You are here

First versus second stage C/S maternal and neonatal morbidity: a systematic review and meta-analysis

European Journal of Obstetrics & Gynecology and Reproductive Biology

Abstract

The rates of cesarean section at full cervical dilatation (second stage cesarean sections) are currently increasing. The purpose of the present study is to compare maternal and neonatal morbidity and mortality among cases offered cesarean section at full dilatation to those offered cesarean section prior to full dilatation. We searched Medline, Scopus, Clinicaltrials.org, Popline, Cochrane CENTRAL, and Google Scholar search engines, along with reference lists from all included studies. The RevMan 5.0 software was used for all analyses. Primary maternal outcomes were defined as death, ICU admission and need for transfusion, while primary neonatal outcomes were defined as death, neonatal unit admission and 5 min Apgar score less than 7. Ten studies were finally retrieved involving 23,104 singleton childbearing women (18,160 operated in the first stage and 4944 in the second stage of labor). Second stage cesarean section seems to lead to higher maternal admissions to ICU (OR 7.41, 95% CI 2.47–22.5) and higher transfusion rates (OR 2.60, 95% CI 1.49–2.54). Neonatal death rates were also increased (OR 5.20, 95% CI 2.49–10.85) along with admissions to neonatal unit (OR 1.63, 95% CI 0.91–2.91) and rates of Apgar score less than 7 in 5 min (OR 2.77, 95% CI 1.02–7.50). Second stage cesarean section seems to result significantly increased morbidity for both mothers and neonates. It seems that a direct evaluation with forceps and vacuum extractors is imperative in order to establish its place in modern evidence-based practice.

Abbreviations: C/S - cesarean section, NICU - neonatal intensive care unit, ICU - intensive care unit, OR - odds ratio, CI - confidence interval, FEM - fixed effects model, REM - random effects model, RCT - randomized controlled trial.

Keywords: Second stage, Cesarean section, Neonatal, Maternal, Morbidity.

1. Introduction

Cesarean section (C/S) rates are increasing at present [1] . Numerous factors during parturition contribute to the decision to perform C/S, and these include maternal, fetal and placental pathology, fetal malpresentation and maternal wish.

During the last three decades it appears that assisted vaginal delivery such as forceps and vacuum extraction, traditionally used for arrest of descent, have been replaced by C/S during the second stage of labor [2] and [3]. Full cervical dilatation is referred hereinafter as second stage of labor. Cesarean section nowadays is usually performed as an alternative to operative vaginal delivery [3] . Its impact on maternal and neonatal morbidity during the active phase of labor and especially after the achievement of full dilatation has gained interest during the last decade. Cesarean section during the second stage of labor with an engaged head is generally thought to carry higher maternal morbidity, usually resulting from tearing of the lower uterine segment, extension of the uterine incision and incision of the urinary bladder. Furthermore, the delay in the decision to perform an emergency cesarean section for reasons of fetal distress puts the fetus at increased risk of developing hypoxia, thus risking brain damage that leads to varying forms of disability in its life [4] .

While several studies have compared maternal and neonatal morbidity arising from cesarean sections during the first and second stage of labor, to the best of our knowledge this is the first systematic review and meta-analysis performed in this field.

2. Methods

2.1. Study design

The present study was designed according to the MOOSE guidelines [5] . All investigators came to agreement regarding eligibility criteria and definition of primary and secondary outcomes. Prospective and retrospective observational studies were included in this systematic review. Two authors (TN and VD) abstracted predetermined data from each selected article to three tables while the rest reviewed the tabulated data independently in order to increase reliability.

We searched Medline (1966–2013), Scopus (2004–2013), Clinicaltrials.org (1997–2013), Popline (1973–2013), Cochrane CENTRAL (1999–2013) and Google Scholar (2004–2013) search engines along with reference lists from all initially retrieved studies. Providing many keywords is not foolproof and may lead to article loss. On the other hand manually reviewing a long list of articles is not feasible. Search criteria were restricted only to humans. No language or date restrictions were applied.

Medline was searched via PubMed using the MeSH terms (second[All Fields] AND stage[All Fields] AND cesarean[All Fields]) AND (“humans”[MeSH Terms]).

Scopus was also searched using the terms: second, stage, cesarean, section. The same strategy was applied in CENTRAL and Popline.

For Google Scholar we used an extended search strategy using the terms: first vs second, stage, cesarean section, Apgar, tears, NICU, transfusion. This way we were able to retrieve a number of articles that we were able to review manually.

The bibliographies of the electronically retrieved articles that were selected in this review were also searched in order to retrieve potentially missed articles from the electronic search. All the articles that met or were presumed to meet the inclusion criteria were retrieved in full text.

We could not retrieve any randomized control trials, as this field is for reasons of ethics not eligible for conducting such trials. Therefore, only cohort studies, either prospective or retrospective, were included.

2.3. Definitions

Second stage of labor was defined as full cervical dilatation. The number of patients requiring transfusion refers to women that were transfused either intraoperatively or postoperatively. “Incision extension” refers to either T, or J or other types of extension of the common transverse uterine incision. The “lower segment tears” index describes unintentional extension of the uterine incision toward the lower uterine segment and/or cervix. Two indices regarding Apgar scores were used (<7 and <3), both of them reflecting possible future adverse outcomes on neurodevelopment [6] . Intubation was designated for reasons other than meconium aspiration. Neonatal department admission was recorded only for cases hospitalized in a neonatal unit for more than 24 h.

