Intrapartum care
DOI: 10.1111/1471-0528.13394 www.bjog.org
Uterine rupture: trends over 40 years n,a,d A-K Daltveit,e,f S Vangena,d I Al-Zirqi,a,b B Stray-Pedersen,b,c L Forse a Norwegian National Advisory Unit on Women’s Health, Oslo, Norway b Women and Children’s Division Rikshospitalet, Oslo University Hospital, Oslo, Norway c Faculty of Medicine, University of Oslo, Oslo, Norway d Norwegian Institute of Public Health, Oslo, Norway e Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway f Medical Birth Registry of Norway, Norwegian Institute of Public Health, Bergen, Norway Correspondence: Dr I Al-Zirqi, Norwegian National Advisory Unit on Women’s Health, Division of Women and Children, Rikshospitalet, Oslo University Hospital, PO Box 4950 Nydalen, 0424 Oslo, Norway. Email [email protected]
Accepted 2 February 2015. Published Online 2 April 2015.
Objective To follow trends of uterine rupture over a period of
40 years in Norway. Design Population-based study using data from the Medical Birth
Registry, the Patient Administration System, and medical records. Setting Norway. Sample Women giving birth in 21 maternity units in Norway
during the period 1967–2008 (n = 1 441 712 maternities). Methods The incidence and outcomes of uterine rupture were
compared across four decades: 1967–1977; 1978–1988; 1989–1999; and 2000–2008. Multivariable logistic regression was used to determine the odds ratio (OR) for uterine rupture in each decade compared with the second decade. Main outcome measure Trends in uterine rupture. Results We identified 359 uterine ruptures. The incidence rates per 10 000 maternities in the first, second, third, and fourth decade were 1.2, 0.9, 1.7, and 6.1, respectively. The ORs for complete and partial ruptures in the fourth versus the second decade were 6.4 (95% confidence interval, 95% CI 3.8–10.8) and 7.2 (95% CI
4.2–12.3), respectively. Significant contributing factors to this increase were the higher rates of labour augmentation with oxytocin, scarred uteri from a previous caesarean section, and labour induction with prostaglandins or prostaglandins combined with oxytocin. After adjusting for risk factors, the ORs for complete and partial ruptures were 2.2 (95% CI 1.3–3.8) and 2.8 (95% CI 1.6–4.8), respectively. Severe postpartum haemorrhage, hysterectomy, intrapartum death and infant death after complete uterine ruptures decreased significantly over time. Conclusions A sharply increasing trend of uterine rupture was
found. Obstetric interventions contributed to this increase, but could not explain it entirely. Keywords Outcome, risk factors, trends, uterine rupture. Tweetable abstract A sharply increasing trend of uterine ruptures has been found in Norway in recent years. Linked article This article is commented on by MB Landon pp. 676–677 in this issue. To view this commentary visit http:// dx.doi.org/10.1111/1471-0528.13482.
Please cite this paper as: Al-Zirqi I, Stray-Pedersen B, Forsen L, Daltveit A, Vangen S. Uterine rupture: trends over 40 years. BJOG 2016;123:780–787.
Introduction Uterine rupture is a rare peripartum complication that occurs in around 7/10 000 individuals, but this rate increases to 20–80/10 000 in mothers with uterine scars, mostly as a result of previous caesarean section.1,2 The use of caesarean section is increasing worldwide, and thus the rate of uterine rupture is also expected to rise.3,4 With the potential for catastrophic outcomes, a study is warranted to determine the trend of uterine rupture. Detecting changes in incidence and outcome of uterine rupture and examining the possible role of certain risk fac-
780
tors may provide deeper insight into the safety of our current obstetric practice. The increasing use of caesarean section, induction of labour with specific methods, and augmentation of labour with oxytocin all play a role in modern obstetrics. The caesarean section rate in Norway increased from 1.8% in 1967 to 16.8% in 2012,5 but there was also a high rate of trial of labour with vaginal births in 51% of mothers with a previous caesarean section.6 This study aimed to determine trends in incidence, risk factors, and outcomes of uterine rupture over a period of 40 years (1967–2008), based on two registries and medical records from 21 maternity units in Norway.
