Objective
The purpose of this study was to describe trends and outcomes of planned births.
Study Design
Data from linked birth and hospital records for 779,521 singleton births at ≥33 weeks’ gestation from 2001-2009 were used to determine trends in planned births (prelabor cesarean section and labor inductions). Adverse outcomes were composite indicators of maternal and neonatal morbidity/death.
Results
From 2001-2009, there were increases in labor inductions and prelabor cesarean deliveries at <40 weeks’ gestation, but no decrease in the stillbirth rate (trend P = .34). By 2009, 14.9% of live births at ≥33 weeks’ gestation were prelabor cesarean deliveries before the due date; 11.4% were inductions. As planned births increased, maternal risks shifted, which included a decline in inductions with maternal hypertension from 31.9-23.9%. Earlier birth was contemporaneous with increases (trend P < .001) in neonatal and maternal morbidity rates from 3.0-3.2% and 1.1-1.5%, respectively.
Conclusion
Planned birth before the due date is increasing without a contemporaneous reduction of stillbirths.
Planned birth is indicated if the perceived risk to the mother or baby of continuing the pregnancy outweighs the risk of early birth. Planned birth follows considered clinical decision-making about both the timing and the method of delivery. There is a widely held perception of a change in obstetric decision-making that has lowered the threshold at which and for which planned birth occurs. In the United States, both planned cesarean delivery and induction have increased, and the mean gestational age has decreased to 39 weeks’ gestation. At term, there is a consensus that induction at <39 weeks’ gestation should be minimized because of the attendant increased risk of neonatal respiratory morbidity if performed earlier. Preterm birth carries even greater short- and long-term morbidity. Worldwide, preterm birth is increasing; this increase, in some populations, has been attributed to planned preterm birth. Whether increased planned early birth rates have led to any increases in neonatal morbidity or reductions in stillbirth rates and the extent to which the implementation of restrictions on planned birth at <39 weeks’ gestation could affect stillbirth rates is uncertain. Two studies have reported no increase in stillbirth rates after the implementation of guidelines that restricted delivery at <39 weeks’ gestation ; however, one hospital reported that a fall in deliveries at <39 weeks’ gestation from 33-26% resulted in an increase in stillbirths at 38-39 weeks’ gestation.
Clinicians require population-based data to assess more fully the tradeoffs in morbidity for both the mother and baby after planned birth. Such data allow for informed decisions and more precisely articulate the risks of birth and the potential benefits of deferring delivery for even a short period. The aim of this study was to inform clinical decision-making about the timing of birth by describing recent trends in planned births and by determining the absolute risk of neonatal and maternal morbidity and death by the type of planned birth and gestational age. For the purpose of this study, we examined all planned births that occurred after 32 weeks’ gestation up until the estimated date of birth at 40 weeks’ gestation.
Methods and Materials
The study population included all singleton births delivered at >32 completed weeks of gestation in New South Wales (NSW) from January 1, 2001, to December 31, 2009. NSW is the most populous state in Australia with approximately 7 million people and approximately one-third of all Australian births in >100 hospitals. During the study period, only 0.1% of women had home births. In Australia, guidelines require that wherever possible deliveries at <33 weeks’ gestation are undertaken in tertiary perinatal centers.
Data for this study were obtained from 2 linked population datasets, birth data from the NSW Perinatal Data Collection and hospital data from the NSW Admitted Patients Data Collection. The Perinatal Data Collection is a statutory population-based collection that covers all births in NSW of at least 400 g birthweight or at least 20 weeks’ gestation and includes information on maternal characteristics, pregnancy, labor, delivery, and condition of the infant. The Admitted Patients Data Collection is a census of all admissions in NSW public and private hospitals. Diagnosis and procedures for each admission are coded according to the 10th revision of the International Classification of Diseases, Australian Modification and the Australian Classification of Health Interventions. The NSW Centre for Health Record Linkage performed probabilistic data linkage between the 2 datasets. The validity of the probabilistic record linkage is extremely high with <1% of records having an incorrect match. The NSW Ministry of Health provided anonymized data for mothers and babies for the birth admission and any subsequent hospital-to-hospital transfers. Ethics approval for the use of the linked data was obtained from the NSW Population and Health Services Research Ethics Committee.