Primary maternal outcomes were defined as death, neonatal department admission and need for transfusion. Primary neonatal outcomes were defined as death, neonatal unit admission and 5 min Apgar score less than 7 and/or less than 3. Whereas certain studies defined neonatal departments as NICUs, others failed to do so; we, therefore, included all possible admissions under the term “neonatal department” rather than “NICU”.

Secondary maternal outcomes included wound infection, need for incision extension, uterine hypogastric artery ligation or obstetric hysterectomy (included in the same index as major operations), wound infection, post-operative pyrexia, postpartum hemorrhage, formation of hematoma, cystotomy, re-admission to the hospital and re-look laparotomy. Secondary neonatal outcomes included birth weight (as a risk factor for performing second stage cesarean section), intubation for reasons other than meconium aspiration, neonatal asphyxia, septicemia, fetal injury and neonatal seizures.

Demographic indices such as maternal age, gestational age, duration of augmentation of labor by oxytocin infusion, operating time and length of hospitalization were also assessed in this review.

2.4. Statistical analysis

Statistical meta-analysis was performed using the RevMan 5.1 software (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011). Confidence intervals were set at 95%. Heterogeneity between studies was assessed using Chi-square and I2-statistic test. p value <0.05 denoted statistical significance in the analysis of heterogeneity. Pooled odds ratios (OR) and 95% confidence intervals (CI) for all primary and secondary outcomes were calculated by using both the Mantel–Haenszel fixed effect and the DerSimonian–Laird random effect model. Specifically, for all analyses, results from the fixed effect model (FEM) are presented when both Chi-square and I2 tests denoted no heterogeneity between studies. When statistically significant heterogeneity was present, results were analyzed using the random effects model (REM).

Publication bias was not assessed due to the small number of studies included in the present meta-analysis [7] .

3. Results

3.1. Included studies

Overall, ten articles were finally selected in the present review [8], [9], [10], [11], [12], [13], [14], [15], [16], and [17], including seven retrospective studies [8], [9], [11], [12], [13], [15], and [16], and three prospective studies [10], [14], and [17]. These studies encompassed 23,104 pregnant women, among whom 18,106 were operated during the first stage of labor while the remaining 4998 had a procedure at full cervical dilatation.

3.2. Excluded studies

Seven articles were excluded from the present review. Three articles were found to be irrelevant to the studied question [18], [19], and [20]. Another represented a clinical opinion [21] . Two more had no control group [22] and [23] and the last one compared two different groups of women offered cesarean section (C/S) at full dilatation [24] .

3.3. Potential bias among included studies

Garret et al. included in their analysis singleton pregnancies at term (more than 37 weeks of gestation) [8] . They excluded cases with abnormal fetal heart rates, placental abruption, chorioamnionitis, excessive bleeding, and any other factor other than labor dysfunction or presumed cephalopelvic disproportion.

Allen et al. conducted their study based on a perinatal database that was oriented toward perinatal, both maternal and neonatal, morbidity that arises from cesarean section in labor [9] . The database was based on data administered from Nova Scotia hospitals. They stated that their population was homogeneous and predominantly Caucasian. Inclusion criteria were live singleton pregnancies at term (37–42 weeks). Cases were excluded if there was a major fetal anomaly, if presentation was non-vertex and if mothers had pre-existing diseases and/or antenatal complications.

Cebekulu et al. conducted a non-consecutive matched study because, according to their explanation, extraction of data was performed only during the days the author was in the hospital [10] . They included only singleton live pregnancies with cephalic presentation. All cases were at a gestation of 36 weeks or more. Women offered previous C/S were excluded. Each patient offered C/S at full dilatation was matched with a suitable control patient who had C/S in the first stage of labor due to poor progress. However, the criteria for matching the participants were not reported.

Alexander et al. conducted a multi-institutional study [11] . Data extraction was performed directly from maternal and infant charts and was entered into a computer. They included in their study all women undergoing a primary cesarean delivery with a gestational age of at least 20 weeks of gestation and infant weight of more than 500 g.

Selo-Ojeme et al. included in their study only nulliparous women at 37–42 weeks of gestation who had an emergency C/S at a single hospital [12] . They stated that they excluded women with multiple pregnancies, fetal anomalies, intrauterine growth restriction, premature rupture of membranes, non-vertex presentation, pregnancy complications or pre-existing maternal diseases.

Winovitch et al. included in their study only women undergoing their first C/S [13] . Prior C/S and/or fetal anomalies were criteria for exclusion.

Govender et al. included in their study all women that were offered an emergency C/S [14] .

Seal et al. included in their study nulliparous women with vertex presentation at term (37–41 completed weeks) [15] . They excluded women with fetal anomalies, malpresentations, pregnancy complications and/or pre-existing maternal disease.

Rabiu et al. included all women offered intrapartum C/S [16] . Inclusion criteria were singleton live pregnancies at term with vertex presentation. Women offered previous C/S, carrying fetuses with anomalies or having pre-existing diseases and/or antenatal complications were excluded.

Sucak et al. included in their study all nulliparous women with vertex presentation with gestational age older than 36 weeks [17] . Pre-existing maternal diseases and antenatal complications were criteria for exclusion.

3.4. Primary maternal outcomes

Maternal deaths were significantly more prevalent among women operated during the second stage of labor (OR 7.96, 95% CI 1.61–39.39, Fig. 1 ).

gr1

Fig. 1 Forest plot of maternal deaths. Results reached statistical significance (p = 0.01). Heterogeneity was not noted among included studies (I2 = 18%).

