Hypertensive disorders of pregnancy
Pereira Bernardes, Thomas Patrick Custodio Heinrich
DOI:
10.33612/diss.99788387
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Pereira Bernardes, T. P. C. H. (2019). Hypertensive disorders of pregnancy: occurrence, recurrence, and management. University of Groningen. https://doi.org/10.33612/diss.99788387
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
neonatal outcomes after induction of
labor versus expectant management
in women with an unripe cervix:
a secondary analysis of
the HYPITAT and DIGITAT trials
Thomas P. Bernardes Kim Broekhuijsen Corine Koopmans Kim Boers Linda van Wyk Parvin Tajik Mariëlle G. van Pampus Sicco A. Scherjon Ben W. Mol Maureen T. Franssen Paul P. van den Berg Henk Groen
ABSTRACT
ObjectiveTo evaluate caesarean section and adverse neonatal outcome rates after induction of labour or expectant management in women with an unripe cervix at or near term.
Design
Secondary analysis of data from two randomised clinical trials. Setting
Data were collected in two nationwide Dutch trials. Population
Women with hypertensive disease (HYPITAT trial) or suspected fetal growth restriction (DIGITAT trial) and a Bishop score ≤ 6.
Methods
Comparison of outcomes after induction of labour and expectant management. Main outcome measures
Rates of caesarean section and adverse neonatal outcome, defined as 5-minute Apgar score ≤ 6 and/or arterial umbilical cord pH < 7.05 and/or neonatal intensive care unit admission and/or seizures and/or perinatal death.
Results
Of 1172 included women with an unripe cervix, 572 had induction of labour and 600 had expectant management. We found no significant difference in the overall caesarean rate (difference -1.1%, 95% CI -5.4 to 3.2). Induction of labour did not increase caesarean rates in women with Bishop scores from 3 to 6 (difference -2.7%, 95% CI -7.6 to 2.2) or adverse neonatal outcome rates (difference -1.5%, 95% CI -4.3 to 1.3). However, there was a significant difference in the rates of arterial umbilical cord pH < 7.05 favouring induction (difference -3.2%, 95% CI -5.6 to -0.9). The number needed to treat to prevent one case of umbilical arterial pH < 7.05 was 32.
Conclusions
We found no evidence that induction of labour increases the caesarean rate or compromises neonatal outcome as compared with expectant management. Concerns over increased risk of failed induction in women with a Bishop score from 3 to 6 seem unwarranted.
This article is commented on by WM Gilbert.
6
INTRODUCTION
Induction of labor (IOL) is a common practice in obstetrics, with estimates ranging from 20
to 25% of all pregnancies.1,2 IOL is indicated in situations in which the outcomes for mother
and child are better if the pregnancy is not further prolonged.1 The effectiveness of IOL is
subject to considerable debate, especially due to concerns over associated high rates of
cesarean sections (CS) in women with an unripe cervix.3 The ripeness of a cervix is often
characterized through the Bishop score (BS).4 However, the use of the BS for the prediction
of the delivery mode in women scheduled for IOL at term has been recently questioned.5
Nonetheless, IOL is identified by multiple studies as a risk factor for CS for both nulliparous
and multiparous women.6–12 However, these studies compared CS rates after IOL with
those after spontaneous labor. This comparison has been criticized as inappropriate because spontaneous labor is not a management choice of the attending physician. In complicated pregnancies the two available options are: (1) expectant management (EM) until spontaneous labor starts or clear indications for immediate delivery appear and (2)
delivery by IOL or CS.13 In a randomized trial with EM as the comparison group, IOL has been
shown to produce lower CS rates in post-term women.14 This trial was included in a recent
meta-analysis that showed reduced CS risk after IOL in women with intact membranes.15
Furthermore, two recent randomized trials, HYPITAT (Hypertension and Pre-eclampsia Intervention Trial At Term) and DIGITAT (Disproportionate Intrauterine Growth Intervention Trial At Term), found comparable CS rates after IOL in term pregnancies complicated by
hypertensive disease or intrauterine growth restriction (IUGR).16,17 The HYPITAT trial showed
that IOL lowers the risk of severe hypertension while the DIGITAT study showed it not to
be harmful, thus possibly allowing for the prevention of stillbirth.18,19 Short and long term
consequences of near (35-37) and early term (37-38) birth should also be taken into account
as IOL inevitably produces lower gestational ages at delivery.20–22
The appropriate management of women with an indication for IOL who also present with an unripe cervix remains elusive. In these women, IOL is not regarded as an appealing option because of the possibility of a long cervical ripening process, which conflicts with the pressing need for rapid delivery. The prospect of a higher risk of CS in case IOL fails may further influence the preference for immediate CS. In this study we evaluated CS and adverse neonatal outcome rates in women whose pregnancies were complicated by hypertensive disease or IUGR and who presented with an unripe cervix in a combined dataset from two randomized trials that compared IOL with EM.
METHODS
Our analyses were based on the combined data from the multicenter, open label, controlled
randomized trials HYPITAT and DIGITAT.16,17 The trials were approved by the Institutional
Review Board of the University of Leiden (p04.210). Both compared EM with IOL in women with a singleton fetus in cephalic presentation recruited between 36 and 41 weeks for either gestational hypertension or mild pre-eclampsia in the HYPITAT trial or suspected IUGR in the DIGITAT trial. Recruitment periods of HYPITAT and DIGITAT overlapped (October/2005-March/2008 and November/2004-November/2008, respectively) and women were assigned preferentially to DIGITAT if they presented with hypertension and suspected IUGR. Further details on each trial have been described previously.
Gestational hypertension was defined as two measurements of diastolic blood pressure ≥ 95 mmHg at least six hours apart. Mild pre-eclampsia was defined as diastolic blood pressure ≥ 90 mmHg measured twice at least six hours apart and proteinuria (two or more positives on a dipstick, > 300 mg total protein on a 24h urine sample or protein to creatinine ratio > 30
mg/mmol).23 Suspected IUGR was defined as one or more of fetal abdominal circumference,
weight below the 10th percentile or flattening of the growth curve in the third trimester.24
Maternal exclusion criteria were previous CS, pre-existing hypertension, diabetes mellitus or gestational diabetes requiring insulin therapy, HIV seropositivity, HELLP syndrome, renal or heart disease, oliguria (< 500 ml/24h), pulmonary edema, cyanosis and prelabor rupture of membranes. The use of intravenous antihypertensive drugs was also an exclusion criterion as well as severe gestational hypertension and pre-eclampsia that were defined as systolic blood pressure ≥ 170 mmHg, diastolic blood pressure ≥ 110 mmHg or proteinuria ≥ 5 g/24h. Fetal abnormalities suspected on ultrasound or during fetal heart-rate monitoring, and decreased or absent fetal movements were also exclusion criteria.
