A clinical prediction model to assess the risk of operative delivery

Is fetal cerebroplacental ratio an independent predictor of intrapartum fetal compromise and neonatal unit admission?

OBJECTIVE

We sought to evaluate the association between fetal cerebroplacental ratio (CPR) and intrapartum fetal compromise and admission to the neonatal unit (NNU) in term pregnancies.

STUDY DESIGN

This was a retrospective cohort study in a single tertiary referral center over a 14-year period from 2000 through 2013. The umbilical artery pulsatility index, middle cerebral artery pulsatility index, and CPR were recorded within 2 weeks of delivery. The birthweight (BW) values were converted into centiles and Doppler parameters converted into multiples of median (MoM), adjusting for gestational age using reference ranges. Logistic regression analysis was performed to identify, and adjust for, potential confounders.

RESULTS

The study cohort included 9772 singleton pregnancies. The rates of operative delivery for presumed fetal compromise and neonatal admission were 17.2% and 3.9%, respectively. Doppler CPR MoM was significantly lower in pregnancies requiring operative delivery or admission to NNU for presumed fetal compromise (P < .01). On multivariate logistic regression, both CPR MoM and BW centile were independently associated with the risk of operative delivery for presumed fetal compromise (adjusted odds ratio [OR], 0.67; 95% confidence interval [CI], 0.52-0.87; P = .003 and adjusted OR, 0.994; 95% CI, 0.992-0.997; P < .001, respectively). The latter associations persisted even after exclusion of small-for-gestational-age cases from the cohort. Multivariate logistic regression also demonstrated that CPR MoM was an independent predictor for NNU admission at term (adjusted OR, 0.55; 95% CI, 0.33-0.92; P = .021), while BW centile was not (adjusted OR, 1.00; 95% CI, 0.99-1.00; P = .794). The rates of operative delivery for presumed fetal compromise were significantly higher for appropriate-for-gestational-age fetuses with low CPR MoM (22.3%) compared to small-for-gestational-age fetuses with normal CPR MoM (17.3%).

CONCLUSION

Lower fetal CPR, regardless of the fetal size, was independently associated with the need for operative delivery for presumed fetal compromise and with NNU admission at term. The extent to which fetal hemodynamic status could be used to predict perinatal morbidity and optimize the mode of delivery merits further investigation.

Antenatal depressive symptoms and the risk of preeclampsia or operative deliveries: a meta-analysis

BACKGROUND

The purpose of the study was to investigate the association between depression and/or depressive symptoms during pregnancy and the risk of an operative delivery or preeclampsia, and to quantify the strength of the association.

METHODS

A search of the PubMed, SCI/SSCI, Proquest PsycARTICLES and CINAHL databases was supplemented by manual searches of bibliographies of key retrieved articles and review articles. We aimed to include case control or cohort studies that reported data on antenatal depression and /or depressive symptoms and the risk of an operative delivery and/or preeclampsia.

RESULTS

Twelve studies with self-reported screening instruments were eligible for inclusion with a total of 8400 participants. Seven articles that contained 4421 total participants reported the risk for an operative delivery, and five articles that contained 3979 total participants reported the risk for preeclampsia. The pooled analyses showed that both operative delivery and preeclampsia had a statistically significant association with antenatal depressive symptoms (RR = 1.24; 95% CI, 1.14 to 1.35, and OR = 1.63, 95% CI, 1.32 to 2.02, respectively). When the pre-pregnancy body mass indices were controlled in their initial design, the risk for preeclampsia still existed (OR = 1.48, 95% CI, 1.04 to 2.01), while the risk for an operative delivery became uncertain (RR = 1.01, 95% CI, 0.85 to 1.22).

CONCLUSIONS

Antenatal depressive symptoms are associated with a moderately increased risk of an operative delivery and preeclampsia. An abnormal pre-pregnancy body mass index may modify this association.

Association between levator hiatal dimensions on ultrasound during first pregnancy and mode of delivery

OBJECTIVES

To determine the association between levator hiatal dimensions, measured using transperineal ultrasound, in women during their first pregnancy and the subsequent mode of delivery, stratified by the indication for intervention.

