Acta Pædiatrica ISSN 0803-5253
REGULAR ARTICLE
Adverse respiratory outcome after premature rupture of membranes before viability Eric Verspyck1,2, Violene Bisson1, Horace Roman1, Ste´phane Marret ([email protected])2,3 1.Department of Obstetrics and Gynecology, Rouen University Hospital, Rouen, France 2.Region-INSERM Team (ERI28) Neovasc, Perinatal Handicap, Institute for Biomedical Research and Innovation (IRIB), School of Medicine, University of Rouen, Rouen, France 3.Department of Neonatal Medicine and Intensive Care, Rouen University Hospital, Rouen, France
Keywords Perinatal outcome, Prematurity, Preterm premature rupture of membranes
ABSTRACT Aim: To determine whether preterm premature rupture of membranes (PPROM) before
Correspondence phane Marret, Laboratory of Microvascular Ste Endothelium and Neonate Brain Lesions, Department of Neonatal Medicine and Intensive Care, Institute for Biomedical Research and Innovation, Rouen University Hospital and RegionINSERM Team (ERI28) “NeoVasc”, University of Rouen, 1, Rue de Germont, 76031 Rouen Cedex, France. Tel: +33 232 888 097 | Fax: +33 232 888 633 | Email: [email protected]
Methods: A retrospective comparative cohort study was conducted, including viable
Received 18 June 2013; revised 4 October 2013; accepted 28 October 2013. DOI:10.1111/apa.12489
24 weeks is an independent risk factor for poor outcome in preterm neonates. premature infants born between 25 and 34-weeks gestation. Each preterm case with early PPROM was matched with two preterm controls of the same gestational age at birth, sex and birth date and who were born spontaneously with intact membranes. Logistic regression was performed to identify independent risk factors associated with composite respiratory and perinatal adverse outcomes for the overall population of preterm infants. Results: Thirty-five PPROM cases were matched with 70 controls. Extreme prematurity (26–28 weeks) was an independent risk factor for composite perinatal adverse outcomes [odds ratio (OR) 43.9; p = 0.001]. Extreme prematurity (OR 42.9; p = 0.001), PPROM (OR 7.1; p = 0.01), male infant (OR 5.2; p = 0.02) and intrauterine growth restriction (IUGR, OR 4.8; p = 0.04) were factors for composite respiratory adverse outcomes. Conclusion: Preterm premature rupture of membranes before viability represents an independent risk factor for composite respiratory adverse outcomes in preterm neonates. Extreme prematurity may represent the main risk factor for both composite respiratory and perinatal adverse outcomes.
INTRODUCTION Preterm premature rupture of membranes (PPROM) occurring before 24 weeks of gestation is a rare complication, occurring in 0.37% of pregnancies (1). However, maternal and perinatal complications resulting from this condition are frequent and include chorioamnionitis, placental abruption, perinatal death and neonatal morbidities among surviving infants (2). Neonatal mortality and severe morbidities are mainly related to extreme prematurity with a reported mean gestational age at delivery occurring at 23–27 weeks in previous retrospective studies (3). PPROM before 24 weeks of gestation may also be associated with a 40% rate of chorioamnionitis, which is a major risk factor in severe neonatal complications. The latter include intraventricular haemorrhage and/or periventricular leukomalacia, chronic lung disease, and neonatal sepsis (4–6). Persistent oligohydramnios and prolonged latency period of pregnancy in women with early PPROM appear to be two major specific risks for pulmonary hypoplasia and chronic lung disease (7). In
Key notes
Abbreviations IUGR, Intrauterine growth restriction; PPROM, Premature rupture of membranes.
256
fact, chorioamnionitis, persistent oligohydramnios and extreme prematurity may contribute to neonatal complications in surviving infants. Our objective was to determine whether early PPROM may represent an independent risk factor for poor neonatal outcomes in neonates born prematurely. We therefore conducted a retrospective comparative cohort study to compare rates of neonatal complications in severe preterm infants who were born after pregnancies complicated, or not, by early PPROM.
