- Email: [email protected]
The effect of mode of delivery on the perinatal outcome in fetuses with abdominal wall defects
Shirodkar operation and outcome
Volume 163 Number 3
et al! cites fetal survival rates after Shirodkar cerclage of75% to 85%. In our series 58 of the 66 treated pregnancies (88%) were carried for at least 37 weeks, with 61 surviving infants for an uncorrected perinatal survival rate of 92%. This is the largest published series of patients treated for incompetent cervix with the Shirodkar operation by a single surgeon. It demonstrates that the operation is effective when used for patients with documented cervical incompetence. There was no morbidity from the procedure, and removal of the Mersilene band as soon as labor began permitted vaginal birth without significantly increased risk to the mother or infant. The Shirodkar operation should be part of resident training and should have a prominent place in the treatment of incompetent cervix. Perhaps a prospective randomized
trial that compares the Shirodkar and McDonald procedures should be undertaken to determine the relative merits of these techniques. We thank Dennis S. Chi and Sara Truesdell for their assistance in preparation of the manuscript. REFERENCES 1. Lash AF, Lash SR. Habitual abortion: the incompetent internal os of the cervix. AM J OBSTET GYNECOL 1950;59:6876. 2. Shirodkar VN. A new method of operative treatment for habitual abortions in the second trimester of pregnancy. Antiseptic 1955;52:299-300. 3. McDonald IA. Suture of the cervix for inevitable miscarriage. J Obstet Gynaecol Br Emp 1957;64:346-50. 4. Golan A, Barnan R, Wexler S, Langer R, Bukovsky I, David MP. Incompetence of the uterine cervix. Obstet Gynecol Surv 1989;44:96-107.
The effect of mode of delivery on the perinatal outcome in fetuses with abdominal wall defects Michael Moretti, MD, Aldo Khoury, MD, Jaime Rodriquez, MD, Thorn Lobe, MD, David Shaver, MD, and Baba Sibai, MD Memphis, Tennessee A descriptive study of 125 infants with abdominal wall defects was undertaken to determine the effect of mode of delivery on outcome. Fifty-six infants had gastroschisis and 69 had omphalocele. Overall, there were no differences between the omphalocele and the gastroschisis groups in either cesarean section rate (22% vs 26%) or prematurity rate (26% vs 30%). However, the omphalocele group had a significantly higher infant death rate (22% vs 7%, P < 0.001), a significantly higher incidence of associated major congenital anomalies (29% vs 5%, P < 0.001), and a higher inCidence of long-term infant morbidity (14.5% vs 8.9%). Within either group there was no significant difference between vaginal and cesarean delivery regarding either infant mortality, acute or long-term infant outcome, or frequency of associated major anomalies. We conclude that vaginal delivery of infants with abdominal wall defects does not adversely affect infant outcome. (AMJ OSSTETGvNECOL 1990;163:833-8.)
Key words: Congenital abdominal wall defects, vaginal delivery, outcome
The incidence of omphalocele is approximately one in 5,800 to one in 5,130 live births, while that of gastroschisis is about one in 10,000 to one in 15,000 live births.' Omphalocele represents the persistence of the body stalk in an area normally occupied by the differFrom the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Tennessee, Memphis. Presented at the Tenth Annual Meeting of the Society of Perinatal Obstetricians, Houston, Texas, January 23-27, 1990. Reprint requests: Michael Moretti MD, Crump Women's Hospital, Room E100, 853 Jefferson Ave., Memphis TN 38163. 616122583
entiated abdominal wall. A membranous sac usually covers the herniated viscera. Omphalocele is associated with major chromosomal abnormalities. This is reported to be as high as 40% of patients where amniocentesis is performed. In addition, there is an increased incidence of other major structural abnormalities (45%) Gastroschisis involves a defect in the abdominal wall usually to the right of the umbilicus. It has been suggested that this results from abnormal involution ofthe right umbilical vein that leads to a paraumbilical defect through which small bowel prolapses at approximately 37 days of embryonic life. In gastroschisis, viscera float
833
834
Moretti et al.
uncovered in the amniotic fluid and lie exposed at birth. The incidence of major chromosomal abnormalities or other structural anomalies is low in this group of fetuses «5%). Prematurity and low birth weight have been reported to be more common than in omphalocele. 2.' There has been an increase in the survival of infants with these defects over the past 10 years. Mortality is related to a number of factors: gestational age at the time of delivery, the condition of the infant before operation, the presence of other anomalies, the size of the lesion, and the type of repair.' Now, with the widespread use of obstetric ultrasonography, congenital defects of this sort are being detected in utero with increasing frequency.6 Since prenatal diagnosis offers the ability to alter the obstetric management, it is important to focus attention on other perinatal factors that may influence neonatal and long-term outcome. Debate has arisen as to the appropriate mode of delivery. Early case reports advocated the use of cesarean section to avoid fetal injury.7-9 Lenke et al. 10 recommended the use of cesarean section in fetuses with gastroschisis while Nakayama et al. 11 and Kirk and Wah' reported that vaginal delivery did not appear to adversely affect fetal outcome for these fetus. In addition, Carpenter et al. 12 reported that vaginal delivery is not contraindicated for the fetus with ventral wall defects. The purpose of this report is to review our experience of 125 cases of anterior abdominal wall defects and to determine the effect of mode of delivery on neonatal survival and long-term outcome. Material and methods
The study group consisted of infants with om ph alocele or gastroschisis who were referred to the neonatal surgical team at the University of Tennessee, Memphis, between August 1978 and July 1989. Data were collected by chart review performed by the authors. The source of the data included office records of the pediatric surgical team, inpatient and outpatient records of the hospital of origin, records from the neonatal intensive care unit, and, when applicable, autopsy reports. Details of long-term follow-up were obtained from the pediatric clinic that the patients attended and by direct contact with the parents of the infants. Records were reviewed for birth weight, gestational age, mode of delivery, presence of other major anomalies, size of the defect, and presence of extracorporeal organs, as well as the presence of other neonatal complications. Special attention was given to the ascertainment of trauma to the extracorporeal organs at delivery and, in the case of omphalocele, rupture of the sac. In those cases of omphalocele with a ruptured sac records were carefully reviewed to determine those that appeared to occur at delivery. This was determined principally by observations of the bowel serosa and the pres-
September 1990 Am J Obstet Gynecol