After written informed consent was obtained, women were randomly allocated to either IOL or EM. For this study, we selected the women with an unripe cervix at inclusion, defined
as a BS ≤ 6.4 In the IOL group, labor was induced within 24h of randomization in HYPITAT
and within 48h in DIGITAT. Cervical ripening was promoted with intracervical or intravaginal prostaglandins or a balloon catheter, according to local protocols.
Women in the EM group were monitored until spontaneous onset of labor. Twice weekly blood pressure measurements and screening for proteinuria using a dipstick or protein to creatinine ratio in a urine sample were performed as well as laboratory tests of liver and kidney function and full blood count. Positive screening of proteinuria was followed by the collection of urine for 24h to quantify proteinuria levels. Assessment of the fetal conditions consisted of electronic fetal heart-rate monitoring, ultrasound examination, and
6
daily reports on fetal movement by the woman. IOL was performed in this group if duringmonitoring a woman exhibited systolic blood pressure ≥ 170 mmHg, diastolic blood pressure ≥ 110 mmHg, proteinuria ≥ 5g/24h, eclampsia, HELLP syndrome, suspected fetal distress, pre-labor rupture of membranes > 48h, meconium stained amniotic fluid, fetus gestational age beyond 41 weeks as well as at the obstetrician’s discretion.
The primary outcomes of this study were CS and a composite adverse neonatal outcome. The composite adverse neonatal outcome was defined as the presence of at least one of the following: 5 minute Apgar score ≤ 6, arterial umbilical cord pH < 7.05, neonatal intensive care (NICU) admission, seizures or perinatal death. Each of these was defined individually as a secondary outcome. If the umbilical artery pH was missing, in both trials the neonatal outcome was classified as normal if all other components of the composite outcome were normal. For the present analyses we did the same for cases of missing data on NICU admission and seizures. We performed subgroup analyses of the primary outcomes by each BS level as well as through the use of the median BS level. Mann-Whitney U-tests, χ² statistics and Wilson’s method for differences in proportions were used as appropriate. To check for potential confounders in the analyses we performed logistic regressions.
RESULTS
During the original trial periods, 756 and 650 eligible women were randomized in the HYPITAT and DIGITAT trials, respectively. Of these 1406 women, 1172 (83.3%) had a BS ≤6 and attended all other criteria for these analyses. After random allocation, 572 had been assigned to IOL and 600 to EM (see Figure S1).
Baseline characteristics are presented in Table 1. The groups were comparable with respect to their median maternal age, BMI at booking, gestational age, BS and cervical length at randomization. The proportion of women selected from the two trials in both intervention and control groups was similar, as well as the number of women with hypertensive disease. The proportion of nulliparas, caucasians and smokers was also comparable between the two groups.
Table 1. Baseline characteristics
Induction of Labor (n = 572) Expectant Management (n = 600) Difference in median and p-value or percentage (95% CI) Maternal age 28.0 (24.0 - 32.0) 28.0 (24.0 - 32.0) 0.0 0.296 BMI at booking † 24.1 (21.1 - 29.0) 24.0 (21.3 - 28.3) 0.1 0.955 Gestational age at randomization ⱡ 37.9 (37.1 - 38.9) 38.0 (37.1 - 39.0) -0.1 0.396 Bishop score at randomization 3.0 (1.0 - 4.0) 3.0 (1.0 - 4.0) 0.0 0.646
Cervical Length (milimeters)* 30.0 (23.3 - 37.0) 30.0 (24.0 - 38.0) 0.0 0.808
Study HYPITAT 309 54% 322 53.7% 0.4% (-5.3% - 6.0%) DIGITAT 263 46% 278 46.3% Hypertensive disease Normotensive 241 42.1% 239 39.8% 2.3% (-3.3% - 7.9%) Gestational hypertension 211 36.9% 230 38.3% -1.5% (-7.0% - 4.1%) Pre-eclampsia 120 21.0% 131 21.8% -0.9% (-5.5% - 3.9%) Nulliparous 368 64.3% 394 65.7% -1.3% (-6.9% - 4.2%) Caucasian 465 81.3% 460 76.7% 4.6% (-0.2% - 9.4%) Smoker 158 29.5% 155 27.2% 2.3% (-3.0% - 7.6%)
Table shows medians, 25th and 75th percentiles, differences in medians and corresponding p-values or category numbers, percentages and 95% confidence intervals for percentage differences.
† Induction n = 505, expectant management n = 539 ⱡ Induction n = 572, expectant management n = 597 * Induction n = 548, expectant management n = 581
6
Table 2 shows labor and delivery data for both groups. In the IOL group, 17 (3%) womenhad spontaneous labor before induction was started. In the EM group, labor was induced in 288 (48%) women, either according to the monitoring protocol (n=235, 81.6%) or electively (n=53, 18.4%). Planned CS was indicated for three women in the induction group; one due to fetal distress, another due to maternal complications and the third was an elective procedure. Of the 15 planned CS that occurred in the EM group, 11 were indicated due to fetal distress, one due to maternal complications and three were elective procedures. CS was performed in 88 (15.4%) and 99 (16.5%) women in the IOL and EM groups, respectively (diff. -1.1%; 95% CI -5.4% to 3.2%). Fetal distress was the indication of 40 CS in the IOL group, identical in proportion to the 45 in the EM group (45.4% vs. 45.4%). We also found similar rates for instrumental vaginal deliveries (10.1% vs. 11.7%) in both groups (diff. -1.6%; 95% CI -5.2% to 2.2%).
Data on overall time from randomization to delivery as well as by levels of BS and parity are presented in Table S1. Of the 572 women in the IOL group, 426 (75%) delivered within 2.5 days. As expected, there was a significant difference in median time from randomization to delivery (1.7 vs. 8.9 days; P-value < 0.001). This difference was reflected in the median gestational age at delivery (268.1 vs. 278.3 days; P-value < 0.001). As expected, the difference in time from randomization to delivery decreased with increasing BS at randomization. In the subgroup of 65 women randomized with a BS of 6, it decreased to 3.5 days (1.4 vs. 4.9 days; P-value < 0.001). The median number of hours from induction start to birth was 3.3 hours lower in the EM group (12.2 vs. 8.9 hours; P-value < 0.001). Induction in the EM group was performed at a median of 8.8 (IQR: 4.7 - 14.1) days from randomization.
