Sinds de vorige richtlijn preventie vroeggeboorte uit 2007 zijn een aantal nieuwe studies verschenen die het nut van progesteron (vaginaal, oraal, of intramusculair), cerclage en pessarium hebben onderzocht bij vrouwen met een vroeggeboorte in de voorgeschiedenis en al dan niet een korte cervixlengte. De huidige module is erop gericht om een eenduidig advies te geven aan hulpverleners zodat praktijkvariatie wordt verminderd, en de juiste behandeling voor de juiste patiëntengroep wordt toegepast. De module progesteron is onlangs gepubliceerd (21-09-2019) en zal daarom grotendeels buiten de beschouwing van deze richtlijnmodule blijven. (https://richtlijnendatabase.nl/richtlijn/preventie_recidief_spontane_vroeggeboorte/progesteron_voor_het_voorkomen_van_vroeggeboorte.html).

Conclusions PICO 1- Cerclage versus no cerclage/pessary/progesterone

1.1 Cerclage compared to no cerclage in women with a cervical length < 25 mm

1. Preterm birth <28 weeks and previous preterm birth

Preterm birth < 28 weeks and no previous preterm birth

2. Preterm birth <34 weeks (previous history not specified)

Preterm birth < 32 weeks, and previous preterm birth

Preterm birth < 34 weeks and no previous preterm birth 

3. Preterm birth <37 weeks (previous history not specified)

Preterm birth and previous preterm birth

Preterm birth < 37 weeks and no previous preterm birth

4. Neonatal death

5. Composite outcome neonatal mortality/morbidity and previous preterm birth

Composite outcome neonatal mortality/morbidity and no previous preterm birth

6. Respiratory distress syndrome

7. Necrotizing enterocolitis

8. Intraventricular haemorrhage 

9. Neonatal sepsis

1.2 Cerclage compared to Progesterone

Conclusions PICO-2 Pessary compared to no pessary/cerclage/progesterone

2.1 Pessary compared to no pessary in women with a cervical length < 25 mm

1. Spontaneous birth < 32 weeks and preterm birth not considered

2. Spontaneous preterm birth < 34 weeks and preterm birth not considered

Spontaneous preterm birth < 34 weeks and no previous preterm birth

Spontaneous preterm birth < 34 weeks and previous preterm birth

3. Spontaneous birth < 37 weeks and preterm birth not considered

4. Perinatal death

5. Composite morbidity/mortality

6. RDS

7. NEC

8. IVH

9. Sepsis

2.2 Pessary compared to Progesterone

PICO 1- Cerclage versus no cerclage/pessary/progesterone

Description of studies

The systematic review and network meta-analysis of Jarde (2019) was conducted 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.

Medline, EMBASE, CINAHL, Cochrane CENTRAL, and Web of Science were searched up to January 2018. Randomised trials of progesterone, cerclage or pessary for preventing preterm birth in at-risk singleton pregnancies were included (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Otsuki, 2016; Keeler, 2009).

Jarde (2019) reported separate analyses for women with a previous preterm birth and women with a cervical length ≤25 mm.

The systematic review of Conde-Agudelo (2018) was an update of a meta-analysis (Conde-Agudelo, 2013), an adjusted indirect comparison meta-analysis of vaginal progesterone vs cerclage using placebo/no cerclage as the common comparator. The aim of the review was to compare the efficacy of vaginal progesterone and cerclage in preventing preterm birth and adverse perinatal outcomes in women with a singleton gestation, previous spontaneous preterm birth, and a midtrimester sonographic short cervix. The review included data (post hoc analysis; subgroups) from studies that reported on women with cervical length < 25mm and previous preterm birth. MEDLINE, EMBASE, LILACS, and CINAHL were searched from their inception to March 2018 as well as Cochrane databases, bibliographies, and conference proceedings. The review included randomized controlled trials comparing vaginal progesterone to placebo/no treatment or cerclage to no cerclage in women with a singleton gestation, previous spontaneous preterm birth, and a sonographic cervical length <25 mm. Primary outcomes were preterm birth <35 weeks of gestation and perinatal mortality.

In the analysis of the literature data from five studies were included that compared cerclage versus no cerclage (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Owen 2009).

The systematic review of Berghella (2017) was an individual patient data (IPD) meta-analysis of five RCTs (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Otsuki, 2016), reporting on 419 women. The aim of this systematic review and meta-analysis was to quantify the efficacy of cervical cerclage in preventing preterm birth (PTB) in asymptomatic singleton pregnancies with a short mid-trimester cervical length (CL) on transvaginal sonography (TVS) and without prior spontaneous preterm birth.

Electronic databases were searched from inception of each database until February 2017. All randomized controlled trials of asymptomatic singleton pregnancies without prior spontaneous PTB, found to have short cervical length < 25 mm on mid-trimester TVS and then randomized to management with either cerclage or no cerclage, were included.

Trials included in the systematic reviews (Jarde, 2019; Conde-Agudelo, 2018; Berghella, 2017)

Otsuki (2016) is a randomised controlled trial to assess the effects of ultrasound indicated cervical cerclage in the mid-trimester to prevent preterm birth in women who have no signs of infection or inflammation of the lower genital tract. Women with a short cervical length < 25 mm between 16 and 26 weeks of gestation were randomly assigned to receive a Shirodkar cerclage (n=34), McDonald cerclage (n=34), or bedrest (no cerclage) (n=30). History of preterm birth in the cerclage group was 13% (n/N =9/68) and 17% in the control group (n/N= 5/30).

Owen (2009) is a randomised controlled trial to assess cerclage to prevent recurrent preterm birth in women with short cervix. Women with prior spontaneous preterm birth <34 weeks were screened for short cervix, and randomly assigned to cerclage (n=158) or no cerclage (n=153) if cervical length was <25 mm.

