Methods

Randomised trial. No mention of method of randomisation. Not placebo‐controlled or blinded.

Participants

82 women treatment 43 control 39. Inclusions: 20‐34 weeks' pregnant. PPROM confirmed by sterile speculum. Singleton pregnancy only.

Interventions

Treatment group: ampicillin 1 g IV every 6 hours for 24 hours. Maintained on oral 500 mg ampicillin 6‐hourly until delivery. In labour they were recommenced on 1 g IV ampicillin.

Outcomes

Death only included.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Unclear risk

No information given.

Blinding (performance bias and detection bias)
All outcomes

High risk

Not blinded.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information available.

Methods

Randomised trial ‐ no mention of the method of randomisation.

Participants

31 women with premature rupture of the membranes between 28‐34 weeks gestation. PPROM confirmed by speculum. Exclusions: women who go into active labour within 24 hours or who need induction of labour. Multiple pregnancy and fetal malformations. Women with serious medical conditions or who need antibiotic treatment for a known infection. Women who have received antibiotics in the last 10 days or who are allergic to penicillin.

Interventions

Treatment group oral ampicillin 1 g 4 x daily.
No placebo arm or tocolysis used.

Outcomes

Death only included.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Unclear risk

No information given.

Blinding (performance bias and detection bias)
All outcomes

High risk

Not blinded.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information available.

Methods

Randomised trial. Using sequentially numbered sealed envelopes. Not placebo‐controlled or blinded. The control group received IV fluids without antibiotics for first 24 hours.

Participants

94 women randomised 48 treatment, 46 control. Inclusions: singleton pregnancies 20‐34 weeks with PPROM confirmed by sterile speculum.
Exclusions: penicillin allergy. Prior antibiotic therapy. Clinical evidence of intra‐amniotic infection. Evidence of labour or fetal distress.

Interventions

Treatment: 24 hours IV ampicillin 2 g every 6 hours for 4 doses; gentamycin 90 mg loading dose 60 mg every 8 hours for 3 doses. Then oral amoxicillin + clavulanic acid. 500 mg 3 x day for 7 days. Control IV fluids without antibiotics for 24 hours.

Outcomes

Death only included.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Low risk

Using sequentially numbered sealed envelopes.

Blinding (performance bias and detection bias)
All outcomes

High risk

Not blinded.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised controlled trial.

Participants

62 women PPROM between 24 and 29 weeks' pregnant. Not stated whether multiple pregnancy included.

Interventions

Co‐amoxiclav 3 g 6‐hourly for 4 doses then co‐amoxiclav 500 mg 6‐hourly for 5 days or matching placebo.

Outcomes

Prolongation of pregnancy.
Neonatal mortality and morbidity.

Notes

Data extracted from abstract only. Further data requested from Dr Cox but not made available.
Study took place between May 1991 and April 1994 in Dallas, Texas.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Unclear risk

No information given.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double blind study.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial. A table of random numbers was used. Drugs and placebo were prepared by research nurses. The authors specify that participants and caregivers were blinded as to group.

Participants

148 women at 21‐37 weeks with premature rupture of the membranes preterm confirmed with positive nitrazine test and 'ferning' of amniotic fluid or by seeing vaginal pool of amniotic fluid from os. No tocolytics or steroids given. Multiple pregnancies included.
Exclusions are not clearly stated.

Interventions

4‐hourly IV 1 million units benzylpenicillin for 12‐24 hours ‐ oral 250 mg penicillin twice daily before delivery or a matched placebo.

Outcomes

Latency period, infection complications and neonatal
outcomes studies. Data on death not included.

Notes

Study conducted from March 2 1989 to May 29 1991, in a single site (North Carolina, USA). 148 women.
71 placebo.
77 treatment.
4 women were excluded because of protocol violation in placebo arm (antibiotics given).

Information on neonatal death not given.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Table of random numbers.

Allocation concealment (selection bias)

Low risk

Stated that nurses were not involved in the preparation or release of either antibiotic or placebo.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Patients and staff blinded.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data excluded for 4 women who were treated with antibiotics outside the protocol.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial ‐ multicentre.

Participants

105 pregnant women with PROM between 24+0 and 32+6 weeks.