Neonatal outcomes can be found in Table 3. Median birthweight was lower in the IOL group (2888 vs. 2998 grams; diff. -110 grams; P-value < 0.001). The highest difference in median birthweight was found in the subgroup of women who presented hypertensive disease (3225 vs. 3480 grams; diff. -255; P-value < 0.001). In the 61 (5%) pregnancies that were complicated by both hypertensive disease and IUGR we found no significant difference in median birthweight between the groups. In these women, the median time from randomization to delivery was 1.3 days (IQR: 0.9 - 2.7) in the induction group and 6.7 days (IRQ: 3.0 - 14.0) in the EM group.
Table 2. Labor and delivery characteristics Induction of Labor (n = 572) Expectant Management (n = 600) Difference in percentage (95% CI); p-value Onset of labor Spontaneous 17 3.0% 296 49.3% -46.3% (-50.6% – -41.9%) Induction 551 96.3% 288 48.0% 48.3% (43.8% – 52.6%)
Planned caesarean section 3 0.5% 15 2.5% -2.0% (-3.7% – -0.4%)
Mode of Delivery
Vaginal 426 74.5% 431 71.8% 2.7% (-7.8% – 2.6%)
Instrumental 58 10.1% 70 11.7% -1.6% (-5.2% – 2.2%)
Caesarean section 88 15.4% 99 16.5% -1.1% (-5.4% – 3.2%)
Caesarean section indication
Arrest of first stage of labor 17 3.0% 22 3.7% -0.7% (-2.8% – 1.4%)
Arrest of second stage of labor 3 0.5% 7 1.2% -0.6% (-1.7% – 0.4%)
Failed instrumental delivery 4 0.7% 1 0.2% 0.5% (-0.2% – 1.3%)
Fetal distress 40 7.0% 45 7.5% -0.5% (-3.5% – 2.5%)
Failure to progress and fetal
distress 18 3.1% 14 2.3% 0.8% (-1.1% – 2.7%)
Maternal complication 4 0.7% 6 1.0% -0.3% (-1.4% – 0.8%)
Elective 2 0.3% 4 0.7% -0.3% (-1.1% – 0.5%)
Congenital abnormalities † 20 4.2% 23 4.9% -0.7% (-3.4% – 2.0%)
Antibiotic treatment
Before onset of labor 6 1.0% 3 0.50% 0.55% (-0.5% – 1.5%)
During labor and delivery 43 7.5% 45 7.5% 0.0% (-3.0% – 3.0%)
Unknown 3 0.5% 1 0.5% 0.0% (-1.0% – 0.4%) Analgesia Pethidine/Phenergan/Nubaine 112 19.6% 98 16.3% 3.2% (-1.1% – 7.7%) Epidural 136 23.8% 116 19.3% 4.4% (-0.2% – 9.2%) Spinal 21 3.7% 20 3.3% 0.3% (-1.8% – 2.4%) Other 23 4.0% 17 2.8% 1.2% (-0.9% – 3.3%) Unknown 2 0.35% 1 0.17% 0.18% (-0.4% – 0.8%)
Table shows category numbers, percentages and 95% confidence intervals for percentage differences. † Induction n = 476, expectant management n = 472
6
There was no perinatal mortality. Two newborns exhibited seizures, one in each group.The number of neonates with an adverse neonatal outcome in the induction group was 32 (5.7%), comparable to the 42 (7.2%) in the EM group (diff. -1.5%; 95% CI -4.3% to 1.3%). The groups did not differ with respect to NICU admission rates as 14 (2.6%) in each group were admitted. Twelve presented 5 minute Apgar scores below 7 in the IOL group, one more than in the EM group. A significant difference was found in the number of newborns with umbilical arterial pH below 7.05, as there were 8 in the IOL group and 24 in the EM group with this condition (1.7% vs. 4.8%; diff. -3.2%; 95% CI -5.6% to -0.9%). With the limit at 7.01, there was one child in the IOL group and 14 in the EM group (0.2% vs. 2.8%, diff. -2.6%, 95% CI -4.2% to -1.1%). To avoid one case of umbilical artery pH < 7.05, 32 (95% CI 18 to 112) women would need to be induced and this number rises to 39 (95% CI 23 to 90) for pH < 7.01.
Table 3. Neonatal outcomes
Induction of
Labour ManagementExpectant
Difference in median; P-value or percentage (95% CI) Birthweight, overall (grams) 2888 (2411 – 3294) 2998 (2540 – 3520) -110 <0.0001 Hypertensive disease 3225 (2872 – 3592) 3480 (3087 – 3790) -250 <0.0001 Intrauterine growth restriction 2425 (2230 – 2662) 2575 (2320 – 2870) -150 <0.0001 Both 2377 (1972 – 2685) 2470 (2155 – 2755) -93 0,855 Composite adverse neonatal outcome 32/563 5,7% 42/587 7,2% -1.5% (-4,3% - 1.3%) Admission to NICU 14/532 2,6% 14/540 2,6% 0,0% (-2,0% - 2.1%) Umbilical arterial pH < 7.05 8/478 1,7% 24/495 4,8% -3,2% (-5,6% - 0.9%) Apgar score < 7 12/570 2,1% 11/600 1,8% 0,3% (-1,4% - 2.0%) Seizures 1/476 0,2% 1/472 0,2% 0% (-0,6% - 0.6%) Fetal death 0/572 0% 0/600 0% 0% -Neonatal death 0/572 0% 0/600 0% 0%
-Table shows medians, 25th and 75th percentiles, differences in medians and corresponding P-values or total and
Table 4 presents the main outcomes by different BS subgroups. We used the median BS at randomization of 3 to divide the sample in two subgroups: BS from 0 to 2 and BS from 3 to 6. In the subgroup with BS 0-2 there were 59 (22.7%) CS in the IOL group and 62 (21.2%) in the EM group (diff. 1.5%; 95% CI -5.4% to 8.4%). The CS rates remained comparable between IOL and EM if the analysis was restricted to the women randomized with BS 3-6, although the numbers were approximately halved to 29 (9.3%) and 37 (12%) (diff. -2.7%; 95% CI -7.6% to 2.2%). Comparison within the IOL group showed that the higher CS rate found in women with BS 0-2 was significantly different from women with scores from 3-6 (22.7% vs. 9.3%; diff. 13.4%; 95% CI 7.5% to 19.3%). Comparable results were found in the EM group (21.2% vs. 12%; diff. 9.2%; 95% CI 3.3% - 15.1%). Rates of adverse neonatal outcomes were comparable in all subgroups. Further analyses of the main outcomes by parity and BS are shown in Table S2.