Berghella (2004) is a randomised controlled trial comparing the efficacy of cerclage and bed rest versus bed rest only for the prevention of preterm birth in women with a short cervix found on transvaginal ultrasound examination. Women with at least 1 high-risk factor for preterm birth (≥ 1 preterm birth at < 35 weeks of gestation, ≥ 2 curettages, diethylstilbestrol exposure, cone biopsy, Müllerian anomaly, or twin gestation) were screened with transvaginal ultrasonography of the cervix every 2 weeks from 14 weeks of gestation to 23 weeks 6 days of gestation. Enrolment was offered to both asymptomatic women who were at high risk and who were identified to have short cervix (<25 mm) or significant funnelling (>25%) and non-screened women who were at low risk and who were identified incidentally. Women were assigned randomly to receive McDonald cerclage (n=31) or bed rest only (n=30).

To (2004) is a multicentre randomised controlled trial to investigate whether, in women with a short cervix identified by routine transvaginal scanning at 22-24 weeks' gestation, the insertion of a Shirodkar suture reduces early preterm delivery. Cervical length was measured in 47.123 women. The cervix was 15 mm or less in 470, and 253 (54%) of these women participated in the study and were randomised to cervical cerclage (n=127) or to expectant management (n=126). Primary outcome was the frequency of delivery before 33 completed weeks (231 days) of pregnancy.

Althuisius (2001) is a randomised controlled trial that compared preterm delivery rates (before 34 weeks of gestation) and neonatal morbidity and mortality in patients with risk factors or symptoms of cervical incompetence managed with therapeutic McDonald cerclage and bed rest (n=19) versus bed rest alone (n=16). Risk factors for cervical incompetence included previous preterm delivery before 34 weeks of gestation that met clinical criteria for the diagnosis of cervical incompetence, previous preterm premature rupture of membranes before 32 weeks of gestation, history of cold knife conization, diethylstilbestrol exposure, and uterine anomaly. When a cervical length of <25 mm was measured before a gestational age of 27 weeks, a randomization for therapeutic cerclage and bed rest (cerclage group) or bed rest alone (bed rest group) was performed.

Rust (2001) is a randomised controlled trial included patients between 16 and 24 weeks of gestation with dilation of the internal os and either prolapse of membranes >25% or cervix <25 mm (history of PTB not considered). Patients were randomly assigned to McDonald cerclage (n=55) or no cerclage (n=58). 

Before randomization, all patients were treated identically with an amniocentesis, multiple urogenital cultures, and therapy with indomethacin and clindamycin for 48 to 72 hours.

Keeler (2009) included patients with a short cervix (≤25 mm) on transvaginal ultrasound between 16 and 24 weeks’ gestation.  Patients were randomly assigned to treatment with McDonald cerclage (n=42) or weekly intramuscular injections of 17 ahydroxyprogesterone caproate (17OHP-C, 250mg/week) (n=37). The primary outcome of this study was spontaneous preterm birth (PTB) prior to 35 weeks’ gestation. A post hoc analysis of patients with a prior PTB was performed.

Results

1.1 Comparison Cerclage versus no cerclage

Outcome 1- preterm birth < 28 weeks

1. Preterm birth < 28 weeks, cervical length ≤25mm (previous preterm birth not considered)

Jarde (2019) did not report the outcome preterm birth < 28 weeks.

2. Preterm birth < 28 weeks, CL < 25mm, history of preterm birth ≥1

Conde-Agudelo (2018) included 5 studies (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Owen 2009) that reported the outcome preterm birth <28 weeks. The use of cerclage in women with a singleton gestation, previous spontaneous preterm birth, and a cervical length <25 mm in the midtrimester was associated with a lower risk of preterm birth <28 weeks of gestation (pooled RR 0.64, 95% CI 0.43 to 0.96)

c. Preterm birth <28 weeks, cervical length <25mm, no history of preterm birth

Berghella (2017) included 5 studies (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Otsuki, 2016) that reported the outcome preterm birth < 28 weeks. Preterm birth <28 weeks was reported in 11.6% (n/N= 26/224) in the cerclage group and in 11.3% (n/N=22/195) in the control group (pooled RR 1.55, 95%CI 0.68 to 1.93).

Outcome 2- Preterm birth <34 weeks

1. Preterm birth < 34 weeks, cervical length ≤25mm (previous preterm birth not considered)

The systematic review of Jarde (2019) included 2 studies (Althuisius, 2001; Otsuki, 2016) (n=136) that reported the outcome preterm birth <34 weeks. Preterm birth < 34 weeks was reported in 7/87 in the group cerclage and 11/49 in the control group (pooled RR 0.28 95%CI 0.02 to 4.19).

Figure 5.1 Outcome spontaneous preterm birth <34 weeks comparison: cerclage versus no cerclage in women with singleton gestations and a cervical length ≤25 mm

Source: Jarde, 2019, Z: p-value of the pooled effect; df: degrees of freedom; I²: statistical heterogeneity; CI: confidence interval.

2. Preterm birth <32 weeks, cervical length <25mm, history of preterm birth ≥1

Conde-Agudelo (2018) did not report the outcome preterm birth < 34 weeks, therefore the outcome < 32 weeks is described. Five trials (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Owen 2009) reported the outcome preterm birth <32 weeks. The use of cerclage in women with a singleton gestation, previous spontaneous preterm birth, and a cervical length <25 mm in the midtrimester was associated with a lower risk of preterm birth <32 weeks (pooled RR 0.66 95% CI 0.48 to 0.91), with 19% (n/N=48/250) in the cerclage group and 30% (n/N=75/254) in the control group.

3. Preterm birth <34 weeks, cervical length <25mm, no history of preterm birth

Berghella (2017) included five studies (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Otsuki, 2016) that reported the outcome preterm birth < 34 weeks. There was a small difference in the risk of preterm birth with 20.1% (n/N= 45/224) in the group cerclage and 25.1% (n/N=49/195) in the group no cerclage (pooled RR 0.89, 95%CI 0.63 to 1.27).