METHODS

In this prospective observational study, 280 nulliparous women with a singleton pregnancy were invited for transperineal ultrasound examination at 12 and 36 weeks' gestation. Their levator hiatal dimensions were measured at rest, on pelvic floor muscle contraction and on Valsalva maneuver. The subsequent mode of delivery was classified into five categories: spontaneous vaginal delivery, instrumental vaginal delivery owing to fetal distress, instrumental vaginal delivery owing to failure to progress, Cesarean section owing to fetal distress and Cesarean section owing to failure to progress. Levator hiatal dimensions according to mode of delivery were compared by analysis of variance and Tukey's post-hoc test. Since multiple comparison tests were performed, the statistical significance level was corrected using the Bonferroni method.

RESULTS

Of the 252 women included, those who delivered by Cesarean section because of failure to progress had a significantly smaller levator hiatal transverse diameter on pelvic floor contraction at 12 weeks' gestation than did women who had a spontaneous vaginal delivery (Tukey's post-hoc test, P < 0.001). There was also a trend towards a smaller hiatal area on pelvic floor contraction at 12 weeks' gestation in women who delivered by Cesarean section because of failure to progress compared to women who had a spontaneous vaginal delivery (Tukey's post-hoc test, P = 0.005). In women who had an instrumental vaginal delivery because of failure to progress there was a trend towards a smaller levator hiatal anteroposterior diameter on pelvic floor contraction at 36 weeks' gestation compared with women who had a spontaneous vaginal delivery (Tukey's post-hoc test, P = 0.033).

CONCLUSIONS

Smaller levator hiatal dimensions on pelvic floor contraction during first pregnancy are associated with a subsequent instrumental vaginal delivery or a Cesarean section owing to failure to progress.

Imputation of systematically missing predictors in an individual participant data meta-analysis: a generalized approach using MICE

Individual participant data meta-analyses (IPD-MA) are increasingly used for developing and validating multivariable (diagnostic or prognostic) risk prediction models. Unfortunately, some predictors or even outcomes may not have been measured in each study and are thus systematically missing in some individual studies of the IPD-MA. As a consequence, it is no longer possible to evaluate between-study heterogeneity and to estimate study-specific predictor effects, or to include all individual studies, which severely hampers the development and validation of prediction models. Here, we describe a novel approach for imputing systematically missing data and adopt a generalized linear mixed model to allow for between-study heterogeneity. This approach can be viewed as an extension of Resche-Rigon's method (Stat Med 2013), relaxing their assumptions regarding variance components and allowing imputation of linear and nonlinear predictors. We illustrate our approach using a case study with IPD-MA of 13 studies to develop and validate a diagnostic prediction model for the presence of deep venous thrombosis. We compare the results after applying four methods for dealing with systematically missing predictors in one or more individual studies: complete case analysis where studies with systematically missing predictors are removed, traditional multiple imputation ignoring heterogeneity across studies, stratified multiple imputation accounting for heterogeneity in predictor prevalence, and multilevel multiple imputation (MLMI) fully accounting for between-study heterogeneity. We conclude that MLMI may substantially improve the estimation of between-study heterogeneity parameters and allow for imputation of systematically missing predictors in IPD-MA aimed at the development and validation of prediction models.

Comparison of childbirth care models in public hospitals, Brazil

OBJECTIVE

To compare collaborative and traditional childbirth care models.

METHODS

Cross-sectional study with 655 primiparous women in four public health system hospitals in Belo Horizonte, MG, Southeastern Brazil, in 2011 (333 women for the collaborative model and 322 for the traditional model, including those with induced or premature labor). Data were collected using interviews and medical records. The Chi-square test was used to compare the outcomes and multivariate logistic regression to determine the association between the model and the interventions used.

RESULTS

Paid work and schooling showed significant differences in distribution between the models. Oxytocin (50.2% collaborative model and 65.5% traditional model; p < 0.001), amniotomy (54.3% collaborative model and 65.9% traditional model; p = 0.012) and episiotomy (collaborative model 16.1% and traditional model 85.2%; p < 0.001) were less used in the collaborative model with increased application of non-pharmacological pain relief (85.0% collaborative model and 78.9% traditional model; p = 0.042). The association between the collaborative model and the reduction in the use of oxytocin, artificial rupture of membranes and episiotomy remained after adjustment for confounding. The care model was not associated with complications in newborns or mothers neither with the use of spinal or epidural analgesia.