Logistic regression was used to adjust for possible confounders and several candidates (age, parity, trial, ethnicity, presence and type of hypertensive disease, dipstick proteinuria level, BMI, cervical length, gestational age and BS at inclusion) were considered but none achieved statistical significance or changed the odds ratio of the intervention variable by more than 10%. From 0 to 2 59/260 22.7% 62/292 21.2% 1.5% (‐5,41% ‐ 8.4%) 17/259 6.6% 19/286 6.6% 0. 0% (‐4,17% ‐ 4.2%) From 3 to 6 29/312 9.3% 37/308 12.0% ‐2.7% (‐7,55% ‐ 2.2%) 15/304 4.9% 23/301 7.6% ‐2 .7% (‐6,56% ‐ 1.2%) 6 5/36 13.9% 1/29 3.4% 10.5% (‐5,4% ‐ 25.5%) 3/34 8.8% 2/28 7.1% 1.7% (‐14, 9% ‐ 16.7%) 5 3/59 5.1% 12/82 14.6% ‐9.5% (‐19,4% ‐ 1.2%) 3/56 5.4% 4/78 5.1% 0.3% (‐7, 9% ‐ 9.9%) 4 8/96 8.3% 11/98 11.2% ‐2.9% (‐11,6% ‐ 5.8%) 3/96 3.1% 8/97 8.2% ‐5.1% (‐1 8,2% ‐ 12.6%) 3 13/121 10.7% 13/99 13.1% ‐2.4% (‐13,1% ‐ 10.7%) 6/118 5.1% 9/98 9.2% ‐4. 1% (‐11,9% ‐ 2.9%) 2 28/103 27.2% 20/118 16.9% 10.3% (‐0,1% ‐ 21.1%) 8/103 7.8% 9/114 7.9% ‐0 .1% (‐7,6% ‐ 7.6%) 1 21/106 19.8% 19/96 19.8% 0.0% (‐10,9% ‐ 11.1%) 5/105 4.8% 7/95 7.4% ‐2.6 % (‐10,1% ‐ 4.4%) 0 10/51 19.6% 23/78 29.5% ‐9.9% (‐23,7% ‐ 5.8%) 4/51 7.8% 3/77 3.9% 3.9% (‐4,5% ‐ 14.9%) Equal or less than 6 88/572 15.4% 99/600 16.5% ‐1.1% (‐5,4% ‐ 3.2%) 32/563 5.7% 42/587 7.2% ‐1.5% (‐4,3% ‐ 1.4%) 5 83/536 15.5% 98/571 17.2% ‐1.7% (‐6,06% ‐ 2.7%) 29/529 5.5% 40/559 7.2% ‐1.7% (‐4,6% ‐ 1.3%) 4 80/477 16.8% 86/489 17.6% ‐0.8% (‐5,56% ‐ 4.0%) 26/473 5.5% 36/481 7.5% ‐2.0% (‐5,2% ‐ 1.2%) 3 72/381 18.9% 75/391 19.2% ‐0.3% (‐5,84% ‐ 5.2%) 23/377 6.1% 28/384 7.3% ‐1.2% (‐4,8% ‐ 2.4%) 2 59/260 22.7% 62/292 21.2% 1.5% (‐5,41% ‐ 8.4%) 17/259 6.6% 19/286 6.6% 0.0% (‐4,3% ‐ 4.3%) 1 31/157 19.7% 42/174 24.1% ‐4.4% (‐13,34% ‐ 4.5%) 9/156 5.8% 10/172 5.8% 0.0% (‐5.0% ‐ 5.0%) 0 10/51 19.6% 23/78 29.5% ‐9.9% (‐25,3% ‐ 5.5%) 4/51 7.8% 3/77 3.9% 3.9% (‐4,5% ‐ 14.9%) Expectant Management Cesarean Section Composite adverse neonatal outcome Induction of Labour Expectant Management Induction of Labour Difference in percentage (95% CI) Difference in percentage (95% CI) Bishop score Table 4. Main out comes b y Bishop sc or e a t r andomiz ation Table sho w s t ot al and out come number s a t each le vel, per cen tag es and 95% c on fidence in ter vals f or the diff er ence in per cen t of ag es
6
Table 4 presents the main outcomes by different BS subgroups. We used the median BS atrandomization of 3 to divide the sample in two subgroups: BS from 0 to 2 and BS from 3 to 6. In the subgroup with BS 0-2 there were 59 (22.7%) CS in the IOL group and 62 (21.2%) in the EM group (diff. 1.5%; 95% CI -5.4% to 8.4%). The CS rates remained comparable between IOL and EM if the analysis was restricted to the women randomized with BS 3-6, although the numbers were approximately halved to 29 (9.3%) and 37 (12%) (diff. -2.7%; 95% CI -7.6% to 2.2%). Comparison within the IOL group showed that the higher CS rate found in women with BS 0-2 was significantly different from women with scores from 3-6 (22.7% vs. 9.3%; diff. 13.4%; 95% CI 7.5% to 19.3%). Comparable results were found in the EM group (21.2% vs. 12%; diff. 9.2%; 95% CI 3.3% - 15.1%). Rates of adverse neonatal outcomes were comparable in all subgroups. Further analyses of the main outcomes by parity and BS are shown in Table S2.