Outcome 3- preterm birth < 37 weeks

1. Preterm birth < 37 weeks, cervical length <25mm (preterm birth not considered)

The systematic review of Jarde (2019) included a single trial (n=101) that reported the outcome preterm birth <37 weeks (Otsuki, 2016). Otsuki (2016) reported a small difference in preterm birth <37 weeks between the group cerclage and the control group (RR 0.87 95%CI 0.46 to 1.68).

Figure 5.2 Outcome spontaneous preterm birth <37 weeks comparison: cerclage versus no cerclage in women with singleton gestations and a cervical length ≤25 mm

Source: Jarde, 2019, Z: p-value of the pooled effect; df: degrees of freedom; I²: statistical heterogeneity; CI: confidence interval.

2. Preterm birth < 37 weeks, CL < 25mm, history of preterm birth ≥1

Conde-Agudelo (2018) included 5 studies (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Owen 2009) that reported on the outcome preterm birth <37 weeks. The use of cerclage in women with a singleton gestation, previous spontaneous preterm birth, and a cervical length <25 mm in the midtrimester was associated with a lower risk of preterm birth <37 weeks (pooled RR 0.70 95% CI, 0.58-0.83), with 42% (105/250) in the cerclage group and 61% (154/254) in the control group.

3. Preterm birth <37 weeks, cervical length <25mm, no history of preterm birth

Berghella (2017) included 5 studies (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Otsuki, 2016) that reported the outcome preterm birth < 37 weeks. Preterm birth <37 weeks was reported in 36.2% (n/N= 481/224) in the cerclage group and in 41.0% (n/N=80/195) in the control group (pooled RR 0.93, 95%CI 0.73 to 1.18). This is no clinically relevant difference.

Outcome 4-  Neonatal death

1. Neonatal death, cervical lenght ≤ 25mm (preterm birth not considered)

The systematic review of Jarde (2019) included 3 studies (n=389) that reported the outcome neonatal death (Althuisius, 2001; To, 2005; Otsuki, 2016). Neonatal death was reported in 2% (n/N=5/214) in the cerclage group and in 5% (n/N=9/175) in the control group (pooled RR 0.56, 95%CI 0.19 to 1.67).

Figure 5.3 Outcome neonatal death comparison: cerclage versus no cerclage in women with singleton gestations and a cervical length ≤25 mm

Source: Jarde, 2019, Z: p-value of the pooled effect; df: degrees of freedom; I²: statistical heterogeneity; CI: confidence interval.

2. Perinatal mortality, cervical length <25mm, history of preterm birth

Conde-Agudelo (2018) included 5 studies (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Owen 2009) that reported the outcome perinatal mortality. Perinatal mortality was reported in 9% (22/250) in the cerclage group and in 14% (35/254) in the control group (pooled RR 0.65, 95%CI 0.40 to 1.07).

3. Neonatal death, cervical length <25mm, no history of preterm birth

Berghella (2017) included 4 studies (Rust, 2001; Berghella, 2004; Althuisius, 2001; Otsuki, 2016) that reported the outcome neonatal death. Neonatal death was reported in 5.9% (n/N= 7/118) in the cerclage group and in 6.5% (n/N=6/92) in the control group (pooled RR 1.08, 95%CI 0.41 to 2.86).

Outcome 5- Composite outcome neonatal mortality/morbidity

1. Composite outcome neonatal mortality/morbidity, cervical length <25mm, history of preterm birth not considered

Not reported

2. Composite outcome neonatal mortality/morbidity, cervical length <25mm, history of preterm birth

Conde-Agudelo (2018) included 5 studies that reported the outcome perinatal morbidity and mortality (Rust, 2001; Berghella, 2004; Althuisius, 2001; To, 2004; Owen 2009). The composite outcome  was defined as the occurrence of any of the following neonatal morbidities; respiratory distress syndrome; grade III/IV intraventricular haemorrhage; necrotizing enterocolitis; neonatal sepsis; bronchopulmonary dysplasia, or perinatal death. The use of cerclage in women with a singleton gestation, previous spontaneous preterm birth, and a cervical length <25 mm in the midtrimester was associated with a significantly lower risk of composite perinatal morbidity and mortality with 16% in the group cerclage (n/N=39/250), and   25% in the group no cerclage (n/N=63/254) (RR 0.64, 95% CI 0.45 to 0.91).

3. Composite outcome neonatal mortality/morbidity, cervical length <25mm, no history of preterm birth

Not reported

Outcome 6- Respiratory distress syndrome

1. Respiratory distress syndrome, cervical length <25mm, history of preterm birth not considered

Not reported

2. Respiratory distress syndrome, cervical length <25mm, history of preterm birth

Conde-Agudelo (2018) included 4 studies […] that reported the outcome RDS. RDS was reported in 6% (13/207) in the cerclage group and in 11% (21/196) in the control group (pooled RR 0.61, 95%CI 0.32 to 1.19).

3. Respiratory distress syndrome, cervical length <25mm, no history of preterm birth

Berghella (2017) included 2 studies (Berghella, 2004; Althuisius, 2001) that reported the outcome respiratory distress syndrome. RDS was reported in 14.3% (n/N= 2/14) in the cerclage group and in 12.5% (n/N=2/16) in the group no cerclage (pooled RR 1.33 95%CI 0.23 to 7.74).

Outcome 7- Necrotizing enterocolitis

1. NEC, cervical length <25mm, history of preterm birth not considered

Not reported

2. NEC,  cervical length <25mm, history of preterm birth

Conde-Agudelo (2018) included 4 studies […] that reported the outcome NEC. NEC was reported in 1% (1/207) in the cerclage group and in 1% (2/196) in the control group (pooled RR 0.62 95%CI 0.08 to 4.67).

