions for short cervices,
confirm the reduction in neonatal death with 17-OHPC,
and randomly investigate combinations of therapies.
Conclusions
Overall, we found that progesterone, particularly PV pro-
gesterone, was a consistently effective intervention to
10
prevent preterm birth as well as neonatal death, in women
with a singleton pregnancy at risk overall and in women at
risk due to a previous preterm birth. In the subpopulation
of women with a short cervix there was no clear evidence
of benefit. The quality of evidence varied between low and
high.
Disclosure of interests
All authors declare that they have no competing interests.
Completed disclosure of interest forms are available to view
online as supporting information.
Contribution to authorship
SDM, AJ, and JB conceived, planned, and designed the
study. AJ and OL executed the search strategy, study selec-
tion, data collection, assessment of the risk of bias, and the
quality of the evidence. AJ conducted the statistical analy-
ses, supervised by JB. AJ and SDM drafted the manuscript.
AJ, OL, JB, and SDM contributed to the interpretation of
the results and commented on the manuscript. AJ is the
guarantor for the study.
Details of ethics approval
No ethical approval was required for this study.
Funding
JB holds the John D. Cameron Endowed Chair in the
Genetic Determinants of Chronic Diseases, Department of
Clinical Epidemiology and Biostatistics, McMaster Univer-
sity. SDM is supported by a Canadian Institutes of Health
Research (CIHR) Tier II Canada Research Chair, Sponsor
Award #950-229920. None of the agencies had any influ-
ence on the design and conduct of the study; collection,
management, analysis, and interpretation of the data;
preparation, review, or approval of the manuscript; and
decision to submit the manuscript for publication.
Acknowledgements
We would like to thank Sugee Korale Liyanage and Kiran
Ninan for their administrative support on the project.
Supporting Information
Additional supporting information may be found online in
the Supporting Information section at the end of the
article.
Figure S1. Network graphs for preterm birth <34 weeks,
preterm birth <37 weeks, and neonatal death (separating
progesterone by route of administration: per vagina (PV),
per os (PO), and intramuscular 17-OHPC).
Figure S2. Forest plot of each comparison (separating
progesterone by route of administration) on preterm birth
<34 weeks in women overall at risk (reported network
ª 2018 Royal College of Obstetricians and Gynaecologists

estimates differ from these results due to the influence of
indirect estimates).
Figure S3. Forest plot of each comparison (separating
progesterone by route of administration) on preterm birth
<37 weeks in women overall at risk (reported network esti-
mates differ from these results due to the influence of indi-
rect estimates).
Figure S4. Forest plot of each comparison (separating
progesterone by route of administration) on neonatal death
in women overall at risk (reported network estimates differ
from these results due to the influence of indirect estimates).
Figure S5. Forest plot of each comparison (separating
progesterone by route of administration) on preterm birth
<34 weeks in women with a previous preterm birth.
Figure S6. Forest plot of each comparison (separating
progesterone by route of administration) on preterm birth
<37 weeks in women with a previous preterm birth (re-
ported network estimates differ from these results due to the
influence of indirect estimates). One of the studies (Shah-
gheibi et al.30) did not specify the number of cases in each
group for this subgroup, but reported instead only OR
(95% CI).
Figure S7. Forest plot of each comparison (separating
progesterone by route of administration) on neonatal death
in women with a previous preterm birth.
Figure S8. Forest plot of each comparison (separating
progesterone by route of administration) on preterm birth
<34 weeks in women with a cervical length ≤25 mm.
Figure S9. Forest plot of each comparison (separating
progesterone by route of administration) on preterm birth
<37 weeks in women with a cervical length ≤25 mm.
Figure S10. Forest plot of each comparison (separating
progesterone by route of administration) on neonatal death
in women with a cervical length ≤25 mm.
Table S1. Effectiveness of progesterone, cerclage, and
pessary compared with control for primary outcomes and
secondary outcomes for which network meta-analysis was
performed.
Table S2. Effectiveness of progesterone, cerclage, and
pessary compared to control for secondary outcomes for
which meta-analysis was performed.
Table S3. Effectiveness of progesterone, cerclage, and pes-
sary compared to control after separating the interventions
by type.
Table S4. Effectiveness of progesterone, cerclage, and
pessary compared with control in women with a previous
preterm birth.
Table S5. Effectiveness of progesterone, cerclage, and
pessary compared to control in women with a cervical
length ≤25 mm.
Table S6. Effectiveness of progesterone, cerclage, and
pessary compared with control in studies with a low risk of
bias.