A study of 35 preterm infants with preterm premature rupture of membranes (PPROM) and 70 matched controls found an association between early PPROM and composite respiratory outcomes. Extreme prematurity may represent the main risk factor for both composite respiratory and perinatal adverse outcomes. Despite the absence of observed clastic cerebral lesions, long-term follow-up is crucial to determine the neurobehavioural development of the children.
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
Early premature rupture of membranes
Verspyck et al.
PATIENTS AND METHODS This retrospective study was performed between January 2000 and December 2010 in the Department of Obstetrics and Gynecology at Rouen University Hospital. The study was approved by the regional Institutional Ethics Committee of Rouen. All data for women with PPROM between 14 and 24-weeks gestational age were extracted from our database and then analysed (Fig. 1). Multiple gestation and infants with congenital or chromosomal abnormalities were excluded from the study. The diagnosis of PPROM was based on findings of amniotic fluid on sterile speculum examination or positive nitrazine/IGFBP-1 tests if necessary. Antibiotics were systematically administered at PPROM diagnosis (oral 2 g of ampicillin for 5 days) and further reintroduced depending on weekly results of bacteriological vaginal samples. In addition, all women were monitored on a weekly basis by ultrasound examination to evaluate growth and amount of amniotic fluid. Cervical culture, C-reactive protein and white blood cell counts were performed twice a week. After a short hospitalisation for initial management and counselling, discharge was encouraged
for those women with a stable medical condition. Women were readmitted to Rouen University Hospital at foetal viability, mostly after 25 weeks of gestation. At this time, corticosteroid regimen, that is, intramuscular betamethasone, was administered at 12 mg twice at 24 h interval for 2 days. Only live infants born between 25 and 34-weeks gestational age with PPROM before 24 weeks of gestation were finally recruited for the study (n = 35) and considered to be the cases (Fig. 1). Each index case was matched to two nonmalformed controls born after idiopathic preterm labour, defined as spontaneous onset of labour before rupture of membranes (8). Controls were matched for gestational age at birth (3 days), sex and birth date (6 months). Gestational age was determined using the first trimester routine ultrasound examination or, if unavailable, the last menstrual period. Maternal and neonatal records including demographic information, medical treatment, delivery data and neonatal outcomes were reviewed for both cases and controls. Clinical chorioamnionitis was considered if two or more of the following appeared: maternal temperature
Figure 1 Flow chart of pregnancies affected by PPROM before 24-weeks’ gestation.
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
257
Early premature rupture of membranes
Verspyck et al.
>38.5°C with no other infectious reasons, uterine contractions and pelvic tenderness, foetal tachycardia with reduced oscillations with no other reasons, or biological abnormalities (C-reactive protein > 20, white blood cells count > 15 000/mm3). Intrauterine growth restriction (IUGR) was defined as birth weight below the 10th percentile using customised French standards (9). Intraventricular haemorrhages were graded I–IV according to the criteria of Gardner et al. (10). Necrotising enterocolitis was defined according to Bell’s criteria (11). Neonatal sepsis was defined in neonates with fever and isolation of organisms from blood or cerebrospinal fluid culture. Persistent pulmonary hypertension was defined as persistent hypoxaemia (PaO2 < 50 mmHg in 100% oxygen) requiring NO administration. Chronic lung disease was defined as oxygen dependency at 36-weeks corrected gestational age (12). Immediate neonatal death was defined as neonatal death occurring before 12 h of life and related to respiratory failure alone. In these latter cases, pulmonary hypoplasia was diagnosed when histological measurement of the post-mortem lung-to-bodyweight was below 0.017 (13). Composite perinatal adverse outcomes were defined as having one or more of the following: neonatal death, grade III or IV intraventricular haemorrhage or periventricular leukomalacia, stage 2 or more necrotising enterocolitis, or chronic lung disease. Composite respiratory adverse outcomes were defined as having one or more of the following: immediate neonatal death, chronic lung disease, persistent pulmonary hypertension or pneumothorax. Composite outcomes and their variables were defined by protocol prior to onset of dataset analysis. Variables were selected according to previously published data which reported increased major morbidities of clinical interest in premature infants born after early PPROM (1,3). Maternal and neonatal characteristics and outcomes were initially compared between the cases and controls using univariate analysis. Comparison of continuous variables was performed using the Student’s t-test. Qualitative variables were compared using the chi-square test and the Fisher’s exact test, where appropriate. Regression models were also used to assess factors associated with the two composite adverse outcomes, that is, perinatal and respiratory. The variables of clinical interest and those whose p-value was inferior to 0.20 in univariate analysis were included in the model. A difference of p < 0.05 was considered as being statistically significant. Analysis was performed using Stata 9.0 (Stata Corporation, College Station, TX, USA).
