- Email: [email protected]
Active induction management of labor for diabetic pregnancies at term; mode of delivery and fetal outcome—a single center experience
European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
Active induction management of labor for diabetic pregnancies at term; mode of delivery and fetal outcome—a single center experience Yariv Yogeva,b,*, Avi Ben-Harousha,b, Rony Chena,b, Hagit Glickman, Boris Kaplana,b, Moshe Hoda,b a
Perinatal Division and WHO Collaborating Center, Department of Obstetrics and Gynecology, Rabin Medical Center, Beilinson Campus, Petah Tiqva 49100, Israel b Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
Received 1 January 2003; received in revised form 1 September 2003; accepted 21 October 2003
Abstract Objectives: To evaluate the mode of delivery in diabetic pregnancies at term following induction of labor with vaginal application of prostaglandin E2 (PGE2), and to identify possible predictors of successful vaginal delivery. Patients and methods: The study group consisted of 105 women with diabetic pregnancies at term admitted for induction of labor; 84 (80%) had gestational diabetes (GDM) and 21 (20%) type 1 diabetes. Findings were compared with women who underwent elective induction of labor (n ¼ 115), and women with normal spontaneous onset of labor (n ¼ 510). Women with previous cesarean section (CS) were excluded from both study and control groups. Results: Maternal age and gravidity were significantly higher in the study group than the control groups (age: 31:4 5, 28 5:0 and 28:1 4:8 years, respectively; gravidity: 3:0 1:9, 2:5 1:6, and 2:1 1:4, respectively; P < 0:001 for both) and gestational age and nulliparity rate were significantly lower (gestational age: 38:6 1:1, 40:2 1:3 and 39:3 2:7 weeks, respectively; nulliparity: 34.6, 45.2, 51.6%, respectively; P < 0:002 for both). There were no between-group differences in the incidence of oligohydramnios, number of PGE2 applications used, birth weight, rate of non-reassuring fetal heart rate pattern leading to CS, and rate of low 5 min Apgar score (<7). The rate of CS in the study group (18.2%) was significantly higher than in the spontaneous labor group (9%) but similar to the elective induction group (14.8%). On stepwise analysis, only nulliparity (OR 4.56, 95% CI 1.11–18.67, P ¼ 0:035) was independently and significantly associated with increased risk of CS. Within the study group (R2 ¼ 0:257, P ¼ 0:002), type 1 diabetes (OR 2.4, 95% CI 1.04–5.51) was independently and significantly associated with increased risk of CS. Conclusion: In diabetic pregnancies, induction of labor at term with vaginal PGE2 is successful in approximately 82% of patients, but yields a significantly higher CS rate compared to uncomplicated pregnancies. Nulliparity and diagnosis of type 1 diabetes are independently and significantly associated with increased risk of CS. Condensation: In diabetic pregnancies, induction of labor at term is successful in 82% of patients, but yields higher CS rates compared to uncomplicated pregnancies. # 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Gestational diabetes; Labor induction; PGE2; Cesarean section
1. Introduction The timing of delivery in pregnancies complicated with diabetes at term remains controversial. The rationale for elective labor induction include the avoidance of fetal demise and the prevention of excessive fetal growth and its concomitant conditions, shoulder dystocia and cesarean delivery. However, there is very little evidence to support either elective delivery or expectant management. Kjos et al. [1] in a randomized control study evaluated women at 38 weeks of gestation who had insulin-treated gestational * Corresponding author. Tel.: þ972-3-9377400; fax: þ972-3-6417324. E-mail address: [email protected] (Y. Yogev).
diabetes mellitus (GDM) and fetuses believed to be of appropriate weight for gestational age. They found no difference in cesarean delivery and shoulder dystocia rates between those who underwent active induction of labor and those who were randomized for expectant management. However, the rate of large-for-gestational age (LGA) infants was lower in the induction group. A recent systematic review showed that compared with placebo or no treatment, induction of labor with vaginal application of prostaglandin E2 (PGE2) in a heterogeneous group of women with term pregnancies did not increase the rate of CS [2]. Our departmental management protocol is to offer labor induction at 38–39 weeks of pregnancy to women with type 1 or insulin-treated GDM and to women with diet-treated
0301-2115/$ – see front matter # 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2003.10.017
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
GDM and a LGA fetus. In all diabetic women, the sonographically estimated fetal weight should be less than 4000 g. If estimated fetal weight is greater, patients are offered CS. The aim of the present study was to evaluate the mode of delivery in diabetic pregnancies following induction of labor with vaginal PGE2 and to identify possible predictors of successful vaginal delivery.
