Is internal podalic version a lost art? Optimum mode of delivery in transverse lie.AR Chauhan, TT Singhal, VS Raut
Department of Obstetrics & Gynaecology, Seth G. S. Medical College & K. E. M. Hospital, Parel, Mumbai - 400 012, India. , India
AIM: To study the changing trend in the delivery of transverse lie, and its effect on neonatal outcome, in a developing country. SUBJECTS AND METHOD: This is a retrospective study involving records of 12 years of all patients with transverse lie. Neonatal outcome of births by internal podalic version (IPV) and lower segment caesarean section (LSCS) were compared. RESULTS: In the first six years, 37.3% of transverse lie underwent IPV and 62.7%, LSCS. In the next six years, 15.8% underwent IPV and 84.2%, LSCS. 87.7% and 12.3% of live babies were delivered by LSCS and IPV respectively. 52% of the live born IPV were discharged compared to 95% of LSCS babies. Neonatal outcome was best when IPV was performed on second twin. CONCLUSION: IPV has a role in the delivery of second twin, pre-viable and dead babies.
Keywords: Adult, Cesarean Section, methods,Female, Fetal Death, epidemiology,etiology,Gestational Age, Human, India, epidemiology,Infant, Newborn, Labor Complications, epidemiology,mortality,Labor Presentation, Pregnancy, Pregnancy Outcome, Retrospective Studies, Twins, Version, Fetal, methods,
Internal Podalic Version (IPV) is an ancient procedure and was extensively practised by Hippocrates, who recommended cephalic version for all presentations other than the head. Aetius, Celsus and others at different times pointed out the fallacies of the Hippocratic teaching and the advantages of podalic version. Supported by Galen, IPV continued to be in favour till the sixteenth century. IPV had a revival that was initiated by Ambroise Pare (1510-90). Pare was the first to describe clearly and to point out the possibilities and the advantages of podalic version. About a hundred years later, an alteration in the technique of the operation was suggested by Portal – the bringing down of one leg instead of both, as was the custom up till then.
In modern obstetrics, caesarean section is the method of choice for the delivery of babies in transverse lie, and IPV is performed less frequently. While this is true in most cases, does IPV still have a role to play in modern obstetrics, or is it indeed a lost art?
There has been a drastic decrease in the number of IPVs performed on babies with transverse lie in the past decade. Hence, this study was carried to evaluate the changing trend in the mode of delivery of transverse lie, maternal and neonatal outcomes with respect to the mode of delivery and whether IPV has a role to play in the management of second twin.
The relavant data from a tertiary care centre and teaching hospital over a period of twelve years i.e. from 1986 to1997 were analysed retrospectively. All women who presented with transverse lie in labour were included in the study. Age, parity, antenatal complications, and condition of the foetus on admission were noted. The mode of delivery whether IPV or lower segment caesarean section (LSCS) was recorded. Intrapartum complications, maternal and neonatal outcome were noted. For the purpose of comparison, the study was divided into two parts covering six years each: Part 1 from 1986 to 1991 and Part 2 from 1992 to 1997. Data was analysed using paired “t” test.
During 1986 to 1991 (Part 1) there were a total of 24,456 deliveries of which 99 (0.4%) presented with transverse lie. During 1992 to 1997 (Part 2) there were a total of 29,275 deliveries of which 95 (0.32%) presented with transverse lie. The total number of IPVs in both parts was 50 and the total number of LSCS was 144. Though the number of transverse lie has remained the same over the years, the number of IPVs has halved. There were a total of 50 IPVs performed, of which in Part 1, 35.3% of cases with transverse lie were delivered by IPV, while in Part 2 only 15 (15.8%) were delivered by IPV, which is statistically significant (P < 0.001).
Most IPVs were performed on patients between 20-30 years. IPV was most often performed on second gravidae in both parts of the study. However in the earlier part of the study (Part 1), as many as eight IPVs were performed on primigravidae, though all were for stillbirths.
