- Email: [email protected]
Short-term maternal outcomes that are associated with the exit procedure, as compared with cesarean delivery
Short-term maternal outcomes that are associated with the EXIT procedure, as compared with cesarean delivery Melinda M. Scully Noah, MD,a Mary E. Norton, MD,a Per Sandberg, MD,a Tania Esakoff,b Jody Farrell, RN,c and Craig T. Albanese, MDc San Francisco, Calif OBJECTIVE: The ex utero intrapartum treatment procedure was developed to treat iatrogenic tracheal obstruction that resulted from in utero fetal therapy for congenital diaphragmatic hernia. The ex utero intrapartum treatment procedure allows for controlled intubation while the neonate is maintained on placental circulation, prolonging operative time. This study evaluates whether there is increased maternal morbidity associated with this procedure compared with routine cesarean deliveries. STUDY DESIGN: Maternal outcomes on 34 patients who underwent the ex utero intrapartum treatment procedure were evaluated. Infection rate, estimated blood loss, need for transfusion, and length of postoperative hospital stay were compared to maternal outcomes from 52 nonlaboring patients who underwent nonemergent primary cesarean delivery of singleton fetuses during the same time interval. RESULTS: The rate of chorioamnionitis was similar between groups (26% vs 21%; P = .57). Postpartum wound complications were more common in patients who underwent ex utero intrapartum treatment (15% vs 2%; P = .03), although the rate of endometritis was similar (15% vs 10%; P = .50). Estimated blood loss was higher in the patients who underwent ex utero intrapartum treatment (1104 mL vs 883 mL; P < .001), but there was no difference between groups in hematocrit level change or postpartum hospital stay. CONCLUSION: Women who undergo the ex utero intrapartum treatment procedure experience more wound complications but no difference in postoperative hematocrit level change or postpartum length of stay. (Am J Obstet Gynecol 2002;186:773-7.)
Key words: EXIT procedure, morbidity, surgery, cesarean delivery
Although increasing data exists regarding the fetal outcomes that are associated with in utero surgical procedures, little information is available about the maternal outcomes that are associated with these procedures. The ex utero intrapartum treatment (EXIT) procedure was developed as a means to maintain fetoplacental circulation until the airway is secured in fetuses with known tracheal obstruction.1 Most of these cases are due to balloon or clip placement across the fetal trachea for in utero treatment of congenital diaphragmatic hernia, although the EXIT procedure has been used in noniatrogenic cases of airway obstruction, such as cystic hygroma, large goiter, or laryngeal atresia.2 As described by Mychaliska et
From the Department of Obstetrics, Gynecology, and Reproductive Sciences, University of California, San Francisco,a the University of California, San Francisco Medical School,b and the Division of Pediatric Surgery, Fetal Treatment Center.c Received for publication June 7, 2001; revised November 26, 2001; accepted December 12, 2001. Reprint requests: Melinda M. Scully Noah, MD, Department of Obstetrics, Gynecology, and Reproductive Sciences, University of California, San Francisco, Box 0550 HSW 1452, 505 Parnassus Ave, San Francisco, CA 94143. E-mail: [email protected] Copyright 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/122249 doi:10.1067/mob.2002.112249
al,3 this surgical procedure requires the pregnant woman to be placed under general endotracheal anesthesia by rapid sequence induction and subsequently maintained on high-dose isoflurane. This regimen is continued while the hysterotomy is performed and the neonate is delivered to the level of the trachea. At this point, airway control is obtained or the airway obstruction is overcome; pressure-controlled bag ventilation is instituted, and surfactant is administered (if indicated) while the neonate is still attached to the placenta. The airway is then secured; the baby is delivered, and the umbilical cord is clamped. After delivery of the neonate, the anesthetic regimen is returned to routine levels of general anesthesia, and oxytocic is added to the intravenous fluids. The surgical portion of the EXIT procedure generally follows a standard obstetric cesarean delivery but is markedly prolonged by the neonatal intubation. During this part of the procedure, uterine tone is purposely absent, and the hysterotomy site is open longer, thus potentially increasing the chance of maternal infection and bleeding. Blood loss from the edges of the hysterotomy is minimized with the use of a specially developed uterine stapler and uterine clamps. In addition, if the initial in utero surgical procedure was an open procedure in which a hysterotomy was performed, then that same sur773
774 Scully Noah et al
April 2002 Am J Obstet Gynecol
Table I. Baseline characteristics Characteristic Maternal age (y) Gravidity Parity Uterine surgery† (n) Gestational age at admission (wk) Gestational age at delivery (wk)
EXIT procedure (n = 34) 28.3 2.6 0.65 3 (9%) 31.1 31.4
Cesarean delivery (n = 52) 28.0 3.0 1.0 12 (23%) 30.5 31.6
P value* .82 .38 .17 .15 .47 .75
*Student t test; χ2 tests, with Fisher exact test. †Uterine surgical procedures include previous cesarean delivery and uterine operation.