2. Methods The study patients were derived from a cohort of 1860 consecutive women admitted to our Perinatal Division for induction of labor for variable indications from January 1998 to December 2000 (8% of the total number of women who gave birth at Rabin Medical Center during this period). Women with multiple pregnancies, gravidity more than 5 or non-cephalic presentation, were referred for a non-PGE2 induction (i.e. with intravenous oxytocin or intracervical balloon). Women with spontaneous uterine contractions, a Bishop score >7, previous CS or refused labor induction were excluded. In addition, all women (in the study group) with a sonographically estimated fetal weight above 4000 g were offered a CS. The remaining 1541 women underwent induction of labor with PGE2 tablets. Of these, 105 had a diabetic pregnancy, and this group served as the sample for the present study. Eighty-four patients had GDM (80%) and 21 (20%) had type 1 diabetes. Type 1 diabetes was diagnosed before the onset of the current pregnancy. GDM was diagnosed according to the criteria of the Fourth International WorkshopConference on Gestational Diabetes Mellitus [3]. Importantly, diabetes was the only obstetric complication in the study group. Findings were compared with two control groups. The first consisted of healthy women with normal pregnancies admitted at term for elective labor induction due to maternal request; (group 2, n ¼ 115), and this group represented labor induction in uncomplicated pregnancies. The second control group consisted of healthy women with normal pregnancies and spontaneous onset of labor (group 3, n ¼ 510). The latter group was randomly recruited from delivery room logbook for the study period. Exclusion criteria for group 3 were multiple pregnancy, non-cephalic presentation, prior induction of labor, elective CS, maternal demand for CS, previous CS and contraindications for vaginal delivery, such as placenta previa. Prior to PGE2 application, all patients received a detailed explanation of the procedure and provided written informed consent. All underwent physical and vaginal examination, non-stress test, and detailed ultrasound assessment for estimated fetal weight, amniotic fluid index and biophysical profile. The usual regimen of the commercially available PGE2 induction was intravaginal application of a 3 mg tablet of Dinoprostone (Pharmacia Upjohn, Puurf, Belgium)
167
to the posterior fornix. If after one dose, cervical change was inadequate with minimal uterine activity; additional tablets were given at intervals of 6–8 h. Those in whom the Bishop score rose to 7 or more were transferred to delivery room, where labor was further augmented with oxytocin or artificial rupture of the membranes, as necessary. Induction was considered a failure if there was no significant change in the Bishop score (Bishop’s score 7 or more) after four applications of PGE2. Demographic and antenatal data were recorded on a computerized database. Antenatal outcome data, including mode of delivery, indications for CS, gestational age, birth weight, Apgar score and admission to a neonatal intensive care unit were collected retrospectively from the labor and delivery logbook, the department’s computerized database or patient files. A low Apgar score was defined as 7 at 5 min. Data were analyzed with the SPSS for Windows (version 10.0) statistical package. Results are expressed as means S:D: or rate. Comparisons between the groups and subgroups were performed with analysis of variance (ANOVA) with Tukey’s multiple comparisons or Student’s ttest for continuous data, and with chi-square or Fisher’s exact test for categorical data. The association between CS and possible predictors was calculated by logistic regression and forward likelihood analysis, controlling for related variables. A P value of <0.05 was considered significant. The odds ratio (OR) and 95% confidence interval (CI) were calculated where appropriate.
3. Results Table 1 presents the antenatal and postnatal data of the study and control groups. Maternal age and gravidity were significantly higher in the study group compared to the two control groups, and gestational age and nulliparity rate were significantly lower. Gestational age at delivery was also significantly lower in the study group. There were no between-group differences in the incidence of oligohydroamnios, number of PGE2 applications, birth weight, rate of operative vaginal deliveries, or rate of low 5 min Apgar score. Analysis of the indications for CS (Table 2) yielded no significant difference in rate of non-reassuring fetal heart rate pattern, failed induction or labor dystocia between the study and control groups. The CS rate was significantly higher in the study group (18.2%) compared with spontaneous delivery group (9%) but not with the elective induction group (14.8%). Separate comparison within the study group between the women with gestational or type 1 diabetes showed no difference in maternal age, gestational age, gravidity, mean number of PGE2 applications, or birth weight. Table 3 shows the antenatal data of the study patients with vaginal delivery or CS. The CS subgroup was characterized by a higher rate of use of more than one application of PGE2. A logistic regression model (R2 ¼ 0:213,
168
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
Table 1 Antenatal and postnatal data in study and control groups Study group (n ¼ 104)
Group 2— ELECTIVE induction (n ¼ 115)
Group 3— spontaneous onset of labor (n ¼ 510)
P value
Age (years) Gravidity Parity Nulliparity rate Incidence of oligohydroamnios Number of PGE2 applications Rate of more than one PGE2 application Sonographically estimated fetal weight Birth weight (g) Gestational age at delivery (weeks) Cesarean section rate