ence of thickened matted bowel. It was thought that those infants who had these associated bowel findings most likely sustained rupture of the omphalocele sac before delivery. The changes in the bowel serosa were attributed to exposure of the bowel to the alkaline amniotic fluid, similar to that of gastroschisis. Gestational age at birth was estimated from the mother's last menstrual period and confirmed where possible by a dating ultrasonogram and Dubowitz examination. Definition of the extent of the ventral wall defect and evidence of any trauma were based on records of the examination by the pediatric surgical team on admission to their unit. The criteria defined by the pediatric surgical team for the diagnosis of omphalocele included those patients with a midline abdominal wall defect with a covering membrane or a remnant of the membrane. Additional criteria included the presence of a dilated umbilical ring containing umbilical vessels passing through the membranous sac without an intervening bridge. The criteria for the diagnosis of gastroschisis were a paraumbilical full-thickness defect of the abdominal wall and a normal umbilical cord insertion independent of the defect with a skin bridge between the ventral defect and the umbilical ring. The analysis was focused on perinatal outcome, longterm outcome, incidence of surgical complications, and cause of death. The infants were divided into four groups according to mode of delivery for analysis: gastroschisis vaginal group versus gastroschisis cesarean section group and omphalocele vaginal group versus omphalocele cesarean section group. Analysis of data was performed with X2 or Fisher's exact test when appropriate and by Student's t test. A p value of <0.05 was considered significant. Results
A total of 125 infants was studied: gastroschisis (n = 56) and omphalocele n = 69). Thirty patients (23.5%) were delivered by cesarean section (21 % in omphalocele group and 26% in gastroschisis group). The indications for cesarean section are listed in Table I. One hundred four infants (83%) were delivered at this center. Prenatal diagnosis of the defects was made in 10 (8%) of the cases. All of the infants with ventral wall defects that were diagnosed prenatally were delivered at this center [gastroschisis (n = 4) omphalocele (n = 6)]. All of the infants with gastroschisis that was diagnosed prenatally survived. Five of six infants with prenatal diagnosis of omphalocele survived. The prematurity rates overall were 26% and 30%, respectively. Table II summarizes the neonatal and long-term outcome of infants with gastroschisis according to the mode of delivery. There were no significant differences between the vaginal and cesarean section groups re-
Volume 163 Number 3
garding any of the categories studied. The overall mortality rate in infants with gastroschisis who were delivered by cesarean section was 6.6% and it was 7.3% in the infants who were delivered vaginally. In the gastroschisis group overall, the delivery-operation interval was an average of 6 hours. There was little difference in this interval between the cesarean section group (6.0 hours) and the vaginal delivery group (6.2 hours). Gastroschisis-vaginal delivery group: Extracorpo real organs other than bowel were present in 16 (28%) of these infants. These included stomach, bladder, fallopian tubes, ovary, omentum, and testis. Extracorporeal liver was not found in any infant within this group. There were no cases of delivery trauma to the extracorporeal organs noted. Intestinal atresia was present in three infants. Overall, 21 infants in this group had primary closure of the defect and 20 required a two-stage repair. Additional bowel surgery was necessary in four infants: bowel resection for volvulus or obstruction (n=3), and ileostomy closure n= 1). It is interesting to note that meconium aspiration syndrome did not occur in any of these neonates. There were three (7.3%) infant deaths in this group: one caused by fungal sepsis at 17 days, one caused by sudden infant death syndrome at 6 weeks, and a late death that was secondary to bowel obstruction and sepsis at 3 years. Long-term follow-up was available for all 41 infants. Three infants suffered from significant morbidity: One infant had long-term nutritional problems and was on a modified diet as a result of short bowel syndrome, one infant had biliary cirrhosis, and the third had failure to thrive. Gastroschisis-cesarean delivery group. Associated major anomalies were present in two (13%) of these infants: One infant had occipital encephalocele and one infant had trisomy 13 and a large ventricular septal defect. This latter case represented the only infant death among cases of gastroschisis with delivery by cesarean section. Extracorporeal organs present in this group of infants included bladder, stomach, gallbladder, ovaries, and fallopian tubes. None had an extracorporeal liver. Delivery trauma to the extracorporeal organs was not noted in any infant. Associated intestinal atresia was noted in two infants in this group. Primary closure was performed in eight and seven infants required a two-stage repair. Additional bowel surgery (closure of jeju .!Ostomy) was necessary in only two infants. All 15 infants had long-term follow-up, and none had any long-term morbidity. Table III summarizes the clinical characteristics of infants with omphalocele according to the mode of delivery. There were no significant differences for any of the categories studied. The overall mortality rate in infants with omphalocele was 20% in those delivered by cesarean section and 22% in infants delivered vag-
Delivery mode in abdominal wall defects
835
Table I. Indications for cesarean section Omphalocele
(n = 15,22%)
Failure to progresscephalopelvic disproportion Elective repeat Fetal distress Breech Other
Gastroschisis
(n = 15,26%)
4
3
4 3
6
2 2*
2
2
2t
*Premature rupture of membranes and chorioamnionitis
(n = 1), maternal genital herpes (n = 1). tPreeciampsia (n = 2).
in ally. The causes of infant deaths in the omphalocele group include multiple congenital anomalies (n = 9), cardiac anomalies (n = 4), and complications of prematurity (n = 2). In the omphalocele group the delivery-operation interval was an average of 6.5 hours. There was little difference in the interval between the cesarean section group (6.6 hours) and the vaginal delivery group (5.8 hours). Omphalocele-vaginal delivery group. There were 16 (30%) infants with major associated anomalies and 22 (41 %) with extracorporeal organs other than bowel present. These included stomach, pancreas, spleen, ovary, and fallopian tubes. Eleven infants had extracorporeal liver (partial or complete). There were three (27%) infant deaths among the babies with herniated liver delivered vaginally. Severe cardiac anomalies were the cause of death in two of these infants (Epstein's anomaly, transposition of the great vessels). The third infant died of complications of prematurity. Delivery trauma to the liver was not noted in any of the II cases. Eight infants with omphalocele who were delivered vaginally had ruptured sacs. On the basis of examination of the bowel, it was determined that two of the eight must have sustained rupture of the sac before delivery. Among the eight infants with ruptured sacs, there were two infant deaths, both of which were attributed to multiple congenital anomalies. There were no cases of sepsis among these eight infants. Among the infants with omphalocele delivered by the vaginal route, there were three infants with associated intestinal atresia. Thirty-four infants had primary repair and 20 required a two-stage operation. Overall, among the cases with omphalocele and vaginal delivery there were 12 infant deaths, 10 of which were associated with major anomalies and two that were secondary to complications of prematurity. There were 36 surviving babies with isolated omphalocele that underwent vaginal delivery. Thirty-five survived without long-term morbidity, and one died with complications of bowel obstruction at 2 years. Long-term follow-up was available for
836
Moretti et al.