gical site is used for delivery, which further decreases blood loss. Oxytocin and antibiotics, usually a firstgeneration cephalosporin, are given intravenously to the mother after the cord is clamped. Because a pediatric surgeon is required at delivery to treat the infant, an emergent or “crash” cesarean delivery is not possible. Therefore, patients are delivered at the earliest evidence of labor or fetal compromise. This study sought to evaluate the short-term morbidities that are associated with the EXIT procedure and to compare them with those morbidities that are incurred by women who undergo prelabor cesarean deliveries at a similar gestational age for standard obstetric, nonemergency conditions at the University of California, San Francisco. Material and methods A retrospective cohort study was performed that evaluated the outcomes of patients who underwent the EXIT procedure from March 1994 through December 1999. Patients who were enrolled in an ongoing clinical trial that evaluated the efficacy of tracheal occlusion for congenital diaphragmatic hernia were excluded. Eight patients who were included in this study were described in a previously published review of the EXIT procedure.3 The comparison group included women who underwent a cesarean delivery before labor for obstetric indications during the same time period and were matched for gestational age range. Exclusion criteria for the comparison cohort were emergent cesarean delivery, placenta previa, any maternal medical disorder (such as preeclampsia, hypertension, renal disease, or diabetes mellitus), multiple gestations, and presence of cerclage. We chose the exclusion criteria because we have not performed fetal operation with these coexisting conditions. The final cohort included women who were delivered for the following indications: preterm premature rupture of membrane (PPROM) with or without preterm labor before 37 weeks of gestation, fetal anomalies (worsening hydrocephalus, newly diagnosed fetal intra-abdominal mass), malpresentation, or nonreassuring fetal status (oligohydramnios [3 cases],
fetal cardiac arrhythmia unresponsive to medical therapy [1 case], hydrops [1 case], sacrococcygeal teratoma [1 case], and intrauterine growth retardation [1 case]). All underwent a nonemergent cesarean delivery; therefore, “fetal distress” as a delivery indication was excluded. Baseline characteristic data that were extracted from the medical records included maternal age, parity, and history of uterine surgical procedures (excluding the fetal therapy procedure). Predictors of morbidity included prolonged rupture of the membranes (PPROM >24 hours), indication for delivery that included chorioamnionitis, type of anesthesia, type of uterine incision, and length of procedure, as defined in the anesthesia record. Chorioamnionitis was defined clinically for both EXIT and control patients. Mothers were evaluated and diagnosed with chorioamnionitis if they had an elevated temperature of >37.8°C and at least 1 of the following signs or symptoms: uterine tenderness, uterine contractions, foul vaginal discharge, maternal tachycardia, fetal tachycardia, or elevated white blood cell count. Amniocentesis is not routinely used in this institution as a diagnostic tool for chorioamnionitis. Betamethasone, antibiotics, and/or tocolytic usage was determined within 24 hours of delivery. Outcome variables that were evaluated were estimated blood loss (EBL) as determined by the attending obstetrician, hematocrit level change as the difference between postoperative day 1 and the immediate preoperative values, infectious morbidity (endomyometritis, wound infection and cellulitis, urinary tract infection, and pneumonia ascertained from chart review), clinically apparent thromboembolic events, and hospital stay. Statistical analyses were performed with Stata 6.0 (Stata Corp, College Station, Tex). Chi-square with Fisher exact test, where appropriate, for small numbers and 2-tailed Student t tests were used to evaluate categoric and continuous variables, respectively. Results were considered significant at a probability value of <.05. The University of California San Francisco Committee on Human Research approved the study.