31.4 5.0 3.0 1.9 1.5 1.5 36 (34.6%) 13 (12.5%) 1.9 1.0 56 (54.3%) 3234 485 (22434000) 3270 490 (18004720) 38.6 1.1 (3542) 19 (18.2%)
28.5 5.0 2.5 1.6 1.0 1.2 52 (45.2%) 18 (5.6%) 1.6 1.0 73 (63.4%) 3402 401 (25004295) 3317 425 (24004400) 40.2 1.3 (3742) 17 (14.8%)
28.1 4.8 2.1 1.4 0.85 1.1 263 (51.6%) – – – – 3258 456 (17804560) 39.3 2.7 (3242) 46 (9%)
Operative vaginal delivery rate (%) Macrosomia (>4000 g) rate (%) 5 min Apgar score 7
7 (7.1%) 6 (5.7%) 3 (2.9%)
13 (11.3%) 6 (5.2%) 3 (2.6%)
50 (10.7%) 30 (5.8%) 6 (1.2%)
<0.001a,b <0.001a,b <0.001a,b NSc, 0.002d NS NS NS NS NS <0.01a,e NSc, 0.007d, OR ¼ 2.01 (1.22–3.32) NS NS NS
a
ANOVA. Tukey’s multiple comparison test (P < 0:05), group 1 > group 2 ¼ group 3. c For the difference between groups 1 and 2. d For the difference between groups 1 and 3. e Tukey’s multiple comparison test (P < 0:05), group 1 < group 3 < group 2. b
Table 2 Indications for cesarean section in study and control groups
Failed inductionb Non-reassuring fetal heart rate Labor dystociac Otherd Missing data
Study group (N ¼ 19)a
Elective induction (N ¼ 17)a
Spontaneous onset of delivery (N ¼ 46)a
P value
2 2 5 8 2
5 7 8 1 2
– 16 (34.7%) 20 (43.5%) 8 (17.4%) 5 (10.8%)
NS NS NS
(10.5%) (10.5%) (26.3%) (42.1%) (10.5%)
(29.4%) (41.2%) (47%) (5.8%) (11.7%)
a
Some patients had more than one indication for cesarean section. Failed induction—four or more applications of PGE2 and a Bishop score of <7. c Labor dystocia—dysfunctional active phase and prolonged second stage not suitable for instrumental delivery. d Severe preeclampsia or eclampsia, prolapse of umbilical cord, vaginal bleeding, placental abruption, maternal request. b
P < 0:001) was used to control for maternal and gestational age, gravidity, parity, nulliparity rate, induction of labor, number of PGE2 tablets used, presence of oligohydramnios, birth weight, and diagnosis of diabetes as predictors of mode
of delivery. On stepwise analysis, only nulliparity (OR 4.56, 95% CI 1.11–18.67, P ¼ 0:035) was independently and significantly associated with an increased risk of CS. A similar analysis within the study group (R2 ¼ 0:257,
Table 3 Antenatal data within the study group of women who delivered vaginally or by cesarean section
Age (years) Gravidity Parity Nulliparity rate Incidence of oligohydramnios Number of PGE2 tablets used Percentage of patients requiring more than one application Gestational age at delivery (weeks) Birth weight (g) Summary data are presented as mean S:D: or incidence.
Cesarean section (N ¼ 19)
Vaginal delivery (N ¼ 85)
P value
31.2 6.5 2.8 2.0 1.2 1.6 44.4% 16.7% 2.0 0.9 61.1% 38.4 1.0 3259 513
31.6 4.7 3.1 1.9 1.6 1.5 32.1% 12.5% 1.9 1.0 51.9% 38.6 1.1 3267 487
NS NS NS NS NS NS 0.03 NS NS
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
P ¼ 0:002), showed that a diagnosis of type 1 diabetes (OR 2.4, 95% CI 1.04–5.51) was independently and significantly associated with an increased risk of CS.
4. Discussion It is generally recommended that in diabetic pregnancies, delivery should be delayed until fetal lung maturation is completed, provided that the diabetes is well controlled and findings on antepartum surveillance remain normal. In these cases, elective induction is usually planned at 38–40 weeks of gestation and CS is performed only if induction of labor fails or if fetal macrosomia is suspected. In our study, we induced diabetic pregnancies only after completing full 38 weeks of pregnancy in order to achieve fetal lung maturation. Although labor induction is recommended for various subgroups in diabetic pregnancies near or at term [4,5], information on outcome is limited. According to one comprehensive review of the literature [6], only one relevant randomized controlled trial has been conducted so far [1]. In a protocol comparison between two time periods, Lurie and colleagues [4] reported no significant difference in macrosomia or cesarean delivery rates by management policy, although the patients in the expectant management group who gave birth after 40 weeks gestation had a higher rate of shoulder dystocia (10%) than the patients in whom labor was induced at 38–39 weeks gestation (1.4%). Although this difference was significant, these data have not been confirmed by other studies. Hod et al. [5] showed that maintaining strict control of maternal diabetes combined with elective delivery at 38–39 weeks of gestation, depending on the estimated fetal weight, leads to a significant reduction in the rate of macrosomia, thereby potentially decreasing the incidence of both traumatic births and cesarean section (CS). It is noteworthy that none of these studies included women with GDM who were not treated with insulin. Furthermore, the methods of cervical ripening and induction of labor, and the indications for CS, were either not controlled or not specified. Further assessments of the mode of delivery in this patient population are therefore needed to determine the optimal obstetrical management. Objectively evaluating the risks and benefits of labor induction in diabetic women is potentially confounded by a number of variables. Among these is the status of the cervix at the time of initiation of induction, the conduct of labor and the reliance on estimates fetal weight to determine route of delivery. The fact that the present study was based on the experience of a single center, over a 3-year period, with the same team of specialists, using the same strict protocol minimizes the potential confounding factors characteristic of studies with different management protocols applied in different medical centers.