Planned live births were those in which the onset of labor was not spontaneous (ie, labor induction or prelabor cesarean delivery), which indicated a considered decision was made about the timing of the birth. Births were identified for all gestational weeks before the due date (at <40 weeks’ gestation) and were stratified further into births with spontaneous onset of labor or by the method of planned birth: labor induction (regardless of mode of delivery) and prelabor cesarean delivery. Onset of labor was reported in the birth data by check-box (spontaneous, induction, or no labor) and compared with the medical records; a validation study found 98% agreement with the medical record (kappa statistic, 0.95). Maternal characteristics, gestational age, mode of delivery, reason for labor induction (check-box), fetal distress as indication for cesarean delivery, infant characteristics, and perinatal deaths were obtained from the birth data; additional information on maternal conditions (any hypertension, any diabetes mellitus, placenta previa, antepartum hemorrhage, rupture of membranes) were obtained from linked maternal hospital data. Gestational age is reported in completed weeks of gestation and as determined by the best clinical estimate including early (<20 weeks’ gestation) ultrasound scan (>97%) and last menstrual period. Small-for-gestational-age and large-for-gestational-age were defined as <10th percentile and >90th percentile birthweight for gestational age, respectively. Only factors that are well and accurately reported were included in the analyses. Missing data were infrequent; only 0.02% records were missing a gestational age, and 0.02% were missing mode of onset of labor.
Maternal and neonatal morbidity and deaths were assessed on the linked birth-hospital data, from the birth admission record to the first discharge from the hospital system, or death (never discharged), which included diagnoses and procedures that occurred subsequent to maternal or neonatal transfer. The primary outcomes were maternal and neonatal morbidity or death as determined by validated composite indicators. The Neonatal Adverse Outcome Indicator includes mortality rates and a comprehensive list of procedures and diagnoses that indicate severe neonatal morbidity, such as mechanical ventilation, respiratory distress syndrome, parenteral nutrition, sepsis, and hypoxic ischemic encephalopathy. Maternal morbidity and death were measured with a similar validated composite indicator that related to serious adverse maternal health outcomes (such as transfusion, pulmonary embolism, hysterectomy, and mechanical ventilation).
Statistical analyses
All singleton births >32 weeks’ gestation were used to test for trends in perinatal deaths (stillbirths and neonatal deaths) over time. Thereafter, analyses were limited to live births because the occurrence of fetal death may be a reason for planned birth.
Trends in the rates of inductions and prelabor cesarean deliveries and in maternal conditions were assessed with the use of a Cochrane-Armitage test for linear trend in proportions with the significance level set at a probability value of < .01 because of the large number of births. The list of maternal conditions that are associated with planned birth is not intended to be exhaustive but rather to capture trends in the conditions most usually associated with planned birth. The analysis of changes in maternal conditions that are associated with prelabor cesarean delivery was limited to nulliparous women as the first birth delivery mode is strongly determinative of subsequent delivery; prelabor cesarean delivery for multiparous women is usually elective repeat cesarean delivery.
Results
From 2001-2009, there were 779,521 singleton births of ≥33 weeks’ gestation in NSW, of whom 1643 infants (0.21%) were stillborn. Of the 777,813 live births, 403,473 infants (51.9%) were delivered at <40 weeks’ gestation (before the due date), which included 224,180 births after spontaneous labor. There were 179,206 planned births (23.0%) before the due date that included 101,292 prelabor cesarean deliveries (13.0%) and 77,914 labor inductions (10.0%); 87 of births (0.02%) before the due date had unknown labor onset.