ICU admission rates of patients undergoing C/S in the second stage were higher compared to those in the first stage (OR 7.41, 95% CI 2.47–22.5, Fig. 2 ).

gr2

Fig. 2 Forest plot of maternal ICU admission rates. Results reached statistical significance (p < 0.001). Heterogeneity was not noted among included studies (I2 = 46%).

Transfusion rates of patients undergoing CS in the second stage were higher compared to those in the first stage (OR 2.60, 95% CI 1.49–2.54, Fig. 3 ).

gr3

Fig. 3 Forest plot of maternal transfusion rates. Results reached statistical significance (p < 0.001). Heterogeneity was noted among included studies (I2 = 75%).

3.5. Primary neonatal outcomes

Death rates of neonates from mothers who underwent C/S in the second stage were higher compared to those in the first stage (OR = 5.20, 95% CI 2.49–10.85, Fig. 4 ).

gr4

Fig. 4 Forest plot of neonatal deaths. Results reached statistical significance (p < 0.001). Heterogeneity was not noted among included studies (I2 = 26%).

Neonatal admission rates were comparable among the two groups (OR 1.63, 95% CI 0.91–2.91, Fig. 5 ).

gr5

Fig. 5 Forest plot of neonatal department admission rates. Results did not reach statistical significance (p = 0.10). Heterogeneity was noted among included studies (I2 = 92%).

Apgar scores less than 7 were comparable among neonates born either during the first or second stage of labor (OR 2.65, 95% CI 0.90–7.78, Fig. 6 ). These results were also reproduced in the case of Apgar scores less than 3 (OR 2.95, 95% CI 0.77–11.28, Fig. 7 ).

gr6

Fig. 6 Forest plot of 5 min Apgar score less than 7 rates. Results were close, but did not reach statistical significance (p = 0.08). Heterogeneity was noted among included studies (I2 = 60%).

gr7

Fig. 7 Forest plot of 5 min Apgar score less than 3 rates. Results did not reach statistical significance (p = 0.11). Heterogeneity was noted among included studies (I2 = 75%).

3.6. Secondary maternal outcomes

Wound infection rates were comparable among women undergoing C/S in the second stage, when compared to those who underwent C/S in the first stage (17,573 patients, REM, OR 1.81, 95% CI 0.81–4.05, data from 7 studies [9], [10], [11], [12], [14], [15], and [16]).

Incision extension rates were higher among women undergoing CS in the second stage (15,323 patients, REM, OR 9.42, 95% CI 3.22–27.59, data from 5 studies [10], [11], [14], [15], and [16]).

Obstetric maneuvers for the avoidance of massive bleeding including uterine artery ligation and obstetric hysterectomy were analyzed together and were more prevalent among women undergoing C/S in the second stage of labor (18,962 patients, REM, OR 3.44, 95% CI 1.34–8.85, data from 8 studies [9], [10], [11], [12], [14], [15], [16], and [17]).

Cystotomy incidence was increased among patients undergoing C/S in the second stage of labor (16,634 patients, FEM, OR 4.16, 95% CI 2.58–6.70, data from 5 studies [11], [14], [15], [16], and [17]).

Post-operative hemorrhage was significantly more prevalent among women having a C/S in the second stage of labor (15,430 patients, REM, OR 2.28, 95% CI 1.01–5.16, data from 3 studies [9], [11], and [15]).

Post-operative pyrexia was significantly more prevalent among women offered C/S in the second stage of labor (3061 patients, FEM, OR 3.94, 95% CI 2.78–5.57, data from 3 studies [9], [14], and [16]).

Lower uterine segment tears were also significantly more frequent among women at full dilatation (1438, FEM, OR 10.19, 95% CI 4.59–22.63, data from 2 studies [14] and [16]).

Broad ligament hematoma was not significantly affected from the stage of labor (13,807 patients, REM, OR 4.91, 95% CI 0.08–290.10, data from 2 studies [11] and [15]).

Readmissions, also, were not affected by the stage of labor (1052 patients, OR 0.97, 95% CI 0.37–2.50, data from 3 studies [10], [12], and [16]).

Relook laparotomies, however, were significantly more prevalent among women offered C/S at full dilatation (1516 patients, OR 3.13, 95% CI 1.01–9.66, data from 3 studies [10], [14], and [16]).

Postpartum hemorrhage, postoperative pyrexia, lower segment tears, hematoma formation, readmission to the hospital and relook laparotomy were indices that were tabulated but were not meta-analyzed, because fewer than four studies were finally included in the present review.

3.7. Secondary neonatal outcomes

Birth weights were significantly higher among women delivering by C/S in the second stage of labor (19,482 patients, REM, MD 130.56, 95% CI 49.99–211.14, data from 6 studies [8], [9], [11], [13], [16], and [17]).

Intubation for reasons other than meconium aspiration was more frequent among neonates of mothers delivering by C/S the second stage of labor (18,203 patients, REM, OR 3.02, 95% CI 1.24–7.31, data from 4 studies [11], [13], [15], and [16]).

Neonatal septicemia was comparable among the two groups (4501 patients, REM, OR 1.93, 95% CI 0.63–5.88, data from 5 studies [9], [10], [12], [15], and [16]).

Fetal injury was more common among women delivering by C/S in the second stage of labor (14,232 patients, REM, OR 5.76, 95% CI 1.27–26.18, data from 4 studies [10], [11], [15], and [16]).