ª 2018 Royal College of Obstetricians and Gynaecologists
Progesterone, cerclage, and pessary to prevent preterm birth
Table S7. Comparison of the results assuming that none
of the twins or all of the twins included in Crowther et al.4
had the given outcome: Network meta-analyses.
Table S8. Comparison of the results assuming that none
of the twins or all twins included in Crowther et al.4 had
the given outcome: pairwise meta-analysis.
Appendix S1. Search strategy.&
References
1 World Health Organization. Preterm Birth 2014. [http://www.who.
int/mediacentre/factsheets/fs363/en/]. Accessed 19 February 2018.
2 Butler AS, Behrman RE. Preterm Birth: Causes, Consequences, and
Prevention. Washington, DC: National Academies Press; 2007.
3 Jarde A, Lutsiv O, Park CK, Beyene J, Dodd JM, Barrett J, et al.
Effectiveness of progesterone, cerclage and pessary for preventing
preterm birth in singleton pregnancies: a systematic review and
network meta-analysis. BJOG 2017;124:1176–89.
4 Crowther CA, Ashwood P, McPhee AJ, Flenady V, Tran T, Dodd JM,
et al. Vaginal progesterone pessaries for pregnant women with a
previous preterm birth to prevent neonatal respiratory distress
syndrome (the PROGRESS Study): a multicentre, randomised,
placebo-controlled trial. PLoS Med 2017;14:e1002390.
5 Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of
Interventions. Hoboken: John Wiley & Sons; 2011.
6 Jansen JP, Crawford B, Bergman G, Stam W. Bayesian meta-analysis
of multiple treatment comparisons: an introduction to mixed
treatment comparisons. Value Health 2008;11:956–64.
7 Gelman A, Rubin DB. Inference from iterative simulation using
multiple sequences. Statist Sci 1992;7:457–72.
8 DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials 1986;7:177–88.
9 Puhan MA, Schu €nemann HJ, Murad MH, Li T, Brignardello-Petersen
R, Singh JA, et al. A GRADE Working Group approach for rating the
quality of treatment effect estimates from network meta-analysis.
BMJ 2014;349:g5630.
10 Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring
inconsistency in meta-analyses. BMJ 2003;327:557.
11 Valkenhoef G, Dias S, Ades A, Welton NJ. Automated generation of
node-splitting models for assessment of inconsistency in network
meta-analysis. Res Synth Methods 2016;7:80–93.
12 Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau J, et al.
Recommendations for examining and interpreting funnel plot
asymmetry in meta-analyses of randomised controlled trials. BMJ
2011;343:d4002.
13 Duval S, Tweedie R. Trim and fill: a simple funnel-plot-based
method of testing and adjusting for publication bias in meta-
analysis. Biometrics 2000;56:455–63.
14 Salanti G, Ades A, Ioannidis JP. Graphical methods and numerical
summaries for presenting results from multiple-treatment meta-
analysis: an overview and tutorial. J Clin Epidemiol 2011;64:163–71.
15 van Valkenhoef G, Kuiper J. gemtc: Network Meta-Analysis Using
Bayesian methods. R package version 06-2 ed2015
16 Akbari S, Birjandi M, Mohtasham N. Evaluation of the effect of
progesterone on prevention of preterm delivery and its
complications. Scientific J Kurdistan Univ Med Sci 2009;14:11–9.
17 O’Brien J, Adair C, Lewis D, Hall D, Defranco E, Fusey S, et al.
Progesterone vaginal gel for the reduction of recurrent preterm
birth: primary results from a randomized, double-blind, placebo-
controlled trial. Ultrasound Obstet Gynecol 2007;30:687–96.
11
 Jarde et al.
18 Hassan S, Romero R, Vidyadhari D, Fusey S, Baxter J, Khandelwal M,
et al. Vaginal progesterone reduces the rate of preterm birth in
women with a sonographic short cervix: a multicenter, randomized,
double-blind, placebo-controlled trial. Ultrasound Obstet Gynecol
2011;38:18–31.
19 Azargoon A, Ghorbani R, Aslebahar F. Vaginal progesterone on the
prevention of preterm birth and neonatal complications in high risk
women: a randomized placebo-controlled double-blind study. Int J
Reprod BioMed 2016;14:309.
20 Fonseca EB, Celik E, Parra M, Singh M, Nicolaides KH. Progesterone
and the risk of preterm birth among women with a short cervix. N
Engl J Med 2007;357:462–9.
21 Norman JE, Marlow N, Messow C-M, Shennan A, Bennett PR,
Thornton S, et al. Vaginal progesterone prophylaxis for preterm
birth (the OPPTIMUM study): a multicentre, randomised, double-
blind trial. Lancet 2016;387:2106–16.