September 1990 Am J Obstet Gyneco1
Table II. Neonatal and long-term outcome in infants with gastroschisis by mode of delivery
Gestational age (wk) Birth weight (gm) Preterm birth (s;37 wk) (No.) 5 min Apgar score (average) Cord pH (average) Meconium fluid (No.) Extracorporeal organs (No.) Visceral trauma (No.) Major anomalies (No.) Infant death (No.) Follow-up (yr) (mean and range) Morbidity (No.)
Vaginal (n = 41)
Cesarean section (n = 15)
37.5 ± 2.6 2557 ± 564 16 (39%) 7 7.28 12 (29%) 16 (39%)
37 ± 3.1 2575 ± 502 4 (27%) 8 7.33 5 (30%) 8 (53%)
o
1 3 3 5
o
(2.4%) (7.3%) (0.2-11) (12%)
2 (13%) 1 (6.7%) 2 (0.5-10)
o (%)
Table III. Neonatal and long-term outcome in infants with omphalocele by mode of delivery
Gestational age (wk) Birth weight (gm) Preterm birth (s;37 wk) (No.) 5 min Apgar score (average) Cord pH (average) Meconium fluid (No.) Extracorporeal organs (No.) Visceral trauma (No.) Major anomalies (No.) Infant death (No.) Follow-up (yr) (mean and range) Morbidity (No.) Ruptured sac (No.)
Vaginal (n = 54)
Cesarean section (n = 15)
38.2 ± 2.2 3020 ± 919 14 (26%) 8 7.26 5 (9%) 22 (41%)
37.8 ± 2.6 3081 ± 724 4 (27%) 8 7.31 1 (6.5%) 6 (40%)
o
16 12 4 7 8*
(30%) (22%) (0.2-11) (13%) (15%)
o
4 3 3 3 1
(27%) (20%) (0.2-11) (20%) (7%)
*Two ruptures were determined to have occurred several days before delivery.
all the infants in the omphalocele-vaginal delivery group. Six infants had significant morbidity: Two had other congenital anomalies (severe scoliosis, ureteral pelvic obstruction), two suffered from complications of preterm birth (chronic respiratory problems), one infant had recurrent urinary tract infections, and the last infant had short bowel syndrome and cirrhosis. This infant had associated intestinal atresia. Omphalocele-cesarean section group. There were four (27%) infants with major congenital anomalies. Six infants had extracorporeal organs other than bowel. These included stomach, spleen, and bladder. There were five cases with extracorporeal liver. Of these five infants, four survived and were without long-term morbidity. The fifth infant had pentalogy of Cantrell. This infant· died after operation. One infant in the omphalocele-cesarean section group had a ruptured sac at delivery. This was reported specifically to have occurred at the time of delivery. The status of the bowel did not indicate prior rupture. This infant survived without long-term morbidity. One infant had an associated intestinal atresia. Primary closure was performed in eight patients in this group and a two-stage repair was necessary in seven. Overall, there were three infant deaths in this group (20%). Two were due to major cardiac anomalies and the third was attributed to
extreme prematurity and respiratory complications. Long-term follow-up was available for all 15 patients. Three infants (20%) had long-term complications, all of which were attributed to associated congenital anomalies: atrial septal defect and obstructed ureteropelvic junction, in perforate anus and absent sacrum, and meningomylocele. Comment
The overall survival rate of 80% in this study is consistent with that reported in other recent studies. 5 The higher mortality rate with omphalocele is also consistent with that of other reports. 3, 12 This increased mortality in the omphalocele group is attributed to the higher rate of associated major anomalies in the omphalocele group as opposed to those infants with gastroschisis (29% vs 5.0%). There was an increased trend toward low birth weight and pre term delivery in the infants with gastroschisis as opposed to those with omphalocele, but the difference did not reach significance. In no case was cesarean section performed solely because of the prenatal diagnosis of abdominal wall defect. This study did not focus on the prenatal diagnosis of these anomalies. Prenatal diagnosis of the defect was made in only 8% of the cases. Although there was no difference in survival between those infants in whom
Volume 163 Number 3
the defect was diagnosed prenatally and those with diagnosis at delivery, it has been shown in other reports that prenatal diagnosis improves management by allowing maternal transport for labor and delivery at a center specializing in high-risk pregnancy and pediatric intensive care. A key factor in postnatal survival is the clinical condition of the newborn on admission to the pediatric surgical team. Delays in neonatal transport after birth can increase the chances of profound dehydration and risk of sepsis, both causes of poor outcome. ll In addition, prenatal assessment can determine other associated lethal anomalies and prevent some cesarean deliveries. All the infants were cared for by the same neonatal surgical team. Thus treatment was consistent in all cases. The most commonly reported postoperative complications in infants with gastroschisis are sepsis, small bowel obstruction, and infarction. 3 It is hypothesized that the combination of uterine contractions and vaginal delivery may lead to vascular compromise and compression of the bowel, thus increasing the chances for these complications. This theory could explain the need for more secondary operations and the greater postoperative complication rate in infants born via the vaginal route. lO In this study three infants with gastroschisis delivered vaginally required secondary operation for bowel obstruction, but two of the three had associated bowel atresia and required more involved initial surgery. Thus these complications are not due to the delivery process itself but rather are caused by the prenatal state of the organs, such as the presence of secondary bowel atresia. In one other infant massive fungal sepsis resulted in neonatal death at 17 days of age. This infant had an associated major anomaly (bladder extrophy), and sepsis developed late in the postoperative period. Thus it is unlikely that such a complication was associated with the delivery process. Since one of the major issues in selecting the mode of delivery in these infants is the risk of delivery trauma to the extracorporeal organs, it is important to note that in the 41 infants with gastroschisis delivered vaginally no cases of trauma to the extracorporeal organs at delivery were reported. None of the infants in the gastroschisis group had extracorporealliver. Thus the effect of mode of delivery in this particular situation cannot be addressed in this study. Ten of the 12 infant deaths in the omphalocelevaginal delivery group were due to associated major congenital anomalies and two were due to severe hyaline membrane disease. It is of interest that there were eight cases of ruptured sac in the omphalocele-vaginal delivery group, none of which were complicated by sepsis. Although extracorporeal liver was a relatively common finding in the omphalocele group (22%), there was no improvement in survival rate when the infant was delivered by cesarean section rather than vaginally. This is in contrast to the findings of Sermer