Neonatal asphyxia rates were comparable among the two groups (2597 patients, REM, OR 1.49, 95% CI 0.49–4.52, data from 3 studies [9], [12], and [16]).

Neonatal seizures were significantly more prevalent among neonates delivered at the second stage of labor (14,244 patients, FEM, OR 2.04, 95% CI 1.16–3.56, data from 3 studies [11], [15], and [16]).

3.8. Other indices

Gestational age at the time of delivery was comparable among the two groups (23,104 patients, REM, MD 0.18, 95% CI −0.13 to 0.48, data from 9 studies [8], [9], [10], [11], [12], [13], [15], [16], and [17]).

Augmentation of labor with oxytocin infusion was more frequent among women delivering by C/S in the second stage of labor (7782 patients, REM, OR 1.16, 95% CI 0.74–1.80, data from 6 studies [8], [10], [13], [15], [16], and [17]).

4. Discussion

Cesarean section as a mode of delivery is gaining ground nowadays due to various indications that present either antepartum or intrapartum. This is the first systematic review and meta-analysis that investigates the impact of the stage of labor on maternal and neonatal outcomes among women delivering by cesarean section. We aimed to study several indices reflecting the impact of stage on maternal and neonatal outcome, analyzing together the characteristics of the women, and comparing mean operative duration and length of hospitalization.

Maternal intraoperative complications that were affected by the deeply engaged fetal head, such as cystotomy, uterine incision extension and uterine artery ligation or obstetrical hysterectomy, were significantly increased among women who underwent cesarean section in the second stage of labor. What is important to note is that studies included in the present review did not describe forceps extraction of the fetal head during the second stage and prior to cesarean section. The finding of higher rates of admission to an ICU department seems fairly reasonable given the fact that transfusion rates, obstetric hysterectomies and uterine artery ligations were also higher among women delivering in the second stage of labor. Other indices that are related to this observation are the significantly increased operative times and the prolonged length of hospitalization among those women.

Maternal mortality was also significantly increased among cases offered C/S during the second stage of labor. The majority of studies did not provide actual results for this index, although we can assume that they did not have any deaths in either group.

Neonatal mortality was increased among neonates born by C/S at the second stage of labor. The same was observed for intubation (excluding cases that were intubated for meconium aspiration), fetal injury and neonatal seizures. Differences in neonatal septicemia and neonatal asphyxia rates did not, however, reach statistical significance among investigated groups.

Various antenatal and peripartum maternal and fetal characteristics seem to influence the occurrence of adverse outcomes. Maternal age and gestational age were comparable in all but one [17] studies included in the present review. On the other hand, dystocia was more prevalent among cases offered cesarean section in the second stage of labor, whereas fetal distress was significantly more prevalent among women that had cesarean section prior to full dilatation ( Table 1 ). We also observed discrepancies among studies in the case of labor augmentation, with some of them reporting higher rates among women offered second stage cesarean section [8], [10], and [16] and others giving contradicting results [13], [15], and [17].

Table 1 Maternal demographics, labor characteristics and length of hospitalization (first vs second stage).

Author, date Type of study Total no patients First stage vs second stage Age Nulliparity Gestational age Duration of rupture of membranes Oxytocin augmentation Fetal distress Dystocia Duration of labor Skin to delivery Operative duration Length of hospitalization
Garrett, 2005 Retrosp 93 62 (66.7%) vs 31 (33.3%) 27.5 ± 6.7 vs 27.4 ± 5.3 75.8% vs 80.6% 40.2 ± 1.3 vs 40.1 ± 1.2 N/A 28/62 (45.2%) vs 4/31 (12.9%) N/A N/A N/A N/A N/A N/A
Allen, 2005 Retrosp 1623 1074 (66.2%) vs 549 (33.8%) 28 ± 5.4 vs 29 ± 5.0 N/A 39.9 ± 1.3 vs 39.2 ± 1.3 N/A N/A 12% vs 24% 84% vs 71% 11 vs 12 N/A N/A N/A
Cebekulu, 2006 Prospec 78 39 (50%) vs 39 (50%) 22.6 ± 4.7 vs 23.7 ± 6.3 N/A 39.3 ± 1.7 vs 39.3 ± 1.5 12 (1–52) vs 8.5 (2.50) 23/39 (64%) vs 15/39 (47%) N/A N/A N/A N/A 30 vs 45 2 vs 2
Alexander, 2007 Prospec 11,981 9265 (77.3%) vs 2716 (22.7%) 26.1 ± 6.5 vs 26.7 ± 6.4 7134 (77%) vs 2268 (84%) 40 ± 1.3 vs 40 ± 1.2 N/A N/A 14% vs 33% 84% vs 59% 17.2 ± 11.7 vs 18.5 ± 9.7 8.8 ± 5.5 vs 9.4 ± 6.0 48.3 ± 19.1 vs 51.3 ± 20.2 3.4 ± 1.5 vs 3.4 ± 1.3
Ojeme, 2008 Retrosp 627 428 (68.3%) vs 199 (31.7%) 29.3 ± 5.6 vs 29.4 ± 4.3 N/A 39.9 ± 1.9 vs 39.9 ± 1.5 N/A N/A 62% vs 52% 32% vs 64% N/A N/A N/A N/A
Winovitch, 2009 Retrosp 4049 3098 (60.8%) vs 951 (18.7%) 28.7 ± 6.6 vs 29.6 ± 6.5 2334 (75.3%) vs 744 (78.2%) 39.8 ± 1.2 vs 40.0 ± 1.1 N/A 1073/3098 (34.6%) vs 497/951 (52.3%) N/A N/A N/A N/A N/A N/A
Govender, 2010 Prospec 1091 975 (89.4%) vs 116 (10.6%) 25.7 vs 25.2 N/A 37.5 vs 38.4 N/A N/A N/A N/A N/A N/A 35.5 vs 41.6 3.7 vs 3.6
Seal, 2010 Retrosp 1826 1702 (93.2) vs 124 (6.8%) 24 vs 24.2 N/A 38.5 ± 1.16 vs 39.3 ± 0.97 N/A 1356/1702 (80.20%) vs 102/124 (82.26%) 12.1% vs 24% 77.4 vs 67% N/A 6.2 vs 8.4 35 vs 52 5.2 vs 6.4
Rabiu, 2011 Retrosp 347 245 (70.6%) vs 102 (29.4%) 29.0 ± 4.6 vs 28.2 ± 4.5 N/A 39.5 ± 1.4 vs 39.4 ± 1.4 N/A 92/245 (62.2%) vs 46/102 (58.2%) N/A N/A 20.0 ± 13.9 vs 28.8 ± 18.1 N/A 41.8 ± 14.7 vs 49.6 ± 14.7 8.01 ± 3.06 vs 10.97 ± 6.77
Sucak, 2011 Prospec 1389 1218 (87.7%) vs 171 (12.3%) 22.3 ± 4.8 vs 34.2 ± 5.2 N/A 38.8 ± 1.2 vs 40.2 ± 1.7 N/A 438/1218 (36%) vs 91/171 (54%) N/A N/A N/A N/A N/A 3.3 ± 0.8 vs 2.2 ± 0.4