Logistic regression was used to adjust for possible confounders and several candidates (age, parity, trial, ethnicity, presence and type of hypertensive disease, dipstick proteinuria level, BMI, cervical length, gestational age and BS at inclusion) were considered but none achieved statistical significance or changed the odds ratio of the intervention variable by more than 10%. 59/260 22.7% 62/292 21.2% 1.5% (‐5,41% ‐ 8.4%) 17/259 6.6% 19/286 6.6% 0. 0% (‐4,17% ‐ 4.2%) 29/312 9.3% 37/308 12.0% ‐2.7% (‐7,55% ‐ 2.2%) 15/304 4.9% 23/301 7.6% ‐2 .7% (‐6,56% ‐ 1.2%) 5/36 13.9% 1/29 3.4% 10.5% (‐5,4% ‐ 25.5%) 3/34 8.8% 2/28 7.1% 1.7% (‐14, 9% ‐ 16.7%) 3/59 5.1% 12/82 14.6% ‐9.5% (‐19,4% ‐ 1.2%) 3/56 5.4% 4/78 5.1% 0.3% (‐7, 9% ‐ 9.9%) 8/96 8.3% 11/98 11.2% ‐2.9% (‐11,6% ‐ 5.8%) 3/96 3.1% 8/97 8.2% ‐5.1% (‐1 8,2% ‐ 12.6%) 13/121 10.7% 13/99 13.1% ‐2.4% (‐13,1% ‐ 10.7%) 6/118 5.1% 9/98 9.2% ‐4. 1% (‐11,9% ‐ 2.9%) 28/103 27.2% 20/118 16.9% 10.3% (‐0,1% ‐ 21.1%) 8/103 7.8% 9/114 7.9% ‐0 .1% (‐7,6% ‐ 7.6%) 21/106 19.8% 19/96 19.8% 0.0% (‐10,9% ‐ 11.1%) 5/105 4.8% 7/95 7.4% ‐2.6 % (‐10,1% ‐ 4.4%) 10/51 19.6% 23/78 29.5% ‐9.9% (‐23,7% ‐ 5.8%) 4/51 7.8% 3/77 3.9% 3.9% (‐4,5% ‐ 14.9%) 88/572 15.4% 99/600 16.5% ‐1.1% (‐5,4% ‐ 3.2%) 32/563 5.7% 42/587 7.2% ‐1.5% (‐4,3% ‐ 1.4%) 83/536 15.5% 98/571 17.2% ‐1.7% (‐6,06% ‐ 2.7%) 29/529 5.5% 40/559 7.2% ‐1.7% (‐4,6% ‐ 1.3%) 80/477 16.8% 86/489 17.6% ‐0.8% (‐5,56% ‐ 4.0%) 26/473 5.5% 36/481 7.5% ‐2.0% (‐5,2% ‐ 1.2%) 72/381 18.9% 75/391 19.2% ‐0.3% (‐5,84% ‐ 5.2%) 23/377 6.1% 28/384 7.3% ‐1.2% (‐4,8% ‐ 2.4%) 59/260 22.7% 62/292 21.2% 1.5% (‐5,41% ‐ 8.4%) 17/259 6.6% 19/286 6.6% 0.0% (‐4,3% ‐ 4.3%) 31/157 19.7% 42/174 24.1% ‐4.4% (‐13,34% ‐ 4.5%) 9/156 5.8% 10/172 5.8% 0.0% (‐5.0% ‐ 5.0%) 10/51 19.6% 23/78 29.5% ‐9.9% (‐25,3% ‐ 5.5%) 4/51 7.8% 3/77 3.9% 3.9% (‐4,5% ‐ 14.9%) Expectant Management Cesarean Section Composite adverse neonatal outcome Induction of Labour Expectant Management Induction of Labour Difference in percentage (95% CI) Difference in percentage (95% CI)
DISCUSSION
Main findings
Through analyses of a combined dataset of the HYPITAT and DIGITAT trials we were able to show that IOL when compared to EM is not associated with increased rates of CS or adverse neonatal outcomes in term or near term pregnancies complicated by IUGR, gestational hypertension or pre-eclampsia even in women with an unripe (BS ≤ 6) or very unripe (BS ≤ 2) cervix and an indication for ‘short-term’ delivery. Although we confirmed the expected relation between a low BS and a longer labor and higher CS rate, this relation occurred irrespective of whether labor was induced srtaightaway, or EM was adopted. Women with BS between 3 and 6 had a CS rate of 9.3% after IOL. These results suggest that concerns over a high risk of failed induction in women with BS between 3 and 6 seem unwarranted.
Strengths and limitations
We were able to analyze a large data set of women with term pregnancies, whose management was determined by randomization to either IOL or EM. BS did not play a role in the allocation. Therefore, selection bias in our study is unlikely.
A possible limitation may be that our study was based on women with complicated pregnancies, so the results cannot be immediately generalized to uncomplicated pregnancies, in which induction may be considered for other reasons. Nonetheless, a recently published retrospective cohort of 5090 women compared medically indicated with elective inductions,
and rates of maternal and neonatal outcomes were found to be similar.25
Interpretation
The BS was initially developed for cervical assessment in multiparous women with
uncomplicated pregnancies.4 BS of 5 or lower were the predominant risk factor for CS in a
prospective study of 1389 women who presented rates after IOL twice higher than those of
women with spontaneous labor.3 However, a systematic review of 40 studies showed that
BS of 4, 5 and 6 were poor predictors of the success of IOL and as such should not be used
in the decision to induce or not.5 Our results are in line with this assessment as women
randomized with BS in that range who were managed expectantly had CS rates similar to the women in the IOL group with comparable scores.
That notwithstanding, our results stand in clear contrast to several other studies that
show higher risks of CS for nulliparas and multiparas after IOL.6–12 These studies have been
criticized for their use of spontaneous labor as the comparison group.13 Comparing outcomes
after spontaneous labor to those after IOL does not present a credible clinical scenario as spontaneous labor is not a management choice of the obstetrician. The appropriate
6
comparison group consists of women who were managed expectantly and later on wentinto labor spontaneously or labor was induced when it was clearly medically indicated. This is found in a retrospective study of 19.377 women in which CS risk for nulliparas in the
EM group was higher than in the IOL group at every term gestational age.26 Furthermore,
the largest randomized trial to date that compared IOL to EM showed lower CS rates after
induction but only evaluated women with gestational ages of 41 weeks or more.14 Another
small trial studied pregnancies at and beyond 39 weeks and found similar rates of CS for
both strategies; however, only women with a ripe cervix were included.27 Finally, a recent
systematic review of IOL versus EM concluded that CS risk was lower after IOL in women
with intact membranes.15 Our results correspond to these latter four studies as we found
comparable CS rates in women with an unripe cervix over a gestational age range of 36 to 41 weeks.
The other key finding of these analyses is that IOL does not lead to higher rates of adverse neonatal outcomes when compared to EM in women with an unripe cervix. Furthermore, we showed a significant benefit of IOL in women with an unripe cervix as the difference in absolute rates of umbilical arterial pH <7.05 was estimated to be lowered by 3.2%. In an observational cohort of 51.519 term neonates, the relative risk of encephalopathy with seizures or death for those that presented umbilical arterial pH levels between 7.05 and
7.01 was 3.6.28 This relative risk rose to 18.2 for pH levels ≤ 7.00. We showed that 32 women
would need to be induced to avoid one case of umbilical arterial pH < 7.05, and 39 to avoid one case of pH < 7.01.
Studies of uncomplicated pregnancies presented longer times to delivery and higher costs
associated with IOL.6,8,13 In contrast, economic analyses of HYPITAT data found IOL to be
less costly when antepartum resource use was taken into account and the same was found
in DIGITAT data for gestational ages from 38 weeks on.29,30 Although IOL in our selected
population may result in higher costs due to the longer time necessary to mature the cervix, we have shown that 75% of the women in the IOL group delivered within 2.5 days after randomization and within 27h of induction start.