3. NEC,  cervical length <25mm, no history of preterm birth

Berghella (2017) included 2 studies (Berghella, 2004; Althuisius, 2001) that reported the outcome NEC. There were no cases of NEC reported in both groups; 0% (n/N= 0/14) in the group cerclage and 0% (n/N=0/16) in the group no cerclage.

Outcome 8- Intraventricular haemorrhage 

1. Intraventricular haemorrhage, cervical length <25mm, history of preterm birth not considered

Jarde (2019) included 1 study (To, 2004) that reported the outcome intraventricular haemorrhage/periventricular haemorrhage. To (2004) reported 1/123 (1%) IVH/PVH in the cerclage group and 2/121 (2%) IVH/PVH in the expectant group (RR 0.49 95%CI 0.04 to 5.45, p=0.620).

2. Intraventricular haemorrhage, cervical length <25mm, history of preterm birth

Conde-Agudelo (2018) included 4 studies […] that reported the outcome Grade III/IV intraventricular haemorrhage. There was no statistically significant difference in IVH between the group cerclage 0% (0/207) and the control group 2% (4/196) (pooled RR 0.28 95%CI 0.05 to 1.64).

3. Intraventricular haemorrhage, cervical length <25mm, no history of preterm birth

Berghella (2017) included 2 studies (Berghella, 2004; Althuisius, 2001) that reported the outcome IVH. There was no significant difference in the risk of IVH with 7.1% (n/N= 1/14) in the group cerclage and 0% (n/N=0/16) in the group no cerclage (RR 3.90 95%CI 0.18 to 85.93).

Outcome 9- Neonatal sepsis

1. Neonatal sepsis, cervical length <25mm, history of preterm birth not considered

Not reported

2. Neonatal sepsis, cervical length <25mm, history of preterm birth

Conde-Agudelo (2018) included 4 studies […] that reported the outcome neonatal sepsis. Neonatal sepsis was reported in 4% (8/207) in the cerclage group and in 9% (17/196) in the control group (Pooled RR 0.47 95%CI 0.21 to 1.05).

3. Neonatal sepsis, cervical length <25mm, no history of preterm birth

Berghella (2017) included 2 studies (Berghella, 2004; Althuisius, 2001) that reported the outcome neonatal sepsis. Neonatal sepsis was reported in 14.3% (n/N= 2/14) in the cerclage group and in 12.5% (n/N=2/16) in the group no cerclage (RR 1.33 95%CI 0.23 to 7.74).

Level of evidence of the literature

Studies with a randomized, placebo-controlled design start at a high GRADE, post hoc analysis are considered observational data and start at a low GRADE. 

Outcome 1. Spontaneous preterm birth < 28 weeks

The level of evidence regarding the outcome measure spontaneous preterm birth <28 weeks in women with a cervical length ≤25 mm, and history of preterm birth started low and was downgraded by to a very low GRADE because of imprecision (-1) (confidence interval crossing the level of clinical relevance).

The level of evidence regarding the outcome measure spontaneous preterm birth <28 weeks in women with a cervical length ≤25 mm, and no history of preterm birth was downgraded to a low GRADE because of risk of bias and imprecision (-2) (wide confidence interval).

Outcome 2. Spontaneous preterm birth < 34 weeks

The level of evidence regarding the outcome measure spontaneous preterm birth <34 weeks in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of heterogeneity (-1) and imprecision (-2) (wide confidence intervals, and low number of events).

The level of evidence regarding the outcome measure spontaneous preterm birth <32 weeks in women with a cervical length ≤25 mm, and history of preterm birth  was downgraded by to a low GRADE because of risk of bias and imprecision (-1) (wide confidence intervals).

The level of evidence regarding the outcome measure spontaneous preterm birth <34 weeks in women with a cervical length ≤25 mm, and no history of preterm birth was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 3. Spontaneous preterm birth < 37 weeks

The level of evidence regarding the outcome measure spontaneous preterm birth <37 weeks in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by two levels to a low GRADE because of imprecision (-2) (single study, wide confidence intervals, and low number of events).

The level of evidence regarding the outcome measure spontaneous preterm birth <37 weeks in women with a cervical length ≤25 mm, and history of preterm birth was downgraded to a low GRADE because of risk of bias and imprecision (-1) (low number of events).

The level of evidence regarding the outcome measure spontaneous preterm birth <37 weeks in women with a cervical length ≤25 mm, and no history of preterm birth was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 4. Neonatal death

The level of evidence regarding the outcome measure neonatal death in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals, and low number of events).

The level of evidence regarding the outcome measure perinatal mortality in women with a cervical length ≤25 mm, and history of preterm birth was downgraded to a very low GRADE because of risk of bias (-1) and imprecision (-2) (wide confidence intervals and low number of events).

The level of evidence regarding the outcome measure neonatal death in women with a cervical length ≤25 mm, and no history of preterm birth was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals and low number of events).

Outcome 5. Composite outcome neonatal mortality/morbidity

The level of evidence regarding the outcome measure composite outcome neonatal mortality/morbidity in women with a cervical length ≤25 mm, and previous preterm birth started low and was downgraded to a very low GRADE because of imprecision (-1) (confidence interval crossing the borders of clinical relevance).

Outcome 6. Respiratory distress syndrome

The level of evidence regarding the outcome measure respiratory distress syndrome  in women with a cervical length ≤25 mm, and previous preterm birth started low and was downgraded to a very low GRADE because of imprecision (-1) (confidence interval crossing the borders of clinical relevance).

The level of evidence regarding the outcome measure respiratory distress syndrome  in women with a cervical length ≤25 mm, and no previous preterm birth started low and was downgraded to a very low GRADE because of risk of bias (-1) and imprecision (-2) (low number of events, confidence interval crossing the borders of clinical relevance).