All indices are tabulated as first vs second stage. Age was measured in years, gestational age in weeks, duration of rupture of the membranes in hours, duration of labor in hours, skin to delivery time in minutes, operative duration in minutes, length of hospitalization in days.

Fetal station was not recorded among studies included in the present review. Similarly, although some studies reported the duration of the second stage of labor of the group that was offered cesarean section at full dilatation, they did not provide results for controls, precluding comparison. Previous studies showed that there is a possible correlation between the prolonged duration of the second stage and adverse perioperative outcomes during cesarean delivery [24] . Low fetal station has also been implicated as a risk factor for unintentional lower uterine segment tears [25] . We strongly feel that these indices should be evaluated in future studies, as they may contribute to the occurrence of both maternal and neonatal complications (Table 2 and Table 3).

Table 2 Maternal morbidity (first vs second stage).

Author, date Wound infection Transfusion Postpartum hemorrhage Post op pyrexia Hypogastric ligation or hysterectomy (¥) Incision extension Lower segment tears Hematoma Cystotomy Re-admission Re-look laparotomy ICU Death
Garrett, 2005 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
Allen, 2005 10/1074 (0.9%) vs 6/549 (1.1%) 5/1074 (0.5%) vs 5/549 (0.9%) 31/1074 (1.3%) vs 26/549 (4.7%) 3/1074 (0.3%) vs 6/549 (1.1%) 1/1074 (0.2%) vs 1/549 (0.1%) N/A N/A N/A N/A N/A N/A N/A N/A
Cebekulu, 2006 3/39 (7.7%) vs 3/39 (7.7%) 0/39 vs 3/39 (7.7%) N/A N/A 0/39 vs 1/39 (2.6%) ¥ 0/39 vs 1/39 N/A N/A N/A 2/39 (5.1%) vs 3/39 (7.7%) 1/39 (2.6%) vs 0/39 0/39 vs 2/39 (5.2%) 0/39 vs 1/39 (2.6%)
Alexander, 2007 102/9265 (1%) vs 32/2716 (1%) 226/9265 (2%) vs 79/2716 (3%) 690/9265 (7%) vs 253/2716 (9%) N/A 80/9265 (0.9%) vs 33/2716 (1%) 14/9265 (0.2%) vs 10/2716 (0.4%) N/A 10/9265 (0.1%) vs 2/2716 (0.1%) 9/9265 (0.1%) vs 10/2716 (0.4%) N/A N/A N/A N/A
Ojeme, 2008 9/428 (1.7%) vs 0/199 27/428 (5.3%) vs 17/199 (14.3%) N/A N/A 1/428 (0.2%) vs 0/199 ¥ N/A N/A N/A N/A 9/428 (1.7%) vs 1/199 (0.8%) N/A 1/428 (0.2%) vs 0/199 N/A
Winovitch, 2009 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
Govender, 2010 27/975 (2.8%) vs 4/116 (3.5%) 13/975 (1.3%) vs 6/116 (5.2%) N/A 79/975 (8.1%) vs 34/116 (29.3%) 5/975 (0.5%) vs 2/116 (1.7%) 18/975 (1.8%) vs 26/116 (22.4%) 10/975 (1.0%) vs 11/116 (9.5%) N/A 4/975 (0.4%) vs 4/116 (3.5%) N/A 5/975 (0.5%) vs 2/116 (1.7%) N/A N/A
Seal, 2010 4/1702 (0.2%) vs 5/124 (4.03%) 3/1702 (0.17%) vs 6/124 (4.83%) 3/1702 (0.17%) vs 4/124 (3.23%) N/A 1/1702 (0.06%) vs 4/124 (3.23%) 5/1702 (0.29%) vs 10/124 (8.06%) N/A 1/1702 (0.06%) vs 3/124 (2.42%) 1/1702 (0.06%) vs 3/124 (2.42%) N/A N/A 2/1702 (0.1%) vs 5/124 (4%) 0/1702 vs 2/124 (1.6%)
Rabiu, 2011 20/245 (8.2%) vs 27/102 (26.5%) 24/245 (9.8%) vs 20/102 (19.6%) N/A 34/245 (13.9%) vs 35/102 (34.5%) 0/245 vs 3/102 (2.9%) 2/245 (0.8%) vs 9/102 (8.8%) 2/245 (0.8%) vs 8/102 (7.8%) N/A 9/245 (3.7%) vs 11/102 (10.8%) 2/245 (0.8%) vs 3/102 (2.9%) 1/245 (0.4%) vs 4/102 (3.9%) 1/245 (0.4%) vs 3/102 (2.9%) 1/245 (0.4%) vs 2/102 (2.0%)
Sucak, 2011 N/A 22/1218 (1.8%) vs 7/171 (4.1%) N/A N/A 5/1218 (0.4%) vs 2/171 (1.2%) N/A N/A N/A 15/1218 (1.2%) vs 7/171 (4.1%) N/A N/A N/A N/A

