Objective
We sought to determine the factors associated with selection of rotational instrumental vs cesarean delivery to manage persistent fetal malposition, and to assess differences in adverse neonatal and maternal outcomes following delivery by rotational instruments vs cesarean delivery.
Study Design
We conducted a retrospective cohort study over a 5-year period in a tertiary United Kingdom obstetrics center. In all, 868 women with vertex-presenting, single, liveborn infants at term with persistent malposition in the second stage of labor were included. Propensity score stratification was used to control for selection bias: the possibility that obstetricians may systematically select more difficult cases for cesarean delivery. Linear and logistic regression models were used to compare maternal and neonatal outcomes for delivery by rotational forceps or ventouse vs cesarean delivery, adjusting for propensity scores.
Results
Increased likelihood of rotational instrumental delivery was associated with lower maternal age (odds ratio [OR], 0.95; P < .01), lower body mass index (OR, 0.94; P < .001), lower birthweight (OR, 0.95; P < .01), no evidence of fetal compromise at the time of delivery (OR, 0.31; P < .001), delivery during the daytime (OR, 1.45; P < .05), and delivery by a more experienced obstetrician (OR, 7.21; P < .001). Following propensity score stratification, there was no difference by delivery method in the rates of delayed neonatal respiration, reported critical incidents, or low fetal arterial pH. Maternal blood loss was higher in the cesarean group (295.8 ± 48 mL, P < .001).
Conclusion
Rotational instrumental delivery is often regarded as unsafe. However, we find that neonatal outcomes are no worse once selection bias is accounted for, and that the likelihood of severe obstetric hemorrhage is reduced. More widespread training of obstetricians in rotational instrumental delivery should be considered, particularly in light of rising cesarean delivery rates.
Fetal head malposition in the second stage of labor is a significant risk factor for adverse maternal and neonatal outcomes, and is associated with high rates of both instrumental delivery and cesarean delivery. While some women will spontaneously deliver a malpositioned fetus, most require obstetric intervention. In cases of persistent malposition, the obstetrician must choose between a potentially difficult rotational instrumental delivery and a second-stage cesarean delivery.
Instrumental rotation of the fetal head has fallen out of favor in modern obstetric practice in much of the world, despite data showing low complication rates. It has recently been demonstrated that, while the majority of obstetricians considered rotation of the fetal head to be an acceptable intervention (97%), less than half (41%) had performed it within the previous year. Second-stage cesarean delivery is an increasingly common alternative, but carries a significant burden of maternal morbidity.
A small number of studies have compared the morbidity associated with different instruments used to effect rotational delivery, and have found low prevalence of adverse maternal and neonatal outcomes, as well as increased risk of some adverse events with emergency cesarean delivery. However, any comparison of delivery outcomes by rotational instruments vs second-stage cesarean delivery must confront the possibility that obstetricians systematically select more difficult cases for cesarean delivery, thereby introducing a selection bias. This study has 2 main objectives: first, to illuminate the factors that make an attempt at rotational instrumental delivery more likely, by modeling the obstetrician’s decision-making process; and second, to use propensity score stratification to create comparable groups that allow differences in maternal and fetal outcomes by delivery type to be tested reliably.
Materials and Methods
A cohort of 25,886 women with vertex-presenting, single, liveborn infants at term (37-42 completed weeks of gestation), aiming for vaginal delivery was identified over a 5-year period (January 2008 through October 2013) in a single tertiary obstetrics center in the United Kingdom. A subcohort of 868 women was identified with a confirmed cephalic fetal malposition in the second stage of labor. Of these, 833 underwent either cesarean delivery (n = 534) or successful instrumental delivery (n = 299), and 35 underwent failed instrumental delivery, followed by second-stage cesarean delivery.
