Obstetric and perinatal outcomes for twin pregnancies in adolescent girls.

This was a nine-year retrospective cohort study to investigate obstetric and perinatal outcomes in a cohort of adolescent girls with twin pregnancies from a major Australian tertiary centre in Brisbane, Australia. The adolescent cohort was aged <19 years and the control group was aged 20-24 years. The total study cohort comprised of 183 women. Of these, the adolescent cohort contained 29 girls (15.8%) and the control group comprised of 154 women (84.2%). Adolescent girls were less likely to delivery via an elective caesarean section compared to women in the control group (10.3% vs. 25.7%, p < 0.001). There were no differences in duration of labour, post-partum haemorrhage or perineal trauma rates. After controlling for the confounding effects of parity, chronicity and birth weight, birth <28 weeks remained significant (aOR 11.20, 95% CI 2.97-42.18, p < 0.001) for the adolescent cohort. There was a higher proportion of adolescents whose babies had an adverse composite perinatal outcome (87.9% vs. 69.5%, OR 3.20 95% CI: 1.40-7.31, p = 0.01) however significance was lost after adjusting for parity, chorionicity, birthweight and gestation at birth (aOR 3.27 95% CI: 0.95-11.31, p = 0.06). Our results show that obstetric and perinatal outcomes for twin pregnancies in teenagers were broadly similar compared to controls although the risk of extreme preterm birth was increased after controlling for confounders.


Results
Over the study period the total study cohort comprised of 183 women. The adolescent cohort comprised 29 girls (15.8%) and 154 women (84.2%) in the control group. There were 17 (58.6%) dichorionic diamniotic (DCDA) and 12 (41.4%) monochorionic diamniotic (MCDA) twins in the adolescent group and 65 (42.2%) DCDA and 89 (57.8%) MCDA twins in the control group respectively. Compared to the control group, adolescents were more likely to be nulliparous and not be married (Table 1).
After adjusting for parity, chorionicity, birth weight and gestation at birth intrapartum outcomes between the two groups were very similar. (Table 2) Although there were high overall rates of emergency caesarean section in both cohorts with almost one in two adolescent girls requiring this intervention this difference was not significant (48.3% vs. 38.6%, p = 0.17). Adolescent girls were less likely to delivery via an elective caesarean section compared to women in the control group (10.3% vs. 25.7%, p < 0.001). There were no differences in duration of labour, post-partum haemorrhage or perineal trauma rates.

Discussion
The key finding of this study of multiple pregnancy outcomes in adolescents is the risk of extreme preterm birth (<28 weeks) after controlling for the potentially confounding effects of parity, chorionicity and birthweight. Although there was a higher proportion of neonates with the composite adverse outcome in the adolescent group this did not reach statistical significance after adjusting for confounders. We also found a decreased risk of low birth weight (birth weight < 10 th centile for gestation and gender) (aOR 0.41 95% CI: 0.17-0.95, p = 0.04). Although approximately 90% of the adolescent cohort was primiparous, there were no differences in outcomes when this cohort was separately analysed.
There is good evidence that the risk of perinatal death is substantially higher in adolescents compared to women aged 20 to 24 years of age 5 . One reason for this, is likely to be the higher rates of preterm birth as  seen in our study. The causes for the higher rate of birth <28 weeks in the adolescent cohort are not immediately apparent. Furthermore, despite the higher rates of preterm birth in adolescent girls in this study appeared to be less likely to deliver a baby under the 10 th centile after controlling for parity and gestation at birth. This is in contrast to outcomes in teenage girls with singletons where there is a higher rate of low birth weight babies 6 . Although we demonstrate overall comparable neonatal outcomes for adolescents with twin pregnancies albeit with some specific differences, earlier studies have suggested that rates of adverse outcomes in teenagers are no higher compared to controls 7 . Although there is evidence showing that overall perinatal outcomes for singletons are poorer in teenagers 8 our results suggest that when confounders such as parity, chorionicity and gestation at birth are taken into account, outcomes are also poor in adolescents with multiple pregnancy. It is possible that these poorer pregnancy outcomes may be attributable to the sub-optimal pre-pregnancy health status of teenage girls as well as factors consistent with higher prevalence of poor socio-economic status in this cohort.
We did not observe poorer obstetric or intrapartum outcomes in our study. Overall caesarean section rates were high in both cohorts and there were no differences in total length of labour, rates of perineal trauma or postpartum haemorrhage. This is consistent with other published data 7 . Adolescents generally have higher rates of pregnancy complications including hypertensive disorders, antepartum haemorrhage, cephalopelvic disproportion and intervention for obstructed labour 1,2 . Consistent with other studies 9 we found lower rates of elective caesarean section in the adolescent cohort suggesting that they were more amenable to attempting a vaginal twin delivery. This is important as there is evidence that adolescent younger mothers are far more likely to have further children in adolescence and thus more children overall in their lifetime 10 . As adolescents are more likely to develop post-natal mental health issues 11 and difficulties with breastfeeding 12 which are known to be associated with caesarean birth, any reduction in operative rates could potentially mitigate these important post-partum issues.
While rates of adolescent pregnancy have declined 13 , complications of pregnancy and childbirth are the second leading cause of death for girls aged 15-19 years old 13 . As such, the importance of education and universal access to contraception is an important priority in addressing overall disparities in obstetric and perinatal   outcomes regardless of the number of fetuses. Indeed, there is evidence to suggest that the risks for adverse pregnancy outcomes in teenage pregnancies can be mitigated by high-quality maternity care 14 .
The strengths of our study are the inclusion of clinically relevant outcomes and the adjusting for relevant confounders such as chorionicity, parity, gestation at birth and birthweight. The limitations were related to the relatively small number of cases of only 29 teenagers, however the literature is extremely limited with only one paper from 1990 7 addressing this subject. Although overall, pregnancies in adolescent are not uncommon, multiple pregnancy is and there is a lack of information regarding pregnancy outcomes. We were unable to ascertain termination of pregnancy and miscarriage rates, outcomes that are pertinent to the study cohort. Our results are clinically relevant, highlighting the importance of engaging adolescents and supporting them both during and outside of pregnancy.