Outcome 7- Necrotizing enterocolitis

The level of evidence regarding the outcome measure necrotizing enterocolitis  in women with a cervical length ≤25 mm, and previous preterm birth started low and was downgraded to a very low GRADE because of imprecision (-2) (low number of events).

No level of evidence was determined for the outcome measure necrotizing enterocolitis in women with cervical length ≤25 mm, and no previous preterm (no GRADE); no cases of NEC reported in both groups (cerclage versus no cerclage). 

Outcome 8. Intraventricular haemorrhage 

The level of evidence regarding the outcome measure intraventricular haemorrhage in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by two levels to a very low GRADE because of imprecision (-3) (one trial. wide confidence intervals, and low number of events).

The level of evidence regarding the outcome measure intraventricular haemorrhage in women with a cervical length ≤25 mm, and history of preterm birth was downgraded to a very low GRADE because of risk of bias (-1) and imprecision (-2) (wide confidence intervals and low number of events).

The level of evidence regarding the outcome measure intraventricular haemorrhage in women with a cervical length ≤25 mm, and no history of preterm birth was downgraded by two levels to a very low GRADE because of imprecision (-3) (wide confidence intervals and low number of events).

Outcome 9. Neonatal sepsis

The level of evidence regarding the outcome measure neonatal sepsis in women with a cervical length ≤25 mm, and history of preterm birth was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals and low number of events).

The level of evidence regarding the outcome measure neonatal sepsis in women with a cervical length ≤25 mm, and no history of preterm birth was downgraded to a very low GRADE because of risk of bias (-1) and imprecision (-2) (wide confidence intervals and low number of events).

1.2 Cerclage versus progesterone

A single study (Keeler, 2009) was included in the systematic review of Jarde (2019) that compared treatment with cerclage to progesterone in women with a short cervix (≤25 mm).

Outcome 1- Spontaneous preterm birth <28 weeks

Keeler (2009) did not report the outcome spontaneous preterm birth <28 weeks

Spontaneous preterm birth prior to 28 weeks' gestation occurred in 23.8% (10/42) in the cerclage group and in 18.9% (7/37) of the 17OHP-C group (RR 0.79 95% CI 0.34 to 1.88).

Outcome 2- Spontaneous preterm birth <34 weeks

Keeler (2009) did not report the outcome spontaneous delivery before 34 weeks of gestation, therefore the outcome preterm birth before 32 weeks is described.

PTB <32 weeks, cervical length <25mm, history of preterm birth not considered

Spontaneous PTB prior to 35 weeks' gestation occurred in 35.7% (15/42) in the cerclage group and in 35.1% (13/37) of the 17OHP-C group (RR 0.98 95% CI 0.54 to 1.79)

Outcome 3- Spontaneous preterm birth <37 weeks

One trial (Keeler, 2009) reported the spontaneous preterm birth prior to 37 weeks' gestation. PTB <37 weeks occurred in 52% (22/42) in the cerclage group and in 63% (22/35) in the 17OHP-C group (RR 0.75 95%CI 0.31 to 1.83).

1. PTB <35 weeks, cervical length <25mm, history of preterm birth not considered

Spontaneous PTB prior to 35 weeks' gestation occurred in 38.1% (16/42) in the cerclage group and in 43.2% (16/37) of the 17OHP-C group (RR 1.14 95% CI 0.67 to 1.93).

2. PTB < 35 weeks, cervical length <25mm, history of preterm birth

A post hoc analysis of patients with a prior PTB showed no difference in spontaneous PTB <35 weeks between groups (no data reported, Keeler, 2009).

Outcome 4- Perinatal mortality

Keeler (2009) reported the outcome perinatal death, which included any stillbirth or neonatal death during the study period. Perinatal death was reported in 11.9% in the cerclage group (n/N=5/42) and in 10.8% in the group 17OHP-C (n/N=4/37) (RR 1.10 95% CI 0.32 to 3.80).

Outcome 5- Composite outcome neonatal morbidity

Keeler (2009) reported the outcome severe morbidity defined as life threatening morbidity including respiratory distress syndrome requiring mechanical ventilation 24h, intraventricular haemorrhage, neonatal sepsis, or necrotizing enterocolitis.

Severe morbidity was reported in 21.4% in the cerclage group (n/N=9/42) and in 18.9% in the group 17OHP-C (n/N=7/37) (RR 1.17 95%CI 0.39 to 3.53).

Level of evidence of the literature

Studies with a randomized, placebo-controlled design start at a high GRADE.

The level of evidence regarding the outcome measure spontaneous preterm birth <34 weeks in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of risk of bias (-1) (early termination of the trial), indirectness (preterm birth <32 weeks was reported), and  imprecision (-2) (single trial, n=77, wide confidence intervals).

The level of evidence regarding the outcome measure spontaneous preterm birth <37 weeks in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of risk of bias (-1), and imprecision (-2) (single trial, n=77, wide confidence intervals).

The level of evidence regarding the outcome measure spontaneous preterm birth <37 weeks in women with a cervical length ≤25 mm, and history of preterm birth started low (posthoc analysis) and was downgraded to a very low GRADE because of risk of bias (-1), indirectness (-1) (outcome <35 weeks was reported) and imprecision (-2) (single trial, wide confidence intervals).

The level of evidence regarding the outcome measure perinatal mortality in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of risk of bias (-1), and imprecision (-2) (single trial, n=77, wide confidence intervals).

The level of evidence regarding the outcome measure composite outcome neonatal morbidity in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of risk of bias (-1), and imprecision (-2) (single trial, n=77, wide confidence intervals).

