If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Complete uterine rupture is a rare peripartum complication often associated with a catastrophic outcome for both mother and child. However, little has been written based on large data sets about maternal and infant outcome after complete ruptures. This is partly due to the rarity of the event and the serious maternal and infant outcome; it is also partly due to the use of international diagnostic codes that do not differentiate between the less catastrophic partial rupture and more catastrophic complete uterine rupture. As uterine rupture is expected to increase due to increased cesarean delivery rates worldwide, it is important to know more completely about the outcome following complete uterine rupture.
We sought to explore risk factors associated with poor infant outcome in cases of complete uterine rupture.
This population-based study used data from the Medical Birth Registry of Norway, the Patient Administration System, and medical records. We included births with complete uterine rupture after start of labor in all maternity units in Norway during the period 1967 through 2008 (n = 244 births), identified among 2,455,797 births. Uterine ruptures were identified and further studied through a review of medical records. We estimated the associations between infant outcomes and demographic and labor risk factors using logistic regression analyses. Odds ratios with 95% confidence intervals for each risk factor were determined after adjustment for demographic factors and period of birth. The main outcome measure was infant outcome: healthy infant, intrapartum/infant deaths, hypoxic ischemic encephalopathy, and admission to the neonatal intensive care unit.
We identified 109 (44.7%) healthy infants, 56 (23.0%) infants needing neonatal intensive care unit admission, 64 (26.2%) intrapartum/infant deaths, and 15 (6.1%) infants with hypoxic ischemic encephalopathy. The highest number of intrapartum/infant deaths occurred in 1967 through 1977 (51.6%) and the fewest in 2000 through 2008 (15.0%). Unscarred uterine ruptures did not significantly increase intrapartum/infant deaths compared to scarred uterine ruptures. Placental separation and/or fetal extrusion had the highest odds ratio for intrapartum/infant deaths (odds ratio, 17.9; 95% confidence interval, 7.5–42.4). Time-to-delivery interval <20 minutes resulted in fewest intrapartum/infant deaths (9.9%), although there were 2 deaths at 10-minute interval. Time to delivery >30 minutes vs <20 minutes increased risk of death (odds ratio, 16.7; 95% confidence interval, 6.4–43.5).
Intrapartum/infant death after complete uterine rupture decreased significantly over the decades. Time to delivery >30 minutes and placental separation and/or fetal extrusion had the highest association with intrapartum/infant deaths after complete uterine rupture. Time to delivery <20 minutes limited the incidence of intrapartum/infant deaths.
Previous studies describing maternal and perinatal outcomes after complete uterine rupture are limited, most likely due to the rarity of the event. Most previous studies were based on registries using international diagnostic codes that did not differentiate between complete and partial rupture. Basing studies on clinical records often results in a small sample underpowered for detecting any associations between risk factors and outcomes following the rare event of uterine rupture. Previous studies have generally concentrated on the outcome of uterine rupture only in scarred uteri, and few have described the outcome in unscarred uteri.
To achieve a large sample, we studied complete rupture of scarred and unscarred uteri after the start of labor over 41 years based on population-based registry data. All medical records were reviewed for accuracy of diagnosis. In Norway, all mothers with 1 previous CS are offered a trial of labor unless there is an absolute contraindication against vaginal delivery. The trial of labor rate after previous CS is high with 64%.
Among those with a trial of labor, 80% undergo vaginal birth. We published earlier an article about risk factors for complete uterine rupture using a validated population of 22 out of a total of 48 maternity units in Norway in 1967 through 2008.
In this current study, we used the total pregnant population of all 48 units, to get a larger sample of complete uterine ruptures. Our aim was to explore risk factors of poor infant outcomes after complete uterine rupture.
This study was conducted to determine the risk factors associated with poor infant outcome in cases with complete uterine rupture.
Suspected diagnosis to delivery time <20 minutes limited the incidence of perinatal deaths, although there were 2 deaths at 10-minute interval, and no deaths <10 minutes.
What does this add to what is known?