Exclusion criteria not clearly stated nor whether multiple pregnancy included.

Interventions

Metzlocillin 2 g given 3 x day for 7 days or placebo.

All women given corticosteroids and tocolytics IV.

Outcomes

Prolongation of pregnancy and neonatal mortality and morbidity.

Notes

5 centres in Germany ‐ dates not given.

47 women in treatment arm and 58 in control.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Unclear risk

No information given.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial.

Participants

60 singleton pregnancy women. Preterm PROM under 36 weeks' pregnant. Ruptured membranes confirmed by sterile speculum examination, ferning test and nitrazine test.
No steroids or tocolytics given after randomisation.

Exclusions: > 37/40.
Discrepancy of over 2 standard deviations between scan and dates EDD.
Bleeding.
Contractions.
Fetal distress.
Fetal malformation.
Fetal death.
Chorioamnionitis on admission.
Antibiotics given during previous 10 days.

Interventions

Erythromycin 500 mg 6‐hourly orally until delivery. Matched placebo given until delivery.

Outcomes

Maternal morbidity. Neonatal mortality and morbidity.

Notes

60 women recruited during 1992 from single centre in Madrid, Spain.
No losses to follow‐up.

Paper in Spanish and data extracted with help from Dr Pigem.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated.

Allocation concealment (selection bias)

Unclear risk

No information given.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

60 women randomised to double blind placebo controlled trial. Randomisation based on random numbers tables with blocks providing 1:1 ratio and balancing every 6 women. Randomisation conducted in pharmacy.

Participants

60 women randomised. Inclusions <= 35 weeks with documented PPROM.
Exclusions: digital examination of cx, non‐reassuring stress test, presence of chorioamnionitis, abruptio placenta, pre‐eclampsia, multiple pregnancy and penicillin allergy.

Interventions

IV ampicillin 2 g every 6 hours for 24 hours followed by 500 mg oral ampicillin until delivery or discharge. Matched placebos.

Outcomes

Maternal morbidity. Neonatal mortality and morbidity.

Notes

Study divided into GBS positive and negative patients. Unclear whether clinician knew of positive culture.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomisation based on random numbers tables with blocks providing 1:1 ratio and balancing every 6 women.

Allocation concealment (selection bias)

Low risk

Randomisation and preparation of drugs conducted in pharmacy.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial. Randomisation table generated by consecutive coin toss, the randomisation schedule kept in pharmacy.

Participants

85 women randomised. Inclusions: mothers with singleton gestations between 20‐34 weeks with PPROM confirmed by sterile speculum for pooling, ferning and nitrazine paper testing.
Exclusions: penicillin allergy, taking antibiotics at the time of PPROM, had fever > 100.4 degrees Fahrenheit, had signs of chorioamnionitis, were in active labour (defined by 3 or more contractions per 10‐minute period for 1 hour or presented with cervical dilatation > 3 cm confirmed at the time of sterile speculum. Fetal indications for exclusion were the presence of fetal distress, defined as repetitive late deceleration or sustained bradycardia, or congenital abnormality on ultrasound.

Interventions

IV mezlocillin for 48 hours followed by oral ampicillin until delivery or matched (IV + oral) placebo.
No doses noted. After randomisation no tocolytic steroids given.
Study drugs discontinued if infection diagnosed.

Outcomes

Not clearly defined other than maternal or perinatal morbidity and mortality.
Outcomes looked at included length of pregnancy, maternal infectious morbidity, mode of delivery. Neonatal outcomes ‐ stillbirth, neonatal death, birthweight Apgar, cord pH, positive blood culture, RDS, IVH, NEC, NICU stay over 30 days.

Notes

Single centre ‐ University Medical Centre ‐ Jacksonville Florida.
85 women randomised.
All women had infection screen on admission. No digital examination allowed.
No comment as to losses to follow‐up or recruitment period.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomisation table generated by consecutive coin toss.

Allocation concealment (selection bias)

Low risk

The randomisation schedule kept in pharmacy.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Randomisation schedule stated as not being available to clinicians.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial.

Participants

4826 women under 37 weeks' pregnant with PROM. Multiple pregnancies included.