PICO-2 Pessary compared to no pessary/cerclage/progesterone

Description of studies

The systematic review and meta-analysis of Conde-Agudelo (2020) is a Cochrane review  that searched MEDLINE, EMBASE, POPLINE, LILACS, CINAHL, the Cochrane Central Register of Controlled Trials, clinical trial registries (all from their inception to October 31,

2019) without language restriction. The review included RCTs comparing cervical pessary with standard care (no pessary) or alternative interventions (such as vaginal progesterone or cervical cerclage) in asymptomatic women at high risk for preterm birth (such as those with a midtrimester sonographic short cervix, history of preterm birth, multiple gestation, and uterine anomalies or excisional cervical procedures) with the aim of preventing preterm birth and/or adverse perinatal outcomes. In the literature analysis six studies from this systematic review are described that included and reported data of women with a cervical length ≤25mm (Goya, 2012; Hui, 2013; Nicolaides, 2016a; Saccone, 2017;  Dugoff, 2018; Cruz-Melguizo, 2018).

In three studies (Dugoff, 2018; Nicolaides, 2016a; Saccone 2017) comparing pessary to no pessary, the intervention and control included concomitant use of progesterone in varying percentages of the study population (44-91%). 

Pessary compared to no pessary

Goya (2012) is a randomized controlled trial, that included women undergoing cervical length screening at 18-22 weeks of gestation and allocated those with a cervical length ≤25 mm to the Arabin pessary (n=192) or expectant management (n=193). At randomization the mean cervical length was 19 mm and mean gestational age 22.3 weeks. Overall, 11% of women had a prior preterm birth.

Hui (2013) is a randomized controlled trial that enrolled women with a singleton gestation and a cervical length <25 mm (at 20 – 24 weeks) and randomly allocated them to Arabin pessary (n=53) or no pessary (n=55). At randomization the mean cervical length was 20.1 mm and the mean gestational age 21.9 weeks. There was no concomitant use of progesterone in either group.

Nicolaides (2016), a multinational study, enrolled 924 women across 16 hospitals and studied women with a cervical length of ≤25 mm (at 20-25 weeks) of gestation randomly allocated to pessary (n=460) or expectant management (n= 464). The mean cervical length at study entry was 20 mm. Women with a cervical length <15 mm were also treated with 200 mg of vaginal progesterone nightly, regardless of treatment group assignment, with 45% of overall participants receiving progesterone (concomitant use of vaginal progesterone in the pessary group 44%, and 47% in the no pessary group).

Overall, 17% of the participants had a history of preterm birth. While per protocol the pessary was to remain in place until 37 weeks of gestation, in 47 women (11%), the pessary was removed prior to 34 weeks of gestation because of vaginal discharge, discomfort, or bleeding.

Saccone (2017) was a randomized controlled trial that enrolled women with a singleton gestation, no previous spontaneous preterm birth, and a cervical length ≤25 mm (at 18 – 24 weeks), and randomly allocated them to Arabin pessary (n=150) or no pessary (n=150). At randomization the mean cervical length was 12.0 mm and the mean gestational age 22.4 weeks. Concomitant use of progesterone in the pessary group was 89% and 83% in the no pessary group.

Dugoff  (2018) was a randomized controlled trial and enrolled women with a singleton gestation, no previous spontaneous preterm birth, and a cervical length ≤25 mm (at 18 – 24 weeks), and randomly allocated them to Bioteque cup pessary (n=60) or no pessary (n=58). At randomization the median cervical length was 17.6 mm in the pessary group and 19.0 in the no pessary group, and mean gestational age 21.1 weeks. Concomitant use of progesteron in the pessary group was 84%, and 91% in the no pessary group.

Pessary compared to progestrone

Cruz-Melguizo (2018) is a randomized non-inferiority trial that compared the efficacy of Arabin pessary (n=125) and vaginal progesterone 200 mg/day (n=118) in women with a singleton gestation and a cervical length ≤25 mm at 19-22 weeks of gestation.

Concomitant use of vaginal progesterone in the pessary group was 5%.

Results

Pessary compared to no pessary

Outcome 1 -Spontaneous preterm birth <28 weeks

Three of the included studies in the review of Conce-Agudelo (2020) reported the outcome spontaneous preterm birth <28 weeks (Nicolaides, 2016; Saccone, 2017, Dugoff, 2018, n=4160). Spontaneous preterm birth <28 weeks with the use of pessary was reported in 5.1% (44/865) compared to no pessary 6.0% (52/865) (pooled RR 0.76 95%CI 0.37 to 1.54, p= 0.44, I²=65%).

Outcome 2- spontaneous preterm birth <34 weeks

Five of the included studies in the review of Conde-Agudelo (2020) reported the outcome spontaneous preterm birth <34 weeks (Goya, 2012; Hui, 2013; Nicolaides, 2016a; Saccone, 2017; Dugoff, 2018). There was no significant reduction of the risk of spontaneous preterm birth < 34 weeks with the use of pessary compared to no pessary in women with a singleton pregnancy with a cervical length ≤25 mm (pooled data 11% pessary group compared to 15% no pessary group; RR 0.72 95% CI 0.36 to 1.45, P=0.36, I²=84%) (Figure 5.4)

Figure 5.4. Outcome spontaneous preterm birth <34 weeks comparison: cervical pessary versus no pessary in women with singleton gestations and a cervical length ≤25 mm

Source: Conde-Agudelo, 2020, Z: p-value of the pooled effect; df: degrees of freedom; I²: statistical heterogeneity; CI: confidence interval.

1.1 Subgroup obstetric history- no previous preterm birth

Three studies reported the outcome measure spontaneous preterm birth < 34 weeks in a (subgroup of) women with no previous preterm birth (Nicolaides, 2016; Saccone, 2017; Dugoff, 2018). Spontaneous preterm birth < 34 was reported in 12.9% (78/605) in the group pessary, and 12.2% (72/591) in the group no pessary (RR 0.97 95%CI 0.54 to 1.76).