RESULTS Comparison between cases and controls for maternal and neonatal characteristics is reported in Table 1. The rate of complete course of antenatal steroids, chorioamnionitis, severe FHR abnormalities, caesarean delivery and IUGR was significantly higher in women with PPROM in comparison with controls. Comparison between cases and
258
Table 1 Maternal and neonatal characteristics Characteristics Maternal age (year) Body mass index (kg/m2) Parity Complete course of antenatal steroids Tocolytic administration Clinical chorioamnionitis Caesarean delivery Severe FHR abnormalities GA at birth (week) Birth weight (g) IUGR Male infant
Cases (n = 35)
Controls (n = 70)
p
28.6 6.2 23.02 6.9 2.05 0.9 31 (88.8)
27.9 5.5 23.01 5.7 1.48 0.8 34 (48.5)
0.55 0.98 0.001 0.001
13 (37.1) 12 (34.2) 13 (37.1) 10 (28.5) 29.4 2.4 1346.2 493.7 10 (28.5) 20 (57.1)
64 (91.4) 0 5 (7.1) 5 (7.1) 29.4 2.4 1411.3 476.5 9 (12.8) 40 (57.1)
0.001 0.001 0.001 0.001 0.92 0.5 0.04 1
GA = gestational age; FHR = foetal heart rate; IUGR = intrauterine growth restriction. Data are mean standard deviation or n (%).
controls for immediate neonatal outcomes is reported in Table 2. Respiratory outcomes, that is, pneumothorax, chronic lung disease, duration of respiratory support, immediate neonatal death and composite respiratory adverse outcomes was more frequently reported in the cases in comparison with controls. Immediate neonatal death occurred in four cases and autopsies were performed in two of them, which revealed pulmonary hypoplasia with a lung-to-body-weight ratio of 0.016. In contrast, grade III or IV intraventricular haemorrhage and periventricular leukomalacia was only diagnosed in the control group and no neonatal sepsis occurred in either the cases or controls. Logistic regression model only showed an association between composite perinatal adverse outcomes and very preterm birth [premature birth 26–28 weeks = OR 43.9 (8.37–222.9); p = 0.001 and 29–30 weeks = OR 9.9 (1.75–55.9), p = 0.009] (Table 3). Factors identified by logistic regression as independently associated with composite respiratory adverse outcomes were very preterm birth [premature birth 26–28 weeks = 42.9 (6.02–306.5); p = 0.001 and 29–30 weeks = OR 11.5 (1.53–87.4), p = 0.02], PPROM [OR 7.1 (1.46, 34.9); p = 0.01], IUGR [OR 4.85 (1.04–22.5); p = 0.04] and male infant [OR 5.02 (1.18–21.3); p = 0.02] (Table 4).
DISCUSSION This is, to our knowledge, the first cohort study comparing two groups of premature birth with or without early PPROM in order to report the various independent risk factors (including PPROM) associated with two composite adverse outcomes. We have shown that PPROM prior to viability represents an independent risk factor for composite respiratory adverse outcomes in preterm neonates, which is similar to very preterm birth, IUGR and male sex
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
Early premature rupture of membranes
Verspyck et al.
