Objective
The purpose of this study was to review systematically the efficacy of transabdominal amnioinfusion (TA) in early preterm premature rupture of membranes (PPROM).
Study Design
We conducted a literature search of EMBASE, MEDLINE, and ClinicalTrials.gov databases and identified studies in which TA was used in cases of proven PPROM and oligohydramnios. Risk of bias was assessed for observational studies and randomized controlled trials. Primary outcomes were latency period and perinatal mortality rates.
Results
Four observational studies (n = 147) and 3 randomized controlled trials (n = 165) were eligible. Pooled latency period was 14.4 (range, 8.2–20.6) and 11.41 (range −3.4 to 26.2) days longer in the TA group in the observational and the randomized controlled trials, respectively. Perinatal mortality rates were reduced among the treatment groups in both the observational studies (odds ratio, 0.12; 95% confidence interval, 0.02–0.61) and the randomized controlled trials (odds ratio, 0.33; 95% confidence interval, 0.10–1.12).
Conclusion
Serial TA for early PPROM may improve early PPROM-associated morbidity and mortality rates. Additional adequately powered randomized control trials are needed.
Preterm premature rupture of membranes (PPROM) complicates approximately 3% of all pregnancies. It is a major cause of neonatal death and morbidity, primarily because of preterm birth. Lack of amniotic fluid may lead to pulmonary hypoplasia, infection, and restrictive joint deformities. Chorioamnionitis negatively affects neonatal prognosis at all gestational ages and warrants prompt delivery. The standard management approach to mid-trimester PPROM includes antibiotic treatment and corticosteroids to accelerate fetal lung maturity between 24 and 32 weeks of gestation. Delivery is warranted if there is clinical evidence of chorioamnionitis or fetal distress. Termination of pregnancy may be offered for previable PPROM (<22-23 weeks of gestation) because of the poor prognosis. Despite the relatively high frequency of this condition, controversy regarding the optimal management persists.
For Editors’ Commentary, see Contents
In recent years, attempts to decrease neonatal mortality and morbidity rates were undertaken with different strategies that included intracervical fibrin application, amniopatch, fetal endoscopic tracheal occlusion, antioxidant treatment, gelatin sponge, and progesterone treatment. None of these strategies have proved to be consistently effective, reproducible, or applicable for most centers.
Amnioinfusion or instillation of physiologic solution into the amniotic cavity was attempted initially to reduce intrapartum variable decelerations. Later, it was suggested as a treatment modality to prolong the latency period and prevent the oligohydramnios-related sequelae in cases of early PPROM. Both transcervical and transabdominal routes have been attempted. One of the hypothetic disadvantages of the transcervical route is the nonsterile environment through which the infusion catheter passes, therefore increasing the risk of the introduction of infectious organisms from the vaginal flora into the amniotic sac. Transabdominal amnioinfusion (TA) theoretically surmounts this pitfall. Several articles have described serial TA as a plausible treatment modality to prolong the latency period between rupture of membranes and birth. Recently, a Cochrane review assessed the efficacy of TA for PPROM with the use of data from 2 randomized trials and concluded that the small number of subjects in those studies precluded a definitive answer in regards to the efficacy of the intervention. However, additional data are available from observational studies of TA that can shed further light on this topic.
Our objective was to review systematically and metaanalyze studies that have assessed efficacy and safety of TA in women with PPROM. This systematic review provides results of separate qualitative and quantitative analyses of randomized controlled trials (RCTs) and observational studies.
Methods
Search strategy
We performed a comprehensive literature search, assisted by an experienced librarian, using the MEDLINE from 1950 to December 2011 and EMBASE from 1980 to December 2011. We also searched the ClinicalTrials.gov database for studies that finished recruitment. We used the terms fetal membranes, premature rupture, rupture, membrane*, pregnancy, amnioinfus*, premature fetus membrane rupture, and amnioinfusion . There were no language or geographic restrictions. Bibliography of identified articles was used to screen for additional related articles.
Study selection
We included both comparative observational and RCTs in which TA and conventional treatment were compared with conventional treatment alone. Case reports, case series, and abstract publications were excluded. Studies that included patients with a confirmed diagnosis of PPROM-associated oligohydramnios were included. Studies that included oligohydramnios from other causes (eg, intrauterine growth restriction, renal anomalies) were excluded. Two reviewers (S.P. and H.A.) independently evaluated studies for inclusion; disagreements were resolved through consensus among the authors.