UK follow‐up at 7 years of age of the 4378 children of the 4148 eligible women who joined the ORACLE trial using a parental questionnaire. Exclusions 661 women (246 due to perinatal death, 376 randomised outside UK and 39 women withdrew).

Interventions

Co‐amoxiclav 375 mg QDS, erythromycin 250 mg QDS orally for 10 days or until delivery matched placebo (2 x 2 factorial design).

Outcomes

Primary outcome: neonatal death or abnormal brain scans on discharge from hospital or oxygenation at 36 weeks' postconceptual age.
Secondary outcomes include prolongation of pregnancy, neonatal infection, respiratory outcomes.

Functional impairment was assessed using the Mark III Multi‐Attribute Health Status classification system. Primary outcome was defined as any level of functional impairment (severe, moderate or mild). Other outcomes included death, behaviour (using the Strengths and Difficulties questionnaire) prespecified questions on respiratory symptoms, hospital admissions, convulsions, other prespecified medical conditions and demographic data. Educational attainment was evaluated for children in England using data from National Cirriculum Tests at 7 years of age (Key Stage 1).

Notes

Multicentre trial (161 centres, 135 in the UK). Randomised 4826 women. 2 women lost to follow‐up and 15 women were excluded due to protocol violations. 4809 women analysed. For twin pregnancies adverse outcomes were considered present if one twin affected. Consumers involved in drawing up of protocol and information for women.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

By computer using randomly generated blocks of 4.

Allocation concealment (selection bias)

Low risk

Sequentially numbered drug boxes of identical appearance.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated that clinicians remained blind to treatment allocation in all but 9 cases and that all staff and participants remained blind to treatment allocation.

For the follow‐up study all participants bar 1 women and all Study staff remained blind to treatment allocation.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

2 women lost to follow‐up and 15 protocol violations.

In the follow‐up study outcome data were determined for 75% of eligible children.

Selective reporting (reporting bias)

Low risk

No selective reporting.

Protocol published for follow‐up study.

Other bias

Low risk

The study appears to be free of other sources of bias.

Methods

Randomised double‐blind, placebo‐controlled trial.

Participants

101 women randomised between 23‐36 weeks' pregnant with visible leakage of amniotic fluid who did not go into labour within 12 hours of admission. Sterile speculum, digital examination and infection screening was performed on admission. Multiple pregnancies included.

Interventions

2 doses of IV penicillin (5 mu) or matched placebo.

Outcomes

Prolongation of pregnancy. Infection, neonatal morbidity and mortality. Long‐term development at 2 years.

Notes

Department of Obstetrics and Gynaecology, Helsinki, Finland.
No mention of where the study was conducted. Sealed envelope randomisation.
Results in 76 women not randomised but admitted during the same period are also reported.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Low risk

Stated as being by sealed envelope.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised trial looking at 3 or 7 days antibiotic therapy. Randomised using table of arbitrary numbers in blocks of 10. Indicator cards placed in sealed envelopes which were sequentially numbered and stored on an area away from the enrolment site.

Participants

84 singleton pregnancies were randomised between 24‐34 weeks' gestation. Exclusions included known infection, absence of cervical cerclage, not penicillin allergic. Corticosteroids given to all participants.

Interventions

Ampicillin‐sulbactam 3 g intravenously every six hours for either 3 or 7 days.

Outcomes

Primary outcome was latency period between membrane rupture and delivery. Infection and neonatal morbidity and mortality.

Notes

3 study sites in Tennessee USA.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomised using table of arbitrary numbers in blocks of 10.

Allocation concealment (selection bias)

Low risk

Indicator cards placed in sealed envelopes which were sequentially numbered and stored on an area away from the enrolment site

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated that all carers were unaware of the randomisation process.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind placebo‐controlled trial.

Participants

75 women randomised with a single fetus at 24‐34 completed weeks (accurate gestational age), admitted with PROM. No digital examination unless active labour. Women had infection screening.
Exclusions: abruption, lethal fetal abnormalities clinical chorioamnionitis, maternal illness, diabetes, PIH, lupus, severe maternal disease, fetal growth retardation, fetal distress, cervical cerclage, active herpes. Women having received antibiotics for existing infection were also excluded.