ª 2015 Royal College of Obstetricians and Gynaecologists
Uterine rupture: trends over 40 years
Methods
Variables
Design and study population
Main outcome measure The main outcome measures were complete and partial uterine ruptures. A complete uterine rupture was defined as tearing in all layers of the uterine wall, including the serosa and amniotic membranes. A partial uterine rupture was defined as tearing in the muscular layers, with intact serosa or amniotic membranes.1 Uterine rupture was reported as plain text in the MBRN registration forms by the midwife in charge, and was then recorded by MBRN personnel as a code in the electronic file: code 71 for complications during delivery before 1999, and with the 10th revision of International Statistical Classification of Diseases (ICD 10) diagnostic codes O710 and O711 from 1999 onwards.9 Uterine rupture was identified in the PAS by ICD–8 code 95610 (1967–1978), ICD–9 codes 6650 and 665111 (1979–1998), and ICD–10 codes O710 and O711 (2000–2008).9 These codes did not specify rupture type. The type of rupture, whether complete or partial, was identified in the medical records.
This was a population-based registry study, complemented with information from the medical records. We used two independent data sources: first, the Medical Birth Registry of Norway (MBRN), established in 1967,7 which is a national registry that contains information on all births in Norway from 16 weeks of gestation; and second, the local Patient Administration System (PAS), which is a local registry from each maternity unit that has recorded all diagnoses for inpatients since 1970. We requested a PAS registry search from all maternity units (n = 48) in Norway, and 21 units agreed to participate.8 The 21 units were distributed throughout Norway, with delivery rates ranging from 1500 ml within 24 hours of delivery, or the need for blood transfusion, regardless of the volume of blood loss; hysterectomy, defined as hysterectomy within 7 days postpartum for uterine rupture; intrapartum/ infant death, defined as intrapartum fetal death from 23 weeks of gestation and infant deaths within 1 year of birth (excluding congenital malformations); post-hypoxic encephalopathy,12 defined as signs of cerebral irritation, depression, or seizures in the presence of severe asphyxia (asphyxia with a 1–minute Apgar score of 0–3); and admission to a neonatal intensive care unit (NICU).
Statistical analysis Cross-tabulation was used to compute the overall incidence and the incidence of different types of uterine rupture among all maternities, and among those with intact or scarred uteri, in each decade. It was also used to compute the percentages
of maternities with risk factors for uterine rupture, the characteristics of ruptures, and the maternal and infant outcomes of complete ruptures in each decade. We tested for a linear trend in the incidences of uterine rupture from the second to the fourth decade, as the second decade was used as the reference. On the other hand, the trends in the risk factors, in the characteristics, and in the outcomes of uterine rupture were tested from the first to the fourth decade. The calculations were based on a general linear model. We used logistic regression to calculate unadjusted and adjusted odds ratios (ORs and aORs, respectively) of complete and partial uterine ruptures in the first, third, and fourth decade, compared with the second decade. When building the multivariable statistical models, we included risk factors that were associated with uterine rupture and had changed through the decades. The ORs of uterine ruptures in different decades were calculated after being
Table 1. Incidence of uterine rupture in Norway over four decades (1967–2008), demonstrating the trend from the second to the fourth decade All maternities n = 1 441 712
1967–1977 n = 316 641
All ruptures n = 359 Complete ruptures n = 196 Partial ruptures n = 163
1978–1988 n = 343 261
1989–1999 n = 430 839
2000–2008 n = 350 971
P* (trend)
n
Per 104 (95% CI)
n
Per 104 (95% CI)
n
Per 104 (95% CI)
n
Per 104 (95% CI)
39
1.2 (0.8, 1.6)
31
0.9 (0.6, 1.2)
74
1.7 (1.3, 2.1)
215
6.1 (5.3, 7.0)
DOI: 10.1111/1471-0528.13394 www.bjog.org
Uterine rupture: trends over 40 years n,a,d A-K Daltveit,e,f S Vangena,d I Al-Zirqi,a,b B Stray-Pedersen,b,c L Forse a Norwegian National Advisory Unit on Women’s Health, Oslo, Norway b Women and Children’s Division Rikshospitalet, Oslo University Hospital, Oslo, Norway c Faculty of Medicine, University of Oslo, Oslo, Norway d Norwegian Institute of Public Health, Oslo, Norway e Department of Global Public Health and Primary Care, University of Bergen, Bergen, Norway f Medical Birth Registry of Norway, Norwegian Institute of Public Health, Bergen, Norway Correspondence: Dr I Al-Zirqi, Norwegian National Advisory Unit on Women’s Health, Division of Women and Children, Rikshospitalet, Oslo University Hospital, PO Box 4950 Nydalen, 0424 Oslo, Norway. Email [email protected]
Accepted 2 February 2015. Published Online 2 April 2015.