Scully Noah et al 775
Volume 186, Number 4 Am J Obstet Gynecol
Table II. Antepartum predictor variables Variable Betamethasone PPROM, >24 h Tocolysis Antibiotics
EXIT procedure (n = 34) 17 (50%) 17 (50%) 31 (91%) 18 (53%)
Cesarean delivery (n = 52)
P value*
13 (25%) 27 (52%) 16 (31%) 25 (48%)
.02 1.00 <.0001 .83
*χ2 tests, with Fisher exact test.
Table III. Indications for delivery Indication Fetal indication Chorioamnionitis Preterm labor Abruptio placentae
EXIT procedure (n = 34)
Cesarean delivery (n = 52)
P value*
11 (32%) 9 (26%) 9 (26%) 5 (15%)
32 (61%) 11 (21%) 4 (8%) 5 (10%)
.008 .57 .03 .51
*χ2 tests, with Fisher exact test.
Results Thirty-eight patients underwent the EXIT procedure between 1994 and 1999, of whom 4 patients were excluded from analysis because of their inclusion in an ongoing congenital diaphragmatic hernia trial. As seen in Table I, the groups were well matched with respect to maternal age, gravidity, parity, and history of previous uterine surgical procedures (including myomectomy and cesarean delivery), except that the EXIT patients had undergone either hysterotomy or laparoscopic fetal operation in the index pregnancy. Nine patients required a hysterotomy for the fetal procedure; 22 patients underwent a laparoscopic fetal procedure. The mean gestational age at which the EXIT patients underwent their initial in utero procedure was 27.9 weeks (±2.7 weeks [SD]). The average time from open fetal operation to delivery was 28.5 days versus 23.8 days for those who underwent laparoscopic fetal operation. All patients were discharged from the hospital after the fetal operation, except for 7 patients who experienced PROM or delivery after the fetal intervention. There was no difference in gestational age between EXIT and cesarean delivery groups at the time of the admission that resulted in delivery or in the gestational age at delivery. Of the antepartum predictor variables evaluated, the EXIT patients were more likely to have received betamethasone (50% vs 25%; P = .02) and tocolysis within 24 hours of delivery (91% vs 31%; P ≤ .0001). The women in the cesarean delivery group all received betamethasone during the index hospitalization if they were admitted at <34 weeks. After 34 weeks, corticosteroid administration varied, depending on the clinical situation. There was no difference between groups in the receipt of antibiotics within 24 hours of delivery or PPROM of >24 hours
(Table II). Preterm labor as an indication for delivery was more common in the EXIT patients (26% vs 8%; P = .03), and delivery for fetal indications was more frequent in the cesarean delivery group (61% vs 32%; P = .008). The rate of chorioamnionitis did not differ between groups (26% vs 21%; P = .57; Table III). The EXIT procedure lasted approximately twice as long as the cesarean delivery (109 vs 57 minutes; P < .0001) from skin incision to skin closure; the fetal portion of the procedure accounted for some of this time (24.2 vs 2.6 minutes; P < .0001). Fifty-three percent of the EXIT patients and 55% of the comparison group underwent a low transverse cesarean delivery; the balance had classic or low vertical uterine incisions. If a hysterotomy was performed for the index procedure, the same site was used for delivery. EXIT patients received intraoperative antibiotics more commonly (91% vs 71%; P = .03; Table IV). The immediate postoperative courses were similar between the 2 groups (Table V). Although the EBL was higher in the EXIT patients (1104 vs 883 mL; P = .0096) the postoperative hematocrit level change was not significantly different (5.3% vs 4.6%; P = .42). Three patients in the EXIT group and 1 of the cesarean delivery patients received blood transfusions. Superficial wound infections were more common in EXIT patients (15% vs 2%; P = .03), but endometritis (15% vs 10%; P = .50) and total infections (21% vs 12%; P = .23) did not differ between groups. One cesarean delivery patient experienced a fascial dehiscence. There were no differences in length of stay after delivery. There were no immediate thromboembolic events in either group, no need for hysterectomy, and no maternal deaths in either group.