Interventions

Piperacillin 3 g IV 6‐hourly 72 hours or placebo.

Outcomes

Prolongation of pregnancy.
Neonatal outcomes.

Notes

Recruitment in 3 centres (USA) from January 1987 to January 1992. 75 women were randomised (treatment 38, placebo 37).
3 babies (1 in the experimental group and 2 in controls) were lost to follow‐up.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated randomisation sequence.

Allocation concealment (selection bias)

Low risk

Same deposited in pharmacy.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated all healthcare providers were blinded to allocation.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised trial not placebo‐controlled. Randomisation by opening sealed consecutive opaque envelopes in admission room.

Participants

171 women randomised 84 in treatment group 87 in no treatment group. Inclusion PROM 26‐36 weeks' gestation drainage of liquor confirmed by sterile speculum. Exclusions: clinical signs of chorioamnionitis, multiple pregnancy, those with any contraindication to continuing the pregnancy and those who had just completed a course of antibiotics for another reason.

Interventions

Co‐amoxiclav for 5 days. No mention of daily frequency or mg of drugs.

Outcomes

Death only included.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Low risk

Stated as being by sealed envelope.

Blinding (performance bias and detection bias)
All outcomes

High risk

Not blinded.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Minimal loss to follow‐up ‐ 2 in the treatment and 1 in the no treatment group.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information available.

Methods

Randomised double‐blind placebo‐controlled trial.
Computer‐generated random number list. Sequentially numbered bottles.

Participants

65 women with singleton pregnancies.
Women between 23‐34 completed weeks' gestation with PROM. Sterile speculum. No corticosteroids administered.
Exclusions: active labour, presence of maternal or fetal complication to necessitate delivery (fetal distress, prolapsed cord, pregnancy‐induced hypertension, abruptio placentae) placenta praevia, cervical cerclage, known infection requiring antibiotic treatment, use of vaginal or oral antibiotics in last 2 weeks, presence of known uterine or fetal abnormality, history of vaginal bleeding in last month, serious existing maternal disease, history of allergy or intolerance to erythromycin.

Interventions

Erythromycin 333 mg 3 x daily or placebo 7 days or until active labour started.

Outcomes

Prolongation of pregnancy. Maternal and neonatal infectious morbidity.

Notes

July 1986‐June 1988 University Hospital Denver.
65 women recruited (10 excluded ‐ 15%).
55 analysed ‐ (28 treatment, 27 placebo).

No replies received to letter requesting breakdown between stillbirths and neonatal deaths and asking if Blanco's paper has ever been published.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated random number list.

Allocation concealment (selection bias)

Low risk

Sequentially numbered bottles.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete after 10 exclusions.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information available.

Methods

Randomised double‐blind, placebo‐controlled trial.
Computerised random number tables. Administered by the pharmacy.
Stratified at 30 weeks. gestational age.

Participants

Inclusions: 220 women 20‐34/6 weeks pregnant with PPROM ‐ sterile speculum and evaluation of cervix. Amniocentesis done for infection screen. Multiple pregnancies included.
Exclusions: PPROM > 72 hours duration, cervical dilatation > 4 cm, progressive labour, vaginal bleeding, temperature 99 degrees Fahrenheit or greater, active infection requiring antibiotic therapy, antibiotic therapy within 1 week prior to admission, active hepatic disease, erythromycin allergy, cervical cerclage or medical condition requiring delivery. IUGR (< 10 centile), congenital abnormalities, evidence of fetal distress, unsuccessful tocolysis on admission for preterm labour.

Interventions

Oral 333 mg erythromycin. 8‐hourly from randomisation to delivery with matched placebo.

Outcomes

Not clearly stated.
Prolongation of pregnancy. Reduction of infectious morbidity.

Notes

Single centre (Memphis, Tennessee, USA).
March 1989‐August 1990.
Women had infection screen before randomisation.
220 randomised, (treatment 106, placebo 114) 3 lost to follow‐up.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computerised random number tables.

Allocation concealment (selection bias)

Low risk

Administered by the pharmacy.