Delivery mode in abdominal wall defects 837
et al.,' who reported a neonatal death rate of 88% in infants with omphalocele and herniated liver. However, in that report no mention was made as to the presence of other congenital anomalies or perinatal complications that may have contributed to the poor outcome in that group of infants. There were 36 infants with isolated omphalocele delivered by the vaginal route. There was only one infant death in this group, the cause of which could not have been delivery mode. This further emphasizes the safety of vaginal delivery in this group of patients. In addition, the data support the fact that the majority of the mortality and morbidity in patients with omphalocele can be attributed to other complicating factors, mainly associated congenital anomalies and chromosomal aberrations. Since the primary admission criterion for this study was referral to the pediatric surgical team, it may be postulated that a number of infants were not referred for care because their conditions were critical and that this could be due to visceral trauma sustained at delivery. We feel that this is unlikely for a number of reasons. All infants in this region requiring neonatal intensive care are transferred to this medical center. It is unlikely that visceral trauma would preclude transfer. Facilities for transfer are available, despite the critical nature of the condition of the infants. All infants that present to the neonatal intensive care unit with this sort of lesion would have a pediatric surgical consultation despite their critical condition. Review of the records of the pediatric surgical team would suggest that few cases have been overlooked. We did not focus our analysis on the incidence of stillbirths among fetuses with ventral wall defects nor did we examine the effect of mode of delivery on survival of the previable infant «24 weeks) with ventral wall defects. We feel there are too many confounding variables to adequately compare these groups. In summary, our findings suggest that vaginal delivery of infants with abdominal wall defects does not have an adverse effect on neonatal outcome. In addition, long-term infant outcome does not appear to be affected by mode of delivery but rather by other perinatal events. Moreover, the presence of extracorporealliver in fetuses with omphalocele does not necessarily preclude a vaginal delivery. Until a large, multicenter, prospective randomized study can be accomplished, we believe that cesarean section should be reserved only for obstetric indications in fetuses with abdominal wall defects. REFERENCES 1. Romero R, Pilu G, Jeanty P, et al. Prenatal diagnosis of congenital anomalies. Norwalk, Connecticut: Appleton and Lange, 1983. 2. deVries PA. The pathogenesis of gastroschisis and omphalocele.J Pediatr Surg 1980;3:245. 3. Sermer M, Benzie RJ, Pitson L, Carr M, Skidmore M. Prenatal diagnosis and management of congenital defects
Moretti et al.
4. 5.
6.
7.
of the anterior abdominal wall. AM J OBSTET GYNECOL 1987; 156:308-12. Lindham J. Omphalocele and gastroschisis in Sweden 1965-1976. Acta Paediatr Scand 1981;70:55. Kirk EP, and Wah RM. Obstetric management of the fetus with omphalocele or gastroschisis: a review and report of one hundred and twelve cases. AM J OBSTET GYNECOL 1983;146:512. Jassani MN, Gauderer MWL, Fanaroff AA, Fletcher B, Merkatz IR. A perinatal approach to the diagnosis and management of gastrointestinal malformations. AMJ OBSTET GYNECOL 1982;59:33. Cameron GM, McQuown DS, Modanlou HD, Zemlyn S, Pillsbury SG Jr. Intrauterine diagnosis of omphalocele by diagnostic ultrasonography. AMJ OBSTETGYNECOL 1978; 131:821.
September 1990 Am J Obstet Gynecol
8. Lomas F, Stafford-Bell M, Tymms A, et al. Prenatal ultrasound in the diagnosis and management of fetal exomphalocele: case reports. Br J Obstet Gynaecol 1979; 86:581. 9. Wilson BRJ, Turner D, Langendoerrfer S, et al. Prenatal diagnosis and subsequent team approach to the management of omphalocele. J Reprod Med 1980;24:134. 10. Lenke RR, Hatch JR, Edwin 1. Fetal gastroschisis: a preliminary report advocating the use of cesarean section. Obstet Gynecol 1986;67:395. 11. Nakayama DK, Harrison MR, Gross B, et al. Management of the fetus with abdominal wall defects. J Pediatr Surg 1984;19:408. 12. Carpenter MW, Curci MR, Dobbins AW, et al. Perinatal management of ventral wall defects. Obstet Gynecol 1984;64:646.
Perinatal survival with expectant management of midtrimester rupture of membranes Carol A. Major, MD, and John L. Kitzmiller, MD San Francisco, California A retrospective study of 70 patients with preterm premature rupture of membranes before 26 weeks of gestation was performed. The purpose of this study was to examine the perinatal outcome and the potential maternal and neonatal morbidity associated with this obstetric condition. The mean gestational age at diagnosis of rupture of membranes was 23.7 weeks. The latency period ranged from 24 hours to 60 days with the mean being 12 days. There was no correlation between gestational age at rupture of membranes and latency period. Seventy-one infants were delivered. The perinatal survival was 63%. Sixty-eight percent of the survivors had normal neurological and physical development at a 1-year follow-up. Amnionitis developed in 43% of patients. The incidence of respiratory distress syndrome in the neonates was 52%. Betamethasone did not appear to reduce this incidence. The perinatal survival in patients with midtrimester rupture of membranes appears to be improving with the advancements in neonatal care of the extremely premature infant. (AM J OesTET GYNECOL 1990;163:838-44.)
Key words: Premature rupture of membranes, expectant management, mid trimester, amnionitis, perinatal survival
Premature rupture of membranes (PROM) before the onset of labor occurs in approximately 10% of all pregnancies. 1 The incidence of preterm PROM is about 1%.".3 Despite the large amount of data generated concerning the topic of preterm PROM, it remains one of the most controversial topics in obstetrics today.4 The management of patients with very early preterm PROM From the Department of Obstetrics and Gynecology, University of California at San Francisco. Presented at the Tenth Annual Meeting of the Society of Perinatal Obstetricians, Houston, Texas, January 23-27, 1990. Reprint requests: Carol A. Major, MD, University of California at Irvine, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, 101 The City Drive, Building 25, Orange, CA 92668.