Outcome measures
The primary outcomes of interest were latency period (interval from PPROM to birth) and perinatal death. Secondary outcomes of interest were pulmonary hypoplasia, neonatal death, gestational age at birth, birthweight, chorioamnionitis, early onset (<72 hours from delivery) neonatal sepsis, bronchopulmonary dysplasia, and cesarean delivery. Two investigators (S.P. and H.A.) independently abstracted the relevant data from selected articles.
Assessment of risk of bias
Risk of bias in observational studies was assessed with the Newcastle-Ottawa scale and in RCTs with the Cochrane collaboration’s tool. For observational studies, the domains of assessment included selection, comparability, and outcome assessment biases. For RCTs, the domains included selection, performance, detection, attrition, reporting, and other biases. Two investigators (S.P. and H.A.) independently assessed risk of bias; discrepancies were resolved through discussion and involvement of third author.
Data extraction
Data were extracted in duplicate from published reports by 2 authors who used a standardized data collection form. A third reviewer was consulted in case of disagreement between the 2 data extractors; discrepancy was resolved by consensus. We did not contact authors for missing information. For continuous outcomes, means and standard deviations were obtained from studies. When they were not reported, they were calculated from range, median, and sample size according to the method described by Hozo et al. For categoric outcomes, event rates were obtained.
Statistical analysis
Statistical analyses were performed with the Review Manager (RevMan) software (version 5.1.4; The Nordic Cochrane Centre, København, Denmark). Metaanalyses were performed separately for cohort studies and the RCTs. Where data were sufficiently homogenous, metaanalysis was conducted with the use of a random effects model, with weighting of studies according to the DerSimonian-Laird method. Random-effect model was used to account for between and within study heterogeneity. Cochran’s Q test was used to test for heterogeneity between studies at the .10 level of significance. The I-squared statistic was used to quantify the degree of heterogeneity.
Report
Results of the metaanalysis of RCTs and the observational studies were reported according to the Preferred Reporting Items for Systematic Reviews and Metaanalyses statement and Metaanalysis of the Observational Studies in Epidemiology guidelines, respectively.
Results
Our initial search yielded 141 citations. After review of titles and abstracts, 126 citations were excluded ( Figure 1 ). After full text review of the remaining 15 articles, 8 articles were excluded (3 because inclusion criteria were not fulfilled ; 4 because of it was not clear whether reported patients in these studies were included in other studies, and 1 because the intervention group and control group were unmatched ). The remaining 7 studies met inclusion criteria and were included in this review. Characteristics of these studies are summarized in Table 1 .
| Study | Characteristics |
|---|---|
| De Santis et al, 2003 | |
| Type of study | Quasi-randomized; patients admitted by chance to 1 of 2 divisions of the same department; both divisions offered expectant treatment; only 1 division offered TA |
| Participants | 37 patients in intervention group; 34 patients in control group |
| Inclusion criteria | Singleton; PPROM at <26 weeks of gestation; severe and persistent oligohydramnios (AFI <30 mm, lasting ≥7 d) |
| Exclusion criteria | Clinical chorioamnionitis; active labor; autoimmune or metabolic disease; history of multiple invasive procedures; declining treatment after informed consent; delivery in the interim 7-day waiting period from admission; transfer from other hospitals after a period of treatment |
| Diagnosis of PPROM | History; sterile speculum examination; vaginal pH >5; AFI measurement by ultrasound scanning |
| Interventions | |
| All | Hospital bed rest; antibiotic prophylaxis (mezlocillin, 2 g intravenously twice daily for at least 7 days) or targeted treatment based on cultures; tocolytic treatment (isoxsuprine, either intravenously or orally) for contractions; betamethasone after 25 weeks; fetal monitoring by daily heart check or cardiocotography after 26 weeks and modified BPP every 3 days; cesarean delivery performed in the presence of chorioamnionitis, abruptio placentae, and/or fetal distress (abnormal fetal monitoring) or at 30 weeks of gestation |
| Intervention | Weekly saline amnioinfusion in a sufficient amount to increase AFI to 10 cm starting 7 days at least after PPROM; antibiotics and tocolysis on the day of amnioinfusion |