CONCLUSIONS

The results suggest that collaborative model may reduce interventions performed in labor care with similar perinatal outcomes.

Influence of parity on fetal hemodynamics and amniotic fluid volume at term

OBJECTIVES

Pregnancy complications, particularly those associated with placental dysfunction, occur more frequently in nulliparous than in parous women. This difference may be a consequence of improved trophoblastic invasion and, as a result, improved placental function following previous pregnancy. Placental dysfunction in cases of fetal growth restriction may be identified by ultrasound assessment of fetoplacental hemodynamics and amniotic fluid volume. In this prospective observational study, we investigated whether differences in these measures of placental function exist between nulliparous and parous women, prior to active labor.

METHODS

Over a 2-year period, 456 nulliparous and 152 parous women with uncomplicated singleton pregnancies were recruited to this prospective observational study. Each participant underwent an ultrasound assessment prior to active labor, during which fetal biometry, umbilical artery, middle cerebral artery and umbilical venous Dopplers, as well as amniotic fluid volume, were assessed. All cases were followed up within 48 h of delivery. Ultrasound parameters and intrapartum outcomes were then compared between the nulliparous and parous groups.

RESULTS

Compared with nulliparous women, parous women had significantly higher fetal middle cerebral artery pulsatility index, cerebroplacental ratio and amniotic fluid volume. In nulliparous women, middle cerebral artery flow rate was also significantly higher and represented a greater percentage of umbilical venous flow than was observed in parous women.

CONCLUSION

Prior to the active phase of labor, ultrasound parameters indicative of placental function differ significantly between nulliparous and parous pregnancy, even amongst an uncomplicated, low-risk cohort.

Epidural analgesia and operative delivery: a ten-year population-based cohort study in The Netherlands

OBJECTIVE

To describe trends in the use of epidural analgesia (EA) and to evaluate the association of EA with operative deliveries.

STUDY DESIGN

In this population-based, retrospective cohort study, women with an intention to deliver vaginally of a term, cephalic, singleton between 2000 and 2009 (n=1378458) were included. Main outcome measures were labor EA rates, unplanned caesarean section (CS), and instrumental vaginal delivery (IVD) including deliveries by either vacuum or forceps. Data were obtained from the Perinatal Registry of The Netherlands and logistic regression analyses were used.

RESULTS

Among nulliparous, EA use almost tripled over the 10-year span (from 7.7% to 21.9%), while rates of CS and IVD did not change much (+2.8% and -3.3%, respectively). Among multiparous, EA use increased from 2.4% to 6.8%, while rates of CS and IVD changed slightly (+0.8% and -0.7%, respectively). Multivariable analysis showed a positive association of EA with CS, which weakened in ten years, from an adjusted OR of 2.35 (95% CI, 2.18 to 2.54) to 1.69 (95% CI, 1.60 to 1.79; p<0.001) in nulliparous, and from an adjusted OR of 3.17 (95% CI, 2.79 to 3.61) to 2.56 (95% CI, 2.34 to 2.81; p<0.001) in multiparous women. A weak inverse association between EA and IVD was found among nulliparous (adjusted OR, 0.76; 95% CI, 0.75 to 0.78), and a positive one among multiparous women (adjusted OR, 2.08; 95% CI, 2.00 to 2.16). Both associations grew slightly weaker over time.

CONCLUSIONS

A near triplication of EA use in The Netherlands in ten years was accompanied by relatively stable rates of operative deliveries. The association between EA and operative delivery became weaker. This supports the idea that EA is not an important causal factor of operative deliveries.

Effects of induction of labour versus expectant management in women with impending post-term pregnancies: the 41 week - 42 week dilemma

BACKGROUND

Post-term pregnancy, a pregnancy exceeding 294 days or 42 completed weeks, is associated with increased perinatal morbidity and mortality and is considered a high-risk condition which requires specialist surveillance and induction of labour. However, there is uncertainty on the policy concerning the timing of induction for post-term pregnancy or impending post-term pregnancy, leading to practice variation between caregivers. Previous studies on induction at or beyond 41 weeks versus expectant management showed different results on perinatal outcome though conclusions in meta-analyses show a preference for induction at 41 weeks. However, interpretation of the results is hampered by the limited sample size of most trials and the heterogeneity in design. Most control groups had a policy of awaiting spontaneous onset of labour that went far beyond 42 weeks, which does not reflect usual care in The Netherlands where induction of labour at 42 weeks is the regular policy. Thus leaving the question unanswered if induction at 41 weeks results in better perinatal outcomes than expectant management until 42 weeks.