All indices are tabulated as first vs second stage. Studies investigating peripartum hysterectomy are marked with ¥. The rest data in this column represent hypogastric artery ligation.

Table 3 Neonatal morbidity (First vs Second stage).

Author, date Birth weight 5 min Apgar <3 5 min Apgar <7 Intubation not meconium Neonatal asphyxia Septicemia Fetal injury Neonatal seizures Neonatal department admission >24 h Death
Garrett, 2004 3714 ± 510 vs 3685 ± 480 N/A 0/62 vs 1/31 (3.2%) N/A N/A N/A N/A N/A N/A N/A
Allen, 2005 3689 ± 508 vs 3720 ± 452 3/1074 (0.3%) vs 3/549 (0.6%) N/A N/A 86/1074 (8.0%) vs 60/549 (11.0%) 13/1074 (1.2%) vs 5/549 (0.9%) N/A N/A N/A N/A
Cebekulu, 2006 3249 vs 3433 N/A 0/39 vs 7/39 (17.8%) N/A N/A 0/39 vs 3/39 (7.7%) 0/39 vs 1/39 (2.6%) N/A 3/39 (7.7%) vs 17/39 (17.8%) Ŧ 0/39 vs 2/39
Alexander, 2007 3482 ± 547 vs 3608 ± 466 32/9265 (0.4%) vs 13/2716 (0.5%) N/A 120/9265 (1.3%) vs 46/2716 (1.7%) N/A N/A 115/9265 (1.2%) vs 77/2716 (2.8%) 29/9265 (0.3%) vs 13/2716 (0.5%) 1992/9265 (22%) vs 477/2716 (18%) 7/9265 (0.08%) vs 4/2716 (0.15%)
Ojeme, 2008 N/A N/A 4/428 (0.8%) vs 0/199 N/A 40/428 (7.9%) vs 9/199 (7.6%) 9/428 (1.8%) vs 0/199 N/A N/A 45/428 (8.5%) vs 8/199 (6.7%) Ŧ N/A
Winovitch, 2009 3511 ± 542 vs 3641 ± 478 N/A 28/3098 (0.9%) vs 12/951 (1.3%) 24/3098 (0.8%) vs 12/951 (1.3%) N/A N/A N/A N/A 176/3098 (5.7%) vs 49/951 (5.2%) Ŧ N/A
Govender, 2010 N/A N/A 0/975 vs 5/116 (0.43%) N/A N/A N/A N/A N/A N/A 0/975 vs 0/116
Seal, 2010 2732.5 vs 2936.4 5/1702 (0.29%) vs 4/124 (3.26%) N/A 7/1702 (0.41%) vs 5/124 (4.03%) N/A 12/1702 (0.70%) vs 6/124 (4.84%) 2/1702 (0.12%) vs 6/124 (4.84%) 8/1702 (0.47%) vs 3/124 (2.41%) 28/1702 (1.65%) vs 12/124 (9.58%) 3/1702 (0.17%) vs 2/124 (1.61%)
Rabiu, 2011 3396 ± 527 vs 3390 ± 487 N/A N/A 8/245 (3.3%) vs 17/102 (16.7%) 23/245 (9.4%) vs 34/102 (33.3%) 29/245 (11.8%) vs 25/102 (24.5%) 2/245 (0.8%) vs 3/102 (2.9%) 1/245 (0.4%) vs 3/102 (3.9%) 67/245 (27.3%) vs 48/102 (47.1%) 3/245 (1.2%) vs 12/102 (11.8%)
Sucak, 2011 3310 ± 455 vs 3780 ± 635 N/A 7/1218 (0.6%) vs 2/171 (1.1%) N/A N/A N/A N/A N/A 50/1218 (4.1%) vs 13/171 (7.6%) Ŧ N/A

All indices are tabulated as first vs second stage. Studies that did not describe neonatal department admissions as NICUs are marked with Ŧ. The rest data in this column represent admission to NICU > 24 h.