Blinding (performance bias and detection bias)
All outcomes

Low risk

States that all involved remained blind to treatment allocation.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial. Urn randomisation scheme (a procedure to increase the likelihood of obtaining an equal number of subjects in each arm), stratified by centre.

Participants

614 women with PPROM at 24‐32 weeks' gestation. Inclusion criteria: membrane rupture within 36 hours of randomisation; cervical dilatation 3 cm or less on usual examination; < 5 contractions in 6 minutes.
Exclusion criteria: non‐reassuring, fetal testing; vaginal bleeding; maternal or fetal indication for delivery, cervical cerclage in place, antibiotics within the last 5 days, corticosteroids within last 7 days, allergy to penicillin or erythromycin, maternal infection or medical disease, ultrasound evidence of placenta praevia, fetal weight < 10th centile for gestational age, malformation. Previous successful tocolysis was not an exclusion criterion.
Tocolysis and corticosteroids were prohibited after randomisation.

Interventions

Ampicillin 2 g 6‐hourly and erythromycin 250 mg 6‐hourly IV for 48 hours, then oral amoxacillin 250 mg every 8 hours and erythromycin 333 mg 8‐hourly for 5 days and a matching placebo regimen.

Outcomes

Composite primary outcome included pregnancies complicated by at least 1 of the following: fetal or infant death, respiratory distress, severe intraventricular haemorrhage, stage 2 or 3 NEC, or sepsis within 72 hours of birth. These perinatal morbidities were also assessed separately and pregnancy prolongation assessed.
For twin pregnancies adverse outcomes considered present if 1 twin affected.

Notes

11 centres ‐ USA.
February 1992‐January 1995.
1867 women screened.
804 eligible.
614 agreed to participate.
29 twin gestations.
GBS positive: 118/614.
3 women lost to follow‐up.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Urn randomisation scheme (a procedure to increase the likelihood of obtaining an equal number of subjects in each arm), stratified by centre.

Allocation concealment (selection bias)

Low risk

Treatment given by pharmacy.

Blinding (performance bias and detection bias)
All outcomes

Low risk

States all involved remained blinded to treatment allocation.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Only 3 withdrawals (2 in placebo and 1 in treatment arm).

Selective reporting (reporting bias)

Unclear risk

No protocol available.

Other bias

Unclear risk

No information given.

Methods

Randomised trial not placebo controlled. RCT of antenatal steroids + ampicillin. 4 groups ‐ GP1 ‐ neither, GP2 steroids only, GP3 antibiotic only, GP4 both. Randomised by using sealed envelopes into 1 of groups.

Participants

Randomised: 41 = GP1, 43 = GP2, 37 = GP3, 44 = GP4.
Inclusion criteria 26‐34 weeks' pregnant singleton gestation. PROM confirmed by sterile speculum L/S ratio less than 2.0.
Exclusions: In labour within 12 hours of randomisation women with uterine tenderness, foul smelling lochia or fetal tachycardia on admission, women allergic to penicillin, with congenital abnormality with L/S ratio greater than 2.0 or not obtained.

Interventions

2 g IV ampicillin every 6 hours until results of cervical cultures negative.

Outcomes

Death only included.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Low risk

States as sealed envelopes into 1 of 4 groups.

Blinding (performance bias and detection bias)
All outcomes

High risk

Not blinded.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial. No comment as to method of randomisation.

Participants

88 women.
Inclusions: women with PPROM 24‐34 weeks, PPROM diagnosed with sterile speculum‐pooling, ferning and nitrazine tests.
No digital examination performed.
Exclusions: labour, significant haemorrhage, abruptio placentae, use of antibiotics within 30 days before screening for study, fetal anomaly or death, multiple gestation, documented allergy to clindamycin or gentamicin, uterine abnormality, presence of IUCD, fetal distress, clinical chorioamnionitis, maternal medical complications necessitating delivery or any condition precluding expectant management and intrauterine growth retardation (< 10th centile for gestational age).

Interventions

Clindamycin 600 mg IV every 6 hours for 48 hours + 4 mg/kg/day gentamycin IV for 48 hours followed by Clindamycin 300 mg orally every 6 hours for 5 days + gentamycin 2 mg/kg/day IM every 12 hours for 5 days.
Matching placebo.