Fetal malposition was defined as any cephalic position >45 degrees from direct occipito-anterior, and was diagnosed by digital examination. The rate of malpositions delivered by each method did not vary significantly across the study years. Deliveries where the obstetrician performed manual rotation of the fetal head followed by direct instrumental delivery were not considered to be cases of persistent fetal malposition, and were not included in the analysis. The indications and procedures for instrumental delivery in our center are defined in the operative vaginal delivery guidance from the Royal College of Obstetricians and Gynaecologists (RCOG), United Kingdom. The classification of and indications for operative vaginal delivery are materially identical to the American Congress of Obstetricians and Gynecologists Practice Bulletin number 17 on operative vaginal delivery.
Rotational instrumental delivery was carried out with either Kjelland forceps or ventouse. Ventouse devices available in the unit include posterior and rotational metal cups, Silastic cups (Dow Corning Corp, Hemlock, MI), and Kiwi Omnicups (Clinical Innovations, Murray, UT). Of the 334 successful instrumental deliveries, 62.0% (n = 207) were conducted with Kjelland forceps and 38.0% (n = 127) using ventouse.
Data regarding each woman’s pregnancy, labor, and delivery were recorded by midwives shortly after birth, and were subsequently obtained from the hospital’s data-recording system. The database is regularly validated by a rolling program of audits where the original case notes are checked against the information recorded in the database. No patient-identifiable data were accessed during this research, which was performed as part of a provision-of-service study for the obstetrics center. Individual medical records were not accessed at any stage. Institutional review board approval was therefore not required.
Characteristics of the maternal-fetal dyad were extracted from the database, including maternal age (at time of delivery), body mass index (BMI) (at first-trimester prenatal booking), parity (prior to delivery), ethnicity, and birthweight to the nearest gram. Also recorded were the time between diagnosis of second stage and delivery (time fully dilated), and the instrument selected. Gestational age (measured by crown-rump length at first-trimester ultrasound) was recorded to the nearest week. Only cases where birth occurred within the interval 37-42 weeks’ completed gestation were included. No adjustment was made for infants found to be small or large for gestational age. The indication for delivery was also classified into those where there was evidence of fetal compromise (including pathological fetal-heart tracing, abnormal fetal-blood sampling result, evidence of sepsis) and those where delivery was undertaken on other grounds (including failure to progress in second stage and maternal exhaustion). Deliveries were conducted under regional anesthesia (epidural or spinal), excepting a small number who required general anesthetic because of time constraints or failure of regional anesthesia during the procedure.
The level of experience of the obstetrician attempting delivery and the time at which the delivery took place were also recorded. Obstetricians were classified into 3 types using years of training as a proxy for experience. Type-1 and -2 obstetricians have 3-5 years and 5-10 years of obstetric training, respectively. Type-3 obstetricians typically have >10 years of clinical obstetric experience. Our study was conducted in a unit where 2 obstetricians are available to perform instrumental deliveries or cesarean deliveries at any time. The first is typically a type-1 obstetrician, and is always supported by an immediately available doctor with >5 years obstetric training: a type-3 obstetrician during the day, or type-2 overnight. All obstetricians had training in at least 1 method of rotational instrumental delivery, in line with RCOG training requirements.
Delay in neonatal respiration was recorded where spontaneous respiration was not achieved within 1 minute of delivery. Umbilical cord blood was obtained immediately following delivery, and the arterial pH recorded. Correlation between arterial and venous pH was checked to confirm accuracy of the measurements. Arterial pH was categorized as ≥7.1 or <7.1. A critical-incident form was generated at delivery in the case of any obstetric or neonatal emergency, including neonatal resuscitation, postpartum hemorrhage, shoulder dystocia, severe perineal trauma, maternal visceral injury, or any other event generating an obstetric emergency call. Maternal blood loss was measured by operating-room staff immediately after delivery, using suction blood collection and weighing of swabs and other pads. Blood loss was treated as a numerical variable to the nearest milliliter, and also categorized as minor (<1500 mL) or major (≥1500 mL). Severe perineal trauma was defined as any disruption to the anal sphincter complex. Simple groupwise comparisons of these outcomes for women undergoing rotational instrumental vs cesarean delivery were carried out using either Student t test or the Mann-Whitney test for numerical data, and Pearson χ 2 test for categorical data.