22 van Os MA, van der Ven AJ, Kleinrouweler CE, Schuit E, Kazemier
BM, Verhoeven CJ, et al. Preventing preterm birth with
progesterone in women with a short cervical length from a low-risk
population: a multicenter double-blind placebo-controlled
randomized trial. Am J Perinatol 2015;32:993–1000.
23 Majhi P, Bagga R, Kalra J, Sharma M. Intravaginal use of natural
micronised progesterone to prevent pre-term birth: a randomised
trial in India. J Obstet Gynaecol 2009;29:493–8.
24 Cetingoz E, Cam C, Sakallı M, Karateke A, Celik C, Sancak A.
Progesterone effects on preterm birth in high-risk pregnancies: a
randomized placebo-controlled trial. Arch Gynecol Obstet
2011;283:423–9.
25 Saghafi N, Khadem N, Mohajeri T, Shakeri MT. Efficacy of 17a-
hydroxyprogesterone caproate in prevention of preterm delivery. J
Obstet Gynaecol Res 2011;37:1342–5.
26 Meis PJ, Klebanoff M, Thom E, Dombrowski MP, Sibai B,
Moawad AH, et al. Prevention of recurrent preterm delivery by 17
alpha-hydroxyprogesterone caproate. N Engl J Med
2003;348:2379–85.
27 Ibrahim M, Ramy ARM, Younis MA-F. Progesterone supplementation
for prevention of preterm labor: a randomized controlled trial.
Middle East Fertil Soc J 2010;15:39–41.
28 Johnson JW, Austin KL, Jones GS, Davis GH, King TM. Efficacy of
17a-hydroxyprogesterone caproate in the prevention of premature
labor. N Engl J Med 1975;293:675–80.
29 Grobman WA, Thom EA, Spong CY, Iams JD, Saade GR, Mercer
BM, et al. 17 alpha-hydroxyprogesterone caproate to prevent
prematurity in nulliparas with cervical length less than 30 mm. Am J
Obstet Gynecol 2012;207:390. e1-. e8.
30 Shahgheibi S, Soofizadeh N, Mojtahedzadeh A, Rezaei M,
Seydoshohadaei F, Moradi G, et al. The effect of 17 alpha-
Hydroxyprogestrone Caproate on prevention of preterm labor in
high-risk pregnant women: a clinical trial study. Int J Med Res
Health Sci 2016;5:261–6.
31 Glover MM, McKenna DS, Downing CM, Smith DB, Croom CS,
Sonek JD. A randomized trial of micronized progesterone for the
prevention of recurrent preterm birth. Am J Perinatol 2011;28:377–
81.
32 Rai P, Rajaram S, Goel N, Ayalur Gopalakrishnan R, Agarwal R,
Mehta S. Oral micronized progesterone for prevention of preterm
birth. Int J Gynecol Obstet 2009;104:40–3.
33 Ezechi O, Kalu B, Nwokoro C. Prophylactic cerclage for the
prevention of preterm delivery. Int J Gynecol Obstet 2004;85:283–4.
34 Rust OA, Atlas RO, Reed J, van Gaalen J, Balducci J. Revisiting the
short cervix detected by transvaginal ultrasound in the second
trimester: why cerclage therapy may not help. Am J Obstet Gynecol
2001;185:1098–105.
12
35 Rust OA, Atlas RO, Jones KJ, Benham BN, Balducci J. A randomized
trial of cerclage versus no cerclage among patients with
ultrasonographically detected second-trimester preterm dilatation of
the internal os. Am J Obstet Gynecol 2000;183:830–5.
36 Rush R, Isaacs S, McPherson K, JONES L, Chalmers I, Grant A. A
randomized controlled trial of cervical cerclage in women at high
risk of spontaneous preterm delivery. BJOG 1984;91:724–30.
37 MacNaughton M, Chalmers I, Dubowitz V, Dunn P, Grant A,
McPherson K, et al. Final report of the Medical Research Council/
Royal College of Obstetricians and Gynaecologists multicentre
randomised trial of cervical cerclage. BJOG 1993;100:516–23.
38 Lazar P, Gueguen S, Dreyfus J, Renaud R, Pontonnier G, Papiernik E.
Multicentred controlled trial of cervical cerclage in women at
moderate risk of preterm delivery. BJOG 1984;91:731–5.
39 Berghella V, Odibo AO, Tolosa JE. Cerclage for prevention of
preterm birth in women with a short cervix found on transvaginal
ultrasound examination: a randomized trial. Am J Obstet Gynecol
2004;191:1311–7.
