Materials and Methods
Eligibility criteria
The research protocol was designed a priori. We performed electronic research in OVID (ie, LWW Health Library and Maternity and Infant Care), Scopus, ClinicalTrials.gov , MEDLINE, the PROSPERO International Prospective Register of Systematic Reviews, EMBASE, and the Cochrane Central Register of Controlled Trials with the use of a combination of text words related to “induction,” “cesarean section,” “cesarean,” “expectant management,” and “pregnancy” from inception of each database through December 2014. All results were then limited to “clinical trial.” No restrictions for language or geographic location were applied.
Study selection
We included all RCTs of asymptomatic and uncomplicated singleton gestations at full term (ie, between 39 0 and 40 6 weeks) with intact membranes randomized to induction of labor or control (ie, expectant management).
Only trials on asymptomatic singleton gestations without premature rupture of membranes or any other indications for induction evaluating the efficacy of induction of labor in full-term singleton gestations were included. Exclusion criteria included quasirandomized trials, and trials in women with premature rupture of membranes, with indication for induction (ie, intrauterine growth restriction, diabetes, gestational hypertension/preeclampsia, oligohydramnios, fetal macrosomia), and with multiple gestations.
The metaanalysis was reported following the Preferred Reporting Item for Systematic Reviews and Metaanalyses (PRISMA) statement. Before data extraction, the review was registered with the PROSPERO International Prospective Register of Systematic Reviews (registration no. CRD42014014261).
Data abstraction
Data abstraction was completed by 2 independent investigators (G.S., V.B.). Each investigator independently abstracted data from each study and analyzed data separately. All analyses were done using an intention-to-treat approach, evaluating women according to the treatment group to which they were randomly allocated in the original trials. The primary outcome was the incidence of cesarean delivery. Secondary outcome included spontaneous vaginal delivery, operative vaginal delivery (forceps or vacuum), chorioamnionitis, postpartum blood loss, and neonatal outcomes including meconium-stained amniotic fluid (MSAF), Apgar score <7 at 5 minutes, birthweight, admission to neonatal intensive care unit (NICU), and perinatal death. For studies that did not stratify data, composite data were extracted. All authors were contacted for missing data. We planned subgroup analyses in women with favorable cervix, in nulliparous women only, in women who received induction between 39 0 and 39 6 , and in women with a previous cesarean delivery.
The risk of bias in each included study was assessed by using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions.
Data analysis
The data analysis was completed independently by the authors using Review Manager 5.3 (2014; The Nordic Cochrane Centre, Cochrane Collaboration, Copenhagen, Denmark). The completed analyses were then compared, and any difference was resolved with review of the entire data. Statistical heterogeneity between studies was assessed using the Cochrane Q statistics and Higgins I 2 statistics. In case of statistically significant heterogeneity, the random effects model of DerSimonian and Laird was used to obtain the pooled risk ratio (RR) estimate, otherwise a fixed effect model was planned. The summary measures were reported as RR with 95% confidence interval (CI). P value < .05 was considered statistically significant. This study had no funding source.
Results
Study selection and study characteristics
We initially identified 16 RCTs evaluating the efficacy of induction in full-term gestations. Eleven studies were excluded. Five RCTs that met inclusion criteria for this metaanalysis were analyzed. Figure 1 shows the flow diagram (PRISMA template) of information through the different phases of the review. Two authors provided unpublished data from their trials.
Most studies had a low risk of bias in selective reporting and incomplete outcome data according with the Cochrane Collaboration tool. No study was double-blind because this was deemed difficult methodologically given the intervention ( Figure 2 ).
The characteristics of the 5 included trials are summarized in Table 1 . Of the 884 women, 444 (50%) were randomized to induction group and 440 (50%) to control. All studies enrolled only uncomplicated full-term vertex singleton gestations. Two studies enrolled only women with favorable cervix defined as a Bishop score of ≥5 in nulliparous or ≥4 in multiparous patients. Only 1 study reported separate data in nulliparous and multiparous women. No studies reported data about prior cesarean delivery.
