We congratulate the effort of Leung et al.1 in the cord blood gas analysis of the 118 twin pairs but we have the strongest reservation about the proposed time limit for twin-to-twin delivery interval. A time limit of 30 minutes was proposed based on the finding of increase in severe fetal acidosis (27%) and risk of fetal distress requiring obstetric intervention after delivery of the first twin. Knowing the limitation of the study (retrospective in nature, possibility of other bias and presence of other obstetric interventions that may affect the result) and authors admitting these, an arbitrary time limit of 30 minutes was still suggested. This was not based on any other clinical outcome indicators and we suspected this limit would not be acceptable by most other obstetricians. This time limit will also set a ‘benchmark’ for the maximum time interval for the delivery of the second twin and has significant medicolegal implication. If the second twin is not delivered by 30 minutes, are we going to stop all manoeuvre and proceed straight for an emergency caesarean section?

Actually looking at the study itself, there are four problems. Firstly, the incidence of combined vaginal–caesarean birth was 16.9%, which was highest among all the studies. Samra et al.2 reported an incidence of 4.3% in Birmingham in 1990. Persad et al.3 reported an increasing incidence of combined vaginal–caesarean birth from 1980 to 1999 (from 2% to 6.2%) in Grace Maternity Hospital, Nova Scotia, Canada. The period of the Leung et al.1 study was not mentioned in the article so comparison with other studies was not possible but the incidence of combined vaginal–caesarean birth for the second twin was at least two times higher than that of Persad et al.'s study3. Secondly, the authors did not try to explain for the significantly higher incidence of combined vaginal–caesarean delivery. It can be related to the ineffective or inadequate use of oxytocin after the delivery of the first twin, the lack of using external cephalic version, internal podalic version and breech extraction and above all the experience of the accoucheur. All these confounding variables or manipulations can directly or indirectly affect the success of the second twin delivery and thus, the delivery time interval, the cord blood result and the baby outcome. The degree of employing the above obstetric interventions was not mentioned in the original paper. Thirdly, the baby outcome was only measured by a single test (i.e. cord blood gas analysis) and there were no clinical parameters for comparison such as low Apgar score, need of intubation, need of neonatal intensive care, etc. What did a low pH at birth mean clinically and was the blood gas repeated 5 to 10 minutes later on those babies found acidotic? Only with these clinical indicators of baby's outcome can we conclude that the 30-minute limit is biochemically and clinically significant. It is unwise to decide for a major obstetric intervention (i.e. caesarean section) purely on the time limit and ignoring other favourable factors for continuation of vaginal delivery (head descending and normal fetal heart pattern). Fourthly, Leung et al.1 found a significantly higher incidence of fetal distress if the intertwin delivery interval was more than 30 minutes (73%) but the definition of ‘fetal distress’ was not clearly stated. It can be an abnormal cardiotocographic trace, deceleration, bradycardia or others, bearing in mind the difficulty in interpreting second stage cardiotocograph. Fetal distress was the indication for emergency caesarean section in 70% of their cases1, which was significantly higher than other series. In Persad et al.'s series3, the indication for emergency caesarean section for the second twin was fetal distress 18%, cord prolapse 18%, malpresentation 40% and failure to descend 22%. Without a clear definition of fetal distress, the conclusions drawn by the authors can be misleading and the intervention proposed for the second twin can be hazardous.

We are not objecting to setting an upper limit for the twin-to-twin delivery interval as a good clinical practice guideline, but the time limit, if ever set, should not be used as an absolute indication for abdominal delivery. Other confounding factors should also be considered: level of experience of staff present, competence of intrauterine manipulation, progress of labour, fetal heart pattern and wish of patients as well. Persad et al.3 proposed a more proactive approach including the use of uterine tocolysis as an adjunct to version and extraction of the second twin in transverse lie or breech extraction, which could reduce the need for casearean delivery. In the same study, a higher incidence of maternal complications (need of general anaesthetic, puerperal infection and longer hospital stay) was noted in women having combined vaginal–caesarean delivery. Delivery of twin pregnancy, especially the second twin, needs special experience and skill of the obstetricians and a clinical protocol should be developed to decide for continuation of vaginal delivery or for emergency caesarean section. The decision sometimes can be difficult but should not be purely based on a time limit.


  1. Top of page
  2. References
  • 1
    Leung TY, Tam WH, Leung TN, Lok IH, Lau TK. Effect of twin-to-twin delivery interval on umbilical cord blood gas in the second twins. Br J Obstet Gynaecol 2002;109: 6367.
  • 2
    Samra JS, Spillane H, Mukoyoko J, Tang L, Obhrai MS. Caesarean section for the birth of the second twin. Br J Obstet Gynaecol 1990;97: 234236.
  • 3
    Persad VL, Baskett TF, O'Connell CM, Scott HM. Combined vaginal–caesarean delivery of twin pregnancies. Obstet Gynecol 2001;98: 10321037.