Objective
To assess the optimal timing of delivery for women with gestational hypertension.
Study Design
A multicenter database that contained 228,668 deliveries was used to extract data on gravidas with gestational hypertension. The week-specific rates of maternal and neonatal morbidity/mortality were calculated after induction of labor. Point wise 95% confidence intervals were calculated around each of these gestational age-specific rates.
Results
After induction of labor, the rate of maternal morbidity/mortality reached a nadir of 89.9 per 1000 live births (95% confidence interval, 68.1–111.8) at 38-38 6/7 weeks’ gestation, although the rate of neonatal morbidity/mortality fell to 10.5 per 1000 live births (95% confidence interval, 2.8–18.2) at 39-39 6/7 weeks. There were only 3 total stillbirths in our study cohort.
Conclusion
In women with gestational hypertension, induction of labor between 38- and 39-weeks’ balances the lowest maternal and neonatal morbidity/mortality.
Hypertension is one of the most common medical disorders in pregnancy and a major cause of maternal and perinatal morbidity and death. Approximately 70% of women diagnosed with hypertension during pregnancy will have gestational hypertension or preeclampsia, which complicates 6% to 8% of all pregnancies. Women with gestational hypertension progress to mild preeclampsia and severe preeclampsia at rates of 46% and 9.6%, respectively.
The majority of gestational hypertensive and mild preeclamptic cases occur after 36-weeks’ gestation, but there is conflicting evidence and controversy regarding the timing for delivery of these women. The decision to induce labor at a given gestational age involves weighing the maternal and fetal risks associated with allowing a pregnancy to continue against the risks after delivery. It was previously thought that women with mild hypertensive disease occurring at 37-weeks’ gestation or later have a pregnancy outcome similar to that found in normotensive women, and previous recommendations have included outpatient management in those who were compliant, with induction of labor near term. However, as recently demonstrated, gravidas with mild disease (gestational hypertension or mild preeclampsia) allocated to expectant monitoring developed adverse maternal outcome in 44% compared with 31% assigned to induction of labor (relative risk [RR], 0.71; 95% confidence interval [CI], 0.59–0.86; P < .0001). Thus, induction of labor has now been suggested for women with mild hypertensive disease who have achieved 37-weeks’ gestation. However, Koopmans et al did not separate those with gestational hypertension from women with mild preeclampsia making it unclear whether morbidities were similar and recommendations should be the same for both. Although early delivery of the fetus may prevent stillbirth and other morbidities associated with ongoing pregnancy, recent evidence from general obstetric populations suggests that there are increased risks of serious neonatal morbidity (eg, prolonged hospitalization, sepsis, adverse respiratory outcomes) associated with delivery at 36, 37, and 38 weeks of gestation compared with 39 or 40 weeks.
The objective of this study was to assess the optimal timing of delivery for women with gestational hypertension by quantifying the risks of adverse maternal and fetal outcomes associated with induction of labor at each gestational week, from 36 to 41 completed weeks, compared with those with ongoing pregnancy. We hypothesized that these risks would decrease at each advancing gestational week at or near term but would increase postterm.
Materials and Methods
This is a retrospective cohort study from the Consortium on Safe Labor (CSL), which was sponsored by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). The primary goal was to establish a comprehensive database from electronic medical records from multiple sites to characterize labor and delivery in a contemporary group of women experiencing current obstetric clinical practices. More detailed information regarding the CSL has been published. The electronic database contained 228,668 deliveries greater than 23 weeks from 12 clinical centers and 19 hospitals representing 9 American College of Obstetricians and Gynecologists districts between 2002 and 2008, with the majority (87%) of births occurring from 2005 through 2007 (reflecting the period when individual institutions initiated their electronic medical record systems). An inhouse obstetrician was available 24 hours per day at 11 of the 12 participating sites. Participating institutions provided data on maternal demographics, medical history, reproductive and prenatal history, labor and delivery information, postpartum, and newborn information. Data inquiries, cleaning, and logic checking were performed on the database. Validation studies were also performed to ensure that the electronic database was a reasonably accurate representation of the medical charts and was noted to be highly consistent (97.3-99.7%). Institutional Review Board approval was obtained by all participating institutions.