Objective To follow trends of uterine rupture over a period of
40 years in Norway. Design Population-based study using data from the Medical Birth
Registry, the Patient Administration System, and medical records. Setting Norway. Sample Women giving birth in 21 maternity units in Norway
during the period 1967–2008 (n = 1 441 712 maternities). Methods The incidence and outcomes of uterine rupture were
compared across four decades: 1967–1977; 1978–1988; 1989–1999; and 2000–2008. Multivariable logistic regression was used to determine the odds ratio (OR) for uterine rupture in each decade compared with the second decade. Main outcome measure Trends in uterine rupture. Results We identified 359 uterine ruptures. The incidence rates per 10 000 maternities in the first, second, third, and fourth decade were 1.2, 0.9, 1.7, and 6.1, respectively. The ORs for complete and partial ruptures in the fourth versus the second decade were 6.4 (95% confidence interval, 95% CI 3.8–10.8) and 7.2 (95% CI
4.2–12.3), respectively. Significant contributing factors to this increase were the higher rates of labour augmentation with oxytocin, scarred uteri from a previous caesarean section, and labour induction with prostaglandins or prostaglandins combined with oxytocin. After adjusting for risk factors, the ORs for complete and partial ruptures were 2.2 (95% CI 1.3–3.8) and 2.8 (95% CI 1.6–4.8), respectively. Severe postpartum haemorrhage, hysterectomy, intrapartum death and infant death after complete uterine ruptures decreased significantly over time. Conclusions A sharply increasing trend of uterine rupture was
found. Obstetric interventions contributed to this increase, but could not explain it entirely. Keywords Outcome, risk factors, trends, uterine rupture. Tweetable abstract A sharply increasing trend of uterine ruptures has been found in Norway in recent years. Linked article This article is commented on by MB Landon pp. 676–677 in this issue. To view this commentary visit http:// dx.doi.org/10.1111/1471-0528.13482.
Please cite this paper as: Al-Zirqi I, Stray-Pedersen B, Forsen L, Daltveit A, Vangen S. Uterine rupture: trends over 40 years. BJOG 2016;123:780–787.
Introduction Uterine rupture is a rare peripartum complication that occurs in around 7/10 000 individuals, but this rate increases to 20–80/10 000 in mothers with uterine scars, mostly as a result of previous caesarean section.1,2 The use of caesarean section is increasing worldwide, and thus the rate of uterine rupture is also expected to rise.3,4 With the potential for catastrophic outcomes, a study is warranted to determine the trend of uterine rupture. Detecting changes in incidence and outcome of uterine rupture and examining the possible role of certain risk fac-
780
tors may provide deeper insight into the safety of our current obstetric practice. The increasing use of caesarean section, induction of labour with specific methods, and augmentation of labour with oxytocin all play a role in modern obstetrics. The caesarean section rate in Norway increased from 1.8% in 1967 to 16.8% in 2012,5 but there was also a high rate of trial of labour with vaginal births in 51% of mothers with a previous caesarean section.6 This study aimed to determine trends in incidence, risk factors, and outcomes of uterine rupture over a period of 40 years (1967–2008), based on two registries and medical records from 21 maternity units in Norway.