Suspected diagnosis to delivery time >30 minutes, placental separation, loss of uterine contractions, and delivery after midnight significantly increased perinatal deaths, which nonetheless were decreasing in recent years.
Materials and Methods
All cases of uterine rupture after the start of labor were identified through diagnostic codes in the Medical Birth Registry of Norway (MBRN) (1967 through 2008, from all 48 maternity units in Norway) and the Patient Administration System (PAS) (1970 through 2008, from 21 units only). Established in 1967, the MBRN contains information on all births in Norway >16 weeks of gestation. The midwives attending a birth complete and send a standardized MBRN form within 7 days after delivery. The PAS is a local registry at each maternity unit that maintains records of all diagnoses for inpatients since 1970.
In the MBRN, the internal code used for uterine rupture before 1999 was 71; from 1999 and onwards, diagnostic codes from the International Statistical Classification of Diseases, 10th Revision
codes O710 and O711 (1999 through 2008). These codes did not specify rupture type. The type of rupture (complete or partial) was identified in the medical records using our definition of complete rupture as rupture of all uterine wall layers, including serosa and amniotic membranes.
All births with uterine rupture identified after the start of labor were studied further by the first author by reading the medical records of mothers, after visiting maternity units in Norway. Only those with complete ruptures were included in the study.
In addition, the first author studied the pediatric notes written about all births with ruptures and the medical records of infants who required follow-up up to the age of 5 years.
The Regional Ethics Committee (2010/1609-4) and the Data Inspectorate of Norway approved the study.
The 4 infant outcome measures, each categorized as Yes or No, included healthy infant (not requiring admission to the neonatal intensive care unit [NICU]); NICU admission due to severe asphyxia without encephalopathy or other causes; intrapartum/infant death, excluding those due to congenital malformations; and hypoxic ischemic encephalopathy (HIE), defined by the pediatrician as signs of cerebral irritation or depression or seizures in the presence of asphyxia with no resulting death. Infants admitted to the NICU for other causes were those admitted for any cause, excluding those with HIE or neonatal deaths. Intrapartum death is intrauterine fetal death during labor; infant death was defined as death after birth until 1 year old (excluding deaths due to congenital malformations). Cerebral irritation was defined by national consensus as a stage-1 encephalopathy, and cerebral depression as stage-2 encephalopathy.
Cerebral sequelae after HIE was defined as a permanent brain injury or impairment due to HIE as diagnosed by a pediatrician later in childhood. Such impairment can include epilepsy, developmental delay, motor impairment, neurodevelopmental delay, and cognitive impairment.
Potential risk factors included periods of birth, grouped into first period (1967 through 1977), second period (1978 through 1988), third period (1989 through 1999), and fourth period (2000 through 2008) (reference); uterine wall integrity, categorized into scarred (reference) and nonscarred; maternal age, categorized into <35 and ≥35 years; parity, categorized into para 1-2 (reference), para 0, and para ≥3; sudden loss of contractions; prolonged second stage of labor (from complete dilatation of cervix to delivery of infant), defined as duration, in nulliparous as >2 hours (without epidural) and >3 hours (with epidural), and in multiparous as >1 hour (without epidural) and >2 hours (with epidural); mode of delivery, categorized into delivery by emergency CS (reference), instrumental vaginal delivery, and noninstrumental vaginal delivery; placental separation and/or fetal extrusion (yes or no); and time of delivery, grouped into 08:00-15:00 (reference), 15.00-24:00, and 24:00-08:00 hours.
In addition, we calculated the time from clinical suspicion of uterine rupture to delivery of the infant in minutes, as identified in the medical records. The symptoms or signs of suspected rupture were: maternal and fetal, fetal signs only on cardiotocography (CTG) (abnormalities), maternal symptoms/signs only (defined as acute abdominal pains, vaginal bleeding, or preshock or shock), or none (defined as undocumented maternal or fetal symptoms or signs). Time to delivery was categorized later into <20 minutes (reference), 20-30 minutes, and >30 minutes. Furthermore, we studied the association between time to delivery stratified into the presence or absence of placental separation and/or fetal extrusion, grouped into the following: no placental separation and/or fetal extrusion and time <20 minutes (reference), placental separation and/or fetal extrusion and time <20 minutes, no placental separation and/or fetal extrusion and time 20-30 minutes, placental separation and/or fetal extrusion and time 20-30 minutes, no placental separation and/or fetal extrusion and time >30 minutes, and placental separation and/or fetal extrusion and time >30 minutes.