For the cohort of interest, we only included the first pregnancy, within the timeframe of data collection, from each subject in the database to avoid intraperson correlation, and excluded multiples, leaving 206,969 deliveries. We included all gravidas who met the definition for gestational hypertension ≥36 gestational weeks. Gestational hypertension was defined as an elevation in blood pressure (BP) ≥140 mm Hg systolic or ≥90 mm Hg diastolic without proteinuria that developed in a woman after 20 weeks of gestation. All participating centers and investigators used standard criteria to define gestational hypertension; absence of proteinuria was diagnosed by urine dip, protein/creatinine ratio, or 24-hour urine collections according to local institutional standards. Pregnancies complicated by diabetes mellitus, cardiac, pulmonary, or renal disease were excluded. We calculated the stillbirth rates bases on this entire cohort of women with gestational hypertension. Gravidas presenting for induction of labor were then extracted from this cohort for calculation of small for gestational age (SGA), stillbirth, and week-specific morbidity and mortality rates.
Maternal outcomes included serious morbidity, which was a composite defined as any of the following: intensive care unit (ICU) admission, abruption, large blood loss (>0.5 L SVD, >1.0 L CS), postpartum hypertension, and mortality. The components of the maternal composite were chosen based on the variables that we found to be significant in this database in our previous investigation. Neonatal outcomes included stillbirth, SGA, serious neonatal morbidity, and neonatal mortality. SGA was defined as less than the 3rd percentile and as less than the 10th percentile of growth. Serious neonatal morbidity was defined as a composite outcome, which included any of the following: seizures, severe respiratory morbidity, sepsis, intracranial hemorrhage, necrotizing enterocolitis, or 5-minute Apgar score ≤3. The components of the neonatal composite were chosen because each of the events was very strongly associated with death and serious disability in long-term follow-up studies. Cesarean delivery after induction of labor was examined as a secondary outcome.
The gestational age-specific rates of stillbirth for ongoing pregnancies of entire cohort of women with gestational hypertension were calculated for each completed week of gestation between 36 and 41 weeks as :
(Number of stillbirths/Number of ongoing pregnancies) × 1000
The gestational week-specific rates of maternal and neonatal morbidity and mortality for live births after induction of labor were calculated for each completed week of gestation between 36 and 42 weeks, which was defined as :
[(Number of mothers (or infants) with serious morbidity and mortality delivered after induction)/(Number of infants delivered after induction)] × 1000
The gestational age-specific rates of SGA for ongoing pregnancies were calculated for each completed week of gestation as:
(Number of SGA/Number of infants delivered after induction) × 1000
Point wise 95% CI were calculated around each of these gestational age-specific rates. Statistical analyses were performed using FREQ, GLM, and LOGISTIC procedures in SAS 9.2 (SAS Institute Inc, Cary, NC).
Results
The electronic database included 228,668 deliveries. After exclusions for comorbidities and multiple gestations, there were a total of 3588 gravidas ≥36 weeks with gestational hypertension that met inclusion criteria. Table 1 describes the maternal characteristics of this cohort. Women with gestational hypertension had a mean admitting BP of 140.3 ± 13.4 mm Hg systolic/86.1 ± 11.7 mm Hg diastolic, and mean arterial BP 104.2 mm Hg ± 11.1.