| Outcomes | Latency period; gestational age at delivery; cesarean delivery; genitourinary infection, amnionitis/endometritis; neonatal weight; orotracheal intubation; survival; deformities; pulmonary hypoplasia; bronchopulmonary dysplasia; early-onset sepsis; early-onset pneumonitis; abnormal neurologic outcome (includes cerebral palsy, spastic diplegia or tetraplegia, deafness, or blindness) |
| Notes | Also included were women who underwent PROM after amniocentesis for prenatal diagnosis: 10 patients (27.0%) in the amnioinfusion group and 8 patients (23.5%) in the control group |
| Tranquilli et al 2005 | |
| Type of study | Randomized controlled trial |
| Participants | 17 patients in each arm |
| Inclusion criteria | Singleton pregnancy with a certain gestational age confirmed by an early second-trimester ultrasonographic examination; gestational age 24-33 weeks; evidence of PPROM within 24 hours of admission; oligohydramnios (amniotic fluid index, <10th percentile); absence of uterine contractions at the time of hospitalization; no evidence of clinical chorioamnionitis at admission; no evidence of placental anomalies or major structural fetal anomalies, and normal cardiotocography at the time of admission |
| Diagnosis of PPROM | PPROM diagnosed on examination by a sterile speculum when obvious leakage of amniotic fluid from the cervical os was confirmed by a positive fibronectin test |
| Interventions | |
| All | Hospital bed rest; antibiotic prophylaxis (sulbactam-ampicillin 3 g, intravenously every 8 hours for 7 days); betamethasone therapy; prophylactic tocolytic (intravenous ritodrine) in the absence of clinical signs of chorioamnionitis or placental abruption; daily fetal heart rate monitoring |
| Intervention | Weekly serial amnioinfusion if the AFI fell <5th percentile and/or a median pocket of amniotic fluid was <2 cm, with a target AFI of >10th percentile; if repeated AFI was ≤5, amnioinfusion repeated weekly until 27 weeks of gestation; a nonstress test performed daily |
| Outcomes | Latency period; gestational age at delivery; birthweight; admission to neonatal intensive care unit; pulmonary hypoplasia; abnormal neurologic outcome |
| Singla et al, 2010 | |
| Type of study | Randomized controlled trial |
| Participants | 30 patients in each arm |
| Inclusion criteria | Singleton pregnancy; PPROM between 26 and 33 + 6 week gestations; AFI <5th percentile for gestational age |
| Exclusion criteria | Women with evidence of clinical chorioamnionitis, placental or fetal anomalies; active labor or AFI >5th percentile |
| Diagnosis of PPROM | Sterile speculum examination or nitrazine/litmus paper test; confirmation by ultrasound scanning |
| Interventions | |
| All | 24 hours of observation after admission before randomization; hospital bed rest; antibiotic prophylaxis (erythromycin: 250-mg tablet 4 times per day for 10 days); betamethasone prophylaxis; daily obstetric examination; weekly BPP, complete blood cell count, and cervical/vaginal cultures |
| Intervention | Amnioinfusion of warmed saline solution in a sufficient amount to maintain the AFI at >5th percentile for gestational age; weekly AFI measurement and repeated amnioinfusion if the AFI fell <5th percentile; labor induction when there was fetal distress or chorioamnionitis |
| Outcomes | Latency period; gestational age at delivery; birthweight; intrapartum fetal distress; early neonatal sepsis; rate and causes of neonatal mortality; type and mode of delivery; postpartum sepsis |
| Vergani et al, 1997 | |
| Type of study | Observational |
| Participants | 18 patients in intervention group and 16 patients in historic cohort group who did not undergo the procedure |
| Inclusion criteria | Singleton pregnancy; PPROM at <25 completed weeks of gestation; no labor; persistent oligohydramnios (maximum pool depth ≤2 cm of cord-free pocket of fluid) at ≥4 days |
| Exclusion criteria | Amniotic fluid leakage after second-trimester amniocentesis; clinical chorioamnionitis; presence of uterine contractions ≥4/hour; sonographic diagnosis of structural fetal abnormalities; maternal immunologic diseases; multiple gestations |
| Diagnosis of PPROM | Observation of vaginal pooling and a positive nitrazine test on sterile speculum examination |
| Interventions | |
| All | Hospital bed rest during the first week, then home bed rest until 25 weeks, after which all patients were admitted until delivery; tocolytic treatment (intravenous ritodrine) given at >25 weeks for uterine contractions in the absence of clinical signs of chorioamnionitis or abruption placentae; betamethasone course at least once between 25 and 32 weeks of gestation; 1-week course of prophylactic antibiotic therapy with sulbactam-ampicillin 3 g intravenously every 8 hours and targeted treatment based on cervical and vaginal cultures; sonographic determination of amniotic fluid volume twice a week for outpatients and daily for inpatients; BPP twice a week at >25 weeks of gestation |