ª 2015 Royal College of Obstetricians and Gynaecologists
Uterine rupture: trends over 40 years
Methods
Variables
Design and study population
Main outcome measure The main outcome measures were complete and partial uterine ruptures. A complete uterine rupture was defined as tearing in all layers of the uterine wall, including the serosa and amniotic membranes. A partial uterine rupture was defined as tearing in the muscular layers, with intact serosa or amniotic membranes.1 Uterine rupture was reported as plain text in the MBRN registration forms by the midwife in charge, and was then recorded by MBRN personnel as a code in the electronic file: code 71 for complications during delivery before 1999, and with the 10th revision of International Statistical Classification of Diseases (ICD 10) diagnostic codes O710 and O711 from 1999 onwards.9 Uterine rupture was identified in the PAS by ICD–8 code 95610 (1967–1978), ICD–9 codes 6650 and 665111 (1979–1998), and ICD–10 codes O710 and O711 (2000–2008).9 These codes did not specify rupture type. The type of rupture, whether complete or partial, was identified in the medical records.
This was a population-based registry study, complemented with information from the medical records. We used two independent data sources: first, the Medical Birth Registry of Norway (MBRN), established in 1967,7 which is a national registry that contains information on all births in Norway from 16 weeks of gestation; and second, the local Patient Administration System (PAS), which is a local registry from each maternity unit that has recorded all diagnoses for inpatients since 1970. We requested a PAS registry search from all maternity units (n = 48) in Norway, and 21 units agreed to participate.8 The 21 units were distributed throughout Norway, with delivery rates ranging from 1500 ml within 24 hours of delivery, or the need for blood transfusion, regardless of the volume of blood loss; hysterectomy, defined as hysterectomy within 7 days postpartum for uterine rupture; intrapartum/ infant death, defined as intrapartum fetal death from 23 weeks of gestation and infant deaths within 1 year of birth (excluding congenital malformations); post-hypoxic encephalopathy,12 defined as signs of cerebral irritation, depression, or seizures in the presence of severe asphyxia (asphyxia with a 1–minute Apgar score of 0–3); and admission to a neonatal intensive care unit (NICU).
Statistical analysis Cross-tabulation was used to compute the overall incidence and the incidence of different types of uterine rupture among all maternities, and among those with intact or scarred uteri, in each decade. It was also used to compute the percentages
of maternities with risk factors for uterine rupture, the characteristics of ruptures, and the maternal and infant outcomes of complete ruptures in each decade. We tested for a linear trend in the incidences of uterine rupture from the second to the fourth decade, as the second decade was used as the reference. On the other hand, the trends in the risk factors, in the characteristics, and in the outcomes of uterine rupture were tested from the first to the fourth decade. The calculations were based on a general linear model. We used logistic regression to calculate unadjusted and adjusted odds ratios (ORs and aORs, respectively) of complete and partial uterine ruptures in the first, third, and fourth decade, compared with the second decade. When building the multivariable statistical models, we included risk factors that were associated with uterine rupture and had changed through the decades. The ORs of uterine ruptures in different decades were calculated after being
Table 1. Incidence of uterine rupture in Norway over four decades (1967–2008), demonstrating the trend from the second to the fourth decade All maternities n = 1 441 712
1967–1977 n = 316 641
All ruptures n = 359 Complete ruptures n = 196 Partial ruptures n = 163
1978–1988 n = 343 261
1989–1999 n = 430 839
2000–2008 n = 350 971
P* (trend)
n
Per 104 (95% CI)
n
Per 104 (95% CI)
n
Per 104 (95% CI)
n
Per 104 (95% CI)
39
1.2 (0.8, 1.6)
31
0.9 (0.6, 1.2)
74
1.7 (1.3, 2.1)
215
6.1 (5.3, 7.0)