Trends over time
From 2001-2009, there was an increase in planned births at <40 weeks’ gestation, as a proportion of all live singleton births ≥33 weeks’ gestation. By 2009, 26.2% of all live births ≥33 weeks’ gestation were both planned and before the due date; 14.9% were prelabor cesarean delivery births, and 11.4% were induced ( Table 1 ). During that period, there was no significant change in the stillbirth rate ( Figure 1 ) for all singletons ≥33 weeks’ gestation (2.1/1000 births; trend P = .34) nor in the neonatal death rate (3.7/100 births; trend P = .44).
| Year | All live births at ≥33 weeks’ gestation, n b | Planned and spontaneous births at 33-39 weeks’ gestation, n (%) | ||
|---|---|---|---|---|
| Labor induction (n = 77,914) | Prelabor cesarean delivery (n = 101,292) | Spontaneous labor (n = 224,180) | ||
| 2001 | 81,515 | 7283 (8.9) | 8381 (10.3) | 22,784 (28.0) |
| 2002 | 81,832 | 7513 (9.2) | 9025 (11.0) | 22,738 (27.8) |
| 2003 | 82,337 | 7489 (9.1) | 10,027 (12.2) | 22,735 (27.6) |
| 2004 | 81,605 | 7652 (9.4) | 10,076 (12.4) | 23,277 (28.5) |
| 2005 | 86,301 | 8653 (10.0) | 11,468 (13.3) | 24,023 (27.8) |
| 2006 | 88,320 | 8700 (9.9) | 12,461 (14.1) | 24,809 (28.1) |
| 2007 | 91,575 | 9993 (10.9) | 12,978 (14.2) | 27,468 (30.0) |
| 2008 | 91,801 | 10,168 (11.1) | 13,187 (14.4) | 28,301 (30.8) |
| 2009 | 92,133 | 10,463 (11.4) | 13,689 (14.9) | 28,045 (30.4) |
a Probability value for annual trend was < .001 for labor induction, prelabor cesarean section delivery, and births with spontaneous labor;
Although the contribution of planned births at the lower gestations was small ( Figure 1 ), planned births at 33-36S weeks’ gestation increased by 30.4%, from 1.5% in 2001 to 1.9% in 2009. During the later gestational ages, there was a differential increase by type of planned births; induction increased by 24% at 38 weeks’ gestation (3.2-4.0%) and by 26% at 39 weeks’ gestation (4.2-5.3%); prelabor cesarean deliveries increased by 25% at 38 weeks’ gestation (4.3-5.4%) and by 68% at 39 weeks’ gestation (4.2-7.1%).
Trends across gestational ages
The rates of planned births by gestational age among all live births at 33-39 weeks’ gestation are given in Figure 2 . At <37 weeks’ gestation, 35-40% of births in each gestational age were planned. At 38 weeks’ gestation, there was a sharp increase in both the number and rate of planned births, to 52.5% of births. This increase was driven primarily by a peak in the rate of prelabor cesarean deliveries, the majority of which (60.4%) were repeat procedures.
Maternal and pregnancy characteristics
Table 2 shows a comparison of pregnancy characteristics of 77,914 women who had labor induction at 33-39 weeks’ gestation over time and shows increasing rates of nulliparity, aged ≥35 years, maternal diabetes mellitus, and premature rupture of membranes (PROM). Pregnancy hypertension was the most common complication, but the proportion of inductions that were associated with pregnancy hypertension fell over time, from 31.9-23.9%, as the number of inductions increased. This was offset by a rise in inductions that were associated with diabetes mellitus and PROM, which resulted in a small rise in the total proportion of labor induction with at least 1 of the 4 listed maternal conditions (61.7-63.5%). The overall rate of vaginal birth after labor induction before the due date was 83.7%, which increased from 74.1% at 33-34 weeks’ gestation to 84.2% at 38-39 weeks’ gestation (including for nulliparous women, 69.2% and 71.5%, respectively, and the multiparous women, 78.3% and 93.2%, respectively). For induction with prostaglandin, the vaginal birth rate was lower overall (80.0% compared with 86.1% when prostaglandin were not used) and at each gestational age.
