Cesarean Section for All Twins?

November 1, 2011

According to B. Russell’s “bad luck” mathematical paradox, a decision is bound to be false if based on the probability of success of another decision. Thus, a decision to perform a cesarean section (CS) based on the probability of a successful vaginal delivery (VD) is likely to be wrong.


Available for download in Word Document format


According to B. Russell’s “bad luck” mathematical paradox, a decision is bound to be false if based on the probability of success of another decision. Thus, a decision to perform a cesarean section (CS) based on the probability of a successful vaginal delivery (VD) is likely to be wrong. Obviously, the converse statement is equally true, leading to a ‘no win situation’, where any decision might prove to be wrong. Since the rate of twins and higher order multiples is reaching epidemic dimensions in most developed countries, the mode of delivery becomes a pertinent question, in which the crucial variable is the probability of a VD in twins. 

The Options 

There are three possibilities: all twins are candidates for a VD, none are candidates for VD (i.e., all should undergo CS), and some are candidates for VD. Any equation that includes probability estimation is prone to serious confounding. An incomplete list of confounders may include:

(1) Fertility history – with increasing rates of twins following assisted reproduction, the proportion of previously infertile women delivering twins is increasing, leading to higher rates of so-called “premium pregnancies” for which, any mode of delivery except CS (justified or not) may be declined. 

(2) Obstetrical history – there is at least a 1:5 to 1:10 chance that a given multiparous parturient carrying twins had a previous CS, which by large excludes her from VBAC (although VBAC in twins is not contraindicated in 50% of the cases).

(3) Obstetrical skills – concerns about singleton breech delivery have been extrapolated to twin births which include at least one breech or transverse lying twin in 50 to 60% of the cases. Operative deliveries, especially for twin B need some manual dexterity and experience. More CSs for twins will decrease the experience in VD, leading to a vicious circle in favor of CS. Clinicians who are less experienced in twin delivery are likely to report on adverse outcome following VD more often, or opt for a combined delivery when encountered with subtle difficulties in delivering twin B. 

(4) Age and size – generally, twins are delivered earlier and are smaller than singletons: two-thirds of twins are delivered before 36 weeks’ gestation (including 14% at <33 weeks) and half are LBW (including 10% VLBW by singleton standards). For these ostensibly friable twins, any mode of delivery except CS (justified or not) may be declined. 

(5) Maternal complications – the increased rate of maternal complications during pregnancy is a frequently used argument in favor of an elective “day-time” CS. 

Taken together, it seems that one may find an indication for CS in almost every twin pregnancy. Admittedly, the higher the CS rate in general, the more negligible is the contribution of CS in all twins to that rate. Thus one may question the wisdom of VD in any pair of twins. 

Evidence-Based Decisions

Research has not quantified the attributed risk of most of the above mentioned confounding variables and, indeed, focused on rather simple ones.

Presentation and size. Combinations of presentation, comprising vertex (Vx), breech and transverse, are usually grouped into four categories: 

Vx-Vx (~40%), Vx-Non-Vx (~30%), Non-Vx-Vx (~20%), and Non-Vx- Non-Vx (~10%). 

- Vx-Vx pairs are considered appropriate candidates for VD (with few exceptions related to size and/or gestational age);

- Vx-Non-Vx are considered conceivable candidates for VD (with many exceptions related to size and/or gestational age);

- Non-Vx-Vx and Non-Vx- Non-Vx are generally considered as an indication for CS, mainly due to lack of evidence about the safety of VD in breech-first pairs. Studies, which showed no difference in outcome for VD were criticized as having low statistical power or being non-randomized. 

Indeed, as with singleton breeches, a randomized study of adequate sample size is not anticipated, and therefore, one must rely on retrospective data. Recently, we have conducted a multi-center collaboration of a large sample of breech-first twin pairs delivered in 13 European centers. The data indicated that when breech-first twin weighed <1500 g, there is
a 2.4-times higher risk of depressed (<7) 5-min Apgar scores and a 9.5-times risk for neonatal mortality in VD as compared with CS. However, CS did not improve outcome when breech-first twin weighed >1500 g. Importantly, no case of “locked twins” has been encountered. 

VBAC. Studies on Vaginal Birth After Cesarean have shown that Vx-Vx and some Vx-Non-Vx pairs may be vaginally delivered. Anticipation for intrapartum manipulations, such as podalic version has generally excluded cases from VBAC.

Discordance. Significant intertwin size differences, per se, do not indicate CS. Nonetheless, discordance has been an argument against VD of Vx-Non-Vx pairs in whom twin B is smaller. 

Labor induction. The over-distended uterus is a relative contraindication for labor induction. However, recent studies suggested that pre-induction ripening of the cervix is effective and safe. Unfavorable cervical conditions seem to be no longer an obstacle for VD in appropriate candidates. 

In summary, twin gestations frequently involve maternal and fetal complications, and are quite often considered as “premium” pregnancies. Hence, many clinicians that follow the cliche “no high risk pregnancy should end with a high risk delivery” may deliver twins for many subtle reasons, other than clear-cut, evidence-based, indications. Thus, the decision for a CS in twins, intentionally or not, is based on qualitative variables that were not quantified and on quantitative variables that suggest no advantage for CS in the majority of cases.




1. Blickstein I, Schwartz Z, Lancet M, Borenstein R. Vaginal delivery of the second twin in breech presentation. Obstet Gynecol 1987; 69:774-6.

2. Blickstein I, Zalel Y, Weissman A. Cesarean delivery of the second twin after the vaginal birth of the first twin - misfortune or mismanagement ? Acta Genet Med Gemellol 1991;40: 389-94.

3. Blickstein I. The definition, diagnosis, and management of growth discordant twins: An international census survey. Acta Genet Med Gemellol 1991; 40:345-51.

4. Blickstein I, Weissman A, Ben-Hur H, Borenstein R, Insler V. Vaginal delivery for breech-vertex twins. J Reprod Med 1993; 38:879-882.

5. Blickstein I, Goldman RD, Smith-Levitin M, Greenberg M, Sherman D, Rydhstroem H. The relation between inter-twin birth weight discordance and total twin birth weight. Obstet Gynecol 1999; 93:113-6.

6. Manor M, Blickstein I, Ben-Arie A, Weissman A, Hagay Z. Case series of labor induction in twin gestations with an intrauterine balloon catheter. Gynecol Obstet Invest 1999; 47:244-6.

7. Blickstein I, Goldman RD, Kuperminc M. Delivery of breech-first twins: a multicenter retrospective study. Obstet Gynecol 1999; in press

8. Blickstein I, Smith-Levitin M. Twinning and twins. In: Current perspectives on the fetus as a patient. Chervenak FA, Kurjak A (eds), Parthenon Publishing, Lancs, 1996; ch. 39, pp. 507-525.

9. Blickstein I. Delivery of twins. In Labor and delivery. Cosmi EV (ed), Parthenon Publishing, Lancs, 1998; pp. 70-73.

10. Blickstein I, Smith-Levitin M. Multifetal pregnancy. In: Fetal disorders: Diagnosis and management. Petrikovsky BM, (ed.), John Wiley and Sons, Inc., New York, 1998; pp. 223-247.