We favor a management strategy of IOL as opposed to elective CS for women with an unripe cervix and an indication for short-term delivery. The CS rate of 15.4% we observed justifies an attempt for vaginal delivery, as this will also result in less complicated future pregnancies. Naturally, the final management decision has to be made in close discussion with the woman herself.
CONCLUSION
In summary, we have shown that for women who present with an unripe cervix and had IOL for hypertensive disease or IUGR the rates of CS and adverse neonatal outcomes are indistinguishable from those found in women who were managed expectantly. We have also presented evidence that BS between 3 and 6 are not associated with elevated rates of induction failure and CS.
We found no evidence that induction of labor increases the cesarean rate or compromises neonatal outcome as compared with expectant management in term or near term pregnancies complicated by IUGR, gestational hypertension or pre-eclampsia. Concerns over increased risk of failed induction in women with a Bishop score from 3 to 6 seem unwarranted.
ACKNOWLEDGEMENTS
The authors wish to acknowledge the contribution of the research nurses from the Dutch Consortium for Obstetric studies to the inclusion of subjects into the original DIGITAT and HYPITAT studies.
6
REFERENCES
1. Calder A. Induction of labour. 2008.
2. Martin JA, Hamilton BE, Ventura SJ, Osterman MJK, Kirmeyer S, Mathews TJ, Wilson EC. Births: final data for 2009. Natl Vital Stat Rep. 2011;60(1):1-70.
3. Vrouenraets FPJM, Roumen FJME, Dehing CJG, van den Akker ES a, Aarts MJB, Scheve EJT. Bishop
score and risk of cesarean delivery after induction of labor in nulliparous women. Obstet Gynecol. 2005;105(4):690-697. doi:10.1097/01.AOG.0000152338.76759.38.
4. Bishop EH. Pelvic scoring for elective induction. Obstet Gynecol. 1964;(24):266-268.
5. Kolkman DGE, Verhoeven CJM, Brinkhorst SJ, van der Post J a M, Pajkrt E, Opmeer BC, Mol BWJ.
The bishop score as a predictor of labor induction success: a systematic review. Am J Perinatol. 2013;30(8):625-630. doi:10.1055/s-0032-1331024.
6. Maslow a S, Sweeny a L. Elective induction of labor as a risk factor for cesarean delivery among low-risk women at term. Obstet Gynecol. 2000;95(6 Pt 1):917-922.
7. Yeast JD, Jones A, Poskin M. Induction of labor and the relationship to cesarean delivery: A review of 7001 consecutive inductions. Am J Obstet Gynecol. 1999;180(3 Pt 1):628-633.
8. Seyb ST, Berka RJ, Socol ML, Dooley SL. Risk of cesarean delivery with elective induction of labor at term in nulliparous women. Obstet Gynecol. 1999;94(4):600-607.
9. Dublin S, Lydon-Rochelle M, Kaplan RC, Watts DH, Critchlow CW. Maternal and neonatal outcomes
after induction of labor without an identified indication. Am J Obstet Gynecol. 2000;183(4):986-994. doi:10.1067/mob.2000.106748.
10. van Gemund N, Hardeman a., Scherjon S a., Kanhai HHH. Intervention Rates after Elective Induction of Labor Compared to Labor with a Spontaneous Onset. Gynecol Obstet Invest. 2003;56(3):133-138. doi:10.1159/000073771.
11. Heffner L. Impact of labor induction, gestational age, and maternal age on cesarean delivery rates. Obstet Gynecol. 2003;102(2):287-293. doi:10.1016/S0029-7844(03)00531-3.
12. Vahratian A, Zhang J, Troendle JF, Sciscione AC, Hoffman MK. Labor progression and risk of cesarean delivery in electively induced nulliparas. Obstet Gynecol. 2005;105(4):698-704. doi:10.1097/01. AOG.0000157436.68847.3b.
13. Osmundson S, Ou-Yang RJ, Grobman W a. Elective induction compared with expectant
management in nulliparous women with an unfavorable cervix. Obstet Gynecol. 2011;117(3):583-587. doi:10.1097/AOG.0b013e31820caf12.
14. Hannah ME. Induction of labour as compared with serial antenatal monitoring in post-term pregnancy. NEJM. 1992;326(24):1587.
15. Wood S, Cooper S, Ross S. Does induction of labour increase the risk of caesarean section? A systematic review and meta-analysis of trials in women with intact membranes. BJOG. July 2013:1-12. doi:10.1111/1471-0528.12328.
16. Koopmans C, Bijlenga D, Groen H. Induction of labour versus expectant monitoring for gestational hypertension or mild pre-eclampsia after 36 weeks’ gestation (HYPITAT): a multicentre, open-label randomised controlled trial. Lancet. 2009;374(9694):979-988. doi:10.1016/S0140-6736(09)60736-4.
for intrauterine growth restriction at term: randomised equivalence trial (DIGITAT). BMJ. 2010;341:c7087. doi:10.1136/bmj.c7087.
18. Pilliod R a, Cheng YW, Snowden JM, Doss AE, Caughey AB. The risk of intrauterine fetal death
in the small-for-gestational-age fetus. Am J Obstet Gynecol. 2012;207(4):318.e1-6. doi:10.1016/j. ajog.2012.06.039.
19. McIntire DD, Leveno KJ. Birth Weight in Relation to Morbidity and Mortality among newborn infants. N Engl J Med. 1999.
20. Stock S, Ferguson E, Duffy A. Outcomes of elective induction of labour compared with expectant management: population based study. BMJ Br Med …. 2012;2838(May):1-13. doi:10.1136/bmj. e2838.
21. Boyle E, Poulsen G. Effects of gestational age at birth on health outcomes at 3 and 5 years of age: population based cohort study. BMJ Br Med …. 2012;896(March):1-14. doi:10.1136/bmj.e896. 22. Kerstjens JM, de Winter AF, Bocca-Tjeertes IF, Bos AF, Reijneveld S a. Risk of developmental delay
increases exponentially as gestational age of preterm infants decreases: a cohort study at age 4 years. Dev Med Child Neurol. 2012;54(12):1096-1101. doi:10.1111/j.1469-8749.2012.04423.x. 23. Roberts JM, Pearson GD, Cutler J a, Lindheimer MD. Summary of the NHLBI Working Group
on Research on Hypertension During Pregnancy. Hypertens Pregnancy. 2003;22(2):109-127. doi:10.1081/PRG-120016792.