Table 2 Neonatal outcomes Outcomes
Cases (n = 35)
Controls (n = 70)
5 min Apgar stage 2 Necrotising enterocolitis Grade 3 or 4 IVH or PVL Neonatal death Immediate neonatal death Composite perinatal adverse outcomes Composite respiratory adverse outcomes Duration of hospitalisation**
15 (42.8) 3 (10.3), n = 29 29 (82.8) 32 (91.4) 3 (9.7)* 8 (25)* 8 (22.8) 27 (87)* 8 (22.8) 14 (48.2) 52.5 57.9 7.2 12.5 0 3 (9.7)* 0* 6 (17.1) 4 (11.4) 20 (57.1) 22 (62.8) 70.5 41.1
14 (20) 5 (7.7), n = 65 41 (58.5) 64 (91.4) 3 (4.2) 17 (24.3) 1 (1.4) 48 (68.5) 1 (1.4) 15 (22.7) 28.9 30.1 3.5 7.2 0 7 (10) 8 (11.4) 4 (5.7) 0 25 (35.7) 16 (22.8) 63.1 31.5
p 0.02 0.69 0.01 0.72 0.37 0.87 0.001 0.09 0.001 0.008 0.01 0.07 1 0.05 0.08 0.01 0.03 0.001 0.37
NICU = neonatal intensive care unit; IVH = intraventricular haemorrhage; PVL = periventricular leukomalacia. Data are mean standard deviation or n (%). *Data were not available for four cases with immediate neonatal death. **Data were not available for six cases and four controls with neonatal death.
Table 3 Factors independently associated with composite perinatal adverse outcome. (logistic regression model)
PPROM* Chorioamnionitis* IUGR* Severe FHR abnormalities* Categorised GA (week)** 26–28 29–30
OR (95% CI)
p
2.6 5.66 5.02 0.92
0.18 0.14 0.05 0.93
(0.64–9.9) (0.56–56.8) (0.98–25.6) (0.16–5.12)
43.9 (8.37–222.9) 9.9 (1.75–55.9)
0.001 0.009
PPROM = preterm premature rupture of membranes; IUGR = intrauterine growth restriction; FHR = foetal heart rate; GA = gestational age. *Reference range was absence of the criteria. **Reference range was gestational age ≥31 weeks.
of the newborn. Only very preterm birth was associated independently with composite perinatal adverse outcomes after logistic regression. A recent review of the literature shows that the various studies published to date were not only retrospective, but also included small series of cases (3). Neonatal complications were variously reported as well as the protocols supported by the authors. Only one study, which is comparable to our study, has been published, which contains, however, some questionable methodological criteria, that is, selection of controls and methodology used
Table 4 Factors independently associated with composite respiratory adverse outcome. (logistic regression model)
PPROM* Chorioamnionitis* IUGR* Severe FHR abnormalities* Tocolytic administration* Male infant* Categorised GA (week)** 26–28 29–30
OR (95% CI)
p
7.1 11.1 4.85 0.92 0.32 5.02
0.01 0.11 0.04 0.93 0.17 0.02
(1.46–34.9) (0.59–210.9) (1.04–22.5) (0.15–5.6) (0.62–1.66) (1.18–21.3)
42.9 (6.02–306.5) 11.5 (1.53–87.4)
0.001 0.02
PPROM = preterm premature rupture of membranes; IUGR = intrauterine growth restriction; FHR = foetal heart rate; GA = gestational age. *Reference range was absence of the criteria. **Reference range was gestational age ≥31 weeks.
for logistic regression (14). Management of PPROM before viability has been standardised in our centre since the year 2000, which allows us to have a homogeneous group of patients for this study. Our attitude is also rather conservative whatever the term of the rupture and the amount of amniotic fluid at ultrasound evaluation except for persistent anhydramnios and maternal complications superimposed. We did not record cases of births occurring after 34 weeks because we wanted to study a population of very and moderate preterm infants. In fact, these infants born after
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
259
Early premature rupture of membranes
Verspyck et al.