169
We studied 105 consecutive women with diabetic pregnancy that underwent induction of labor with vaginal tablets of PGE2; all had an unfavorable cervix prior to labor induction. Importantly, the GDM sub-population was heterogeneous, some being treated by diet alone and other by diet combined with insulin. Induction of labor with vaginal tablets of PGE2 led to successful vaginal delivery in approximately 82% of the patients, with apparently no serious maternal or fetal complications. The CS rate in the diabetic pregnancy group was significantly higher than in the normal, non-induced pregnancy group, but similar to that in the elective induction group. Earlier studies have shown that CS rates are higher in women with diabetic pregnancies compared with controls. However, the difference is not entirely attributable to fetal macrosomia [7,8]. It may be that caregivers are more prone to operate on patients with diabetic pregnancy because of concern about the likelihood of shoulder dystocia. Furthermore, it is generally acknowledged that induction of labor by itself for medical indications is associated with an increased risk of CS [9–13]. For elective induction, some studies found higher rates [14–16], restricted to nulliparous patients [10,15], and others reported lower or equivalent rates [9,10]. In the most recent study, Cammu et al. [17] showed that when compared with spontaneous deliveries, elective labor induction in nulliparous women is associated with significantly more CS and instrumental deliveries. Importantly, in our study, the nulliparity rate was significantly lower in the study than in the control groups; thus, if the nulliparity rate were higher, we would expect an even higher rate of CS in the study group. In the logistic regression model and forward likelihood analysis, only nulliparity and, within the study group, a diagnosis of type 1 diabetes, were independently and significantly associated with an increased risk of CS. Therefore, we cannot claim that the induction process by itself or a diagnosis of diabetic pregnancy per se causes a high rate of CS. The lack of between-group differences in rates of non-reassuring fetal heart rate pattern, labor dystocia leading to CS, and low 5 min Apgar score (<7) indicates that these variables too, cannot explain the higher rate of CS in the study group, which also had non-significantly lower incidence of failed induction than the elective induction group. Our analysis of the antenatal data within the study group by type of delivery yielded a greater need for more than one PGE2 application in the women who underwent CS than in those with vaginal delivery. Thus, it may be that the induction process combined with an unfavorable cervix increased the CS rate in our sample. The impact of epidural analgesia on the rate of CS remains controversial [18]. Data on epidural analgesia were not analyzed specifically in our study. However, on crude analysis of the study period, the rate of use of epidural analgesia in labor was higher in nulliparous (80%) than the multiparous women (50%).
170
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
Importantly, our management protocol yielded a similar rate of macrosomia and birth weight in the diabetic and the non-diabetic groups, although gestational age at delivery was significantly lower in the study group. Thus, using active induction management of labor may prevent higher rates of macrosomia at the price of a higher rate of CS. Although there were no major maternal or fetal complications in the study and control groups, the retrospective study design and the relatively small sample preclude us from ruling out minor maternal or fetal complications or later neonatal complications. Further studies are needed in larger samples. An interventional prospective study is needed to evaluate if prolongation of pregnancy without active induction might lead to a higher rate of spontaneous deliveries without a significant increase in birth weight, macrosomia, traumatic deliveries and higher incidence of stillbirth. In summary, currently available evidence suggests that, while induction of labor for women who have diabetes may not carry maternal or fetal risk, the benefit of this procedure is unclear. References [1] Kjos SL, Henry OA, Montoro M, Buchanan TA, Mestman JH. Insulin-requiring diabetes in pregnancy: a randomized trial of active induction of labor and expectant management. Am J Obstet Gynecol 1993;169:611–5. [2] Kelly AJ, Kavanagh J, Thomas J. Vaginal prostaglandin (PGE2 and PGF2a) for induction of labor at term (Cochrane Review). In: The cochrane library, Issue 3; 2002. Oxford: Update Software. [3] Metzger BE, Coustan DR. Summary and recommendations of the Fourth International Workshop-Conference on Gestational Diabetes Mellitus. Diab Care 1998;21(suppl 2):B161–7. [4] Lurie S, Insler V, Hagay ZJ. Induction of labor at 38–39 weeks of gestation reduces the incidence of shoulder dystocia in gestational diabetic patients Class A2. Am J Perinatol 1996;13:293–6.
[5] Hod M, Bar J, Peled Y, Fried S, Katz I, Itzhak M, et al. Antepartum management protocol. Timing and mode of delivery in gestational diabetes. Diab Care 1998;21(Suppl 2):B113–7. [6] Boulvain M, Stan C, Irion O. Elective delivery in diabetic pregnant women (Cochrane Review). In: The cochrane library, Issue 3; 2000. Oxford: Update Software. [7] Casey BM, Lucas MJ, McIntire DD, Leveno KJ. Pregnancy outcomes in women with gestational diabetes compared with the general obstetric population. Obstet Gynecol 1997;90:869–73. [8] Naylor CD, Sermer M, Chen E, Sykora K. Cesarean delivery in relation to birth weight and gestational glucose tolerance: pathophysiology or practice style? Toronto Trihospital Gestational Diabetes Investigators. JAMA 1996;275:1165–70. [9] 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;180:628–33. [10] Jarvelin MR, Hartikainen-Sorri AL, Rantakallio P. Labour induction policy in hospitals of different levels of specialisation. Br J Obstet Gynaecol 1993;100:310–5. [11] Robson S, Pridmore B, Dobb J. Outcomes of induced labor. Aust N Z J Obstet Gynaecol 1997;31:167–9. [12] Rojansky N, Reubinoff B, Tanos V, Shushan A, Weinstein D. High risk pregnancy outcome following induction of labour. Eur J Obstet Gynecol Reprod Biol 1997;72:153–8. [13] 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;94:600–7. [14] Smith LP, Nagourney BA, McLean FH, Usher RH. Hazards and benefits of elective induction of labor. Am J Obstet Gynecol 1984;148:579–85. [15] Macer JA, Macer CL, Chan LS. Elective induction versus spontaneous labor: a retrospective study of complications and outcome. Am J Obstet Gynecol 1992;166:1690–7. [16] Prysak M, Castronova FC. Elective induction versus spontaneous labor: a case-control analysis of safety and efficacy. Obstet Gynecol 1998;92:47–52. [17] Cammu H, Martens G, Ruyssinck G, Amy JJ. Outcome after elective labor induction in nulliparous women: a matched cohort study. Am J Obstet Gynecol 2002;186:240–4. [18] American College of Obstetricians and Gynecologists. Obstetric, Analgesia and Anesthesia. Practice Bulletin number 36; 2002.