METHODS/DESIGN

In this study we compare a policy of labour induction at 41 + 0/+1 weeks with a policy of expectant management until 42 weeks in obstetrical low risk women without contra-indications for expectant management until 42 weeks and a singleton pregnancy in cephalic position. We will perform a multicenter randomised controlled clinical trial. Our primary outcome will be a composite outcome of perinatal mortality and neonatal morbidity. Secondary outcomes will be maternal outcomes as mode of delivery (operative vaginal delivery and Caesarean section), need for analgesia and postpartum haemorrhage (≥1000 ml). Maternal preferences, satisfaction, wellbeing, pain and anxiety will be assessed alongside the trial.

DISCUSSION

This study will provide evidence for the management of pregnant women reaching a gestational age of 41 weeks.

TRIAL REGISTRATION

Dutch Trial Register (Nederlands Trial Register): NTR3431. Registered: 14 May 2012.

The association between attendance of midwives and workload of midwives with the mode of birth: secondary analyses in the German healthcare system

Background

The continuous rise in caesarean rates across most European countries raises multiple concerns. One factor in this development might be the type of care women receive during childbirth. ‘Supportive care during labour’ by midwives could be an important factor for reducing fear, tension and pain and decreasing caesarean rates. The presence and availability of midwives to support a woman in line with her needs are central aspects for ‘supportive care during labour’.

To date, there is no existing research on the influence of effective ‘supportive care’ by German midwives on the mode of birth. This study examines the association between the attendance and workload of midwives with the mode of birth outcomes in a population of low-risk women in a German multicentre sample.

Methods

The data are based on a prospective controlled multicentre trial (n = 1,238) in which the intervention ‘midwife-led care’ was introduced. Four German hospitals participated between 2007 and 2009.

Secondary analyses included a convenience sample of 999 low-risk women from the primary analyses who met the selection criterion ‘low-risk status’. Participation was voluntary. The association between the mode of birth and the key variables ‘attendance of midwives’ and ‘workload of midwives’ was assessed using backward logistic regression models.

Results

The overall rate of spontaneous delivery was 80.7% (n = 763). The ‘attendance of midwives’ and the ‘workload of midwives’ did not exhibit a significant association with the mode of birth. However, women who were not satisfied with the presence of midwives (OR: 2.45, 95% CI 1.54-3.95) or who did not receive supportive procedures by midwives (OR: 3.01, 95% CI 1.50-6.05) were significantly more likely to experience operative delivery or a caesarean. Further explanatory variables include the type of hospital, participation in childbirth preparation class, length of stay from admission to birth, oxytocin usage and parity.

Conclusion

Satisfaction with the presence of and supportive procedures by midwives are associated with the mode of birth. The presence and behaviour of midwives should suit the woman’s expectations and fulfil her needs. For reasons of causality, we would recommend experimental or quasi-experimental research that would exceed the explorative character of this study.

Factors influencing the likelihood of instrumental delivery success

OBJECTIVE

To evaluate risk factors for unsuccessful instrumental delivery when variability between individual obstetricians is taken into account.

METHODS

We conducted a retrospective cohort study of attempted instrumental deliveries over a 5-year period (2008-2012 inclusive) in a tertiary United Kingdom center. To account for interobstetrician variability, we matched unsuccessful deliveries (case group) with successful deliveries (control group) by the same operators. Multivariate logistic regression was used to compare successful and unsuccessful instrumental deliveries.

RESULTS

Three thousand seven hundred ninety-eight instrumental deliveries of vertex-presenting, single, term newborns were attempted, of which 246 were unsuccessful (6.5%). Increased birth weight (odds ratio [OR] 1.11; P<.001), second-stage labor duration (OR 1.01; P<.001), rotational delivery (OR 1.52; P<.05), and use of ventouse compared with forceps (OR 1.33; P<.05) were associated with unsuccessful outcome. When interobstetrician variability was controlled for, instrument selection and decision to rotate were no longer associated with instrumental delivery success. More senior obstetricians had higher rates of unsuccessful deliveries (12% compared with 5%; P<.05) but were used to undertake more complicated cases. Cesarean delivery during the second stage of labor without previous attempt at instrumental delivery was associated with higher birth weight (OR 1.07; P<.001), increased maternal age (OR 1.03; P<.01), and epidural analgesia (OR 1.46; P<.001).