Outcomes

Prolongation of pregnancy, maternal infection related morbidity, birthweight, neonatal morbidity and admission to neonatal intensive care unit.

Notes

November 1990‐September 1994. 3 sites: 2 Chile, 1 USA.
Women had infection screen.
88 women randomised
(treatment 42, control 46).
1 lost to follow‐up in placebo arm.
Trial stopped after intermediate evaluation showed treatment group had better outcome.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

No information given.

Allocation concealment (selection bias)

Unclear risk

No information given.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete with 1 loss to follow‐up.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

High risk

Trial stopped after intermediate evaluation showed treatment group had better outcome.

Methods

Randomised not placebo‐controlled. Randomised using sealed opaque envelopes determined by computer algorithm.

Participants

118 randomised 1 lost to follow‐up. 59 treatment 58 controls. Inclusions 24 to 34 weeks' gestation. PPROM confirmed by speculum. Exclusions in labour, clinical evidence of infection suspected fetal compromise, membrane rupture over 2 days, fetal abnormality, antibiotics in last 7 days, multiple pregnancy, cervical cerclage, prompt delivery required.

Interventions

IV 1 g ampicillin 6‐hourly for 24 hours then 500 mg ampicillin orally every 6 hours. If allergic to penicillin 500 mg erythromycin used 6‐hourly. Treatment continued with delivery or diagnosis of chorioamnionitis.

Outcomes

Death only included.

Notes

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

By computer algorithm.

Allocation concealment (selection bias)

Low risk

Sealed opaque envelope.

Blinding (performance bias and detection bias)
All outcomes

High risk

Not blinded.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete ‐ 1 woman lost to follow‐up in control group.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial of 3 or 7 days treatment. Pharmacy provided randomisation with a computer‐generated random number table in blocks of 4.

Participants

48 women randomised: 24 in each arm‐analysis on 23 in each arm. Women 24‐33 weeks with clinically confirmed ruptured membranes. Exclusions included penicillin allergy, active labour, suspected infection, multiple pregnancy, known medical maternal or fetal problems and exposure to antibiotics within 1 week before admission.

Interventions

For first 48 hours all women received parenteral ampicillin 2 g every 6 hours. Women were then randomly selected to receive either 3 or 7 days oral ampicillin. Women allocated the 3‐day course received a matching placebo.

Outcomes

Primary outcome of prolongation of pregnancy for at least 7 days. Secondary outcomes included rated of chorioamnionitis, postpartum endometritis and neonatal morbidity and mortality.

Notes

Study took place between September 1999 ‐ December 2001, Pennsylvania USA.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Computer‐generated random number table in blocks of 4.

Allocation concealment (selection bias)

Low risk

Study medicine given by pharmacy in identical packs.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information given.

Methods

Randomised double‐blind, placebo‐controlled trial. Block randomisation done using computer‐generated numbers.

Participants

67 women randomised. 26 + 0 ‐ 33 + 6 rupture of membranes, leakage of amniotic fluid at vaginal speculum examination. Preceding onset of uterine contractions. Singleton pregnancies.

Interventions

Ampicillin 2 g IV 6‐hourly. 24 hours ‐ pivampicillin 500 g orally 8‐hourly for 7 days plus IV metronidazole 500 mg every 8 hours for 24 hours, followed by metronidazole 400 mg orally every 8 hours for 7 days or identical placebo.

Outcomes

Latency period from admission ‐ delivery. Gestational age at delivery. Preterm delivery less than 37/40 maternal ‐ neonatal infection birthweight.

Notes

October 1991‐April 1994. 6 centres around Copenhagen.
67 women randomised.
30 antibiotics.
37 placebo.
Data sent from Mr Svare and extracted from PhD thesis.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Block randomisation done using computer‐generated numbers.

Allocation concealment (selection bias)

Unclear risk

No information available.

Blinding (performance bias and detection bias)
All outcomes

Low risk

Stated as double‐blind trial.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Data appear complete.

Selective reporting (reporting bias)

Unclear risk

Protocol not available.

Other bias

Unclear risk

No information available