The incidences of outcomes were obtained from frequency tables. Cross-tabulation and logistic regression models were used to measure the association between different demographic and labor risk factors and different infant outcomes. Factors that were significant in bivariate analysis were included in separate multiple regression models adjusted for demographic factors (maternal age, parity, unscarred uterus, and periods of birth). Separate logistic regression models were used to estimate the association between time to delivery in minutes (continuous and categorical) and each infant outcome.
Cross-tabulation was used to measure the association between the variable time to delivery and each of the infant outcomes in separate models stratified for placental separation/fetal extrusion. The level of significance was set to P < .05. All analyses were performed using software (SPSS, Version 21; IBM Corp, Armonk, NY).
We identified 253 births with complete ruptures among 2,455,797 births recorded in 1967 through 2008 (0.1/1000). Nine of these births were antepartum fetal deaths and were excluded.
The final sample included 244 births (184 from the 48 units registered in MBRN and 60 from the 21 units in the PAS), comprising 106 live births and 38 intrapartum deaths.
Among 244 infants, 138 (56.5%) were gestational age 37-40 weeks, 94 (38.5%) ≥41 weeks, 11 (4.5%) 28-36 weeks, and 1 (0.4%) 23-28 weeks. The number of complete ruptures was highest (120) in 2000 through 2008 and lowest (20) in 1978 through 1988. Among 164 with scarred uteri, there were 157 who had 1 previous CS, 3 who had 2 previous CS, and 4 had non-CS scars from myomectomy or tubal corneal resection. Eighty mothers had unscarred uteri.
Compared to scarred uteri, unscarred uteri ruptures were detected significantly more often postpartum at laparotomy (51.3% vs 20.7%, P < .001).
In all, 109 (44.7%) healthy infants did not require admission to the NICU, 56 (23.0%) infants required NICU admission for severe asphyxia only or other causes, 64 (26.2%) infants were classified as intrapartum/infant deaths presumably due to the hypoxic effects of uterine rupture, and 15 (6.1%) surviving infants had HIE (Figure). Only 2 of the infants with HIE had cerebral sequelae diagnosed by 5 years of age. Among the 64 deaths, 38 died intrapartum, 25 during the neonatal period (within 28 days), and 1 at 2 months old.
A total of 169 ruptures were detected intrapartum during emergency CS, and 75 were detected postpartum at laparotomy after vaginal delivery. The symptoms and signs of rupture in each group are provided in Table 1. The most CTG abnormality noticed was fetal bradycardia; maternal symptoms were mainly abdominal pain, feeling unwell, and having tachycardia or signs of shock. Few presented with vaginal bleeding (10%), as most of bleeding was intraabdominal (concealed). The majority presented with combined severe pains and CTG changes. Those who had epidural during labor did not present with less pain than those without epidural. Those 9 who had no symptoms or signs documented had CS done due to prolonged labor. Fetal signs only were detected in a significantly larger percentage (60%) in 2000 through 2008 compared with 1967 through 1977 (11.4%). Mothers with detected ruptures postpartum had significantly higher percentage of maternal symptoms or signs vs mothers with detected ruptures intrapartum. The median time from suspected rupture to delivery was 20 minutes (Q1: 15, Q3: 30). From the time of clinically suspected uterine rupture, 91 infants were delivered <20 minutes, 104 delivered 20-30 minutes, and 49 delivered >30 minutes. We found that time-to-delivery interval >30 minutes was significantly higher in 1967 through 1977 at 35.5% (vs 2000 through 2008 at 11.7%; P < .005).