776 Scully Noah et al
April 2002 Am J Obstet Gynecol
Table IV. Intrapartum predictor variables Variable Intraoperative antibiotics (n) General anesthesia (n) Uterine incision (n) Classic (n) Low transverse cesarean delivery (n) Procedure length (min) Surgical operating time Fetal operating time
EXIT procedure (n = 34)
Cesarean delivery (n = 52)
P value*
31 (91%) 34 (100%)
37 (71%) 12 (35%)
.03 <.0001
16 (47%) 18 (53%)
21 (40%) 29 (55%)
.57
110 24.2
57 2.6
<.0001 <.0001
*χ2 tests, with Fisher exact test; Student t test.
Table V. Outcome variables Variable EBL (mL) Hematocrit change (%) Total postoperative infection (n) Wound complications (n) Endometritis (n) Length of stay (d)
EXIT procedure (n = 34) 1104 5.3 8 (24%) 5 (15%) 5 (15%) 3.5
Cesarean delivery (n = 52) 883 4.6 6 (12%) 1 (2%) 5 (10%) 3.5
P value* <.001 .42 .23 .03 .50 .97
*Student t test and χ2 tests, with Fisher exact for small numbers.
Comment This study provides the first complete set of data on maternal outcomes that are associated with the EXIT procedure. The possible trend toward increased wound complications would not be surprising, given 2 surgeries in such a short time period. We did not look at long-term outcomes to know whether morbidity developed later, but this will be evaluated in future studies. Although the indications for delivery differed between groups, the rate of chorioamnionitis was the same, even though the EXIT group had undergone an invasive in utero surgical procedure only weeks before delivery. Because the EXIT procedure is not one that can be performed emergently, given the complexity of the set-up necessary for neonatal care, the patients in this group must be delivered sooner if evidence is found of a developing in utero infection. Yet, one does not want to act too quickly because preterm delivery of an already compromised fetus markedly worsens the chance of the fetus for survival. Thus, the similar rate of chorioamnionitis as the primary indication for delivery may have affected the rate of endometritis postpartum because the EXIT patients would have been identified early in the infection process and treated aggressively. Moreover, it is well known that surgical time directly influences postoperative infection risk. The EXIT patients underwent a lengthier procedure but were also treated more frequently with intraoperative antibiotics. This also may have influenced the postpartum endometritis rate in the EXIT group.
Furthermore, although the markedly prolonged operative times that are associated with the EXIT procedure and general anesthesia with deep relaxation probably contributed to the increased blood loss, these factors did not affect hematocrit level change. It is likely, however, that neither variable reflects the true blood loss experienced by these patients. The EBL often underestimates the amount of blood that has accumulated in the sponges and drapes. Conversely, the hematocrit drawn on the first postoperative day often is affected by hydration status because the patient’s blood volume may not have equilibrated. This study was not large enough to identify the potential morbidities that are associated with general anesthesia and marked uterine relaxation. However, the cesarean delivery group in this study was skewed by a disproportionate number of failed epidural procedures; the general endotracheal anesthesia rate for all cesarean deliveries in this institution is 8%. Women in the EXIT group recovered quickly, even with 2 major operations over a 4-week time period, and were discharged in the same time course as women who underwent nonemergent cesarean deliveries. Importantly, >50% of the patients in the EXIT group had a low transverse hysterotomy and subsequent cesarean delivery. Thus, this group of women can undergo a trial of labor in future pregnancies. Although we did not observe any major maternal morbidities in this study, long-term data on maternal health after these invasive fetal procedures will be collected and evaluated.