[Table - 1] shows antenatal and intrapartum risk factors. Most patients with complications such as placenta praevia, twins, previous LSCS, short stature, fibroid with pregnancy, rheumatic heart disease, bad obstetric history, were sectioned in both groups. Surprisingly, three IPVs were performed on patients with grade I placenta praevia and one IPV was performed on a patient with previous LSCS. However all of these were done in the earlier part of the study i.e. Part 1. A macerated stillbirth with impacted shoulder was delivered vaginally by IPV in Part 1.
Maternal outcome was good in all cases in both parts of the study. There were no uterine ruptures seen. Three mothers had mild post partum haemorrhage, which was managed conservatively. No blood transfusion was required.
Majority of live births in both groups underwent LSCS, 64 of 78 in Part 1 and 75 of 80 in Part 2 [Table - 2]. Another trend noted was LSCS for stillbirths. While no patient with stillbirth was sectioned in Part 1, as many as five of 15 fresh stillbirths were sectioned in Part 2, which was a statistically significant increase. One of these was an 1800 gm baby in 1995, on whom an IPV was unsuccessfully attempted. All live babies on whom IPV was performed were delivered alive.
As seen in [Table - 3], 158 foetuses were live on admission. Of these, 19 underwent IPV; 14 of 78 in Part 1 and only five of 80 in Part 2, showing a declining trend for IPV in the presence of a viable foetus, which was statistically significant. 139 patients were delivered by LSCS, 64 and 75 in the two parts of the study respectively. None of the IPVs resulted in stillbirths, whereas one live baby in each part of the study who underwent LSCS, resulted in a stillbirth.
When neonatal mortality was analysed by the mode of delivery, eight of 19 IPVs resulted in neonatal death, whereas only nine neonates out of 139 died in the LSCS group, indicating a statistically increased risk of neonatal mortality with IPV.
The main causes of neonatal death were birth asphyxia, prematurity and neonatal sepsis, as seen in [Table - 4]. It was found that birth asphyxia led to the death of three of 19 babies (15.78%) delivered by IPV as compared to only seven of 139 babies (5.03%) delivered by caesarean section, which was statistically significant.
On comparing the neonatal outcome of live born babies by weight, it was found that, of the babies delivered by caesarean section, 33% weighing between 1 to1.49kg went home in Part 1 while 50% were discharged in Part 2 [Table:5A]. Similarly 77% of LSCS babies between 1.5 to 1.9kg went home in Part 1 while 93.3% went home in Part 2. A better outcome was seen in Part 2, probably due to improved neonatal outcome in the latter part of the study. The neonatal outcome of babies in our series who underwent IPV has deteriorated over time because IPV was performed only on moribund babies in the latter years. Another possible factor is the decreased skill of the obstetrician in intrauterine manipulations, with greater reliance on LSCS.
[Table:5B] shows the neonatal outcome with respect to gestational age. Of the babies delivered by LSCS, while no baby below 32 weeks went home in Part 1, 50% went home in Part 2. Again 50% of the babies between 32 to 34 weeks went home in Part 1, while 88% went home in Part 2. This showed that better neonatal care facilities improved the outcome, irrespective of the mode of delivery.
While studying the neonatal outcome of IPV alone, it was found that a better outcome was observed when IPV was performed on second twin. [Table - 6] shows that 62.5% of singleton babies were neonatal deaths, but only 27.3% of second twin ended similarly.
No second twin underwent LSCS for transverse lie in Part 1 while three LSCS were performed in Part 2 [Table - 7]. Interestingly it was found that the neonatal outcome in both these groups was identical.
Though the incidence of transverse lie has not changed over the years, the incidence of IPV performed for the same has definitely decreased in the latter half of the study, which is statistically significant.
In earlier years, IPV was performed on all stillbirths, nonviable babies, second twin and in a few singleton pregnancies in which the babies were of low birth weight or were preterm. However with improved neonatal facilities today, low birth weight and preterm babies have a better chance of survival. This, combined with the increased safety of caesarean section is the important reason for the change in the trend of delivery of babies in transverse lie. But, this has resulted in caesarean section being performed even on dead babies and a large number of second twins.