24. Chien PF, Owen P, Khan KS. Validity of ultrasound estimation of fetal weight. Obstet Gynecol. 2000;95(6 Pt 1):856-860.
25. Baud D, Rouiller S, Hohlfeld P, Tolsa J-F, Vial Y. Adverse obstetrical and neonatal outcomes in elective and medically indicated inductions of labor at term. J Matern Fetal Neonatal Med. May 2013. doi: 10.3109/14767058.2013.795533.
26. Caughey AB, Nicholson JM, Cheng YW, Lyell DJ, Washington a E. Induction of labor and cesarean delivery by gestational age. Am J Obstet Gynecol. 2006;195(3):700-705. doi:10.1016/j. ajog.2006.07.003.
27. Nielsen PE, Howard BC, Hill CC, Larson PL, Holland RH, Smith PN. Comparison of elective induction of labor with favorable Bishop scores versus expectant management: a randomized clinical trial. J Matern Fetal Neonatal Med. 2005;18(1):59-64. doi:10.1080/14767050500139604.
28. Yeh P, Emary K, Impey L. The relationship between umbilical cord arterial pH and serious adverse neonatal outcome: analysis of 51,519 consecutive validated samples. BJOG. 2012;119(7):824-831. doi:10.1111/j.1471-0528.2012.03335.x.
29. Vijgen SMC, Boers KE, Opmeer BC, Bijlenga D, Bekedam DJ, Bloemenkamp KWM, de Boer K, Bremer
H a, le Cessie S, Delemarre FMC, Duvekot JJ, Hasaart THM, Kwee A, van Lith JMM, van Meir C a, van Pampus MG, van der Post J a M, Rijken M, Roumen FJME, van der Salm PCM, Spaanderman ME a, Willekes C, Wijnen EJ, Mol BWJ, Scherjon S a. Economic analysis comparing induction of labour and expectant management for intrauterine growth restriction at term (DIGITAT trial). Eur J Obstet Gynecol Reprod Biol. 2013;170(2):358-363. doi:10.1016/j.ejogrb.2013.07.017.
30. Vijgen SMC, Koopmans CM, Opmeer BC, Groen H, Bijlenga D, Aarnoudse JG, Bekedam DJ, van den
Berg PP, de Boer K, Burggraaff JM, Bloemenkamp KWM, Drogtrop a P, Franx a, de Groot CJM, Huisjes a JM, Kwee a, van Loon a J, Lub a, Papatsonis DNM, van der Post J a M, Roumen FJME, Scheepers HCJ, Stigter RH, Willekes C, Mol BWJ, Van Pampus MG. An economic analysis of induction of labour and expectant monitoring in women with gestational hypertension or pre-eclampsia at term (HYPITAT trial). BJOG. 2010;117(13):1577-1585. doi:10.1111/j.1471-0528.2010.02710.x.
6
SUPPLEMENTAL MATERIALS
Figure S1. Flowchart of included patients from DIGITAT and HYPITAT studies
Figure S1. Flowchart of included patients from DIGITAT and HYPITAT studies
572 women analyzed as induction of labour 600 women analysed as expectant management
1406 women eligible for randomisation
2269 eligible women
863 refused to participate
572 women allocated to induction of labour 600 women allocated to expectant management
Ta bl e S1 . T im e fro m ra nd om iza tio n to de liv er y and in duc tio n to d el iv er y Induc ti on of La bour E x p e c ta n t M a n ag em en t Induc ti on of La bour E x p e c ta n t M an ag em e n t N u llip ar o u s Fr om r a ndom iz a ti on (da y s ) Fr om i nduc ti on (hour s ) O v er a ll 1.9 ( 1.2 - 2.8 ) 9.0 ( 5.0 - 15 .1 ) 1 3.8 ( 7.7 - 29 .7 ) 11 .9 ( 7.0 - 23 .2 ) B y B is h o p s c o re s 0 2.3 ( 1.4 - 5.2 ) 9.2 ( 5.1 - 16 .2 ) 1 9.0 ( 11 .7 - 44 .8 ) 17 .5 ( 7.9 - 39 .1 ) 1 2.1 ( 1.5 - 3.1 ) 11 .9 ( 6.0 - 15 .4 ) 1 9.7 ( 9.7 - 36 .6 ) 14 .5 ( 7.8 - 23 .4 ) 2 2.0 ( 1.4 - 2.9 ) 9.7 ( 5.1 - 16 .2 ) 1 7.1 ( 11 .0 - 32 .3 ) 9.7 ( 5.9 - 18 .9 ) 3 1.7 ( 1.1 - 3.2 ) 9.8 ( 6.2 - 16 .6 ) 1 4.0 ( 7.3 - 28 .6 ) 15 .2 ( 8.4 - 24 .8 ) 4 1.5 ( 1.1 - 2.2 ) 8.1 ( 3.6 - 14 .2 ) 9.5 ( 6.5 - 19 .6 ) 10 .9 ( 6.9 - 23 .3 ) 5 1.1 ( 1.0 - 1.8 ) 6.1 ( 4.4 - 12 .1 ) 8.3 ( 6.3 - 12 .8 ) 8.1 ( 6.1 - 13 .5 ) 6 2.0 ( 1.1 - 2.4 ) 4.8 ( 2.6 - 9.3 ) 1 0.3 ( 6.3 - 33 .2 ) 9.9 ( 4.9 - 25 .3 ) M u lt ip ar o u s O v er a ll 1.5 ( 1.1 - 2.2 ) 8.9 ( 4.4 - 12 .9 ) 9.4 ( 5.4 - 17 .5 ) 5.8 ( 3.9 - 10 .6 ) B y B is h o p s c o re s 0 1.8 ( 1.3 - 2.4 ) 10 .3 ( 4.4 - 15 .3 ) 1 4.3 ( 7.6 - 27 .0 ) 11 .9 ( 6.8 - 17 .2 ) 1 2.0 ( 1.3 - 3.3 ) 10 .2 ( 4.8 - 13 .6 ) 2 4.0 ( 12 .8 - 42 .1 ) 9.0 ( 6.8 - 14 .1 ) 2 1.6 ( 1.2 - 2.4 ) 9.8 ( 5.6 - 16 .5 ) 1 4.4 ( 6.7 - 20 .7 ) 4.8 ( 3.9 - 8.4 ) 3 1.5 ( 1.0 - 2.2 ) 9.5 ( 5.5 - 14 .4 ) 9.4 ( 5.3 - 15 .9 ) 6.6 ( 4.0 - 12 .9 ) 4 1.3 ( 1.0 - 2.2 ) 10 .2 ( 5.7 - 12 .7 ) 7.1 ( 4.2 - 9.9 ) 4.3 ( 3.2 - 4.9 ) 5 1.2 ( 0.9 - 1.4 ) 6.2 ( 3.2 - 8.8 ) 5.7 ( 3.8 - 10 .2 ) 4.8 ( 3.7 - 10 .0 ) 6 1.3 ( 1.0 - 1.4 ) 6.7 ( 0.7 - 12 .0 ) 6.4 ( 4.3 - 11 .3 ) 3.7 ( 3.3 - 8.9 ) Ta bl e s ho w s m ed ia ns , 2 5th a nd 7 5th p er cen til es . Ti m e to d eliv ery in clu de d tim e re qu ire d fo r c er vic al rip eni ng w he re appl ica bl e. Table S1. Time fr om r andomiz ation t o deliv er y and induction t o deliv er y Table sho w s medians, 25 th and 75 th per cen tiles. Time t o deliv er y included time r equir ed f or cer vic al ripening wher e applic able.