Active induction management of labor for diabetic pregnancies at term; mode of delivery and fetal outcome—a single center experience Yariv Yogeva,b,*, Avi Ben-Harousha,b, Rony Chena,b, Hagit Glickman, Boris Kaplana,b, Moshe Hoda,b a
Perinatal Division and WHO Collaborating Center, Department of Obstetrics and Gynecology, Rabin Medical Center, Beilinson Campus, Petah Tiqva 49100, Israel b Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
Received 1 January 2003; received in revised form 1 September 2003; accepted 21 October 2003
Abstract Objectives: To evaluate the mode of delivery in diabetic pregnancies at term following induction of labor with vaginal application of prostaglandin E2 (PGE2), and to identify possible predictors of successful vaginal delivery. Patients and methods: The study group consisted of 105 women with diabetic pregnancies at term admitted for induction of labor; 84 (80%) had gestational diabetes (GDM) and 21 (20%) type 1 diabetes. Findings were compared with women who underwent elective induction of labor (n ¼ 115), and women with normal spontaneous onset of labor (n ¼ 510). Women with previous cesarean section (CS) were excluded from both study and control groups. Results: Maternal age and gravidity were significantly higher in the study group than the control groups (age: 31:4 5, 28 5:0 and 28:1 4:8 years, respectively; gravidity: 3:0 1:9, 2:5 1:6, and 2:1 1:4, respectively; P < 0:001 for both) and gestational age and nulliparity rate were significantly lower (gestational age: 38:6 1:1, 40:2 1:3 and 39:3 2:7 weeks, respectively; nulliparity: 34.6, 45.2, 51.6%, respectively; P < 0:002 for both). There were no between-group differences in the incidence of oligohydramnios, number of PGE2 applications used, birth weight, rate of non-reassuring fetal heart rate pattern leading to CS, and rate of low 5 min Apgar score (<7). The rate of CS in the study group (18.2%) was significantly higher than in the spontaneous labor group (9%) but similar to the elective induction group (14.8%). On stepwise analysis, only nulliparity (OR 4.56, 95% CI 1.11–18.67, P ¼ 0:035) was independently and significantly associated with increased risk of CS. Within the study group (R2 ¼ 0:257, P ¼ 0:002), type 1 diabetes (OR 2.4, 95% CI 1.04–5.51) was independently and significantly associated with increased risk of CS. Conclusion: In diabetic pregnancies, induction of labor at term with vaginal PGE2 is successful in approximately 82% of patients, but yields a significantly higher CS rate compared to uncomplicated pregnancies. Nulliparity and diagnosis of type 1 diabetes are independently and significantly associated with increased risk of CS. Condensation: In diabetic pregnancies, induction of labor at term is successful in 82% of patients, but yields higher CS rates compared to uncomplicated pregnancies. # 2003 Elsevier Ireland Ltd. All rights reserved. Keywords: Gestational diabetes; Labor induction; PGE2; Cesarean section
1. Introduction The timing of delivery in pregnancies complicated with diabetes at term remains controversial. The rationale for elective labor induction include the avoidance of fetal demise and the prevention of excessive fetal growth and its concomitant conditions, shoulder dystocia and cesarean delivery. However, there is very little evidence to support either elective delivery or expectant management. Kjos et al. [1] in a randomized control study evaluated women at 38 weeks of gestation who had insulin-treated gestational * Corresponding author. Tel.: þ972-3-9377400; fax: þ972-3-6417324. E-mail address: [email protected] (Y. Yogev).
diabetes mellitus (GDM) and fetuses believed to be of appropriate weight for gestational age. They found no difference in cesarean delivery and shoulder dystocia rates between those who underwent active induction of labor and those who were randomized for expectant management. However, the rate of large-for-gestational age (LGA) infants was lower in the induction group. A recent systematic review showed that compared with placebo or no treatment, induction of labor with vaginal application of prostaglandin E2 (PGE2) in a heterogeneous group of women with term pregnancies did not increase the rate of CS [2]. Our departmental management protocol is to offer labor induction at 38–39 weeks of pregnancy to women with type 1 or insulin-treated GDM and to women with diet-treated
0301-2115/$ – see front matter # 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2003.10.017
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
GDM and a LGA fetus. In all diabetic women, the sonographically estimated fetal weight should be less than 4000 g. If estimated fetal weight is greater, patients are offered CS. The aim of the present study was to evaluate the mode of delivery in diabetic pregnancies following induction of labor with vaginal PGE2 and to identify possible predictors of successful vaginal delivery.
