Owen et al.
10. II.
12. 13. 14.
gel for cervical ripening. Am j Hosp Pharm 1983;40:2195-6. Chalmer I, ed. Oxford database of perinatal trials. London: Oxford University Press, 1988. Rosenthal R. Combining results of independent studies. Psychol Bull 1978;85: 185-93. Mackenzie IZ, Burnet FR, Embrey MP. In vitro studies of the characteristics of the release of prostaglandins from viscous solutions. Br j Obstet Gynaecol 1980;87:292-5. Rosenthal R. The"file drawer problem" and tolerance for null results. Psychol Bull 1979;86:638-41. Schreyer P, Sherman Dj, Ariely S, Herman A, Caspi E. Ripening the highly unfavorable cervix with extra-am-
October 1991 Am J Obstet Gynecol
niotic saline instillation or vaginal prostaglandin E2 application. Obstet Gynecol 1989;73:938-42. 15. johnson IR, MacPherson MBA, Welch CC, Filstie GIM. A comparison of lamicel and prostaglandin E2 vaginal gel for cervical ripening before induction of labor. AM j OBSTET GYNECOL 1985;151:604-7. 16. jagani N, Schulman H, Bleischer A, Mitchell j, Blattner P. Role of prostaglandin-induced cervical changes in labor induction. Obstet Gynecol 1984;63:225-9. 17. Dickerson K. The existence of publication bias and risk factors for its occurrence. JAMA 1990;263:1385-9. A complete list of references is available from the authors on request.
Uterine rupture during trial of labor after previous cesarean section Richard M. Farmer, MD, PhD, Thomas Kirschbaum, MD, Daniel Potter, Thomas H. Strong, MD, and Arnold L. Medearis, MD Los Angeles, California This study was undertaken to determine the incidence and associations of uterine rupture and dehiscence with an attempted vaginal birth after cesarean section. The charts from 137 patients who had uterine scar separation after a previous cesarean section from 1983 to 1989 were examined. Approximately 9.3% of the 119,395 women who were delivered in that interval had a prior cesarean section. Of those, 68.8% underwent a trial of labor with a 79.2% success rate. The uterine rupture rate in this latter group was 0.8%, while an additional 0.7% had a bloodless dehiscence. Bleeding and pain were unlikely findings with a uterine scar separation (3.4% and 7.6%, respectively). The most common manifestation of a scar separation was a prolonged fetal heart rate deceleration leading to operative intervention (70.3%). We conclude that, although the incidence of uterine rupture was low, the event is most often seen as an acute emergency. Prevention should be directed toward timely diagnosis and prompt management of labor dystocias. Staff and facilities for safe management of a uterine scar separation are a requisite for the conduct of a vaginal birth after previous cesarean section. (AM J OBSTET GYNECOL 1991 ;165:996-1001.)
Key words: Uterine rupture, cesarean section, dehiscence The incidence of cesarean section m the United States remains at a high level, approximately 25%.1 Financial and medical considerations and patient preferences have initiated pressures to decrease the number of cesarean sections by promoting vaginal birth after a
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Southern California, and Los Angeles County-University of Southern California Medical Center, Women's Hospital. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetricians, San Francisco, California, January 28-February 2, 1991. Reprint requests: Arnold L. Medearis, MD, Department of Obstetrics and Gynecology, University of Southern California, LAC 1USC Medical Center, Women's Hospital, Room 5K40, 1240 N. Mission Road, Los Angeles, CA 90033. 616131204
996
previous cesarean section. Counter balancing this trend is concern about a possible catastrophic scar separation during labor. Previous studies have estimated the incidence of uterine scar separation to be between 0.2% and 3.5%.2.6 However, interpretation of incidence figures has been complicated by a lack of agreement on definitions. Uterine rupture has not been uniformly distinguished from uterine scar dehiscence. Bloodless uterine scar dehiscence has not been shown to have serious consequences for the mother or infant, whereas the effect of uterine rupture can be devastating to both. Previous studies from this department and others 7. lo have demonstrated the relative safety of a trial of labor after a cesarean section. The low incidence and small numbers of uterine ruptures, however, have precluded a comprehensive examination of this complication.
Uterine rupture in patients undergoing trial of labor 997
Volume 165 Number 4, Part 1
Table I. Occurrence of scar separation in patients with previous cesarean section Dehiscence
All patients with previous cesarean section Elective repeat cesarean section All patients with trial of labor Successful trial of labor Failed trial of labor (cesarean section)
Rupture
n
%
No.
%
No.