Volume 186, Number 4 Am J Obstet Gynecol
REFERENCES
1. Harrison MR, Mychaliska GB, Albanese CT, Jennings RW, Farrell JA, Hawgood S, et al. Correction of congenital diaphragmatic hernia in utero, IX: fetuses with poor prognosis (liver herniation and low lung-to-head ratio) can be saved by fetoscopic temporary tracheal occlusion. J Pediatr Surg 1999; 33:1017-23.
Scully Noah et al 777
2. Crombleholme TM, Albanese CT. The fetus with airway obstruction. In: Harrison MR, Adzick NS, Evans MI, Holzgreve W, editors. The unborn patient: the art and science of fetal therapy. 3rd ed. Philadelphia: Saunders; 2000. p. 357-71. 3. Mychaliska GB, Bealer JF, Graf JL, Rosen MA, Adzick NS, Harrison MR. Operating on placental support: the ex utero intrapartum treatment procedure. J Pediatr Surg 1997;32:227-30.
Key words: EXIT procedure, morbidity, surgery, cesarean delivery
Although increasing data exists regarding the fetal outcomes that are associated with in utero surgical procedures, little information is available about the maternal outcomes that are associated with these procedures. The ex utero intrapartum treatment (EXIT) procedure was developed as a means to maintain fetoplacental circulation until the airway is secured in fetuses with known tracheal obstruction.1 Most of these cases are due to balloon or clip placement across the fetal trachea for in utero treatment of congenital diaphragmatic hernia, although the EXIT procedure has been used in noniatrogenic cases of airway obstruction, such as cystic hygroma, large goiter, or laryngeal atresia.2 As described by Mychaliska et
From the Department of Obstetrics, Gynecology, and Reproductive Sciences, University of California, San Francisco,a the University of California, San Francisco Medical School,b and the Division of Pediatric Surgery, Fetal Treatment Center.c Received for publication June 7, 2001; revised November 26, 2001; accepted December 12, 2001. Reprint requests: Melinda M. Scully Noah, MD, Department of Obstetrics, Gynecology, and Reproductive Sciences, University of California, San Francisco, Box 0550 HSW 1452, 505 Parnassus Ave, San Francisco, CA 94143. E-mail: [email protected] Copyright 2002, Mosby, Inc. All rights reserved. 0002-9378/2002 $35.00 + 0 6/1/122249 doi:10.1067/mob.2002.112249
al,3 this surgical procedure requires the pregnant woman to be placed under general endotracheal anesthesia by rapid sequence induction and subsequently maintained on high-dose isoflurane. This regimen is continued while the hysterotomy is performed and the neonate is delivered to the level of the trachea. At this point, airway control is obtained or the airway obstruction is overcome; pressure-controlled bag ventilation is instituted, and surfactant is administered (if indicated) while the neonate is still attached to the placenta. The airway is then secured; the baby is delivered, and the umbilical cord is clamped. After delivery of the neonate, the anesthetic regimen is returned to routine levels of general anesthesia, and oxytocic is added to the intravenous fluids. The surgical portion of the EXIT procedure generally follows a standard obstetric cesarean delivery but is markedly prolonged by the neonatal intubation. During this part of the procedure, uterine tone is purposely absent, and the hysterotomy site is open longer, thus potentially increasing the chance of maternal infection and bleeding. Blood loss from the edges of the hysterotomy is minimized with the use of a specially developed uterine stapler and uterine clamps. In addition, if the initial in utero surgical procedure was an open procedure in which a hysterotomy was performed, then that same sur773
774 Scully Noah et al
April 2002 Am J Obstet Gynecol
Table I. Baseline characteristics Characteristic Maternal age (y) Gravidity Parity Uterine surgery† (n) Gestational age at admission (wk) Gestational age at delivery (wk)
EXIT procedure (n = 34) 28.3 2.6 0.65 3 (9%) 31.1 31.4
Cesarean delivery (n = 52) 28.0 3.0 1.0 12 (23%) 30.5 31.6
P value* .82 .38 .17 .15 .47 .75
*Student t test; χ2 tests, with Fisher exact test. †Uterine surgical procedures include previous cesarean delivery and uterine operation.