CONCLUSION

Results suggest that birth weight and head position are the most important factors in successful instrumental delivery, whereas the influence of instrument selection and rotational delivery appear to be operator-dependent. Risk factors for lack of instrumental delivery success are distinct from risk factors for requiring instrumental delivery, and these should not be conflated in clinical practice.

How to interpret a small increase in AUC with an additional risk prediction marker: decision analysis comes through

An important question in the evaluation of an additional risk prediction marker is how to interpret a small increase in the area under the receiver operating characteristic curve (AUC). Many researchers believe that a change in AUC is a poor metric because it increases only slightly with the addition of a marker with a large odds ratio. Because it is not possible on purely statistical grounds to choose between the odds ratio and AUC, we invoke decision analysis, which incorporates costs and benefits. For example, a timely estimate of the risk of later non-elective operative delivery can help a woman in labor decide if she wants an early elective cesarean section to avoid greater complications from possible later non-elective operative delivery. A basic risk prediction model for later non-elective operative delivery involves only antepartum markers. Because adding intrapartum markers to this risk prediction model increases AUC by 0.02, we questioned whether this small improvement is worthwhile. A key decision-analytic quantity is the risk threshold, here the risk of later non-elective operative delivery at which a patient would be indifferent between an early elective cesarean section and usual care. For a range of risk thresholds, we found that an increase in the net benefit of risk prediction requires collecting intrapartum marker data on 68 to 124 women for every correct prediction of later non-elective operative delivery. Because data collection is non-invasive, this test tradeoff of 68 to 124 is clinically acceptable, indicating the value of adding intrapartum markers to the risk prediction model.

Uterine electromyography for identification of first-stage labor arrest in term nulliparous women with spontaneous onset of labor

OBJECTIVE

We sought to study whether uterine electromyography (EMG) can identify inefficient contractions leading to first-stage labor arrest followed by cesarean delivery in term nulliparous women with spontaneous onset of labor.

STUDY DESIGN

EMG was recorded during spontaneous labor in 119 nulliparous women with singleton term pregnancies in cephalic position. Electrical activity of the myometrium during contractions was characterized by its power density spectrum (PDS).

RESULTS

Mean PDS peak frequency in women undergoing cesarean delivery for first-stage labor arrest was significantly higher (0.55 Hz), than in women delivering vaginally without (0.49 Hz) or with (0.51 Hz) augmentation of labor (P = .001 and P = .01, respectively). Augmentation of labor increased the mean PDS frequency when comparing contractions before and after start of augmentation. This increase was only significant in women eventually delivering vaginally.

CONCLUSION

Contraction characteristics measured by uterine EMG correlate with progression of labor and are influenced by labor augmentation.

A framework for developing, implementing, and evaluating clinical prediction models in an individual participant data meta-analysis

The use of individual participant data (IPD) from multiple studies is an increasingly popular approach when developing a multivariable risk prediction model. Corresponding datasets, however, typically differ in important aspects, such as baseline risk. This has driven the adoption of meta-analytical approaches for appropriately dealing with heterogeneity between study populations. Although these approaches provide an averaged prediction model across all studies, little guidance exists about how to apply or validate this model to new individuals or study populations outside the derivation data. We consider several approaches to develop a multivariable logistic regression model from an IPD meta-analysis (IPD-MA) with potential between-study heterogeneity. We also propose strategies for choosing a valid model intercept for when the model is to be validated or applied to new individuals or study populations. These strategies can be implemented by the IPD-MA developers or future model validators. Finally, we show how model generalizability can be evaluated when external validation data are lacking using internal-external cross-validation and extend our framework to count and time-to-event data. In an empirical evaluation, our results show how stratified estimation allows study-specific model intercepts, which can then inform the intercept to be used when applying the model in practice, even to a population not represented by included studies. In summary, our framework allows the development (through stratified estimation), implementation in new individuals (through focused intercept choice), and evaluation (through internal-external validation) of a single, integrated prediction model from an IPD-MA in order to achieve improved model performance and generalizability.