11,041
0.6
67
0.6
70
3,443
0.4
13
0.3
9
7,598
0.7
54
0.8
61
6,021 1,577
0.5 1.7
28 26
0.1 3.4
7 54
This study was undertaken to examine the Los Angeles County-University of Southern California Women's Hospital experience with uterine rupture in patients who underwent a trial of labor after previous cesarean section. An attempt was made to define maternal risk factors associated with a uterine rupture. Material and methods
The study population consisted of patients who were delivered at the Los Angeles County-University of Southern California Women's Hospital, the teaching hospital of the University of Southern California School of Medicine. Patients with a previous cesarean section who had a scar separation were identified during the period from January 1983 through December 1989 from coded medical records and department morbidity and mortality review documents. The year 1983 marked the beginning of the systematic introduction of vaginal birth after previous cesarean section. The records of 137 patients with a scar separation were retrospectively reviewed. The patients who underwent a trial of labor had either a low-transverse, low-vertical, or unknown uterine incision. Exclusion criteria from 1983 through 1985 were an unknown uterine scar, a previous classic incision, or transverse lie that failed version. After that time, patients with unknown scars were included. Patients with multiple gestation or breech presentation were not excluded. If the patient refused a trial of labor, a repeat cesarean section was planned. The fetal heart rate was monitored with a fetal scalp electrode or external Doppler technique. Uterine activity was monitored with an intrauterine pressure catheter when cervical dilatation allowed or with an external tocodynamometer. In each case blood was available and intravenous access established. Oxytocin and an epidural anesthetic were administered for obstetric indications. A uterine exploration was performed after each successful vaginal delivery to establish scar status. Patient management was conducted by resident physicians with staff supervision.
Uterine rupture was defined for our purposes by a modification of Donnelly and Franzoni. l l Uterine dehiscence was defined as a uterine wall defect with no emergent laparotomy, no evidence of fetal distress, and no excessive bleeding. A uterine rupture was defined as a uterine wall defect and emergency laparotomy, acute fetal distress necessitating operative intervention, or acute maternal bleeding with an estimated blood loss of> 1 L. Statistical analysis was conducted with the Crunch statistical software package (Crunch Software Corp., Oakland, Calif.). A t test was used to compare groups of patients, with p < 0.05 considered significant. Timetrend analysis was conducted with the BMDP package (BMDP Statistical Software Inc., Los Angeles) statistical software with a Cochran-Mantel-Haenszel analysis. Results
During the period from January 1983 through December 1989 there were 119,395 deliveries at Los Angeles County-University of Southern California Women's Hospital, including 11,041 (9.3%) women with a history of a cesarean section. Of these, 7598 patients (69%) underwent a trial of labor, with 6021 (79.2%) delivered vaginally. The uterine rupture rate with a previous cesarean section was 0.6% (70 cases), whereas the uterine dehiscence rate was 0.6% (67 cases). The distribution of patients into trial of labor and elective repeat cesarean section groups is shown in Table I. In those patients who underwent a trial of labor, the incidence of uterine rupture and dehiscence was 0.8% (61 cases) and 0.7% (54 cases), respectively. The uterine rupture and dehiscence rates over the years studied are shown in Table II for patients who underwent a trial of labor. Fetal distress was the primary indication for cesarean section in patients with a uterine rupture (85.2%, 46/54 cases), with arrest disorders second (13.0%, 7/54 cases). In patients with a uterine dehiscence, arrest disorders were the most common indication for cesarean section (69.2%, 19/26 cases), with fetal distress sec-
998
Farmer et al.
October 1991 Am J Obstet Gynecol
Table II. Incidence of scar separation in patients who underwent trial of labor after previous cesarean section, by year Dehiscence
Rupture
Year
n
%
No.
%
No.
1983 1984 1985 1986 1987 1988 1989
908 1153 1367 1131 730 999 1310
0.1 0.8 0.4 1.3 0.8 0.6 0.8
I 9 6 15 6 6 II
0.4 0.4 0.6 1.0 1.2 0.4 1.5
4 5 7 II 10 4 20
TOTAL
7595
0.7
(54)
0.8
(61)
Cochran-Mantel-Haenszel trend analysis; statistically significant increase in rate of uterine rupture (p nificant increase in rate of uterine dehiscence.
=
0.003) with no sig-
Table III. Occurrence of uterine scar separation as function of type of previous uterine incision in patients who underwent trend of labor Dehiscence (n = 54) Type of previous uterine incision
No.