| Intervention | 1-2 TA/wk to aim to restore AFI >5 cm; delivery in the presence of clinical chorioamnionitis, fetal distress, abruption placentae, or documented fetal lung maturity on amniocentesis after 28 weeks of gestation |
| Outcomes | Gestational age at delivery; latency period; survival rate; pulmonary hypoplasia; chorioamnionitis; fetal distress; placental abruption; preterm labor; in utero death; umbilical cord prolapse |
| Garzetti et al, 1997 | |
| Type of study | Observational |
| Participants | 18 women in each arm; control group recruited from historic data |
| Inclusion criteria | Singleton; PPROM between 25 and 32 weeks of gestation; oligohydramnios (AFI, <5th percentile) |
| Exclusion criteria | Presence of uncontrolled labor; presence of obstetric complications; maternal immunocompromise; uterine fibroid tumors; lack of written consent |
| Diagnosis of PPROM | Observation of gross vaginal pooling of amniotic fluid and a positive nitrazine test on speculum examination |
| Interventions | |
| All | Hospital bed rest until delivery with minimal activity limited to bathroom necessities; weekly complete blood cell count and semiquantitative C-reactive protein measurement; weekly ultrasound scanning and cardiocotography; daily nonstress test; prophylactic tocolytic treatment (intravenous ritodrine) in the absence of chorioamnionitis or abruption placentae; prophylactic antibiotic treatment (ceftazidime 2 g/d intramuscularly) and targeted therapy based on cultures; delivery in the presence of clinical chorioamnionitis, positive amniotic fluid culture, fetal distress, and documented fetal lung maturity |
| Intervention | If AFI <10th percentile for gestational age and deepest pocket of cord-free fluid ≥10 mm, weekly TA of 150-350 mL warmed saline solution; weekly AFI assessment before and after each procedure; biweekly fetal growth assessment |
| Outcomes | Latency period duration; median amniotic fluid volume and short-term variability; relationship between amniotic fluid volume and fetal short-term variability in the amnioinfusion group |
| Ogunyemi and Thompson, 2002 | |
| Type of study | Observational |
| Participants | 12 patients in each group |
| Inclusion criteria | PPROM at gestational age ≤27 weeks; oligohydramnios with AFI <5 cm; normal fetal anatomic scan; absence of gross infection; stable mother and fetus |
| Exclusion criteria | Presence of active labor; clinical chorioamnionitis |
| Diagnosis of PPROM | Observation of vaginal pooling, a positive nitrazine test or ferning on speculum evaluation |
| Interventions | |
| All | Initial hospital bed rest, followed by outpatient follow-up evaluation for stable patients; corticosteroids after 24 weeks of gestation; magnesium sulfate and terbutaline were used for tocolysis as needed if preterm labor suspected in the absence of clinical chorioamnionitis; prophylactic intravenous antibiotics |
| Intervention | Before the procedure, intravenous magnesium sulfate 4 g loading dose followed by 1 g/hr was initiated and discontinued 12 hours after the procedure if preterm labor did not ensue; weekly amnioinfusion of warm 0.9% normal saline solution with ampicillin 1 g/L until 27 weeks of gestation if AFI ≤5 |
| Outcomes | Chorioamnionitis; latency period; cesarean delivery rate; gestational age at delivery; birthweight; neonatal sepsis; neonatal death; perinatal death; total death |
| Gramellini et al, 2003 | |
| Type of study | Observational |
| Participants | 24 patients in intervention group and 29 patients in historic control group |
| Inclusion criteria | Singleton; <34 weeks gestational age; PPROM; oligohydramnios (AFI, ≤5 cm) |
| Exclusion criteria | Active labor; clinical evidence of placental abruption or chorioamnionitis |
| Diagnosis of PPROM | Observation of persistent vaginal pooling and a positive nitrazine paper test |
| Interventions | |
| All | Tocolytic treatment (ritodrine) for >4 uterine contractions/20 min; betamethasone after 24 weeks of gestation; prophylactic antibiotic therapy (erythromycin 2 g/d) given to 90% of patients |
| Intervention | TA of 0.9% normal saline solution or lactated Ringer’s solution according to a volume criterion of 10 mL/gestational week; repeated if AFI measurement >12 hours after the procedure was <5 cm |
| Outcomes | Gestational age at delivery; latency period; birthweight; rate of intrauterine death, vaginal bleeding, cesarean delivery, and postpartum endometritis |
| Notes | Refers to a group of patients in whom the oligohydramnios was not attributed to PPROM; however, all data cited refer only to patients affected by PPROM |