| Variable | Percentage of inductions 2001 (n = 7283) | Percentage of inductions 2009 (n = 10,463) | P for trend over all 9 years (n = 77,914) |
|---|---|---|---|
| Nulliparous | 41.2 | 45.7 | < .001 |
| Maternal age ≥35 y | 18.7 | 24.3 | < .001 |
| Any hypertension | 31.9 | 23.9 | < .001 |
| Any diabetes mellitus | 10.7 | 16.7 | < .001 |
| Prelabor rupture of membranes | |||
| Term | 16.3 | 20.7 | < .001 |
| Preterm | 2.1 | 2.7 | < .001 |
| Small for gestational age a or suspected fetal growth retardation | 12.9 | 13.2 | .82 |
| Any hypertension, diabetes mellitus, preterm rupture of membranes, or small for gestational age | 61.7 | 63.5 | < .001 |
Overall, the 101,292 women birthing by prelabor cesarean section delivery were older and of higher parity; most births (58.1%) were repeat cesareans. Table 3 shows the characteristics of the subgroup of 26,271 nulliparous women who were delivered by prelabor cesarean delivery at 33-39 weeks’ gestation, with fetal distress, placenta previa, and birthweight ≥4000 g added to the list of conditions. The proportion in which fetal distress was an indication declined from 6.2-4.7% as the number of prelabor cesareans increased, although the proportion that was associated with diabetes mellitus and PROM rose. The total proportion that was associated with at least 1 of the listed conditions rose from 50.0-52.6%. The proportion of women who were ≥35 years old (31.9% in 2009) was significantly higher than among nulliparous women whose labor was induced (15.2% in 2009) or who went into spontaneous labor (12.4% in 2009; P < .001 for both tests of 2 proportions).
| Variable | Percentage of prelabor first cesarean deliveries, 2001 (n = 2311) | Percentage of prelabor first cesarean deliveries, 2009 (n = 3564) | P for trend over all 9 years (n = 26,271) |
|---|---|---|---|
| Maternal age ≥35 y | 24.0 | 31.9 | < .001 |
| Any hypertension | 22.2 | 18.2 | .06 |
| Any diabetes mellitus | 8.7 | 11.2 | < .01 |
| Prelabor rupture of membranes | |||
| Term | 1.5 | 3.8 | < .001 |
| Preterm | 1.6 | 2.1 | .11 |
| Small for gestational age a or suspected fetal growth retardation | 13.9 | 13.1 | .84 |
| Birthweight ≥4000 g | 6.4 | 7.5 | < .01 |
| Fetal distress | 6.2 | 4.7 | < .001 |
| Placenta previa | 7.8 | 10.0 | < .01 |
| Any of hypertension, diabetes mellitus, preterm rupture of membranes, small for gestational age, birthweight ≥4000 g, distress, or placenta previa | 50.0 | 52.6 | < .001 |
Neonatal outcomes after planned and spontaneous births
As planned births increased, severe neonatal morbidity/mortality rates among all live singleton births at ≥33 weeks’ gestation increased from 3.0% in 2001 to 3.2% in 2009 (trend, P < .001; Figure 1 ). Among the 179,206 planned births at 33-39 weeks’ gestation, 8107 infants (4.5%) experienced a serious adverse outcome before discharge from the hospital system, which included 765 neonatal deaths (4.3 per 1000). For the 224,180 infants who were born after spontaneous onset of labor before their due date, the morbidity rate was 3.3%, and the neonatal death rate was 4.0 per 1000. When limited to births at 37-39 weeks’ gestation, there was no increase in neonatal morbidity rates that were associated with prelabor cesarean delivery compared with labor induction (relative risk, 0.97; 95% confidence interval, 0.91–1.02). For both planned and spontaneous births, there was a steady decrease in morbidity rates as gestational age advanced ( Figure 3 , A). Not obvious from the figure’s vertical scale is that the rate of neonatal morbidity for planned births continued to drop from 3.0% at 38 weeks’ gestation to 2.1% at 39 weeks’ gestation ( P < .001), which was a relative decrease of 30%.