6/6/22350
838
in past years usually involved termination of the pregnancy in an attempt to prevent potential maternal morbidity and mortality. In these cases the thought was that the maternal risks greatly outweighed the chances for a normal neonatal outcome. With recent improvements in the neonatal care of the extremely premature infant, there has been an increased interest in the practice of expectant management of patients with preterm PROM.' There have only been four published reports examining the maternal and neonatal outcomes of patients with mid trimester PROM. 6 .g The perinatal survival ranged from 24% in the study by Taylor and Garite 6 in 1984 to 51 % in the study by Bengtson and VanMarter9 in 1988. From these studies and ours, it
Volume 163 Number 3
et al! cites fetal survival rates after Shirodkar cerclage of75% to 85%. In our series 58 of the 66 treated pregnancies (88%) were carried for at least 37 weeks, with 61 surviving infants for an uncorrected perinatal survival rate of 92%. This is the largest published series of patients treated for incompetent cervix with the Shirodkar operation by a single surgeon. It demonstrates that the operation is effective when used for patients with documented cervical incompetence. There was no morbidity from the procedure, and removal of the Mersilene band as soon as labor began permitted vaginal birth without significantly increased risk to the mother or infant. The Shirodkar operation should be part of resident training and should have a prominent place in the treatment of incompetent cervix. Perhaps a prospective randomized
trial that compares the Shirodkar and McDonald procedures should be undertaken to determine the relative merits of these techniques. We thank Dennis S. Chi and Sara Truesdell for their assistance in preparation of the manuscript. REFERENCES 1. Lash AF, Lash SR. Habitual abortion: the incompetent internal os of the cervix. AM J OBSTET GYNECOL 1950;59:6876. 2. Shirodkar VN. A new method of operative treatment for habitual abortions in the second trimester of pregnancy. Antiseptic 1955;52:299-300. 3. McDonald IA. Suture of the cervix for inevitable miscarriage. J Obstet Gynaecol Br Emp 1957;64:346-50. 4. Golan A, Barnan R, Wexler S, Langer R, Bukovsky I, David MP. Incompetence of the uterine cervix. Obstet Gynecol Surv 1989;44:96-107.
The effect of mode of delivery on the perinatal outcome in fetuses with abdominal wall defects Michael Moretti, MD, Aldo Khoury, MD, Jaime Rodriquez, MD, Thorn Lobe, MD, David Shaver, MD, and Baba Sibai, MD Memphis, Tennessee A descriptive study of 125 infants with abdominal wall defects was undertaken to determine the effect of mode of delivery on outcome. Fifty-six infants had gastroschisis and 69 had omphalocele. Overall, there were no differences between the omphalocele and the gastroschisis groups in either cesarean section rate (22% vs 26%) or prematurity rate (26% vs 30%). However, the omphalocele group had a significantly higher infant death rate (22% vs 7%, P < 0.001), a significantly higher incidence of associated major congenital anomalies (29% vs 5%, P < 0.001), and a higher inCidence of long-term infant morbidity (14.5% vs 8.9%). Within either group there was no significant difference between vaginal and cesarean delivery regarding either infant mortality, acute or long-term infant outcome, or frequency of associated major anomalies. We conclude that vaginal delivery of infants with abdominal wall defects does not adversely affect infant outcome. (AMJ OSSTETGvNECOL 1990;163:833-8.)
Key words: Congenital abdominal wall defects, vaginal delivery, outcome
The incidence of omphalocele is approximately one in 5,800 to one in 5,130 live births, while that of gastroschisis is about one in 10,000 to one in 15,000 live births.' Omphalocele represents the persistence of the body stalk in an area normally occupied by the differFrom the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Tennessee, Memphis. Presented at the Tenth Annual Meeting of the Society of Perinatal Obstetricians, Houston, Texas, January 23-27, 1990. Reprint requests: Michael Moretti MD, Crump Women's Hospital, Room E100, 853 Jefferson Ave., Memphis TN 38163. 616122583
entiated abdominal wall. A membranous sac usually covers the herniated viscera. Omphalocele is associated with major chromosomal abnormalities. This is reported to be as high as 40% of patients where amniocentesis is performed. In addition, there is an increased incidence of other major structural abnormalities (45%) Gastroschisis involves a defect in the abdominal wall usually to the right of the umbilicus. It has been suggested that this results from abnormal involution ofthe right umbilical vein that leads to a paraumbilical defect through which small bowel prolapses at approximately 37 days of embryonic life. In gastroschisis, viscera float
833
834
Moretti et al.
uncovered in the amniotic fluid and lie exposed at birth. The incidence of major chromosomal abnormalities or other structural anomalies is low in this group of fetuses «5%). Prematurity and low birth weight have been reported to be more common than in omphalocele. 2.' There has been an increase in the survival of infants with these defects over the past 10 years. Mortality is related to a number of factors: gestational age at the time of delivery, the condition of the infant before operation, the presence of other anomalies, the size of the lesion, and the type of repair.' Now, with the widespread use of obstetric ultrasonography, congenital defects of this sort are being detected in utero with increasing frequency.6 Since prenatal diagnosis offers the ability to alter the obstetric management, it is important to focus attention on other perinatal factors that may influence neonatal and long-term outcome. Debate has arisen as to the appropriate mode of delivery. Early case reports advocated the use of cesarean section to avoid fetal injury.7-9 Lenke et al. 10 recommended the use of cesarean section in fetuses with gastroschisis while Nakayama et al. 11 and Kirk and Wah' reported that vaginal delivery did not appear to adversely affect fetal outcome for these fetus. In addition, Carpenter et al. 12 reported that vaginal delivery is not contraindicated for the fetus with ventral wall defects. The purpose of this report is to review our experience of 125 cases of anterior abdominal wall defects and to determine the effect of mode of delivery on neonatal survival and long-term outcome. Material and methods
The study group consisted of infants with om ph alocele or gastroschisis who were referred to the neonatal surgical team at the University of Tennessee, Memphis, between August 1978 and July 1989. Data were collected by chart review performed by the authors. The source of the data included office records of the pediatric surgical team, inpatient and outpatient records of the hospital of origin, records from the neonatal intensive care unit, and, when applicable, autopsy reports. Details of long-term follow-up were obtained from the pediatric clinic that the patients attended and by direct contact with the parents of the infants. Records were reviewed for birth weight, gestational age, mode of delivery, presence of other major anomalies, size of the defect, and presence of extracorporeal organs, as well as the presence of other neonatal complications. Special attention was given to the ascertainment of trauma to the extracorporeal organs at delivery and, in the case of omphalocele, rupture of the sac. In those cases of omphalocele with a ruptured sac records were carefully reviewed to determine those that appeared to occur at delivery. This was determined principally by observations of the bowel serosa and the pres-
September 1990 Am J Obstet Gynecol