Of the 7 included studies, 2 studies were RCTs (47 patients in each group); 1 study was quasirandomized (34 patients in the control group and 37 in the intervention group), and 4 studies were observational studies (75 patients in the control group and 72 in the intervention group). In the quasirandomized study, patients were admitted by chance to one of 2 departments that differed in their management approach toward PPROM: one department provided standard care; the other department provided standard care in addition to serial amnioinfusion to consented patients. We decided to include the quasirandomized study with the other 2 randomized studies because we believed that the risk of bias in that study was not high. The gestational age at inclusion varied from 16-33 weeks. Information on individual patients was provided only in 1 study ; therefore, subgroup analysis by gestational age was not possible. Ascertainment of rupture of membranes was performed in all studies by speculum examination to confirm pooling of amniotic fluid in the posterior fornix. In addition, 6 studies used nitrazine, and 1 study used fetal fibronectin as a confirmatory test. Conventional care for patients with PPROM included bed rest in the hospital and prophylactic antibiotic therapy in all studies. Five studies used tocolysis only when uterine contractions appeared without clinical chorioamnionitis or abruptio placenta ; however, 2 studies used tocolysis as a prophylactic measure for all patients, regardless of the presence of uterine contractions. Targeted antibiotic therapy based on cervical and vaginal cultures was used in 3 studies. Corticosteroids for fetal lung maturation was used after viability in all but 1 study. The number of procedures per patient, success rate, and volumes infused varied among different studies and among different subjects in the same study. The average number of infusions per patient ranged from 1.23 in Singla et al to 4.0 in De Santis et al. Vergani et al reported a median number of 3 infusions per patient with a range of 1–9. Most of the studies did not report success rate; however, Garzetti et al reported a success in 18 of 19 patients, and De Santis et al reported successful amnioinfusion in 143 of 147 procedures. The infused volumes range from 140-350 mL per infusion.
De Santis et al reported on 5 complications that occurred within 24 hours after infusion: 2 cord prolapses (1 cephalic and 1 transverse lie); 2 abruptio placentae, and 1 onset of labor. Gramellini et al reported on a higher rate of vaginal bleeding in the intervention group (21%) compared with the nonamnioinfused group (7%), although this difference did not reach statistical significance. Ogunyemi and Thompson reported on 2 neonatal complications that can be regarded as direct injuries from the procedure: 1 baby had a 2-cm leg laceration that was sutured, and 1 baby had a 0.5 × 0.5–cm superficial chest scar that needed no treatment.
Assessment of the risk of bias in included observational cohort studies and RCTs are shown in Tables 2 and 3 , respectively. The observational studies had minimal risk of bias, whereas 2 of 3 RCTs had moderate risk of bias, and 1 RCT had high risk of bias.
| Study | Selection | Comparability | Outcome | Total stars, n | |||||
|---|---|---|---|---|---|---|---|---|---|
| Representativeness of the exposed cohort | Selection of the nonexposed cohort | Ascertainment of exposure | Demonstration that outcome of interest was not present at start of study | Comparability of cohorts on the basis of the design or analysis | Assessment of outcome | Was follow-up long enough for outcomes to occur? | Adequacy of follow up of cohorts | ||
| Ogunyemi and Thompson (2002) | * | * | * | * | ** | * | * | * | 9 |
| Vergani et al (1997) | * | * | * | * | * | * | * | 7 | |
| Garzetti et al (1997) | * | * | * | * | ** | * | * | * | 9 |
| Gramellini et al (2003) | * | * | * | * | * | * | * | 7 | |
| Study | Bias | |||||||
|---|---|---|---|---|---|---|---|---|
| Selection | Performance: blinding of participants and personnel | Detection: blinding of outcome assessment | Attrition: incomplete outcome data | Reporting: selective reporting | Other sources | Overall | ||
| Random sequence generation | Allocation concealment | |||||||
| Singla et al (2010) | Low risk | Unclear risk | Unclear | Unclear | Low risk | Low risk | Low risk | Moderate risk |
| Tranquilli et al (2005) | Low risk | Low risk | Unclear | Unclear | Low risk | Low risk | Low risk | Moderate risk |
| De Santis et al (2003) | High risk | High risk | Unclear | High risk | Low risk | Low risk | Low risk | High risk |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
