Results
A total 274 women had previous complete uterine ruptures. We studied 195 women after excluding 3 maternal deaths, 56 hysterectomies, and 20 sterilizations. Among these women, 88 women got pregnant, 16 of whom had miscarriages (18.2%) and 72 continued the pregnancy to delivery after 28 weeks (81.8%).
Among the 72 women, 37 (51.4%) had their previous ruptures within the LS and 35 (48.6%) had their ruptures outside the LS (Table 1). Some of previous ruptures included ruptures of scars that were not due to CS; These included three bicornute uterus ruptures, two traumatic ruptures after traffic accident, one hysterotomy rupture at 20 weeks for termination of pregnancy, one myomectomy scar rupture, one rupture after perforation during transcervical resection of myom, and one rupture at tubouterine area related to previous ectopic pregnancy. The previous ruptures that were outside LS included mostly ruptures in vertical scars from classical CS and ruptures in the lateral side of uterine wall; this was followed by ruptures in the posterior wall of uterine corpus as well as uterine fundus. The majority of previous ruptures occurred at 37-40 weeks (47.2%); only 19.4% occurred before the start of labour. Mothers who were <35 years old accounted for 61.1% of the study cohort. Previous ruptures resulted in 34 perinatal deaths (47.2%). The last period (2000-2011) had the highest percentage of new pregnancies.
Table 2 shows the outcomes of new pregnancies among mothers with previous complete uterine rupture. The majority delivered by elective CS (79.2%), whereas only two women delivered vaginally. The two vaginal deliveries occurred in the first period (1967-1977); one was spontaneous uneventful premature delivery at 32 weeks following previous uterine rupture in the LS, and the second was spontaneous premature twin delivery at 36 weeks following previous traumatic rupture in the uterine fundus at 28 weeks. No rupture occurred in these two cases, but the twin delivery ended with hysterectomy due to severe atonic postpartum haemorrhage.
Three new complete ruptures occurred, resulting in a rate of 4.2%. This is significantly higher than the rate of 0.16% in new pregnancies after previous CS without previous rupture (OR: 26.4; 95% CI: 5.3-81.5). The complete ruptures presented with acute abdominal pain, occurring at 29, 31, and 32 weeks in each of the three cases. There were six uneventful partial ruptures (8.3%), and two hysterectomies without the presence of uterine rupture (2.8%). The two hysterectomies included one case of severe postpartum haemorrhage after vaginal delivery as described above, and another case with placenta accreta after four preterm CSs. No maternal deaths were recorded among the 72 mothers, and the corrected perinatal mortality was 1.3%.
Delivery at 37-38 weeks occurred in 61.1% of pregnancies, and 36.1% experienced premature delivery. This is significantly higher than the rate of 9.6% among mothers with previous CS without previous rupture (OR: 5.3; 95% CI: 3.2-8.8).
Seven (9.7%) preterm deliveries occurred at 28-32 weeks. These included the three complete ruptures and two partial ruptures at 28 weeks. The remaining two included the previously described spontaneous premature vaginal delivery at 32 weeks and another emergency CS at 28 weeks due to pain, but no rupture was found; the previous rupture was in the LS at 37 weeks and resulted in perinatal death. There were 19 deliveries at 33-36 weeks without new ruptures, including 12 elective and 7 emergency CSs. The elective CSs included four at 36 weeks (two with twins, and two with previous ruptures at 37 weeks and perinatal deaths). All four previous ruptures were in the LS. In addition, six mothers delivered electively at 35 weeks, all with previous ruptures resulting in perinatal deaths; the first one had a previous rupture in the LS after trial of labour (TOL) at 39 weeks, the second had a previous rupture at 34 weeks after TOL; the third had a large rupture outside the LS after vacuum extraction in the unscarred uterus; the fourth had previous rupture and placenta percreta at 24 weeks; the fifth had previous rupture outside the LS at 39 weeks in an unscarred uterus after TOL; and the sixth had bicornuate uterus and rupture in one corneum after TOL at 37 weeks. There was one elective CS at 34 weeks due to ultrasound finding a very thin lower segment, and one at 33 weeks due to previous prelabour rupture in the LS at 33 weeks; both previous ruptures resulted in perinatal deaths.
The seven emergency CSs at 33-36 weeks were performed due to pain and suspected rupture or premature labour or preeclampsia (PE); they included five at 36 weeks and two at 34 weeks. All previous ruptures were in the LS except for one with previous rupture in the accessory horn.
Only 5 of 19 who delivered between 33-36 weeks had a previous rupture outside the LS (26.3%), whereas 13 (68.4%) had perinatal deaths as a result of their previous ruptures.
Other women had their planned CS at 38 weeks despite having a previous rupture outside the LS, or prelabour or at early gestational age (data not shown).
Table 3 shows that all three complete ruptures occurred in mothers who had previous ruptures outside the LS; the rate of repeat ruptures among those with previous rupture outside the LS was 8.6%. The partial rupture rate in this group was 11.4%, compared to the rate of 5.4% for those with previous rupture in the LS. The difference was not significant (OR: 2.3; 95% CI: 0.3-26.3). Mothers with an inter-delivery interval of 2-3 years did not develop repeat complete ruptures. There was a tendency toward an increased rupture rate when the inter-delivery interval was 1 year or ≥ 4 years vs. 2-3 years, but the difference was not significant.
There was a tendency toward an increased rupture rate when previous rupture occurred before labour start or at gestational age < 37 weeks, but this also did not reach significance.
The details of the three repeat complete ruptures are provided in Table S4A. All had immediate CS upon arrival to the hospital. They resulted in one stillbirth due to asphyxia as a result of rupture, early neonatal death due to severe multiple congenital malformations without asphyxia, and one infant with moderate asphyxia.
The details on mothers who developed partial ruptures are provided in Table S4B. Five presented with mild abdominal pains or irregular contractions, whereas one was discovered coincidentally at elective CS.
Discussion