1.2   Subgroup obstetric history –  ≥1 previous preterm birth

One study reported the outcome measure spontaneous preterm birth < 34 weeks in a subgroup women with previous preterm birth (Nicolaides, 2016). Spontaneous preterm birth < 34  was reported 10.0% (7/70) in the group pessary, and 19.0% (16/84) in the group no pessary (RR 0.53 95%CI 0.23 to 1.20).

Outcome 3- Spontaneous preterm birth < 37 weeks

Four of the included studies in the review of Conde-Agudelo (2020) reported the outcome spontaneous preterm birth <37 weeks (Goya, 2012; Nicolaides, 2016; Saccone, 2017, Dugoff, 2018).

Spontaneous preterm birth <37 weeks with the use of pessary compared to no pessary in women with a singleton pregnancy and cervical length <25mm, was reported in 22.7% (196/865) in the pessary group and in 32.6% (282/865) in the no pessary group (pooled RR 0.71 95% CI 0.41 to 1.24, favouring pessary, p= 0.23, I²=91%)

Outcome 4. Perinatal death

Conde-Agudelo (2020) included five trials (Goya, 2012; Hui, 2013; Nicolaides, 2016; Saccone, 2017, Dugoff 2018) that reported on the outcome perinatal death. Perinatal death was reported in 1.3% (12/918) in the pessary group and 1.6% (15/920) in the control group (pooled RR 0.81 95%CI 0.38 to 1.74, p=0.59) (Figure 5.5).

Figure 5.5. Outcome perinatal death: cervical pessary versus no pessary in women with singleton gestations and a cervical length ≤25 mm

Source: Conde-Agudelo, 2020, Z: p-value of the pooled effect; df: degrees of freedom; I²: statistical heterogeneity; CI: confidence interval.

Outcome 5. Composite outcome neonatal mortality/morbidity

Four studies reported a composite outcome (Goya 2012; Nicolaides, 2016; Saccone, 2017; Dugoff, 2018), defined as any composite adverse neonatal outcome (Conde-Agudelo, 2020).

Composite outcome was reported in the group pessary in 8.0% (69/865) and in the group no pessary in 13.2% (114/865) (pooled RR 0.59 95%CI 0.28 to 1.27, p=0.18, I²=83).

Outcome 6. Respiratory distress syndrome

Conde-Agudelo (2020) included 5 trials (Goya, 2012; Hui, 2013; Nicolaides, 2016; Saccone, 2017; Dugoff, 2018) that reported on the outcome respiratory distress syndrome.

Respiratory distress syndrome was reported in 6.8% (62/918) in the pessary group and 9.8% (90/920) in the control group (pooled RR 0.72 95% CI 0.36 to 1.43, p=0.35). 

Outcome 7. Necrotizing enterocolitis

Conde-Agudelo (2020) included 4 trials (Goya, 2012; Nicolaides, 2016; Saccone, 2017; Dugoff, 2018) that reported on the outcome necrotizing enterocolitis.

Necrotizing enterocolitis was reported in 1.3% (11/865) in the pessary group and 1.2% (10/856) in the control group (pooled RR 1.15 95% CI 0.47 to 2.79, p=0.76). 

Outcome 8. Intraventricular haemorrhage 

Conde-Agudelo (2020) included 5 trials (Goya, 2012; Hui, 2013; Nicolaides, 2016; Saccone, 2017; Dugoff, 2018) that reported on the outcome intraventricular haemorrhage.

Intraventricular haemorrhage was reported in 1.9% (17/918) in the pessary group and 1.5% (14/920) in the control group (pooled RR 1.16 95% CI 0.48 to 2.80, p=0.73). 

Outcome 9. Neonatal sepsis

Conde-Agudelo (2020) included 5 trials (Goya, 2012; Hui, 2013; Nicolaides, 2016; Saccone, 2017; Dugoff, 2018) that reported on the outcome neonatal sepsis.

Neonatal sepsis was reported in 5.3% (49/918) in the pessary group and 6.1% (56/920) in the control group (pooled RR 0.80 95% CI 0.46 to 1.40, p=0.44).

Level of evidence of the literature

Studies with a randomized, placebo-controlled design start at a high GRADE.

Outcome 1.

The level of evidence regarding the outcome measure spontaneous preterm birth <28 weeks in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 2.

The level of evidence regarding the outcome measure spontaneous preterm birth <34 weeks in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of inconsistency (-1), and imprecision (-2) (wide confidence intervals).

The level of evidence regarding the outcome measure spontaneous preterm birth <34 weeks in women with a cervical length ≤25 mm, no history of preterm birth was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

The level of evidence regarding the outcome measure spontaneous preterm birth <34 weeks in women with a cervical length ≤25 mm, and a history of >=1 preterm birth was downgraded by two levels to a  low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 3.

The level of evidence regarding the outcome measure spontaneous preterm birth <37 weeks in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a low GRADE because of inconsistency (-1) and imprecision (-2) (wide confidence intervals).

Outcome 4.

The level of evidence regarding the outcome measure perinatal death in women with a cervical length ≤25 mm was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 5.

The level of evidence regarding the composite outcome measure neonatal mortality/morbidity in women with a cervical length ≤25 mm was downgraded by three levels to a very low GRADE because of inconsistency (-1) and imprecision (-2) (wide confidence intervals).

Outcome 6.

The level of evidence regarding the outcome measure respiratory distress syndrome in women with a cervical length ≤25 mm was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 7.

The level of evidence regarding the outcome measure necrotizing enterocolitis in women with a cervical length ≤25 mm was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 8.

The level of evidence regarding the outcome measure intraventricular hemorrhage in women with a cervical length ≤25 mm was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Outcome 9.