Interestingly, no study participating in this review stated the possible advantageous role of disengagement of the fetal part via the transvaginal route during cesarean section performed with the woman placed in the lithotomy position. We could propose that a third person could easily manage to perform this maneuver. This way the surgeon could avoid deep lacerations of the lower uterine segment that seem to be caused by surgical maneuvers during the transabdominal effort to disengage the fetal part. In a recent review, Blickstein et al. stated that transvaginal disengagement seems to result in higher rates of maternal morbidity [26] . The authors stated, however, that this observation depends on adequate uterine relaxation, the patient's position during C/S and the type of uterine incision. Reverse breech extraction of a fetus presenting with a deeply engaged head (with the vertex at or below zero station) is an alternative maneuver that according to certain researchers is accompanied by less intraoperative blood loss, decreased need for uterine incision extension and lower operating times [27] and [28]. We can assume that this maneuver was carried out in certain cases presented in this review, but this was not described in any study.

Vertical versus low transverse abdominal wall incisions have been evaluated in a recent study of primary and repeat emergency cesarean deliveries. Although delivery of the fetus was found significantly quicker with vertical incisions it was not associated with differences on neonatal outcome (level of evidence II) [29] . Studies included in the present review did not provide data on the type of primary incision. It is our belief that future studies should consistently describe this index in order to evaluate its role in the occurrence of maternal and neonatal complications.

We also strongly suggest the evaluation of cesarean section-specific, low profile, Malmstrom-design vacuum extractors during C/S in the second stage of labor in order to assess their efficacy in reducing lower uterine segment tears. Their use could be useful either as a solo maneuver or simultaneously with the aforementioned transvaginal efforts of disengagement. Single-use plastic vacuum extractors seem to be followed by lower rates of vaginal and cervical lacerations during normal labor and it would be interesting to assess their place during cesarean section with a deeply engaged head [30] .

A previous descriptive study compared operative vaginal delivery with cesarean section in the second stage of labor and concluded that both of them are associated with serious neonatal complications [31] . The authors reported that whereas cesarean section was related to increased maternal morbidity, operative vaginal delivery resulted in higher rates of neonatal complications.

It is our belief that future studies should compare the possible benefits and adverse effects of all methods of operative delivery (instrumental vaginal or second stage cesarean section). This way we will be able to re-evaluate the place of the latter among specific cases of labor dystocia that do not represent absolute contraindications to vaginal delivery. Specific evaluation of the fetal station at the time of operative delivery and analysis of its interrelation with adverse effects in the case of the aforementioned methods would be valuable as knowledge in order to decide which one should be preferred among cases of labor dystocia. Our meta-analysis provides a useful tool for such studies, as we examined consistently all potential maternal and neonatal complications in the field of operative abdominal delivery.

4.1. Study limitations

Our meta-analysis was limited by certain differences in the design of the included studies. Whereas some studies were retrospective, others were prospective ( Fig. 1 ). Two studies were multi-institutional [9] and [11]. In the first one the data were collected from a specific database retrospectively [9] , whereas in the second one the data were retrieved from maternal and neonatal charts prospectively [11] . Criteria for inclusion and exclusion of participants were not always the same among included studies. Whereas nulliparity was predetermined as criterion of inclusion in some of them, others suggested only absence of prior C/S. Similarly, fetal anomalies and co-existing maternal morbidity (either pre-existing or presenting antenatally) were not always defined as criteria for exclusion. Finally, the cut-off for the gestational age at delivery was not the same in all included studies.

5. Conclusion

With respect to the heterogeneity of the included studies we report significantly higher rates of perinatal morbidity, both maternal and neonatal, among cases offered cesarean section in the second stage of labor. It is our belief that the morbidity which arises from these procedures renders imperative their direct evaluation with operative vaginal techniques, among specific cases of arrest of labor that do not represent absolute contraindications to vaginal delivery.

Conflict of interest

The authors report no conflict of interest.

Funding

None for all authors.

