We performed a retrospective (historical) cohort study using data gathered from the obstetric records of women with a singleton pregnancy delivering after 24 weeks of gestation between January 2004 and December 2008 at Liverpool Women's Hospital NHS Foundation Trust, Liverpool, UK. This trust provides maternity care for over 8000 deliveries per annum. Anonymous data were retrieved from the Meditech database and the labour and delivery records, which were entered at booking (approximately 12 weeks of gestation) and immediately after delivery. The maternal variables that we assessed were; age, race, height and weight at booking, parity, smoking status, gestation at delivery, delivery outcome including onset of delivery, mode of delivery, reason for delivery mode, labour length (first, second and third stages), estimated blood loss, second and third degree tears and episiotomy. Neonatal characteristics included sex, birthweight, Apgar score at 1 and 5 minutes after delivery, cord blood pH and the incidence of shoulder dystocia and stillbirth. Data were analysed using Statistical Package for the Social Sciences (SPSS), Version 16 (SPSS Inc., Chicago, IL, USA). Ethical approval was sought before commencing the study (Liverpool REC Ref 08/H1005/13).
Maternal BMI was calculated based upon maternal height and weight measurements provided during pregnancy booking between gestational weeks 10 and 12. Women were grouped into the following six BMI categories: ≤19.9 kg/m2 (underweight), 20–24.9 kg/m2 (normal weight) 25–29.9 kg/m2 (overweight), 30–34.9 kg/m2 (obese), 35–39.9 kg/m2 (very obese) and >40 kg/m2 (morbidly obese). In some analyses, to enable stringent exclusion criteria to be applied (and therefore smaller sample sizes) all women with a BMI of ≥30 kg/m2 were combined and classified as obese. Gestational age records were based upon an ultrasound scan taken at booking. Prolonged pregnancy was defined as delivery on or after 290 days (41+3 weeks) of gestation, as the hospital protocol is to refer women to a dedicated prolonged pregnancy clinic at 290 days. At the clinic, women have a cardiotocogram and ultrasound assessment of liquor volume. If these are normal, then women are given a choice between immediate induction or alternate day fetal surveillance. The majority of women (>90%) choose induction. The hospital protocol for IOL was the same in obese and non-obese women. Term delivery was defined as reaching 37–41+2 weeks (259–289 days), while preterm birth was defined as any birth before 37 completed weeks (259 days) of gestation.
The main outcome measures included length of gestation, prolonged pregnancy and delivery outcomes such as need for caesarean section following IOL. Measures used to assess the rate of complications in obese women with prolonged pregnancy undergoing IOL included length of labour, estimated blood loss and incidence of postpartum haemorrhage (defined as an estimated blood loss of >500 ml for vaginal deliveries and >1000 ml for delivery by caesarean section), second- and third-degree tear, episiotomy, retained placenta (defined as third stage >30 minutes), macrosomia (>4.5 kg), cord blood pH <7.0 and <7.2, Apgar score <7 after 5 minutes, incidence of shoulder dystocia and stillbirth.
Strict exclusion criteria were applied for each question addressed (). For all investigations we only included each mother once so the 5790 subsequent pregnancies to the same mothers during the study period were excluded from the cohort. Women with missing data for gestation, height or weight and those with gestations <168 days or >308 days or faulty data (random input error) (n = 3054) were also excluded.
Number of singleton deliveries during the study period and the application of exclusion and inclusion criteria for each study sample.
To examine the risk of prolonged pregnancy in our population of women we also excluded women with nonspontaneous labour onset before 290 days (n = 8625) because these pregnancies had an intervention preventing the true gestation from being ascertained.
For analysis of mode of delivery following IOL, and to examine the effect of high BMI on delivery complications following IOL for women with a prolonged pregnancy, only women with a pregnancy reaching 290 days were included in this study cohort because this follows the hospital's protocol for IOL for prolonged pregnancy. Therefore we excluded women who would not present to antenatal clinic with prolonged pregnancy because they would have been delivered electively early.
This is a pragmatic analysis of clinical data so the clinically most useful approach was used and comparisons were made between obese and normal weight women. We examined length of gestation by booking BMI group by Kruskal–Wallis test. Bivariate analyses were performed and chi-square statistic was computed to assess the association between maternal characteristics and delivery for each gestational age group. Multivariable logistic regression was performed to assess the risk of prolonged pregnancy with increased BMI while controlling for confounding variables noted from the bivariate analyses. Term pregnancies and women of normal weight were used as the reference categories in the regression model. Variables included in the model were maternal age, race, parity, maternal hypertension or diabetes (gestational and pregestational) and smoking status. Survival curves were constructed for each BMI category where delivery was the outcome event. Curves were compared by the Log-rank test and a statistical difference was taken at P< 0.05 level of significance.
To examine mode of delivery following IOL in women with a prolonged pregnancy and to compare their outcomes with those of their normal weight counterparts, vaginal, assisted vaginal (use of forceps or ventouse) and caesarean section delivery rates after IOL were examined by BMI category and chi-square tests were performed. To examine how the incidence of delivery complications after IOL in obese women with a prolonged pregnancy compared with their normal weight counterparts, Kruskal–Wallis or chi-square tests were performed. Outcomes were assessed according to mode of delivery, i.e. vaginal or caesarean section. For these analyses, all women with a BMI ≥ 30 were classified as obese and only women having IOL ≥ 290 days were included in the analysis. Analyses stratified by parity group (primiparous or multiparous) were also computed provided we retained statistical power. For all statistical analyses P< 0.05 was taken as the level of significance.