34 weeks (n = 9) were otherwise healthy with no reported respiratory, neurological complication or sepsis. The most original aspect of our study is that our population of cases with PPROM was compared with matched controls from spontaneous and idiopathic preterm births with intact membranes. In fact, each of the three major obstetric precursors leading to preterm birth described by Goldenberg et al. (8) (preterm infants born after PPROM, preterm labour with intact membranes or infants born in a vascular context) may be associated with various risk factors for neonatal and long-term morbidities. As we reported previously, preterm small-for-gestationalage infants are at increased risk of neonatal mortality, but at lower risk of cerebral injuries in comparison with infants born following idiopathic preterm labour and PPROM (15). A more recent study, in a population of preterm infants born following premature rupture of membranes, reported a higher rate of intraventricular haemorrhage and a lower rate of bronchopulmonary dysplasia in infants born after a short latency period (
REGULAR ARTICLE
Adverse respiratory outcome after premature rupture of membranes before viability Eric Verspyck1,2, Violene Bisson1, Horace Roman1, Ste´phane Marret ([email protected])2,3 1.Department of Obstetrics and Gynecology, Rouen University Hospital, Rouen, France 2.Region-INSERM Team (ERI28) Neovasc, Perinatal Handicap, Institute for Biomedical Research and Innovation (IRIB), School of Medicine, University of Rouen, Rouen, France 3.Department of Neonatal Medicine and Intensive Care, Rouen University Hospital, Rouen, France
Keywords Perinatal outcome, Prematurity, Preterm premature rupture of membranes
ABSTRACT Aim: To determine whether preterm premature rupture of membranes (PPROM) before
Correspondence phane Marret, Laboratory of Microvascular Ste Endothelium and Neonate Brain Lesions, Department of Neonatal Medicine and Intensive Care, Institute for Biomedical Research and Innovation, Rouen University Hospital and RegionINSERM Team (ERI28) “NeoVasc”, University of Rouen, 1, Rue de Germont, 76031 Rouen Cedex, France. Tel: +33 232 888 097 | Fax: +33 232 888 633 | Email: [email protected]
Methods: A retrospective comparative cohort study was conducted, including viable
Received 18 June 2013; revised 4 October 2013; accepted 28 October 2013. DOI:10.1111/apa.12489
24 weeks is an independent risk factor for poor outcome in preterm neonates. premature infants born between 25 and 34-weeks gestation. Each preterm case with early PPROM was matched with two preterm controls of the same gestational age at birth, sex and birth date and who were born spontaneously with intact membranes. Logistic regression was performed to identify independent risk factors associated with composite respiratory and perinatal adverse outcomes for the overall population of preterm infants. Results: Thirty-five PPROM cases were matched with 70 controls. Extreme prematurity (26–28 weeks) was an independent risk factor for composite perinatal adverse outcomes [odds ratio (OR) 43.9; p = 0.001]. Extreme prematurity (OR 42.9; p = 0.001), PPROM (OR 7.1; p = 0.01), male infant (OR 5.2; p = 0.02) and intrauterine growth restriction (IUGR, OR 4.8; p = 0.04) were factors for composite respiratory adverse outcomes. Conclusion: Preterm premature rupture of membranes before viability represents an independent risk factor for composite respiratory adverse outcomes in preterm neonates. Extreme prematurity may represent the main risk factor for both composite respiratory and perinatal adverse outcomes.
INTRODUCTION Preterm premature rupture of membranes (PPROM) occurring before 24 weeks of gestation is a rare complication, occurring in 0.37% of pregnancies (1). However, maternal and perinatal complications resulting from this condition are frequent and include chorioamnionitis, placental abruption, perinatal death and neonatal morbidities among surviving infants (2). Neonatal mortality and severe morbidities are mainly related to extreme prematurity with a reported mean gestational age at delivery occurring at 23–27 weeks in previous retrospective studies (3). PPROM before 24 weeks of gestation may also be associated with a 40% rate of chorioamnionitis, which is a major risk factor in severe neonatal complications. The latter include intraventricular haemorrhage and/or periventricular leukomalacia, chronic lung disease, and neonatal sepsis (4–6). Persistent oligohydramnios and prolonged latency period of pregnancy in women with early PPROM appear to be two major specific risks for pulmonary hypoplasia and chronic lung disease (7). In
Key notes
Abbreviations IUGR, Intrauterine growth restriction; PPROM, Premature rupture of membranes.
256
fact, chorioamnionitis, persistent oligohydramnios and extreme prematurity may contribute to neonatal complications in surviving infants. Our objective was to determine whether early PPROM may represent an independent risk factor for poor neonatal outcomes in neonates born prematurely. We therefore conducted a retrospective comparative cohort study to compare rates of neonatal complications in severe preterm infants who were born after pregnancies complicated, or not, by early PPROM.
A study of 35 preterm infants with preterm premature rupture of membranes (PPROM) and 70 matched controls found an association between early PPROM and composite respiratory outcomes. Extreme prematurity may represent the main risk factor for both composite respiratory and perinatal adverse outcomes. Despite the absence of observed clastic cerebral lesions, long-term follow-up is crucial to determine the neurobehavioural development of the children.