Any rigorous attempt to compare outcomes for the rotational-instrumental and cesarean-delivery groups is complicated by the fact that obstetricians may systematically select more difficult cases for cesarean delivery. This selection bias may involve physicians’ own training and experience, their immediate concern for fetal well-being, and anticipated fetal weight. An extensive set of these assignment-related variables are available in our data set, allowing us to explicitly model the obstetrician’s decision-making process. This allows us to use propensity score stratification to adjust for factors that influence the decision to move towards cesarean delivery. Propensity score stratification involves 2 stages. First, we build a statistical model for the treatment assignment (instrumental vs cesarean delivery), given a suitable set of predictors. The propensity score is the predicted probability of receiving the treatment derived from this first model. We then build a second set of models to estimate the effect of the treatment on each clinical outcome of interest, conditional on subjects’ propensity scores. This approach generates a balanced cohort of subjects whose baseline characteristics will be statistically similar, regardless of treatment status. For the purpose of estimating treatment effects, it is typically more robust than standard regression modeling, and may be formally justified under the potential-outcomes framework for causal inference. The effect of the covariates themselves on the clinical outcome is captured by the propensity score, and is never explicitly modeled.
For the purpose of estimating propensity scores, the 35 failed instrumental deliveries were included in the instrumental group, as the goal of this first-stage analysis was to model the physician’s initial treatment decision. For the purpose of estimating treatment effects, we ran 2 sets of second-stage analyses: 1 set with the 35 failed instrumental deliveries included, and 1 with them excluded.
Propensity scores were generated using a logistic-regression model predicting assignment to the instrumental-delivery group (the “treatment”). The regression model included 7 covariates found to be significantly different between women undergoing rotational instrumental and cesarean delivery, and which were thought to be clinically relevant: maternal age, maternal BMI, parity, birthweight, evidence of fetal compromise, time of delivery, and degree of experience of the delivering obstetrician. Although birthweight is unknown before delivery, it has been included because it plausibly may be anticipated by the physician and it strongly predicts the decision to move to cesarean delivery.
The resulting propensity scores were then stratified by quintile, and the balance of covariates between cases of rotational instrumental delivery and cesarean delivery checked within each quintile to verify that no significant differences remained. Adverse maternal and fetal outcomes were then modeled using linear and binary logistic regression, including the type of delivery and dummy variables for the propensity score quintiles as predictors. Findings were considered statistically significant at an alpha level of 0.05. Power calculations were performed by Monte Carlo simulation. All data analysis was conducted using the R statistical software package version 2.14.1 (R Foundation for Statistical Computing, Vienna, Austria).
Results
In all, 868 women with confirmed fetal malpositions in the second stage of labor were identified. In all, 534 (61.5%) were delivered directly by second-stage cesarean delivery; and 334 (38.5%) had an attempted rotational instrumental delivery, 299 of which resulted in successful delivery, and 35 of which were converted to second-stage cesarean delivery. Characteristics of the maternal-fetal dyad were compared between the instrumental-delivery and cesarean-delivery groups ( Table 1 ). Women in the cesarean-delivery group were more likely to be older ( P < .01), to have higher BMI ( P < .001), and to have babies with higher birthweights ( P < .01). In terms of events surrounding delivery, women in the cesarean-delivery group were more likely to have had a delivery involving evidence of fetal compromise ( P < .001), to have been delivered during the night ( P < .01), and to have been delivered by a less experienced obstetrician ( P < .001).