6
Ta bl e S2 . M ai n ou tc om es b y pa rit y a nd B ish op sc ore a t r an do m iza tion Nu lli pa ro us Pa rou s Ca es ar ea n sec tio n Ind uc tio n of La bou r Expec ta nt M an ag em en t Di ffer en ce in pe rc enta ge (9 5% C I) Induc tio n of La bou r Ex pec ta nt M ana gem ent Di ffer en ce in per cen ta ge (9 5% C I) Bi sho p sc or e F rom 0 to 2 52/ 176 2 9, 5 % 57/ 206 27 ,7% 1, 8 (-7, 1 - 11, 0) 7/ 8 5 8, 3% 5/ 8 6 5, 8% 2, 5 (-5, 8 - 10, 9) F rom 3 to 6 23/ 192 1 2, 0% 29/ 188 15 ,4% -3, 4 (-10, 5 - 3, 5) 6/ 120 5, 0% 8/ 120 6, 7% -1, 7 (-8, 2 - 4, 7) 6 3/ 21 1 4, 3 % 1/ 18 5, 6% 8, 7 (-13, 5 - 29, 6) 2/ 1 5 13, 3% 0/ 1 1 0, 0% 13 ,3 (-14 ,3 - 37, 9) 5 3/ 44 6, 8% 8/ 52 15, 4% -8, 6 (-21, 5 - 5, 0) 0/ 1 5 0, 0% 4/ 3 0 13 ,3 % -13, 3 (-29 ,7 - 8, 6) 4 6/ 56 1 0, 7 % 9/ 61 14, 8% -4, 1 (-16, 4 - 8, 7) 2/ 4 0 5, 0% 2/ 3 7 5, 4% -0, 4 (-13 ,2 - 11, 7) 3 11 /71 15 ,5% 1 1/ 5 7 19, 3% -3, 8 (-17, 6 - 9, 2) 2/ 5 0 4, 0% 2/ 4 2 4, 8% -0, 8 (-12 ,2 - 9, 3) 2 23 /65 35 ,4% 1 8/ 7 6 23, 7% 11, 7 (-3, 3 - 26, 3) 5/ 3 8 13, 2% 2/ 4 2 4, 8% 8, 4 (-4, 8 - 23, 0) 1 20 /78 25 ,6% 1 8/ 6 7 26, 9% -1, 22 (-15, 6 - 12, 8) 1/ 2 8 3, 6% 1/ 2 9 3, 4% -0, 2 (-14 ,6 - 13, 9) 0 9/ 33 2 7, 3 % 2 1/ 6 3 33, 3% -6, 1 (-23, 4 - 13, 8) 1/ 1 8 5, 6% 2/ 1 5 13 ,3 % -7, 8 (-32 ,7 - 14, 6) Co m po site ad ver se neo na ta l o utc om e Bi sh op sc or e F rom 0 to 2 13/ 175 7, 4% 16/ 202 7, 9% -0, 5 (-6, 0 - 5, 2) 4/ 8 4 4, 8% 3/ 8 4 3, 6% 1, 2 (-5, 8 - 8, 4) F rom 3 to 6 14/ 185 7, 6% 18/ 182 9, 9% -2, 3 (-8, 3 - 3, 6) 1/ 119 0, 8% 5/ 119 4, 2% -3, 4 (-8, 6 - 0, 1) 6 3/ 20 1 5, 0 % 1/ 17 5, 9% 9, 1 (-14, 1 - 30, 7) 0/ 1 4 0, 0% 1/ 1 1 9, 1% -9, 1 (-37 ,7 - 13, 7) 5 3/ 41 7, 3% 4/ 49 8, 2% -0, 9 (-12, 8 - 12, 2) 0/ 1 5 0, 0% 0/ 2 9 0, 0% 0. 0 (-20 ,4 - 11, 7) 4 3/ 56 5, 4% 6/ 60 10, 0% -4, 6 (-15, 4 - 6, 0) 0/ 4 0 0, 0% 2/ 3 7 5, 4% -5, 4 (-17 ,7 - 4, 2) 3 5/ 68 7, 4% 7/ 56 12, 5% -5, 1 (-17, 0 - 5, 6) 1/ 5 0 2, 0% 2/ 4 2 4, 8% -2, 8 (-13 ,9 - 6, 4) 2 4/ 65 6, 2% 8/ 74 10, 8% -4, 6 (-14, 5 - 5, 4) 4/ 3 8 10, 5% 1/ 4 0 2, 5% 8, 0 (-4, 1 - 21, 8) 1 5/ 77 6, 5% 6/ 66 9, 1% -2, 6 (-12, 6 - 6, 6) 0/ 2 8 0, 0% 1/ 2 9 3, 4% -3, 4 (-17 ,2 - 8, 9) 0 4/ 33 1 2, 1 % 2/ 62 3, 2% 8, 9 (-1, 8 - 24, 2) 0/ 1 8 0, 0% 1/ 1 5 6, 7% -6, 7 (-29 ,8 - 11, 7) Ta bl e s ho w s t ot al a nd ou tc om e nu m be rs a t e ac h le ve l, pe rc en ta ge s a nd 9 5% co nf id en ce in te rv al s f or th e dif fe re nc e in p er ce nt a ge s Table S2. Main out comes by parity and Bishop sc
or e a t r andomiz ation Table sho w s t ot al and out come number s a t each le vel, per cen tag es and 95% c on fidence in ter vals f or the diff er ence in per cen t ag es.