Table 1Symptoms or signs of uterine rupture in 244 births with complete uterine rupture
Tables 2 and 3 shows the association between demographic and labor risk factors and infant outcome after complete rupture. Compared to ruptures in 2000 through 2008, the occurrence of ruptures in 1967 through 1977 significantly increased the risk of intrapartum/infant death by 6.0 times, and decreased the percentage of healthy infants and admission to the NICU by 60% each. Ruptures in unscarred uteri resulted in slightly higher intrapartum/infant deaths than ruptures in scarred uteri, but this association was not statistically significant. Intrapartum/infant death was significantly increased by sudden loss of contractions and parity ≥3 (47% with unscarred uterus) vs parity 1-2. Prolonged second stage of labor significantly increased the risk of HIE by 4.5 times. Ruptures detected after noninstrumental vaginal delivery resulted in significantly lower percentage of intrapartum/infant deaths vs ruptures detected during CS (adjusted odds ratio [OR], 0.1; 95% confidence interval, 0.03–0.5). Placental separation and/or fetal extrusion occurred in 34.4% of complete ruptures and increased risk of intrapartum/infant deaths by 17.1 times. Delivery time after midnight significantly increased the risk for intrapartum/infant death by 4.3 times vs delivery between 08:00-15:00 hours (Table 3).
Table 2Association between demographic risk factors and infant outcomes following complete uterine rupture (N = 244)
There were 9 intrapartum/infant deaths at time interval <20 minutes (Table 4); 2 of them were at 10-minute interval, and were associated with placental separation. Time to delivery >30 minutes significantly increased the risk of intrapartum/infant death by 16.7 times vs delivery <20 minutes even after correcting for periods of birth. Regression analysis with time to delivery as a linear term estimated that every additional minute of time until delivery is associated with an approximate 10% increase in intrapartum/infancy death, 5% decrease in delivering a healthy infant, and 5% increase in admission to the NICU.
Table 4Time-to-delivery interval and infant outcomes
We found 84 cases with placental separation and/or fetal extrusion. This resulted in 25 (96.2%) deaths and only 1 healthy infant when time to delivery was >30 minutes (Table 5). When time to delivery was <20 minutes, placental separation and/or fetal extrusion resulted in 5 intrapartum/infant deaths and 4 cases of HIE. When time to delivery was >30 minutes without placental separation and/or fetal extrusion, there were 9 (39.1%) deaths. The OR for having a healthy infant after placental separation and/or fetal extrusion >30 minutes was 0.02 (95% confidence interval, 0.01–0.1) vs no placental separation and/or fetal extrusion and delivery <20 minutes.
Table 5Association between placental separation and/or fetal extrusion combined with time to delivery and infant outcome (N = 244 births)
Complete uterine rupture in Norway was associated with a large percentage of intrapartum/infant deaths (26.2%), especially in 1967 through 1977 (51.6%). The most recent study period (2000 through 2008) was associated with less mortality (15.0%). Unscarred uterine ruptures did not significantly increase intrapartum/infant deaths compared to scarred uterine ruptures. Parity ≥3, sudden loss of contraction, delivery after midnight, placental separation, and/or infant extrusion significantly increased the risk of intrapartum/infant deaths, particularly when the time from rupture to delivery exceeded 30 minutes. Time interval <20 minutes had least intrapartum/infant deaths. For every additional minute, there was a 10% increase in intrapartum/infant deaths and 5% decrease in the delivery of a healthy infant.
Meaning of the findings/clinical implications
Our present results confirmed the result of our previous study performed only on 21 units in Norway, showing increased rupture rates in recent years, yet decreased incidence of intrapartum/infant death following such ruptures with time.
included 1 area in Israel in 1988 through 1999 and found almost a similar percentage as ours (19.0%) for 42 ruptures. Methodological aspects such as sample size and ascertainment of cases and outcomes could affect the precision of results. We cannot exclude, however, that we in reality may have higher perinatal deaths following complete ruptures than these countries of previous studies. Most of our perinatal deaths were in the first period of birth (1967 through 1977), when also the time to delivery >30 minutes was more prevalent, indicating suboptimal management of labor. In our study, the percentage of healthy infants increased as intrapartum/infant death decreased with time, indicating improved obstetric management as well advances in neonatal resuscitation. The prevalence of HIE is expected to increase as a result of fewer deaths. However, the absolute number of HIE cases was very small, and most of them recovered without sequelae.