2. Methods The study patients were derived from a cohort of 1860 consecutive women admitted to our Perinatal Division for induction of labor for variable indications from January 1998 to December 2000 (8% of the total number of women who gave birth at Rabin Medical Center during this period). Women with multiple pregnancies, gravidity more than 5 or non-cephalic presentation, were referred for a non-PGE2 induction (i.e. with intravenous oxytocin or intracervical balloon). Women with spontaneous uterine contractions, a Bishop score >7, previous CS or refused labor induction were excluded. In addition, all women (in the study group) with a sonographically estimated fetal weight above 4000 g were offered a CS. The remaining 1541 women underwent induction of labor with PGE2 tablets. Of these, 105 had a diabetic pregnancy, and this group served as the sample for the present study. Eighty-four patients had GDM (80%) and 21 (20%) had type 1 diabetes. Type 1 diabetes was diagnosed before the onset of the current pregnancy. GDM was diagnosed according to the criteria of the Fourth International WorkshopConference on Gestational Diabetes Mellitus [3]. Importantly, diabetes was the only obstetric complication in the study group. Findings were compared with two control groups. The first consisted of healthy women with normal pregnancies admitted at term for elective labor induction due to maternal request; (group 2, n ¼ 115), and this group represented labor induction in uncomplicated pregnancies. The second control group consisted of healthy women with normal pregnancies and spontaneous onset of labor (group 3, n ¼ 510). The latter group was randomly recruited from delivery room logbook for the study period. Exclusion criteria for group 3 were multiple pregnancy, non-cephalic presentation, prior induction of labor, elective CS, maternal demand for CS, previous CS and contraindications for vaginal delivery, such as placenta previa. Prior to PGE2 application, all patients received a detailed explanation of the procedure and provided written informed consent. All underwent physical and vaginal examination, non-stress test, and detailed ultrasound assessment for estimated fetal weight, amniotic fluid index and biophysical profile. The usual regimen of the commercially available PGE2 induction was intravaginal application of a 3 mg tablet of Dinoprostone (Pharmacia Upjohn, Puurf, Belgium)
167
to the posterior fornix. If after one dose, cervical change was inadequate with minimal uterine activity; additional tablets were given at intervals of 6–8 h. Those in whom the Bishop score rose to 7 or more were transferred to delivery room, where labor was further augmented with oxytocin or artificial rupture of the membranes, as necessary. Induction was considered a failure if there was no significant change in the Bishop score (Bishop’s score 7 or more) after four applications of PGE2. Demographic and antenatal data were recorded on a computerized database. Antenatal outcome data, including mode of delivery, indications for CS, gestational age, birth weight, Apgar score and admission to a neonatal intensive care unit were collected retrospectively from the labor and delivery logbook, the department’s computerized database or patient files. A low Apgar score was defined as 7 at 5 min. Data were analyzed with the SPSS for Windows (version 10.0) statistical package. Results are expressed as means S:D: or rate. Comparisons between the groups and subgroups were performed with analysis of variance (ANOVA) with Tukey’s multiple comparisons or Student’s ttest for continuous data, and with chi-square or Fisher’s exact test for categorical data. The association between CS and possible predictors was calculated by logistic regression and forward likelihood analysis, controlling for related variables. A P value of <0.05 was considered significant. The odds ratio (OR) and 95% confidence interval (CI) were calculated where appropriate.
3. Results Table 1 presents the antenatal and postnatal data of the study and control groups. Maternal age and gravidity were significantly higher in the study group compared to the two control groups, and gestational age and nulliparity rate were significantly lower. Gestational age at delivery was also significantly lower in the study group. There were no between-group differences in the incidence of oligohydroamnios, number of PGE2 applications, birth weight, rate of operative vaginal deliveries, or rate of low 5 min Apgar score. Analysis of the indications for CS (Table 2) yielded no significant difference in rate of non-reassuring fetal heart rate pattern, failed induction or labor dystocia between the study and control groups. The CS rate was significantly higher in the study group (18.2%) compared with spontaneous delivery group (9%) but not with the elective induction group (14.8%). Separate comparison within the study group between the women with gestational or type 1 diabetes showed no difference in maternal age, gestational age, gravidity, mean number of PGE2 applications, or birth weight. Table 3 shows the antenatal data of the study patients with vaginal delivery or CS. The CS subgroup was characterized by a higher rate of use of more than one application of PGE2. A logistic regression model (R2 ¼ 0:213,
168
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
Table 1 Antenatal and postnatal data in study and control groups Study group (n ¼ 104)
Group 2— ELECTIVE induction (n ¼ 115)
Group 3— spontaneous onset of labor (n ¼ 510)
P value
Age (years) Gravidity Parity Nulliparity rate Incidence of oligohydroamnios Number of PGE2 applications Rate of more than one PGE2 application Sonographically estimated fetal weight Birth weight (g) Gestational age at delivery (weeks) Cesarean section rate