| Characteristic | All patients, n = 868 | Rotational instrumental delivery, n = 334 | Second-stage cesarean, n = 534 |
|---|---|---|---|
| Maternal age, y | 30.5 | 29.8 | 31.0 a |
| Maternal BMI, kg/m 2 | 25.5 | 24.4 | 26.2 b |
| Gestation, wk | 39.9 | 39.9 | 39.8 |
| Parity | |||
| 0 | 454 (52.3) | 166 (49.7) | 288 (53.9) |
| ≥1 | 414 (47.7) | 168 (50.3) | 246 (46.1) |
| Time fully dilated, min | 159.1 | 159.8 | 161.2 |
| Epidural | |||
| Yes | 601 (69.2) | 229 (68.6) | 372 (69.7) |
| No | 267 (30.8) | 105 (31.4) | 162 (30.3) |
| Obstetrician type | |||
| 1 | 405 (46.7) | 104 (31.1) | 301 (56.4) b |
| 2 | 366 (42.2) | 159 (47.6) | 207 (38.8) |
| 3 | 97 (11.2) | 71 (21.3) | 26 (4.9) |
| Birthweight, g | 3592 | 3532 | 3630 a |
| Fetal indication | |||
| Yes | 439 (50.6) | 116 (34.7) | 313 (58.6) b |
| No | 429 (49.4) | 218 (65.3) | 221 (41.4) |
| Ethnicity | |||
| White | 784 (90.3) | 303 (90.7) | 481 (90.1) |
| Southeast Asian | 53 (6.1) | 21 (6.3) | 32 (6.0) |
| Black | 9 (1.0) | 5 (1.5) | 4 (0.7) |
| Chinese | 8 (1.0) | 1 (0.3) | 7 (1.3) |
| Other | 14 (1.6) | 4 (1.2) | 10 (1.9) |
| Time of delivery | |||
| Day | 467 (53.8) | 206 (61.7) | 261 (48.9) |
| Night | 401 (46.2) | 128 (38.3) | 273 (51.1) a |
Table 2 shows the results of the logistic regression predicting assignment to rotational instrumental delivery (the “treatment”). Lower birthweight ( P < .01), lower maternal age ( P < .01), lower maternal BMI ( P < .001), higher parity ( P < .1), absence of evidence of fetal compromise ( P < .001), delivery during the daytime ( P < .05), and increased experience of obstetrician ( P < .001) are all significant predictors of assignment to rotational instrumental delivery.
| Variable | Odds ratio (95% CI) |
|---|---|
| Maternal age, y | 0.95 a (0.93–0.98) |
| Maternal BMI, kg/m 2 | 0.94 b (0.91–0.97) |
| Parity | |
| 0 | ref |
| ≥1 | 1.36 c (1.00–1.85) |
| Obstetrician type | |
| 1 | ref |
| 2 | 2.49 b (1.79–3.48) |
| 3 | 7.21 b (4.22–12.64) |
| Birthweight (per 100 g) | 0.95 a (0.92–0.98) |
| Fetal indication | 0.31 b (0.23–0.43) |
| Time of delivery | |
| Night | ref |
| Day | 1.45 d (1.05–2.01) |
The balance of covariates between the 2 groups was then checked within each propensity score quintile. There were no significant differences between the groups on any covariates in any of the quintiles. The results for the fourth quintile are shown as an example ( Table 3 ); note that the propensity score quintiles are based on the combined groups. The characteristics of the maternal-fetal dyad are now much more similar across the 2 groups than they were before stratification ( Table 1 ).
| Characteristic | Rotational instrumental delivery, n = 81 | Second-stage cesarean, n = 92 | P value |
|---|---|---|---|
| Maternal age, y | 29.4 | 29.8 | .60 |
| Maternal BMI, kg/m 2 | 23.8 | 24.0 | .69 |
| Parity | |||
| 0 | 42 (51.9) | 46 (50.0) | .81 |
| ≥1 | 39 (48.1) | 46 (50.0) | |
| Obstetrician type | |||
| 1 | 27 (33.3) | 31 (33.7) | .96 |
| 2 | 45 (55.6) | 58 (63.0) | |
| 3 | 9 (11.1) | 3 (3.3) | |
| Birthweight, g | 3531.7 | 3592.4 | .10 |
| Fetal indication | |||
| Yes | 58 (71.6) | 68 (73.9) | .73 |
| No | 23 (28.4) | 24 (26.1) | |
| Time of delivery | |||
| Day | 48 (59.3) | 52 (56.5) | .72 |
| Night | 33 (40.7) | 40 (43.5) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