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
Early premature rupture of membranes
Verspyck et al.
PATIENTS AND METHODS This retrospective study was performed between January 2000 and December 2010 in the Department of Obstetrics and Gynecology at Rouen University Hospital. The study was approved by the regional Institutional Ethics Committee of Rouen. All data for women with PPROM between 14 and 24-weeks gestational age were extracted from our database and then analysed (Fig. 1). Multiple gestation and infants with congenital or chromosomal abnormalities were excluded from the study. The diagnosis of PPROM was based on findings of amniotic fluid on sterile speculum examination or positive nitrazine/IGFBP-1 tests if necessary. Antibiotics were systematically administered at PPROM diagnosis (oral 2 g of ampicillin for 5 days) and further reintroduced depending on weekly results of bacteriological vaginal samples. In addition, all women were monitored on a weekly basis by ultrasound examination to evaluate growth and amount of amniotic fluid. Cervical culture, C-reactive protein and white blood cell counts were performed twice a week. After a short hospitalisation for initial management and counselling, discharge was encouraged
for those women with a stable medical condition. Women were readmitted to Rouen University Hospital at foetal viability, mostly after 25 weeks of gestation. At this time, corticosteroid regimen, that is, intramuscular betamethasone, was administered at 12 mg twice at 24 h interval for 2 days. Only live infants born between 25 and 34-weeks gestational age with PPROM before 24 weeks of gestation were finally recruited for the study (n = 35) and considered to be the cases (Fig. 1). Each index case was matched to two nonmalformed controls born after idiopathic preterm labour, defined as spontaneous onset of labour before rupture of membranes (8). Controls were matched for gestational age at birth (3 days), sex and birth date (6 months). Gestational age was determined using the first trimester routine ultrasound examination or, if unavailable, the last menstrual period. Maternal and neonatal records including demographic information, medical treatment, delivery data and neonatal outcomes were reviewed for both cases and controls. Clinical chorioamnionitis was considered if two or more of the following appeared: maternal temperature
Figure 1 Flow chart of pregnancies affected by PPROM before 24-weeks’ gestation.
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
257
Early premature rupture of membranes
Verspyck et al.
>38.5°C with no other infectious reasons, uterine contractions and pelvic tenderness, foetal tachycardia with reduced oscillations with no other reasons, or biological abnormalities (C-reactive protein > 20, white blood cells count > 15 000/mm3). Intrauterine growth restriction (IUGR) was defined as birth weight below the 10th percentile using customised French standards (9). Intraventricular haemorrhages were graded I–IV according to the criteria of Gardner et al. (10). Necrotising enterocolitis was defined according to Bell’s criteria (11). Neonatal sepsis was defined in neonates with fever and isolation of organisms from blood or cerebrospinal fluid culture. Persistent pulmonary hypertension was defined as persistent hypoxaemia (PaO2 < 50 mmHg in 100% oxygen) requiring NO administration. Chronic lung disease was defined as oxygen dependency at 36-weeks corrected gestational age (12). Immediate neonatal death was defined as neonatal death occurring before 12 h of life and related to respiratory failure alone. In these latter cases, pulmonary hypoplasia was diagnosed when histological measurement of the post-mortem lung-to-bodyweight was below 0.017 (13). Composite perinatal adverse outcomes were defined as having one or more of the following: neonatal death, grade III or IV intraventricular haemorrhage or periventricular leukomalacia, stage 2 or more necrotising enterocolitis, or chronic lung disease. Composite respiratory adverse outcomes were defined as having one or more of the following: immediate neonatal death, chronic lung disease, persistent pulmonary hypertension or pneumothorax. Composite outcomes and their variables were defined by protocol prior to onset of dataset analysis. Variables were selected according to previously published data which reported increased major morbidities of clinical interest in premature infants born after early PPROM (1,3). Maternal and neonatal characteristics and outcomes were initially compared between the cases and controls using univariate analysis. Comparison of continuous variables was performed using the Student’s t-test. Qualitative variables were compared using the chi-square test and the Fisher’s exact test, where appropriate. Regression models were also used to assess factors associated with the two composite adverse outcomes, that is, perinatal and respiratory. The variables of clinical interest and those whose p-value was inferior to 0.20 in univariate analysis were included in the model. A difference of p < 0.05 was considered as being statistically significant. Analysis was performed using Stata 9.0 (Stata Corporation, College Station, TX, USA).