31.4 5.0 3.0 1.9 1.5 1.5 36 (34.6%) 13 (12.5%) 1.9 1.0 56 (54.3%) 3234 485 (22434000) 3270 490 (18004720) 38.6 1.1 (3542) 19 (18.2%)
28.5 5.0 2.5 1.6 1.0 1.2 52 (45.2%) 18 (5.6%) 1.6 1.0 73 (63.4%) 3402 401 (25004295) 3317 425 (24004400) 40.2 1.3 (3742) 17 (14.8%)
28.1 4.8 2.1 1.4 0.85 1.1 263 (51.6%) – – – – 3258 456 (17804560) 39.3 2.7 (3242) 46 (9%)
Operative vaginal delivery rate (%) Macrosomia (>4000 g) rate (%) 5 min Apgar score 7
7 (7.1%) 6 (5.7%) 3 (2.9%)
13 (11.3%) 6 (5.2%) 3 (2.6%)
50 (10.7%) 30 (5.8%) 6 (1.2%)
<0.001a,b <0.001a,b <0.001a,b NSc, 0.002d NS NS NS NS NS <0.01a,e NSc, 0.007d, OR ¼ 2.01 (1.22–3.32) NS NS NS
a
ANOVA. Tukey’s multiple comparison test (P < 0:05), group 1 > group 2 ¼ group 3. c For the difference between groups 1 and 2. d For the difference between groups 1 and 3. e Tukey’s multiple comparison test (P < 0:05), group 1 < group 3 < group 2. b
Table 2 Indications for cesarean section in study and control groups
Failed inductionb Non-reassuring fetal heart rate Labor dystociac Otherd Missing data
Study group (N ¼ 19)a
Elective induction (N ¼ 17)a
Spontaneous onset of delivery (N ¼ 46)a
P value
2 2 5 8 2
5 7 8 1 2
– 16 (34.7%) 20 (43.5%) 8 (17.4%) 5 (10.8%)
NS NS NS
(10.5%) (10.5%) (26.3%) (42.1%) (10.5%)
(29.4%) (41.2%) (47%) (5.8%) (11.7%)
a
Some patients had more than one indication for cesarean section. Failed induction—four or more applications of PGE2 and a Bishop score of <7. c Labor dystocia—dysfunctional active phase and prolonged second stage not suitable for instrumental delivery. d Severe preeclampsia or eclampsia, prolapse of umbilical cord, vaginal bleeding, placental abruption, maternal request. b
P < 0:001) was used to control for maternal and gestational age, gravidity, parity, nulliparity rate, induction of labor, number of PGE2 tablets used, presence of oligohydramnios, birth weight, and diagnosis of diabetes as predictors of mode
of delivery. On stepwise analysis, only nulliparity (OR 4.56, 95% CI 1.11–18.67, P ¼ 0:035) was independently and significantly associated with an increased risk of CS. A similar analysis within the study group (R2 ¼ 0:257,
Table 3 Antenatal data within the study group of women who delivered vaginally or by cesarean section
Age (years) Gravidity Parity Nulliparity rate Incidence of oligohydramnios Number of PGE2 tablets used Percentage of patients requiring more than one application Gestational age at delivery (weeks) Birth weight (g) Summary data are presented as mean S:D: or incidence.
Cesarean section (N ¼ 19)
Vaginal delivery (N ¼ 85)
P value
31.2 6.5 2.8 2.0 1.2 1.6 44.4% 16.7% 2.0 0.9 61.1% 38.4 1.0 3259 513
31.6 4.7 3.1 1.9 1.6 1.5 32.1% 12.5% 1.9 1.0 51.9% 38.6 1.1 3267 487
NS NS NS NS NS NS 0.03 NS NS
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
P ¼ 0:002), showed that a diagnosis of type 1 diabetes (OR 2.4, 95% CI 1.04–5.51) was independently and significantly associated with an increased risk of CS.
4. Discussion It is generally recommended that in diabetic pregnancies, delivery should be delayed until fetal lung maturation is completed, provided that the diabetes is well controlled and findings on antepartum surveillance remain normal. In these cases, elective induction is usually planned at 38–40 weeks of gestation and CS is performed only if induction of labor fails or if fetal macrosomia is suspected. In our study, we induced diabetic pregnancies only after completing full 38 weeks of pregnancy in order to achieve fetal lung maturation. Although labor induction is recommended for various subgroups in diabetic pregnancies near or at term [4,5], information on outcome is limited. According to one comprehensive review of the literature [6], only one relevant randomized controlled trial has been conducted so far [1]. In a protocol comparison between two time periods, Lurie and colleagues [4] reported no significant difference in macrosomia or cesarean delivery rates by management policy, although the patients in the expectant management group who gave birth after 40 weeks gestation had a higher rate of shoulder dystocia (10%) than the patients in whom labor was induced at 38–39 weeks gestation (1.4%). Although this difference was significant, these data have not been confirmed by other studies. Hod et al. [5] showed that maintaining strict control of maternal diabetes combined with elective delivery at 38–39 weeks of gestation, depending on the estimated fetal weight, leads to a significant reduction in the rate of macrosomia, thereby potentially decreasing the incidence of both traumatic births and cesarean section (CS). It is noteworthy that none of these studies included women with GDM who were not treated with insulin. Furthermore, the methods of cervical ripening and induction of labor, and the indications for CS, were either not controlled or not specified. Further assessments of the mode of delivery in this patient population are therefore needed to determine the optimal obstetrical management. Objectively evaluating the risks and benefits of labor induction in diabetic women is potentially confounded by a number of variables. Among these is the status of the cervix at the time of initiation of induction, the conduct of labor and the reliance on estimates fetal weight to determine route of delivery. The fact that the present study was based on the experience of a single center, over a 3-year period, with the same team of specialists, using the same strict protocol minimizes the potential confounding factors characteristic of studies with different management protocols applied in different medical centers.