ence of thickened matted bowel. It was thought that those infants who had these associated bowel findings most likely sustained rupture of the omphalocele sac before delivery. The changes in the bowel serosa were attributed to exposure of the bowel to the alkaline amniotic fluid, similar to that of gastroschisis. Gestational age at birth was estimated from the mother's last menstrual period and confirmed where possible by a dating ultrasonogram and Dubowitz examination. Definition of the extent of the ventral wall defect and evidence of any trauma were based on records of the examination by the pediatric surgical team on admission to their unit. The criteria defined by the pediatric surgical team for the diagnosis of omphalocele included those patients with a midline abdominal wall defect with a covering membrane or a remnant of the membrane. Additional criteria included the presence of a dilated umbilical ring containing umbilical vessels passing through the membranous sac without an intervening bridge. The criteria for the diagnosis of gastroschisis were a paraumbilical full-thickness defect of the abdominal wall and a normal umbilical cord insertion independent of the defect with a skin bridge between the ventral defect and the umbilical ring. The analysis was focused on perinatal outcome, longterm outcome, incidence of surgical complications, and cause of death. The infants were divided into four groups according to mode of delivery for analysis: gastroschisis vaginal group versus gastroschisis cesarean section group and omphalocele vaginal group versus omphalocele cesarean section group. Analysis of data was performed with X2 or Fisher's exact test when appropriate and by Student's t test. A p value of <0.05 was considered significant. Results
A total of 125 infants was studied: gastroschisis (n = 56) and omphalocele n = 69). Thirty patients (23.5%) were delivered by cesarean section (21 % in omphalocele group and 26% in gastroschisis group). The indications for cesarean section are listed in Table I. One hundred four infants (83%) were delivered at this center. Prenatal diagnosis of the defects was made in 10 (8%) of the cases. All of the infants with ventral wall defects that were diagnosed prenatally were delivered at this center [gastroschisis (n = 4) omphalocele (n = 6)]. All of the infants with gastroschisis that was diagnosed prenatally survived. Five of six infants with prenatal diagnosis of omphalocele survived. The prematurity rates overall were 26% and 30%, respectively. Table II summarizes the neonatal and long-term outcome of infants with gastroschisis according to the mode of delivery. There were no significant differences between the vaginal and cesarean section groups re-
Volume 163 Number 3
garding any of the categories studied. The overall mortality rate in infants with gastroschisis who were delivered by cesarean section was 6.6% and it was 7.3% in the infants who were delivered vaginally. In the gastroschisis group overall, the delivery-operation interval was an average of 6 hours. There was little difference in this interval between the cesarean section group (6.0 hours) and the vaginal delivery group (6.2 hours). Gastroschisis-vaginal delivery group: Extracorpo real organs other than bowel were present in 16 (28%) of these infants. These included stomach, bladder, fallopian tubes, ovary, omentum, and testis. Extracorporeal liver was not found in any infant within this group. There were no cases of delivery trauma to the extracorporeal organs noted. Intestinal atresia was present in three infants. Overall, 21 infants in this group had primary closure of the defect and 20 required a two-stage repair. Additional bowel surgery was necessary in four infants: bowel resection for volvulus or obstruction (n=3), and ileostomy closure n= 1). It is interesting to note that meconium aspiration syndrome did not occur in any of these neonates. There were three (7.3%) infant deaths in this group: one caused by fungal sepsis at 17 days, one caused by sudden infant death syndrome at 6 weeks, and a late death that was secondary to bowel obstruction and sepsis at 3 years. Long-term follow-up was available for all 41 infants. Three infants suffered from significant morbidity: One infant had long-term nutritional problems and was on a modified diet as a result of short bowel syndrome, one infant had biliary cirrhosis, and the third had failure to thrive. Gastroschisis-cesarean delivery group. Associated major anomalies were present in two (13%) of these infants: One infant had occipital encephalocele and one infant had trisomy 13 and a large ventricular septal defect. This latter case represented the only infant death among cases of gastroschisis with delivery by cesarean section. Extracorporeal organs present in this group of infants included bladder, stomach, gallbladder, ovaries, and fallopian tubes. None had an extracorporeal liver. Delivery trauma to the extracorporeal organs was not noted in any infant. Associated intestinal atresia was noted in two infants in this group. Primary closure was performed in eight and seven infants required a two-stage repair. Additional bowel surgery (closure of jeju .!Ostomy) was necessary in only two infants. All 15 infants had long-term follow-up, and none had any long-term morbidity. Table III summarizes the clinical characteristics of infants with omphalocele according to the mode of delivery. There were no significant differences for any of the categories studied. The overall mortality rate in infants with omphalocele was 20% in those delivered by cesarean section and 22% in infants delivered vag-
Delivery mode in abdominal wall defects
835
Table I. Indications for cesarean section Omphalocele
(n = 15,22%)
Failure to progresscephalopelvic disproportion Elective repeat Fetal distress Breech Other
Gastroschisis
(n = 15,26%)
4
3
4 3
6
2 2*
2
2
2t
*Premature rupture of membranes and chorioamnionitis
(n = 1), maternal genital herpes (n = 1). tPreeciampsia (n = 2).
in ally. The causes of infant deaths in the omphalocele group include multiple congenital anomalies (n = 9), cardiac anomalies (n = 4), and complications of prematurity (n = 2). In the omphalocele group the delivery-operation interval was an average of 6.5 hours. There was little difference in the interval between the cesarean section group (6.6 hours) and the vaginal delivery group (5.8 hours). Omphalocele-vaginal delivery group. There were 16 (30%) infants with major associated anomalies and 22 (41 %) with extracorporeal organs other than bowel present. These included stomach, pancreas, spleen, ovary, and fallopian tubes. Eleven infants had extracorporeal liver (partial or complete). There were three (27%) infant deaths among the babies with herniated liver delivered vaginally. Severe cardiac anomalies were the cause of death in two of these infants (Epstein's anomaly, transposition of the great vessels). The third infant died of complications of prematurity. Delivery trauma to the liver was not noted in any of the II cases. Eight infants with omphalocele who were delivered vaginally had ruptured sacs. On the basis of examination of the bowel, it was determined that two of the eight must have sustained rupture of the sac before delivery. Among the eight infants with ruptured sacs, there were two infant deaths, both of which were attributed to multiple congenital anomalies. There were no cases of sepsis among these eight infants. Among the infants with omphalocele delivered by the vaginal route, there were three infants with associated intestinal atresia. Thirty-four infants had primary repair and 20 required a two-stage operation. Overall, among the cases with omphalocele and vaginal delivery there were 12 infant deaths, 10 of which were associated with major anomalies and two that were secondary to complications of prematurity. There were 36 surviving babies with isolated omphalocele that underwent vaginal delivery. Thirty-five survived without long-term morbidity, and one died with complications of bowel obstruction at 2 years. Long-term follow-up was available for
836
Moretti et al.