RESULTS Comparison between cases and controls for maternal and neonatal characteristics is reported in Table 1. The rate of complete course of antenatal steroids, chorioamnionitis, severe FHR abnormalities, caesarean delivery and IUGR was significantly higher in women with PPROM in comparison with controls. Comparison between cases and
258
Table 1 Maternal and neonatal characteristics Characteristics Maternal age (year) Body mass index (kg/m2) Parity Complete course of antenatal steroids Tocolytic administration Clinical chorioamnionitis Caesarean delivery Severe FHR abnormalities GA at birth (week) Birth weight (g) IUGR Male infant
Cases (n = 35)
Controls (n = 70)
p
28.6 6.2 23.02 6.9 2.05 0.9 31 (88.8)
27.9 5.5 23.01 5.7 1.48 0.8 34 (48.5)
0.55 0.98 0.001 0.001
13 (37.1) 12 (34.2) 13 (37.1) 10 (28.5) 29.4 2.4 1346.2 493.7 10 (28.5) 20 (57.1)
64 (91.4) 0 5 (7.1) 5 (7.1) 29.4 2.4 1411.3 476.5 9 (12.8) 40 (57.1)
0.001 0.001 0.001 0.001 0.92 0.5 0.04 1
GA = gestational age; FHR = foetal heart rate; IUGR = intrauterine growth restriction. Data are mean standard deviation or n (%).
controls for immediate neonatal outcomes is reported in Table 2. Respiratory outcomes, that is, pneumothorax, chronic lung disease, duration of respiratory support, immediate neonatal death and composite respiratory adverse outcomes was more frequently reported in the cases in comparison with controls. Immediate neonatal death occurred in four cases and autopsies were performed in two of them, which revealed pulmonary hypoplasia with a lung-to-body-weight ratio of 0.016. In contrast, grade III or IV intraventricular haemorrhage and periventricular leukomalacia was only diagnosed in the control group and no neonatal sepsis occurred in either the cases or controls. Logistic regression model only showed an association between composite perinatal adverse outcomes and very preterm birth [premature birth 26–28 weeks = OR 43.9 (8.37–222.9); p = 0.001 and 29–30 weeks = OR 9.9 (1.75–55.9), p = 0.009] (Table 3). Factors identified by logistic regression as independently associated with composite respiratory adverse outcomes were very preterm birth [premature birth 26–28 weeks = 42.9 (6.02–306.5); p = 0.001 and 29–30 weeks = OR 11.5 (1.53–87.4), p = 0.02], PPROM [OR 7.1 (1.46, 34.9); p = 0.01], IUGR [OR 4.85 (1.04–22.5); p = 0.04] and male infant [OR 5.02 (1.18–21.3); p = 0.02] (Table 4).
DISCUSSION This is, to our knowledge, the first cohort study comparing two groups of premature birth with or without early PPROM in order to report the various independent risk factors (including PPROM) associated with two composite adverse outcomes. We have shown that PPROM prior to viability represents an independent risk factor for composite respiratory adverse outcomes in preterm neonates, which is similar to very preterm birth, IUGR and male sex
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
Early premature rupture of membranes
Verspyck et al.