Ruptures in unscarred, compared to scarred, uteri were not significantly associated with more serious infant outcomes, which is in agreement with some previous studies
respectively). We were expecting deaths to be higher among ruptures of unscarred uteri due to a commonly lower index of suspicion and consequent delay in delivery. However, this was not the case here, which may be explained by many unscarred uterine ruptures in our study being detected postpartum, indicating that rupture occurred shortly before delivering the infant, with a short duration of hypoxia. Parity ≥3 and sudden loss of contractions were associated with increased intrapartum/infant deaths in our study; we speculate that less vigilance from health personnel in such groups may play a role. The highest risk of intrapartum/infant death was when rupture was associated with placental separation and/or fetal extrusion, in agreement with Bujold and Gauthier
Interestingly, delivery after midnight was associated with an increased risk of infant death and morbidity, possibly due to human factors (eg, fewer staff or suboptimal concentration and clinical judgment during night time).
We showed that time to delivery >30 minutes was a significant risk factor for intrapartum/infant death, even in present time as shown in previous studies.
found any perinatal mortality or morbidity when time-to-delivery interval was <18 minutes. Deaths and HE did occur in our study, even when time to delivery was <20 minutes, which is in agreement with Bujold and Gauthier;
we had 2 deaths at 10-minute interval. However, we showed that time to delivery <20 minutes limited the incidence of such deaths. Furthermore, we showed that placental separation and/or fetal extrusion had an important association with worse infant outcome regardless of duration. The association would be more profound as time from rupture to delivery increases ≥20 minutes, and especially >30 minutes. Currently, predicting placental separation is not possible, but one thing seems clear: the longer we wait with delivering the infant, the less chance for a healthy infant.
Strengths and weaknesses
Due to the rarity of the uterine rupture, there is a paucity of literature describing the clinical features and outcomes, especially in unscarred uteri.
This study is the largest thus far, especially regarding the number of unscarred uteri included, thereby increasing the precision of the results. Moreover, all information was extracted from the medical records reviewed by the first author, increasing the validity of the results and ensuring the diagnosis of complete ruptures vs partial ruptures, and accurately identifying studied outcomes and risk factors. Our sample also represents the whole Norwegian pregnant population, avoiding selection bias.
A weakness of this study is that we may have lost additional ruptures that were not recorded in the MBRN, as only 21 units were searched in the PAS. We previously published an article regarding risk factors for complete uterine rupture in a validated population of 1,411,268 births, where 163 complete ruptures were identified (0.11/1000).
In this current study, we used the total pregnant population, even those not fully validated, to get a larger sample of complete uterine ruptures (244 ruptures). To see whether missed cases affected the reliability of our current results, we repeated our analysis among only 163 ruptures from the fully validated population. We found similar results to our findings among the 244 ruptures regarding risk factors for infant outcomes following complete uterine ruptures. Therefore, potentially missed cases did not influence the study results.
The cases were collected from different periods of time. Therefore, we did a sensitivity analysis testing the association between different risk factors in the fourth period of time only (2000 through 2008) (results are not shown in article). The results showed that when limiting the cases to the most recent period, the effect of different risk factors on infant outcome was similar to the effect in the whole study period. Placental separation and/or fetal extrusion effect in 2000 through 2008 was still significant though the OR dropped from 17.1 in the whole study period to 11.4 in 2000 through 2008 only.
Intrapartum/infant deaths after complete uterine ruptures decreased significantly over the decades. Time-to-delivery interval <20 minutes resulted in fewest intrapartum/infant deaths, although there were 2 deaths at 10-minute interval. Time to delivery >30 minutes and placental separation and/or fetal extrusion at complete uterine rupture significantly increased intrapartum/infant deaths.