Our data shows a decrease in incidence of IPV in singleton pregnancy in the later years, both, for viable and previable or dead babies. As many as five LSCS were performed on dead babies in Part 2 i.e. from 1992-1997, including one case where IPV was unsuccessfully attempted. IPV by skilful obstetricians in these patients may have prevented an LSCS. With the decrease in trend for IPV, most obstetricians are not well versed in the art of successful version and many junior obstetricians have not even seen it being performed. It is difficult to compare our data to international literature, since there are no references for IPV in singleton pregnancies in recent literature. However in developing countries IPV may have a role to play in the delivery of previable and dead babies.
The optimal route of delivery for a twin pregnancy with vertex- breech or vertex-transverse presentation has remained a subject of controversy. In our study we found that neonatal outcome of IPV was best when it was performed on second twin. While 62.5% of live born singletons in transverse lie resulted in neonatal deaths, only 27.3% of second twin ended similarly. All second twin were delivered by IPV in Part 1 while as many as three babies were delivered by caesarean section in Part 2. Interestingly we found that the neonatal outcome in both these groups was identical.
A recent CME review article for the delivery of the non-vertex second twin found that a vaginal delivery is no more dangerous than a caesarean delivery for low birth weight twins, regardless of presentation, and in twins with expected birth weights > 1500gm, vaginal delivery is reasonable, provided experienced staff are available.
Dufour et al conducted a retrospective study of 35 cases of IPV followed by breech extraction of second twin. They found that internal version is the only alternative (together with external version) to caesarean section, allowing rapid delivery of the second twin. Maternal prognosis is excellent and foet al prognosis is good if contraindications are avoided.
Rabinovici et al conducted a prospective study for the management of the second non-vertex twin. 60 twin deliveries after the 35th gestational week with vertex- breech, and vertex- transverse presentations were managed according to a randomised protocol of vaginal or abdominal delivery. Of 21 patients with vertex- transverse presentation, 12 were delivered vaginally by IPV and breech extraction while nine were delivered by LSCS. They found that the neonatal morbidity was similar in both the groups. There was no case of birth trauma or intraventricular haemorrhage in either group. However there was a significantly higher incidence of febrile morbidity in the LSCS group. They concluded that in selected twin pregnancies with vertex- breech or vertex- transverse presentations at a gestational age 35 weeks or more, a vaginal delivery could be performed without increased risk to the mother or the infant.
However, in a retrospective study from Poland, Piekarski et al found that vaginal delivery with IPV of the second twin in vertex- transverse presentations was related to an increased risk of lower five minute Apgar score and an increased risk of birth trauma compared to caesarean section.
Drew et al conducted a study to gauge the quality of survival of the neonate, where the second twin was born by breech extraction following internal version. 25 sets of twins were fully assessed as children ranging in age from 2 to 12 years. They found that growth and psychological scores were not significantly different between the first and second twin. Due to small numbers, their results did not have statistical significance but they did show that the majority of infants so born are doing well.
Several newer modalities have been described to improve the outcome of the neonate. Rabinovici et al in a prospective study of IPV on the second twin with unruptured membranes, showed good neonatal outcome and no birth injuries. IPV under intrapartum ultrasound guidance after the first twin has delivered also gives a better neonatal outcome. Dufour et al in a prospective study used intravenous nitroglycerin in high doses (0.1-0.2mg per kg) to relax the uterus while performing IPV. They found that this method gave good maternal and neonatal outcome.
There is a changing trend in the mode of delivery of transverse lie, more in favour of caesarean section. The neonatal outcome is definitely better with caesarean section, especially in view of improved neonatal facilities and better neonatal survival. We believe that though caesarean section is certainly a better option for single viable babies in transverse lie, IPV has a role to play in the delivery of the second twin. It may also be attempted in the delivery of nonviable and dead babies in the absence of contraindications, especially in developing countries.
[Table - 1], [Table - 2], [Table - 3], [Table - 4], [Table - 5], [Table - 6], [Table - 7], [Table - 8]