The level of evidence regarding the outcome measure neonatal sepsis in women with a cervical length ≤25 mm was downgraded by two levels to a low GRADE because of imprecision (-2) (wide confidence intervals).

Pessary compared to progesterone

A single trial (n=246) was included in the review of Conde-Agudelo (2020) that compared pessary to progesterone (Cruz-Melguizo, 2018).

Outcome -1 spontaneous preterm birth <28 weeks

Cruz-Melguizo (2018) reported no difference between the pessary group (8.0%, n/N=10/125) and the progesterone group (7.6%, N/N=9/118) (RR 1.05, 95% CI 0.44- 2.49).

Outcome-2 spontaneous preterm birth <34 weeks

Cruz-Melguizo (2018) reported no difference between the pessary group and the progesterone group in the rate of spontaneous delivery before 34 weeks of gestation with 14.2% (18/127) in the pessary group and 14.3% (17/119) in the progesterone group (RR 0.99, 95% CI 0.54 to 1.83).  The rate of spontaneous delivery before 34 weeks of gestation was 14% (n=18/127) in the pessary group and 14% (n=17/119) in the progesterone group with a risk difference of -0.11% (95% CI -8.85% to 8.62%; P=.99), that is, noninferiority was not shown for the pessary.

Outcome -3 spontaneous preterm birth <37 weeks

Cruz-Melguizo (2018) reported no difference between the pessary group (22%, n/N=27/125) and the progesterone group (21%, N/N=25/118) (RR 1.02, 95% CI 0.63-1.65).

Outcome -4 perinatal death

Cruz-Melguizo (2018) reported fetal and neonatal death in 5% (6/125) in the pessary group and 3% (3/118) the progesterone group (RR 1.89, 95% CI 0.48-7.38).

Outcome -5 Composite adverse neonatal outcome

Cruz-Melguizo (2018) reported no difference between the pessary group (19%, n/N=24/125) and the progesterone group (17%, n/N=20/118) (RR 1.13, 95% CI 0.66-1.94).

Level of evidence of the literature

Studies with a randomized, placebo-controlled design start at a high GRADE.

The level of evidence regarding the outcome measure spontaneous preterm birth (<37, <34, <28 weeks) in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of risk of bias (-1) (no intention to treat analysis) and imprecision (-2) (1 trial, n=246 participants, wide confidence intervals).

The level of evidence regarding the outcome measure perinatal death in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of risk of bias (-1) (no intention to treat analysis) and imprecision (-2) (1 trial, n=246 participants, wide confidence intervals).

The level of evidence regarding the outcome measure composite adverse neonatal outcome in women with a cervical length ≤25 mm (preterm birth not considered) was downgraded by three levels to a very low GRADE because of risk of bias (-1) (no intention to treat analysis) and imprecision (-2) (1 trial, n=246 participants, wide confidence intervals).

A systematic review of the literature was performed to answer the following questions:

1. What are the (un)favourable effects of cerclage compared to no cerclage, pessary or progesterone in the prevention of spontaneous preterm birth in woman with singleton pregnancies and a cervical length ≤25 mm with or without a history of spontaneous preterm birth <37 weeks?

2. What are the (un)favourable effects of pessary compared to no pessary, cerclage or progesterone in the prevention of spontaneous preterm birth in woman with singleton pregnancies and a cervical length ≤25 mm with or without a history of spontaneous preterm birth <37 weeks?

Relevant outcome measures

The guideline development group considered preterm birth <34 weeks, composite outcome of neonatal morbidity and mortality, and neonatal death as critical outcome measures for decision making;

A priori, the working group did define the outcome measures preterm birth: i.e., preterm preterm birth <28 weeks, preterm birth <34 weeks, and preterm birth <37 weeks. When a included study did not report the effect of progesterone on preterm birth <34 weeks, but did report preterm birth <32 weeks or preterm birth <33 weeks, if possible those results were included. The same method was used considering outcome measure preterm birth <37 weeks. If possible, the results of preterm birth <35 weeks or <36 weeks were included.

A priori, the working group did define the outcome measures neonatal morbidity and mortality as a composite outcome measure including respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, proven neonatal sepsis, neonatal death. The working group also reported these outcome measures individually.

The outcome measures neonatal death, perinatal mortality, respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, proven neonatal sepsis were not defined a priori, but the definitions applied in the studies were used.

The working group defined a relative risk ≤0.9 or ≥1.1 as a minimal clinically important difference for preterm birth and ≤0.8 or ≥1.25 as a minimal clinically important difference for all other outcome measures.

Search and select (Methods)

The databases Medline (via OVID) and Embase (via Embase.com) were searched with relevant search terms until October, 16th 2019. The working group was aware of the publication of Conde-Agudelo (2020) who searched until October, 31th 2019. Since the included studies did not change the conclusions and therefore recommendations, the original search until October, 16th 2019 was not updated. The detailed search strategy is depicted under the tab Methods. The systematic literature search resulted in 535 hits. Studies were selected based on the following criteria: women cervical length ≤25 mm, treated with pessary, cerclage compared versus no treatment or progesterone and at least one of the relevant outcome measures reported. Only systematic reviews and meta analyses or randomized controlled trials (RCTs), involving the original study data were included. 17 studies were initially selected based on title and abstract screening. After reading the full text, 13 studies were excluded (see the table with reasons for exclusion under the tab Methods), and four studies were included.

Results

Four systematic reviews and meta-analysis were included in the analysis of the literature (Conde-Agudelo, 2020; Jarde, 2019; Conde-Agudelo; 2018; Berghella, 2017). These reviews included six studies directly comparing cerclage to no cerclage, one study comparing cerclage to progesterone, five studies comparing pessary to no pessary, and one study comparing pessary to progesterone. No studies were included that directly compared cerclage versus pessary. Important study characteristics and results are summarized in the evidence tables. The assessment of the risk of bias is summarized in the risk of bias tables.