| Characteristic | Cole et al, 1975 | Martin et al, 1978 | Tylleskar et al, 1979 | Nielsen et al, 2005 | Miller et al, 2014 |
|---|---|---|---|---|---|
| Location | United Kingdom | United Kingdom | Sweden | United States | United States |
| Sample size, n (induction/control) | 228 (111/117) | 184 (92/92) | 84 (43/41) | 226 (116/110) | 162 (82/80) |
| Inclusion criteria | Singleton, uncomplicated | Singleton, uncomplicated | Singleton, uncomplicated, favorable cervix | Singleton, uncomplicated, favorable cervix | Singleton, uncomplicated, nulliparous, unfavorable cervix |
| Range GA at randomization, wk | 39 0 –40 6 | 38 0 –38 6 | 40 0 –40 6 | 38 0 –38 6 | 38 0 –38 6 |
| Range GA at induction, wk | 39 0 –40 6 | 39 0 –39 6 | 40 0 –40 6 | 39 0 –39 6 | 39 0 –39 6 |
| Induction method | AROM, oxytocin | AROM, oxytocin | AROM, oxytocin | AROM, oxytocin | Misoprostol followed by Foley bulb and oxytocin, or Foley bulb and oxytocin |
| Control group | Expectant management and possible induction at 41 wk | Expectant management and possible induction at 42 wk | Expectant management | Expectant management and possible induction at 42 wk | Expectant management |
| Study primary outcomes | Meconium staining | Serum bilirubin levels, meconium staining | N/R | Cesarean delivery | Cesarean delivery |
Synthesis of results
Uncomplicated full-term singleton gestations receiving induction of labor had similar incidence of cesarean delivery compared to controls (9.7% vs 7.5%; RR, 1.25; 95% CI, 0.75–2.08) ( Figure 3 and Table 2 ). Rates of spontaneous (75.9% vs 80.2%; RR, 0.95; 95% CI, 0.87–1.02) and operative (13.1% vs 10.6%; RR, 1.22; 95% CI, 0.83–1.81) vaginal delivery were also similar. Induction was associated with similar rates of chorioamnionitis (9.6% vs 8.0%; RR, 1.17; 95% CI, 0.38–3.39), but significantly less blood loss (mean difference –57.59 mL; 95% CI, –83.96 to –31.21) ( Figure 4 ) compared to controls. Regarding neonatal outcomes, induction was associated with a significantly lower rate of MSAF (4.0% vs 13.5%; RR, 0.32; 95% CI, 0.18–0.57) ( Figure 5 and Table 2 ), and significantly lower mean birthweight (mean difference –135.51 g; 95% CI, –205.24 to –65.77) compared to control group. There were no differences in other adverse neonatal outcomes, including Apgar <7 at 5 minutes, NICU admission, and perinatal death, between the 2 groups ( Table 2 ).
| Variable | Cole et al, 1975 | Martin et al, 1978 | Tylleskar et al, 1979 | Nielsen et al, 2005 | Miller et al, 2014 | Total | RR (95% CI) |
|---|---|---|---|---|---|---|---|
| Cesarean delivery | 5/111 (4.5%) vs 9/117 (7.7%) | 4/92 (4.3%) vs 1/92 (1.1%) | 1/43 (2.3%) vs 1/41 (2.4%) | 8/116 (6.7%) vs 8/110 (7.8%) | 25/82 (30.5%) vs 14/80 (17.5%) | 43/444 (9.7%) vs 33/440 (7.5%) | 1.25 (0.75–2.08) |
| SVD | 72/111 (64.7%) vs 82/117 (70.1%) | N/R | 41/43 (95%) vs 38/41 (92.7%) | 100/116 (86.2%) vs 93/110 (84.5%) | 54/82 (65.9%) vs 66/80 (82.5%) | 267/352 (75.9%) vs 279/348 (80.2%) | 0.95 (0.87–1.02) |
| Operative vaginal delivery a | 34/111 (30.6%) vs 26/117 (22.2%) | N/R | 1/43 (2.3%) vs 2/41 (4.9%) | 8/116 (6.9%) vs 9/110 (8.2%) | 3/82 (3.7%) vs 0/80 | 46/352 (13.1%) vs 37/348 (10.6%) | 1.22 (0.83–1.81) |
| Chorioamnionitis | N/R | N/R | N/R | 7/116 (6.0%) vs 6/110 (5.5%) | 12/82 (14.6%) vs 9/80 (11.3%) | 19/198 (9.6%) vs 15/190 (8.0%) | 1.17 (0.38–3.39) |
| Mean blood loss, mL | 185 vs 233 | N/R | N/R | 303 vs 312 | 754 vs 900 | – | Mean difference –57.59 mL (–83.96 to –31.21) b |
| Meconium-stained amniotic fluid | 1/111 (0.9%) vs 13/117 (11.1%) | 3/92 (3.3%) vs 13/92 (14.1%) | N/R | 6/116 (5.2%) vs 11/110 (10%) | 6/82 (7.3%) vs 17/80 (21.3%) | 16/401 (4.0%) vs 54/399 (13.5%) | 0.32 (0.18–0.57) b |
| Apgar <7 at 5 min | N/R | N/R | N/R | 0/116 vs 0/110 | 0/82 vs 1/80 | 0/198 (0%) vs 1/190 (0.5%) | 0.33 (0.01–7.87) |
| Mean birthweight, g | 3250 vs 3390 | N/A | 3638 vs 3720 | 3459 vs 3604 | 3401 vs 3513 | – | Mean difference –135.51 (–205.24 to –65.77) b |
| NICU admission | N/R | N/R | N/R | 0/116 vs 0/110 | 5/82 (6.1%) vs 5/80 (6.3%) | 5/198 (3.0%) vs 5/190 (2.6%) | 0.98 (0.47–2.04) |
| Perinatal death | 0/111 vs 1/117 (0.9%) | 0/92 vs 1/92 (1.1%) | N/R | 0/116 vs 0/110 | 0/82 vs 0/80 | 0/401 (0%) vs 2/399 (0.5%) | 0.35 (0.04–3.37) |
Table 3 shows the results for primary outcome in the subgroup analyses. We found no differences in the rate of cesarean delivery in women with favorable cervix, in nulliparous women, and in women who received induction between 39 0 and 39 6 weeks ( Table 3 ). No study stratified data by previous cesarean delivery. Since no studies stratified data for multiparous women with favorable cervices, this subgroup analysis was not feasible.