gical site is used for delivery, which further decreases blood loss. Oxytocin and antibiotics, usually a firstgeneration cephalosporin, are given intravenously to the mother after the cord is clamped. Because a pediatric surgeon is required at delivery to treat the infant, an emergent or “crash” cesarean delivery is not possible. Therefore, patients are delivered at the earliest evidence of labor or fetal compromise. This study sought to evaluate the short-term morbidities that are associated with the EXIT procedure and to compare them with those morbidities that are incurred by women who undergo prelabor cesarean deliveries at a similar gestational age for standard obstetric, nonemergency conditions at the University of California, San Francisco. Material and methods A retrospective cohort study was performed that evaluated the outcomes of patients who underwent the EXIT procedure from March 1994 through December 1999. Patients who were enrolled in an ongoing clinical trial that evaluated the efficacy of tracheal occlusion for congenital diaphragmatic hernia were excluded. Eight patients who were included in this study were described in a previously published review of the EXIT procedure.3 The comparison group included women who underwent a cesarean delivery before labor for obstetric indications during the same time period and were matched for gestational age range. Exclusion criteria for the comparison cohort were emergent cesarean delivery, placenta previa, any maternal medical disorder (such as preeclampsia, hypertension, renal disease, or diabetes mellitus), multiple gestations, and presence of cerclage. We chose the exclusion criteria because we have not performed fetal operation with these coexisting conditions. The final cohort included women who were delivered for the following indications: preterm premature rupture of membrane (PPROM) with or without preterm labor before 37 weeks of gestation, fetal anomalies (worsening hydrocephalus, newly diagnosed fetal intra-abdominal mass), malpresentation, or nonreassuring fetal status (oligohydramnios [3 cases],
fetal cardiac arrhythmia unresponsive to medical therapy [1 case], hydrops [1 case], sacrococcygeal teratoma [1 case], and intrauterine growth retardation [1 case]). All underwent a nonemergent cesarean delivery; therefore, “fetal distress” as a delivery indication was excluded. Baseline characteristic data that were extracted from the medical records included maternal age, parity, and history of uterine surgical procedures (excluding the fetal therapy procedure). Predictors of morbidity included prolonged rupture of the membranes (PPROM >24 hours), indication for delivery that included chorioamnionitis, type of anesthesia, type of uterine incision, and length of procedure, as defined in the anesthesia record. Chorioamnionitis was defined clinically for both EXIT and control patients. Mothers were evaluated and diagnosed with chorioamnionitis if they had an elevated temperature of >37.8°C and at least 1 of the following signs or symptoms: uterine tenderness, uterine contractions, foul vaginal discharge, maternal tachycardia, fetal tachycardia, or elevated white blood cell count. Amniocentesis is not routinely used in this institution as a diagnostic tool for chorioamnionitis. Betamethasone, antibiotics, and/or tocolytic usage was determined within 24 hours of delivery. Outcome variables that were evaluated were estimated blood loss (EBL) as determined by the attending obstetrician, hematocrit level change as the difference between postoperative day 1 and the immediate preoperative values, infectious morbidity (endomyometritis, wound infection and cellulitis, urinary tract infection, and pneumonia ascertained from chart review), clinically apparent thromboembolic events, and hospital stay. Statistical analyses were performed with Stata 6.0 (Stata Corp, College Station, Tex). Chi-square with Fisher exact test, where appropriate, for small numbers and 2-tailed Student t tests were used to evaluate categoric and continuous variables, respectively. Results were considered significant at a probability value of <.05. The University of California San Francisco Committee on Human Research approved the study.
Scully Noah et al 775
Volume 186, Number 4 Am J Obstet Gynecol
Table II. Antepartum predictor variables Variable Betamethasone PPROM, >24 h Tocolysis Antibiotics
EXIT procedure (n = 34) 17 (50%) 17 (50%) 31 (91%) 18 (53%)
Cesarean delivery (n = 52)
P value*
13 (25%) 27 (52%) 16 (31%) 25 (48%)
.02 1.00 <.0001 .83
*χ2 tests, with Fisher exact test.