September 1990 Am J Obstet Gyneco1
Table II. Neonatal and long-term outcome in infants with gastroschisis by mode of delivery
Gestational age (wk) Birth weight (gm) Preterm birth (s;37 wk) (No.) 5 min Apgar score (average) Cord pH (average) Meconium fluid (No.) Extracorporeal organs (No.) Visceral trauma (No.) Major anomalies (No.) Infant death (No.) Follow-up (yr) (mean and range) Morbidity (No.)
Vaginal (n = 41)
Cesarean section (n = 15)
37.5 ± 2.6 2557 ± 564 16 (39%) 7 7.28 12 (29%) 16 (39%)
37 ± 3.1 2575 ± 502 4 (27%) 8 7.33 5 (30%) 8 (53%)
o
1 3 3 5
o
(2.4%) (7.3%) (0.2-11) (12%)
2 (13%) 1 (6.7%) 2 (0.5-10)
o (%)
Table III. Neonatal and long-term outcome in infants with omphalocele by mode of delivery
Gestational age (wk) Birth weight (gm) Preterm birth (s;37 wk) (No.) 5 min Apgar score (average) Cord pH (average) Meconium fluid (No.) Extracorporeal organs (No.) Visceral trauma (No.) Major anomalies (No.) Infant death (No.) Follow-up (yr) (mean and range) Morbidity (No.) Ruptured sac (No.)
Vaginal (n = 54)
Cesarean section (n = 15)
38.2 ± 2.2 3020 ± 919 14 (26%) 8 7.26 5 (9%) 22 (41%)
37.8 ± 2.6 3081 ± 724 4 (27%) 8 7.31 1 (6.5%) 6 (40%)
o
16 12 4 7 8*
(30%) (22%) (0.2-11) (13%) (15%)
o
4 3 3 3 1
(27%) (20%) (0.2-11) (20%) (7%)
*Two ruptures were determined to have occurred several days before delivery.
all the infants in the omphalocele-vaginal delivery group. Six infants had significant morbidity: Two had other congenital anomalies (severe scoliosis, ureteral pelvic obstruction), two suffered from complications of preterm birth (chronic respiratory problems), one infant had recurrent urinary tract infections, and the last infant had short bowel syndrome and cirrhosis. This infant had associated intestinal atresia. Omphalocele-cesarean section group. There were four (27%) infants with major congenital anomalies. Six infants had extracorporeal organs other than bowel. These included stomach, spleen, and bladder. There were five cases with extracorporeal liver. Of these five infants, four survived and were without long-term morbidity. The fifth infant had pentalogy of Cantrell. This infant· died after operation. One infant in the omphalocele-cesarean section group had a ruptured sac at delivery. This was reported specifically to have occurred at the time of delivery. The status of the bowel did not indicate prior rupture. This infant survived without long-term morbidity. One infant had an associated intestinal atresia. Primary closure was performed in eight patients in this group and a two-stage repair was necessary in seven. Overall, there were three infant deaths in this group (20%). Two were due to major cardiac anomalies and the third was attributed to
extreme prematurity and respiratory complications. Long-term follow-up was available for all 15 patients. Three infants (20%) had long-term complications, all of which were attributed to associated congenital anomalies: atrial septal defect and obstructed ureteropelvic junction, in perforate anus and absent sacrum, and meningomylocele. Comment
The overall survival rate of 80% in this study is consistent with that reported in other recent studies. 5 The higher mortality rate with omphalocele is also consistent with that of other reports. 3, 12 This increased mortality in the omphalocele group is attributed to the higher rate of associated major anomalies in the omphalocele group as opposed to those infants with gastroschisis (29% vs 5.0%). There was an increased trend toward low birth weight and pre term delivery in the infants with gastroschisis as opposed to those with omphalocele, but the difference did not reach significance. In no case was cesarean section performed solely because of the prenatal diagnosis of abdominal wall defect. This study did not focus on the prenatal diagnosis of these anomalies. Prenatal diagnosis of the defect was made in only 8% of the cases. Although there was no difference in survival between those infants in whom
Volume 163 Number 3
the defect was diagnosed prenatally and those with diagnosis at delivery, it has been shown in other reports that prenatal diagnosis improves management by allowing maternal transport for labor and delivery at a center specializing in high-risk pregnancy and pediatric intensive care. A key factor in postnatal survival is the clinical condition of the newborn on admission to the pediatric surgical team. Delays in neonatal transport after birth can increase the chances of profound dehydration and risk of sepsis, both causes of poor outcome. ll In addition, prenatal assessment can determine other associated lethal anomalies and prevent some cesarean deliveries. All the infants were cared for by the same neonatal surgical team. Thus treatment was consistent in all cases. The most commonly reported postoperative complications in infants with gastroschisis are sepsis, small bowel obstruction, and infarction. 3 It is hypothesized that the combination of uterine contractions and vaginal delivery may lead to vascular compromise and compression of the bowel, thus increasing the chances for these complications. This theory could explain the need for more secondary operations and the greater postoperative complication rate in infants born via the vaginal route. lO In this study three infants with gastroschisis delivered vaginally required secondary operation for bowel obstruction, but two of the three had associated bowel atresia and required more involved initial surgery. Thus these complications are not due to the delivery process itself but rather are caused by the prenatal state of the organs, such as the presence of secondary bowel atresia. In one other infant massive fungal sepsis resulted in neonatal death at 17 days of age. This infant had an associated major anomaly (bladder extrophy), and sepsis developed late in the postoperative period. Thus it is unlikely that such a complication was associated with the delivery process. Since one of the major issues in selecting the mode of delivery in these infants is the risk of delivery trauma to the extracorporeal organs, it is important to note that in the 41 infants with gastroschisis delivered vaginally no cases of trauma to the extracorporeal organs at delivery were reported. None of the infants in the gastroschisis group had extracorporealliver. Thus the effect of mode of delivery in this particular situation cannot be addressed in this study. Ten of the 12 infant deaths in the omphalocelevaginal delivery group were due to associated major congenital anomalies and two were due to severe hyaline membrane disease. It is of interest that there were eight cases of ruptured sac in the omphalocele-vaginal delivery group, none of which were complicated by sepsis. Although extracorporeal liver was a relatively common finding in the omphalocele group (22%), there was no improvement in survival rate when the infant was delivered by cesarean section rather than vaginally. This is in contrast to the findings of Sermer
Delivery mode in abdominal wall defects 837