Table 2 Neonatal outcomes Outcomes
Cases (n = 35)
Controls (n = 70)
5 min Apgar stage 2 Necrotising enterocolitis Grade 3 or 4 IVH or PVL Neonatal death Immediate neonatal death Composite perinatal adverse outcomes Composite respiratory adverse outcomes Duration of hospitalisation**
15 (42.8) 3 (10.3), n = 29 29 (82.8) 32 (91.4) 3 (9.7)* 8 (25)* 8 (22.8) 27 (87)* 8 (22.8) 14 (48.2) 52.5 57.9 7.2 12.5 0 3 (9.7)* 0* 6 (17.1) 4 (11.4) 20 (57.1) 22 (62.8) 70.5 41.1
14 (20) 5 (7.7), n = 65 41 (58.5) 64 (91.4) 3 (4.2) 17 (24.3) 1 (1.4) 48 (68.5) 1 (1.4) 15 (22.7) 28.9 30.1 3.5 7.2 0 7 (10) 8 (11.4) 4 (5.7) 0 25 (35.7) 16 (22.8) 63.1 31.5
p 0.02 0.69 0.01 0.72 0.37 0.87 0.001 0.09 0.001 0.008 0.01 0.07 1 0.05 0.08 0.01 0.03 0.001 0.37
NICU = neonatal intensive care unit; IVH = intraventricular haemorrhage; PVL = periventricular leukomalacia. Data are mean standard deviation or n (%). *Data were not available for four cases with immediate neonatal death. **Data were not available for six cases and four controls with neonatal death.
Table 3 Factors independently associated with composite perinatal adverse outcome. (logistic regression model)
PPROM* Chorioamnionitis* IUGR* Severe FHR abnormalities* Categorised GA (week)** 26–28 29–30
OR (95% CI)
p
2.6 5.66 5.02 0.92
0.18 0.14 0.05 0.93
(0.64–9.9) (0.56–56.8) (0.98–25.6) (0.16–5.12)
43.9 (8.37–222.9) 9.9 (1.75–55.9)
0.001 0.009
PPROM = preterm premature rupture of membranes; IUGR = intrauterine growth restriction; FHR = foetal heart rate; GA = gestational age. *Reference range was absence of the criteria. **Reference range was gestational age ≥31 weeks.
of the newborn. Only very preterm birth was associated independently with composite perinatal adverse outcomes after logistic regression. A recent review of the literature shows that the various studies published to date were not only retrospective, but also included small series of cases (3). Neonatal complications were variously reported as well as the protocols supported by the authors. Only one study, which is comparable to our study, has been published, which contains, however, some questionable methodological criteria, that is, selection of controls and methodology used
Table 4 Factors independently associated with composite respiratory adverse outcome. (logistic regression model)
PPROM* Chorioamnionitis* IUGR* Severe FHR abnormalities* Tocolytic administration* Male infant* Categorised GA (week)** 26–28 29–30
OR (95% CI)
p
7.1 11.1 4.85 0.92 0.32 5.02
0.01 0.11 0.04 0.93 0.17 0.02
(1.46–34.9) (0.59–210.9) (1.04–22.5) (0.15–5.6) (0.62–1.66) (1.18–21.3)
42.9 (6.02–306.5) 11.5 (1.53–87.4)
0.001 0.02
PPROM = preterm premature rupture of membranes; IUGR = intrauterine growth restriction; FHR = foetal heart rate; GA = gestational age. *Reference range was absence of the criteria. **Reference range was gestational age ≥31 weeks.
for logistic regression (14). Management of PPROM before viability has been standardised in our centre since the year 2000, which allows us to have a homogeneous group of patients for this study. Our attitude is also rather conservative whatever the term of the rupture and the amount of amniotic fluid at ultrasound evaluation except for persistent anhydramnios and maternal complications superimposed. We did not record cases of births occurring after 34 weeks because we wanted to study a population of very and moderate preterm infants. In fact, these infants born after
©2013 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 2014 103, pp. 256–261
259
Early premature rupture of membranes
Verspyck et al.
34 weeks (n = 9) were otherwise healthy with no reported respiratory, neurological complication or sepsis. The most original aspect of our study is that our population of cases with PPROM was compared with matched controls from spontaneous and idiopathic preterm births with intact membranes. In fact, each of the three major obstetric precursors leading to preterm birth described by Goldenberg et al. (8) (preterm infants born after PPROM, preterm labour with intact membranes or infants born in a vascular context) may be associated with various risk factors for neonatal and long-term morbidities. As we reported previously, preterm small-for-gestationalage infants are at increased risk of neonatal mortality, but at lower risk of cerebral injuries in comparison with infants born following idiopathic preterm labour and PPROM (15). A more recent study, in a population of preterm infants born following premature rupture of membranes, reported a higher rate of intraventricular haemorrhage and a lower rate of bronchopulmonary dysplasia in infants born after a short latency period (