Table III. Indications for delivery Indication Fetal indication Chorioamnionitis Preterm labor Abruptio placentae
EXIT procedure (n = 34)
Cesarean delivery (n = 52)
P value*
11 (32%) 9 (26%) 9 (26%) 5 (15%)
32 (61%) 11 (21%) 4 (8%) 5 (10%)
.008 .57 .03 .51
*χ2 tests, with Fisher exact test.
Results Thirty-eight patients underwent the EXIT procedure between 1994 and 1999, of whom 4 patients were excluded from analysis because of their inclusion in an ongoing congenital diaphragmatic hernia trial. As seen in Table I, the groups were well matched with respect to maternal age, gravidity, parity, and history of previous uterine surgical procedures (including myomectomy and cesarean delivery), except that the EXIT patients had undergone either hysterotomy or laparoscopic fetal operation in the index pregnancy. Nine patients required a hysterotomy for the fetal procedure; 22 patients underwent a laparoscopic fetal procedure. The mean gestational age at which the EXIT patients underwent their initial in utero procedure was 27.9 weeks (±2.7 weeks [SD]). The average time from open fetal operation to delivery was 28.5 days versus 23.8 days for those who underwent laparoscopic fetal operation. All patients were discharged from the hospital after the fetal operation, except for 7 patients who experienced PROM or delivery after the fetal intervention. There was no difference in gestational age between EXIT and cesarean delivery groups at the time of the admission that resulted in delivery or in the gestational age at delivery. Of the antepartum predictor variables evaluated, the EXIT patients were more likely to have received betamethasone (50% vs 25%; P = .02) and tocolysis within 24 hours of delivery (91% vs 31%; P ≤ .0001). The women in the cesarean delivery group all received betamethasone during the index hospitalization if they were admitted at <34 weeks. After 34 weeks, corticosteroid administration varied, depending on the clinical situation. There was no difference between groups in the receipt of antibiotics within 24 hours of delivery or PPROM of >24 hours
(Table II). Preterm labor as an indication for delivery was more common in the EXIT patients (26% vs 8%; P = .03), and delivery for fetal indications was more frequent in the cesarean delivery group (61% vs 32%; P = .008). The rate of chorioamnionitis did not differ between groups (26% vs 21%; P = .57; Table III). The EXIT procedure lasted approximately twice as long as the cesarean delivery (109 vs 57 minutes; P < .0001) from skin incision to skin closure; the fetal portion of the procedure accounted for some of this time (24.2 vs 2.6 minutes; P < .0001). Fifty-three percent of the EXIT patients and 55% of the comparison group underwent a low transverse cesarean delivery; the balance had classic or low vertical uterine incisions. If a hysterotomy was performed for the index procedure, the same site was used for delivery. EXIT patients received intraoperative antibiotics more commonly (91% vs 71%; P = .03; Table IV). The immediate postoperative courses were similar between the 2 groups (Table V). Although the EBL was higher in the EXIT patients (1104 vs 883 mL; P = .0096) the postoperative hematocrit level change was not significantly different (5.3% vs 4.6%; P = .42). Three patients in the EXIT group and 1 of the cesarean delivery patients received blood transfusions. Superficial wound infections were more common in EXIT patients (15% vs 2%; P = .03), but endometritis (15% vs 10%; P = .50) and total infections (21% vs 12%; P = .23) did not differ between groups. One cesarean delivery patient experienced a fascial dehiscence. There were no differences in length of stay after delivery. There were no immediate thromboembolic events in either group, no need for hysterectomy, and no maternal deaths in either group.
776 Scully Noah et al
April 2002 Am J Obstet Gynecol
Table IV. Intrapartum predictor variables Variable Intraoperative antibiotics (n) General anesthesia (n) Uterine incision (n) Classic (n) Low transverse cesarean delivery (n) Procedure length (min) Surgical operating time Fetal operating time
EXIT procedure (n = 34)
Cesarean delivery (n = 52)
P value*
31 (91%) 34 (100%)
37 (71%) 12 (35%)
.03 <.0001
16 (47%) 18 (53%)
21 (40%) 29 (55%)
.57
110 24.2
57 2.6
<.0001 <.0001
*χ2 tests, with Fisher exact test; Student t test.