References

  • [1] J.A. Martin, B.E. Hamilton, S.J. Ventura, et al. Births: final data for 2009. Natl Vital Stat Rep. 2011;60:1-70 Crossref.
  • [2] A.F. Gei. Prevention of the first cesarean delivery: the role of operative vaginal delivery. Semin Perinatol. 2012;36:365-373 Crossref.
  • [3] P.E. Bailey. The disappearing art of instrumental delivery: time to reverse the trend. Int J Gynaecol Obstet. 2005;91:89-96 Crossref.
  • [4] T.Y. Leung, T.T. Lao. Timing of caesarean section according to urgency. Best Pract Res Clin Obstet Gynaecol. 2013;27:251-267 Crossref.
  • [5] D.F. Stroup, J.A. Berlin, S.C. Morton, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283:2008-2012 Crossref.
  • [6] V. Ehrenstein. Association of Apgar scores with death and neurologic disability. Clin Epidemiol. 2009;1:45-53 Crossref.
  • [7] J.P. Ioannidis, T.A. Trikalinos. The appropriateness of asymmetry tests for publication bias in meta-analyses: a large survey. CMAJ. 2007;176:1091-1096
  • [8] K. Garrett, A. Butler, W.R. Cohen. Cesarean delivery during second-stage labor: characteristics and diagnostic accuracy. J Matern Fetal Neonatal Med. 2005;17:49-53 Crossref.
  • [9] V.M. Allen, C.M. O’Connell, T.F. Baskett. Maternal and perinatal morbidity of caesarean delivery at full cervical dilatation compared with caesarean delivery in the first stage of labour. BJOG. 2005;112:986-990 Crossref.
  • [10] L. Cebekulu, E.J. Buchmann. Complications associated with cesarean section in the second stage of labor. Int J Gynaecol Obstet. 2006;95:110-114
  • [11] J.M. Alexander, K.J. Leveno, D.J. Rouse, et al. Comparison of maternal and infant outcomes from primary cesarean delivery during the second compared with first stage of labor. Obstet Gynecol. 2007;109:917-921 Crossref.
  • [12] D. Selo-Ojeme, S. Sathiyathasan, M. Fayyaz. Caesarean delivery at full cervical dilatation versus caesarean delivery in the first stage of labour: comparison of maternal and perinatal morbidity. Arch Gynecol Obstet. 2008;278:245-249 Crossref.
  • [13] K.C. Winovitch, D.A. Wing, D.C. Lagrew, J.H. Chung. The risk of acute neonatal morbidities in the delivery room after primary cesarean at term: influence of labor and stage. Am J Perinatol. 2009;26:545-551 Crossref.
  • [14] V. Govender, M. Panday, J. Moodley. Second stage caesarean section at a tertiary hospital in South Africa. J Matern Fetal Neonatal Med. 2010;23:1151-1155 Crossref.
  • [15] S.L. Seal, G. Kamilya, J. Mukherji, S.K. Bhattacharyya, A. De, A. Hazra. Outcome in second- versus first-stage cesarean delivery in a teaching institution in eastern India. Am J Perinatol. 2010;27:507-512 Crossref.
  • [16] K.A. Rabiu, A.A. Adewunmi, O.I. Akinola, A.E. Eti, A.O. Tayo. Comparison of maternal and neonatal outcomes following caesarean section in second versus first stage of labour in a Tertiary Hospital in Nigeria. Niger Postgrad Med J. 2011;18:165-171
  • [17] A. Sucak, S. Celen, E. Akbaba, S. Soysal, O. Moraloglu, N. Danisman. Comparison of nulliparas undergoing cesarean section in first and second stages of labour: a prospective study in a tertiary teaching hospital. Obstet Gynecol Int. 2011;2011:986506
  • [18] A. Boyle, U.M. Reddy, H.J. Landy, C.C. Huang, R.W. Driggers, S.K. Laughon. Primary cesarean delivery in the United States. Obstet Gynecol. 2013;122:33-40 Crossref.
  • [19] N. Jastrow, S. Demers, R.J. Gauthier, N. Chaillet, N. Brassard, E. Bujold. Adverse obstetric outcomes in women with previous cesarean for dystocia in second stage of labor. Am J Perinatol. 2013;30:173-178
  • [20] S. Miot, D. Riethmuller, K. Deleplancque, et al. Cesarean section for failed vacuum extraction: risk factors and maternal and neonatal outcomes. Gynecol Obstet Fertil. 2004;32:607-612 Crossref.
  • [21] C. Spencer, D. Murphy, S. Bewley. Caesarean delivery in the second stage of labour. BMJ. 2006;333:613-614 Crossref.
  • [22] R. Bahl, R.R. Patel, R. Swingler, M. Ellis, D.J. Murphy. Neurodevelopmental outcome at 5 years after operative delivery in the second stage of labor: a cohort study. Am J Obstet Gynecol. 2007;197:147.e1-147.e6 Crossref.
  • [23] A.T. Bleich, J.M. Alexander, D.D. McIntire, K.J. Leveno. An analysis of second-stage labor beyond 3 hours in nulliparous women. Am J Perinatol. 2012;29:717-722
  • [24] J.F. Sung, K.I. Daniels, L. Brodzinsky, Y.Y. El-Sayed, A.B. Caughey, D.J. Lyell. Cesarean delivery outcomes after a prolonged second stage of labor. Am J Obstet Gynecol. 2007;197:306.e1-306.e5 Crossref.
  • [25] T. Bergholt, J.K. Stenderup, A. Vedsted-Jakobsen, P. Helm, C. Lenstrup. Intraoperative surgical complication during cesarean section: an observational study of the incidence and risk factors. Acta Obstet Gynecol Scand. 2003;82:251-256
  • [26] I. Blickstein. Difficult delivery of the impacted fetal head during cesarean section: intraoperative disengagement dystocia. J Perinat Med. 2004;32:465-469
  • [27] S. Chopra, R. Bagga, A. Keepanasseril, V. Jain, J. Kalra, V. Suri. Disengagement of the deeply engaged fetal head during cesarean section in advanced labor: conventional method versus reverse breech extraction. Acta Obstet Gynecol Scand. 2009;88:1163-1166 Crossref.
  • [28] K.A. Frass, A. Al Eryani, A.H. Al-Harazi. Reverse breech extraction versus head pushing in cesarean section for obstructed labor. A comparative study in Yemen. Saudi Med J. 2011;32:1261-1266
  • [29] B.J. Wylie, S. Gilbert, M.B. Landon, et al. Comparison of transverse and vertical skin incision for emergency cesarean delivery. Obstet Gynecol. 2010;115:1134-1140 Crossref.
  • [30] G. Werkoff, O. Morel, P. Desfeux, et al. Kiwi vacuum extractor versus forceps and spatula: maternal and fetal morbidity evaluation in 169 fetal extractions. Gynecol Obstet Fertil. 2010;38:653-659 Crossref.
  • [31] S. Baloch, M. Khaskheli, I.A. Khushk, A. Sheeba. Frequency of second stage intervention and its outcome in relation with instrumental vaginal delivery versus caesarean section. J Ayub Med Coll Abbottabad. 2008;20:87-90

Footnotes

1st Department of Obstetrics and Gynecology, Athens University, Medical School, Alexandra Hospital, 80, Vasilisis Sofias Avenue, Greece

lowast Corresponding author at: 6, Danaidon Street, Halandri 15232, Greece. Tel.: +30 6947326459.