40 Althuisius SM, Dekker GA, Hummel P, Bekedam DJ, van Geijn HP.
Final results of the Cervical Incompetence Prevention Randomized
Cerclage Trial (CIPRACT): therapeutic cerclage with bed rest versus
bed rest alone. Am J Obstet Gynecol 2001;185:1106–12.
41 To MS, Alfirevic Z, Heath VC, Cicero S, Cacho AM, Williamson PR,
et al. Cervical cerclage for prevention of preterm delivery in woman
with short cervix: randomised controlled trial. Lancet
2004;363:1849–53.
42 Szeverenyi M, Chalmels J, Grant A, Nagy T, Nagy J, Balogh I, et al.
Surgical cerclage in the treatment of cervical incompetence during
pregnancy (determining the legitimacy of the procedure). Orv Hetil
1992;133:1823–6.
43 Otsuki K, Nakai A, Matsuda Y, Shinozuka N, Kawabata I, Makino Y,
et al. Randomized trial of ultrasound-indicated cerclage in singleton
women without lower genital tract inflammation. J Obstet Gynaecol
Res 2016;42:148–57.
44 Saccone G, Maruotti GM, Giudicepietro A, Martinelli P. Effect of
cervical pessary on spontaneous preterm birth in women with
singleton pregnancies and short cervical length: a randomized
clinical trial. JAMA 2017;318:2317–24.
45 Nicolaides KH, Syngelaki A, Poon LC, Picciarelli G, Tul N, Zamprakou
A, et al. A randomized trial of a cervical pessary to prevent preterm
singleton birth. N Engl J Med 2016;374:1044–52.
46 Hui S, Chor CM, Lau TK, Lao TT, Leung TY. Cerclage pessary for
preventing preterm birth in women with a singleton pregnancy and
a short cervix at 20 to 24 weeks: a randomized controlled trial. Am
J Perinatol 2013;30:283–8.
47 Goya M, Pratcorona L, Merced C, Rodo  C, Valle L, Romero A, et al.
Cervical pessary in pregnant women with a short cervix (PECEP): an
open-label randomised controlled trial. Lancet 2012;379:1800–6.
48 Keeler SM, Kiefer D, Rochon M, Quinones JN, Novetsky AP, Rust O. A
randomized trial of cerclage vs. 17 a-hydroxyprogesterone caproate
for treatment of short cervix. J Perinatal Med 2009;37:473–9.
49 Wajid R, Zafar M, Waheed F. Effectiveness of vaginal versus
intramuscular progesterone for the prevention of preterm delivery.
Ann King Edward Med Univ 2016;22:284–9.
50 Elimian A, Smith K, Williams M, Knudtson E, Goodman JR,
Escobedo MB. A randomized controlled trial of intramuscular versus
vaginal progesterone for the prevention of recurrent preterm birth.
Int J Gynecol Obstet 2016;134:169–72.
51 Pirjani R, Heidari R, Rahimi-Foroushani A, Bayesh S, Esmailzadeh A.
17-Alpha-hydroxyprogesterone caproate versus vaginal progesterone
suppository for the prevention of preterm birth in women with a
sonographically short cervix: a randomized controlled trial. J Obstet
Gynaecol Res 2017;43:57–64.
ª 2018 Royal College of Obstetricians and Gynaecologists

52 Maher MA, Abdelaziz A, Ellaithy M, Bazeed MF. Prevention of preterm
birth: a randomized trial of vaginal compared with intramuscular
progesterone. Acta Obstet Gynecol Scand 2013;92:215–22.
53 El-Gharib M, El-Hawary T. Matched sample comparison of
intramuscular versus vaginal micronized progesterone for prevention
of preterm birth. J Matern Fetal Neonatal Med 2013;26:716–9.
54 Bafghi AST, Bahrami E, Sekhavat L. Comparative study of vaginal
versus intramuscular progesterone in the prevention of preterm
delivery: a randomized clinical trial. Electron Phys 2015;7:1301.
ª 2018 Royal College of Obstetricians and Gynaecologists
Progesterone, cerclage, and pessary to prevent preterm birth
55 Neumann I, Santesso N, Akl EA, Rind DM, Vandvik PO, Alonso-
Coello P, et al. A guide for health professionals to interpret and use
recommendations in guidelines developed with the GRADE
approach. J Clin Epidemiol 2016;72:45–55.
56 Romero R, Conde-Agudelo A, Da FE, O’Brien JM, Cetingoz E,
Creasy GW, et al. Vaginal progesterone for preventing preterm birth
and adverse perinatal outcomes in singleton gestations with a short
cervix: a meta-analysis of individual patient data. Am J Obstet
Gynecol 2017;38:220–1.
13