| Maternal characteristics | GHTN n = 3588 |
|---|---|
| Maternal age | Mean (± SD) |
| Age, y | 26.6 (± 5.9) |
| Race, n (%) | |
| White | 2218 (61.8) |
| Black/non-Hispanic | 686 (19.1) |
| Hispanic | 484 (13.5) |
| Asian/Pacific Islander | 57 (1.6) |
| Other | 143 (4.0) |
| Parity (mean), n (%) | |
| 0 | 2128 (59.3) |
| 1 | 746 (20.8) |
| 2 | 408 (11.4) |
| 3 | 178 (5.0) |
| ≥4 | 128 (3.6) |
| BMI, mean (± SD) | |
| Prepregnancy BMI, kg/m 2 | 28.2 (± 7.1) |
| Delivery BMI, kg/m 2 | 34.1 (± 6.9) |
| Blood pressure, mm HG, mean (± SD) | |
| Admit systolic BP | 140.3 (± 13.4) |
| Admit diastolic BP | 86.1 (± 11.7) |
| Admit mean arterial BP | 104.2 (± 11.1) |
| Gestational age at delivery, wks | Mean (± SD) |
| Overall | 38.5 (± 1.3) |
| Insurance, n (%) | |
| Private | 2210 (61.6) |
| Public | 1207 (33.6) |
| Self-pay/unknown | 171 (4.8) |
| Substance use, n (%) | |
| Smoking | 161 (4.5) |
| Alcohol | 90 (2.5) |
| Illicit drug use | 36 (1.1) |
| IUGR, n (%) | |
| <3rd percentile | 74 (2.1) |
| <10th percentile | 258 (7.2) |
The numbers of ongoing pregnancies and stillbirth rates for the overall gestational hypertension cohort are provided in Table 2 . For those women undergoing induction of labor, the live births, cases of maternal and neonatal morbidity and mortality, along with rates, and 95% CIs for maternal and neonatal morbidity and mortality are also noted in Table 2 . There were no maternal deaths in our study cohort. There were only 3 stillbirths in our study cohort, 2 occurring at 37-weeks’ and 1 at 38-weeks’ gestation. After induction, the rate of maternal morbidity reached a nadir of 89.9 per 1000 live births (95% CI, 68.1–111.8) at 38-38 6/7 weeks’ gestation, although the rate of neonatal morbidity and mortality fell to 10.5 per 1000 live births (95% CI, 2.8–18.2) at 39-39 6/7 weeks.
| GA, a wks | Ongoing pregnancies, n | Stillbirth, n | Rate of stillbirth b (per 1000 ongoing pregnancies) | Following induction of labor | |||||
|---|---|---|---|---|---|---|---|---|---|
| Serious maternal morbidity/mortality, n | Rate of maternal morbidity/mortality (per 1000 live births) [95% CI] | Live births, n | Neonatal deaths, n | Serious neonatal morbidity/mortality, n | Rate of neonatal morbidity/mortality (per 1000 live births) [95% CI] | ||||
| 36 | 3588 | 0 | — | 24 | 106.2 [66.0–146.4] | 226 | 1 | 24 | 106.2 [66.0–146.4] |
| 37 | 3208 | 2 | 0.6 | 53 | 121.8 [91.1–152.6] | 435 | 0 | 13 | 29.9 [13.9–45.9] |
| 38 | 2521 | 1 | 0.4 | 59 | 89.9 [68.1–111.8] | 656 | 0 | 12 | 18.3 [8.0–28.6] |
| 39 | 1527 | 0 | — | 69 | 103.0 [80.0–126.0] | 670 | 0 | 7 | 10.5 [2.8–18.2] |
| 40 | 595 | 0 | — | 37 | 103.9 [72.2–135.6] | 356 | 0 | 10 | 28.1 [10.9–45.3] |
| 41 | 89 | 0 | — | 7 | 129.6 [40.0–219.2] | 54 | 0 | 2 | 37.0 [−13.3 to 87.4] |
| ≥42 | 7 | 0 | — | 0 | 0 | 4 | 0 | 0 | 0 |
a GA defined as completed weeks (ie, 36 completed weeks includes infants born between 36-0 and 36-6 inclusive);
The rates of SGA after induction of labor are displayed in Table 3 . The rate of SGA ≤10th percentile decreased sharply between 36- and 37-weeks’ gestation, from 115.6 (95% CI, 73.8–157.3) to 53.0 (95% CI, 31.9–74.1). The rate of SGA ≤3rd percentile reached a nadir of 11.9 between 39- and 39-6/7 weeks’ gestation (95% CI, 3.7–20.2) and then sharply rose to a rate of 25.5 between 40- and 40-6/7 weeks’ gestation (95% CI, 9.1–41.9). Table 4 demonstrates the cesarean delivery rate per 1000 inductions. What is notable is that the rate sharply increases from 286.5 (95% CI, 239.6–335.5) to 407.4 (95% CI, 276.4–538.5) between 40- and 41-weeks’ gestation. The majority of indications for the cesarean deliveries at 41-weeks’ gestation were cephalopelvic disproportion and fetal distress.