et al.,' who reported a neonatal death rate of 88% in infants with omphalocele and herniated liver. However, in that report no mention was made as to the presence of other congenital anomalies or perinatal complications that may have contributed to the poor outcome in that group of infants. There were 36 infants with isolated omphalocele delivered by the vaginal route. There was only one infant death in this group, the cause of which could not have been delivery mode. This further emphasizes the safety of vaginal delivery in this group of patients. In addition, the data support the fact that the majority of the mortality and morbidity in patients with omphalocele can be attributed to other complicating factors, mainly associated congenital anomalies and chromosomal aberrations. Since the primary admission criterion for this study was referral to the pediatric surgical team, it may be postulated that a number of infants were not referred for care because their conditions were critical and that this could be due to visceral trauma sustained at delivery. We feel that this is unlikely for a number of reasons. All infants in this region requiring neonatal intensive care are transferred to this medical center. It is unlikely that visceral trauma would preclude transfer. Facilities for transfer are available, despite the critical nature of the condition of the infants. All infants that present to the neonatal intensive care unit with this sort of lesion would have a pediatric surgical consultation despite their critical condition. Review of the records of the pediatric surgical team would suggest that few cases have been overlooked. We did not focus our analysis on the incidence of stillbirths among fetuses with ventral wall defects nor did we examine the effect of mode of delivery on survival of the previable infant «24 weeks) with ventral wall defects. We feel there are too many confounding variables to adequately compare these groups. In summary, our findings suggest that vaginal delivery of infants with abdominal wall defects does not have an adverse effect on neonatal outcome. In addition, long-term infant outcome does not appear to be affected by mode of delivery but rather by other perinatal events. Moreover, the presence of extracorporealliver in fetuses with omphalocele does not necessarily preclude a vaginal delivery. Until a large, multicenter, prospective randomized study can be accomplished, we believe that cesarean section should be reserved only for obstetric indications in fetuses with abdominal wall defects. REFERENCES 1. Romero R, Pilu G, Jeanty P, et al. Prenatal diagnosis of congenital anomalies. Norwalk, Connecticut: Appleton and Lange, 1983. 2. deVries PA. The pathogenesis of gastroschisis and omphalocele.J Pediatr Surg 1980;3:245. 3. Sermer M, Benzie RJ, Pitson L, Carr M, Skidmore M. Prenatal diagnosis and management of congenital defects
Moretti et al.
4. 5.
6.
7.
of the anterior abdominal wall. AM J OBSTET GYNECOL 1987; 156:308-12. Lindham J. Omphalocele and gastroschisis in Sweden 1965-1976. Acta Paediatr Scand 1981;70:55. Kirk EP, and Wah RM. Obstetric management of the fetus with omphalocele or gastroschisis: a review and report of one hundred and twelve cases. AM J OBSTET GYNECOL 1983;146:512. Jassani MN, Gauderer MWL, Fanaroff AA, Fletcher B, Merkatz IR. A perinatal approach to the diagnosis and management of gastrointestinal malformations. AMJ OBSTET GYNECOL 1982;59:33. Cameron GM, McQuown DS, Modanlou HD, Zemlyn S, Pillsbury SG Jr. Intrauterine diagnosis of omphalocele by diagnostic ultrasonography. AMJ OBSTETGYNECOL 1978; 131:821.
September 1990 Am J Obstet Gynecol
8. Lomas F, Stafford-Bell M, Tymms A, et al. Prenatal ultrasound in the diagnosis and management of fetal exomphalocele: case reports. Br J Obstet Gynaecol 1979; 86:581. 9. Wilson BRJ, Turner D, Langendoerrfer S, et al. Prenatal diagnosis and subsequent team approach to the management of omphalocele. J Reprod Med 1980;24:134. 10. Lenke RR, Hatch JR, Edwin 1. Fetal gastroschisis: a preliminary report advocating the use of cesarean section. Obstet Gynecol 1986;67:395. 11. Nakayama DK, Harrison MR, Gross B, et al. Management of the fetus with abdominal wall defects. J Pediatr Surg 1984;19:408. 12. Carpenter MW, Curci MR, Dobbins AW, et al. Perinatal management of ventral wall defects. Obstet Gynecol 1984;64:646.
Perinatal survival with expectant management of midtrimester rupture of membranes Carol A. Major, MD, and John L. Kitzmiller, MD San Francisco, California A retrospective study of 70 patients with preterm premature rupture of membranes before 26 weeks of gestation was performed. The purpose of this study was to examine the perinatal outcome and the potential maternal and neonatal morbidity associated with this obstetric condition. The mean gestational age at diagnosis of rupture of membranes was 23.7 weeks. The latency period ranged from 24 hours to 60 days with the mean being 12 days. There was no correlation between gestational age at rupture of membranes and latency period. Seventy-one infants were delivered. The perinatal survival was 63%. Sixty-eight percent of the survivors had normal neurological and physical development at a 1-year follow-up. Amnionitis developed in 43% of patients. The incidence of respiratory distress syndrome in the neonates was 52%. Betamethasone did not appear to reduce this incidence. The perinatal survival in patients with midtrimester rupture of membranes appears to be improving with the advancements in neonatal care of the extremely premature infant. (AM J OesTET GYNECOL 1990;163:838-44.)
Key words: Premature rupture of membranes, expectant management, mid trimester, amnionitis, perinatal survival
Premature rupture of membranes (PROM) before the onset of labor occurs in approximately 10% of all pregnancies. 1 The incidence of preterm PROM is about 1%.".3 Despite the large amount of data generated concerning the topic of preterm PROM, it remains one of the most controversial topics in obstetrics today.4 The management of patients with very early preterm PROM From the Department of Obstetrics and Gynecology, University of California at San Francisco. Presented at the Tenth Annual Meeting of the Society of Perinatal Obstetricians, Houston, Texas, January 23-27, 1990. Reprint requests: Carol A. Major, MD, University of California at Irvine, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, 101 The City Drive, Building 25, Orange, CA 92668.
6/6/22350
838
in past years usually involved termination of the pregnancy in an attempt to prevent potential maternal morbidity and mortality. In these cases the thought was that the maternal risks greatly outweighed the chances for a normal neonatal outcome. With recent improvements in the neonatal care of the extremely premature infant, there has been an increased interest in the practice of expectant management of patients with preterm PROM.' There have only been four published reports examining the maternal and neonatal outcomes of patients with mid trimester PROM. 6 .g The perinatal survival ranged from 24% in the study by Taylor and Garite 6 in 1984 to 51 % in the study by Bengtson and VanMarter9 in 1988. From these studies and ours, it