Table V. Outcome variables Variable EBL (mL) Hematocrit change (%) Total postoperative infection (n) Wound complications (n) Endometritis (n) Length of stay (d)
EXIT procedure (n = 34) 1104 5.3 8 (24%) 5 (15%) 5 (15%) 3.5
Cesarean delivery (n = 52) 883 4.6 6 (12%) 1 (2%) 5 (10%) 3.5
P value* <.001 .42 .23 .03 .50 .97
*Student t test and χ2 tests, with Fisher exact for small numbers.
Comment This study provides the first complete set of data on maternal outcomes that are associated with the EXIT procedure. The possible trend toward increased wound complications would not be surprising, given 2 surgeries in such a short time period. We did not look at long-term outcomes to know whether morbidity developed later, but this will be evaluated in future studies. Although the indications for delivery differed between groups, the rate of chorioamnionitis was the same, even though the EXIT group had undergone an invasive in utero surgical procedure only weeks before delivery. Because the EXIT procedure is not one that can be performed emergently, given the complexity of the set-up necessary for neonatal care, the patients in this group must be delivered sooner if evidence is found of a developing in utero infection. Yet, one does not want to act too quickly because preterm delivery of an already compromised fetus markedly worsens the chance of the fetus for survival. Thus, the similar rate of chorioamnionitis as the primary indication for delivery may have affected the rate of endometritis postpartum because the EXIT patients would have been identified early in the infection process and treated aggressively. Moreover, it is well known that surgical time directly influences postoperative infection risk. The EXIT patients underwent a lengthier procedure but were also treated more frequently with intraoperative antibiotics. This also may have influenced the postpartum endometritis rate in the EXIT group.
Furthermore, although the markedly prolonged operative times that are associated with the EXIT procedure and general anesthesia with deep relaxation probably contributed to the increased blood loss, these factors did not affect hematocrit level change. It is likely, however, that neither variable reflects the true blood loss experienced by these patients. The EBL often underestimates the amount of blood that has accumulated in the sponges and drapes. Conversely, the hematocrit drawn on the first postoperative day often is affected by hydration status because the patient’s blood volume may not have equilibrated. This study was not large enough to identify the potential morbidities that are associated with general anesthesia and marked uterine relaxation. However, the cesarean delivery group in this study was skewed by a disproportionate number of failed epidural procedures; the general endotracheal anesthesia rate for all cesarean deliveries in this institution is 8%. Women in the EXIT group recovered quickly, even with 2 major operations over a 4-week time period, and were discharged in the same time course as women who underwent nonemergent cesarean deliveries. Importantly, >50% of the patients in the EXIT group had a low transverse hysterotomy and subsequent cesarean delivery. Thus, this group of women can undergo a trial of labor in future pregnancies. Although we did not observe any major maternal morbidities in this study, long-term data on maternal health after these invasive fetal procedures will be collected and evaluated.
Volume 186, Number 4 Am J Obstet Gynecol
REFERENCES
1. Harrison MR, Mychaliska GB, Albanese CT, Jennings RW, Farrell JA, Hawgood S, et al. Correction of congenital diaphragmatic hernia in utero, IX: fetuses with poor prognosis (liver herniation and low lung-to-head ratio) can be saved by fetoscopic temporary tracheal occlusion. J Pediatr Surg 1999; 33:1017-23.
Scully Noah et al 777
2. Crombleholme TM, Albanese CT. The fetus with airway obstruction. In: Harrison MR, Adzick NS, Evans MI, Holzgreve W, editors. The unborn patient: the art and science of fetal therapy. 3rd ed. Philadelphia: Saunders; 2000. p. 357-71. 3. Mychaliska GB, Bealer JF, Graf JL, Rosen MA, Adzick NS, Harrison MR. Operating on placental support: the ex utero intrapartum treatment procedure. J Pediatr Surg 1997;32:227-30.