169
We studied 105 consecutive women with diabetic pregnancy that underwent induction of labor with vaginal tablets of PGE2; all had an unfavorable cervix prior to labor induction. Importantly, the GDM sub-population was heterogeneous, some being treated by diet alone and other by diet combined with insulin. Induction of labor with vaginal tablets of PGE2 led to successful vaginal delivery in approximately 82% of the patients, with apparently no serious maternal or fetal complications. The CS rate in the diabetic pregnancy group was significantly higher than in the normal, non-induced pregnancy group, but similar to that in the elective induction group. Earlier studies have shown that CS rates are higher in women with diabetic pregnancies compared with controls. However, the difference is not entirely attributable to fetal macrosomia [7,8]. It may be that caregivers are more prone to operate on patients with diabetic pregnancy because of concern about the likelihood of shoulder dystocia. Furthermore, it is generally acknowledged that induction of labor by itself for medical indications is associated with an increased risk of CS [9–13]. For elective induction, some studies found higher rates [14–16], restricted to nulliparous patients [10,15], and others reported lower or equivalent rates [9,10]. In the most recent study, Cammu et al. [17] showed that when compared with spontaneous deliveries, elective labor induction in nulliparous women is associated with significantly more CS and instrumental deliveries. Importantly, in our study, the nulliparity rate was significantly lower in the study than in the control groups; thus, if the nulliparity rate were higher, we would expect an even higher rate of CS in the study group. In the logistic regression model and forward likelihood analysis, only nulliparity and, within the study group, a diagnosis of type 1 diabetes, were independently and significantly associated with an increased risk of CS. Therefore, we cannot claim that the induction process by itself or a diagnosis of diabetic pregnancy per se causes a high rate of CS. The lack of between-group differences in rates of non-reassuring fetal heart rate pattern, labor dystocia leading to CS, and low 5 min Apgar score (<7) indicates that these variables too, cannot explain the higher rate of CS in the study group, which also had non-significantly lower incidence of failed induction than the elective induction group. Our analysis of the antenatal data within the study group by type of delivery yielded a greater need for more than one PGE2 application in the women who underwent CS than in those with vaginal delivery. Thus, it may be that the induction process combined with an unfavorable cervix increased the CS rate in our sample. The impact of epidural analgesia on the rate of CS remains controversial [18]. Data on epidural analgesia were not analyzed specifically in our study. However, on crude analysis of the study period, the rate of use of epidural analgesia in labor was higher in nulliparous (80%) than the multiparous women (50%).
170
Y. Yogev et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 114 (2004) 166–170
Importantly, our management protocol yielded a similar rate of macrosomia and birth weight in the diabetic and the non-diabetic groups, although gestational age at delivery was significantly lower in the study group. Thus, using active induction management of labor may prevent higher rates of macrosomia at the price of a higher rate of CS. Although there were no major maternal or fetal complications in the study and control groups, the retrospective study design and the relatively small sample preclude us from ruling out minor maternal or fetal complications or later neonatal complications. Further studies are needed in larger samples. An interventional prospective study is needed to evaluate if prolongation of pregnancy without active induction might lead to a higher rate of spontaneous deliveries without a significant increase in birth weight, macrosomia, traumatic deliveries and higher incidence of stillbirth. In summary, currently available evidence suggests that, while induction of labor for women who have diabetes may not carry maternal or fetal risk, the benefit of this procedure is unclear. References [1] Kjos SL, Henry OA, Montoro M, Buchanan TA, Mestman JH. Insulin-requiring diabetes in pregnancy: a randomized trial of active induction of labor and expectant management. Am J Obstet Gynecol 1993;169:611–5. [2] Kelly AJ, Kavanagh J, Thomas J. Vaginal prostaglandin (PGE2 and PGF2a) for induction of labor at term (Cochrane Review). In: The cochrane library, Issue 3; 2002. Oxford: Update Software. [3] Metzger BE, Coustan DR. Summary and recommendations of the Fourth International Workshop-Conference on Gestational Diabetes Mellitus. Diab Care 1998;21(suppl 2):B161–7. [4] Lurie S, Insler V, Hagay ZJ. Induction of labor at 38–39 weeks of gestation reduces the incidence of shoulder dystocia in gestational diabetic patients Class A2. Am J Perinatol 1996;13:293–6.
[5] Hod M, Bar J, Peled Y, Fried S, Katz I, Itzhak M, et al. Antepartum management protocol. Timing and mode of delivery in gestational diabetes. Diab Care 1998;21(Suppl 2):B113–7. [6] Boulvain M, Stan C, Irion O. Elective delivery in diabetic pregnant women (Cochrane Review). In: The cochrane library, Issue 3; 2000. Oxford: Update Software. [7] Casey BM, Lucas MJ, McIntire DD, Leveno KJ. Pregnancy outcomes in women with gestational diabetes compared with the general obstetric population. Obstet Gynecol 1997;90:869–73. [8] Naylor CD, Sermer M, Chen E, Sykora K. Cesarean delivery in relation to birth weight and gestational glucose tolerance: pathophysiology or practice style? Toronto Trihospital Gestational Diabetes Investigators. JAMA 1996;275:1165–70. [9] 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;180:628–33. [10] Jarvelin MR, Hartikainen-Sorri AL, Rantakallio P. Labour induction policy in hospitals of different levels of specialisation. Br J Obstet Gynaecol 1993;100:310–5. [11] Robson S, Pridmore B, Dobb J. Outcomes of induced labor. Aust N Z J Obstet Gynaecol 1997;31:167–9. [12] Rojansky N, Reubinoff B, Tanos V, Shushan A, Weinstein D. High risk pregnancy outcome following induction of labour. Eur J Obstet Gynecol Reprod Biol 1997;72:153–8. [13] 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;94:600–7. [14] Smith LP, Nagourney BA, McLean FH, Usher RH. Hazards and benefits of elective induction of labor. Am J Obstet Gynecol 1984;148:579–85. [15] Macer JA, Macer CL, Chan LS. Elective induction versus spontaneous labor: a retrospective study of complications and outcome. Am J Obstet Gynecol 1992;166:1690–7. [16] Prysak M, Castronova FC. Elective induction versus spontaneous labor: a case-control analysis of safety and efficacy. Obstet Gynecol 1998;92:47–52. [17] Cammu H, Martens G, Ruyssinck G, Amy JJ. Outcome after elective labor induction in nulliparous women: a matched cohort study. Am J Obstet Gynecol 2002;186:240–4. [18] American College of Obstetricians and Gynecologists. Obstetric, Analgesia and Anesthesia. Practice Bulletin number 36; 2002.