Low-transverse Unknown Classic Vertical
37 16 0 I
I
ond (26.9%, 7/26 cases). The seven cases of fetal distress in the uterine dehiscence group were not believed to be related to the scar separation, because the dehiscences were small «2 em). The type of previous uterine incision in patients who underwent a trial of labor, either low-transverse, lowvertical, classic, or unknown, is shown in Table III. The type of previous scar was identified from operative reports or by intraoperative identification of the scar site. Eleven of 115 patients (9.5%) who underwent a trial of labor had uterine rupture that was distant and separate from the previous scar. The average number of previous cesarean sections in those patients who underwent a trial of labor was 1.4, with only four patients having more than two previous cesarean sections. Patients were not excluded from a trial of labor on the basis of the number of previous cesarean sections, nor was the indication for previous cesarean section a cause for exclusion, with the exception of a previous scar separation. In such patients a repeat elective cesarean section was recommended. The number of successful vaginal deliveries after a cesarean section is shown in Table I for the trialof-labor population. Oxytocin was used during the trial of labor in 73.8% of the patients with a uterine rupture and in 74.1 % of the patients with a uterine dehiscence. Overall use of oxytocin during a trial of labor was 74.0% of patients
Rupture (n = 61)
%
No.
68.5 29.6 0 1.9
29 25 5 2
I
% 47.5 41.0 8.2 3.3
with a scar separation. An epidural anesthetic was administered during labor to 27.9% and 35.2% of the uterine rupture and dehiscence populations, respectively. The concurrent administration of epidural anesthesia and oxytocin was seen in 24.6% of the uterine rupture and 33.3% of the uterine dehiscence groups. Average labor duration in those patients with a uterine rupture was 14.8 ± 10.6 hours, whereas the average duration was 15.0 ± 8.2 hours in patients with a uterine dehiscence. The maternal and infant variables of age, gravidity, parity, estimated blood loss, Apgar scores, birth weight, and gestational age are shown in Table IV for patients with a scar separation. Maternal morbidity and mortality as evidenced by maternal death, transfusion, urologic injury, and hysterectomy are seen in Table V. Thirteen patients were transfused, with an average of 2.7 units per patient. No patient with uterine dehiscence received a transfusion. The 15 urologic injuries consisted predominantly of intentional cystotomy or bladder laceration, with no ureteral injuries reported. Cystotomy with or without the passage of stents was most often done for the determination of ureter position or patency. Fourteen hysterectomies were performed in the uterine rupture group. Neonatal morbidity and mortality, as evidenced by neonatal death, fetal death, and neonatal intensive care unit admission, are also seen in Table V. There were no late infant
Volume 165 Number 4, Part I
Uterine rupture in patients undergoing trial of labor 999
Table IV. Patient characteristics Dehiscence
Age (yr) Gravidity Parity Estimated blood loss (ml) Apgar score at 1 min
10. II.
12. 13. 14.
gel for cervical ripening. Am j Hosp Pharm 1983;40:2195-6. Chalmer I, ed. Oxford database of perinatal trials. London: Oxford University Press, 1988. Rosenthal R. Combining results of independent studies. Psychol Bull 1978;85: 185-93. Mackenzie IZ, Burnet FR, Embrey MP. In vitro studies of the characteristics of the release of prostaglandins from viscous solutions. Br j Obstet Gynaecol 1980;87:292-5. Rosenthal R. The"file drawer problem" and tolerance for null results. Psychol Bull 1979;86:638-41. Schreyer P, Sherman Dj, Ariely S, Herman A, Caspi E. Ripening the highly unfavorable cervix with extra-am-
October 1991 Am J Obstet Gynecol
niotic saline instillation or vaginal prostaglandin E2 application. Obstet Gynecol 1989;73:938-42. 15. johnson IR, MacPherson MBA, Welch CC, Filstie GIM. A comparison of lamicel and prostaglandin E2 vaginal gel for cervical ripening before induction of labor. AM j OBSTET GYNECOL 1985;151:604-7. 16. jagani N, Schulman H, Bleischer A, Mitchell j, Blattner P. Role of prostaglandin-induced cervical changes in labor induction. Obstet Gynecol 1984;63:225-9. 17. Dickerson K. The existence of publication bias and risk factors for its occurrence. JAMA 1990;263:1385-9. A complete list of references is available from the authors on request.
Uterine rupture during trial of labor after previous cesarean section Richard M. Farmer, MD, PhD, Thomas Kirschbaum, MD, Daniel Potter, Thomas H. Strong, MD, and Arnold L. Medearis, MD Los Angeles, California This study was undertaken to determine the incidence and associations of uterine rupture and dehiscence with an attempted vaginal birth after cesarean section. The charts from 137 patients who had uterine scar separation after a previous cesarean section from 1983 to 1989 were examined. Approximately 9.3% of the 119,395 women who were delivered in that interval had a prior cesarean section. Of those, 68.8% underwent a trial of labor with a 79.2% success rate. The uterine rupture rate in this latter group was 0.8%, while an additional 0.7% had a bloodless dehiscence. Bleeding and pain were unlikely findings with a uterine scar separation (3.4% and 7.6%, respectively). The most common manifestation of a scar separation was a prolonged fetal heart rate deceleration leading to operative intervention (70.3%). We conclude that, although the incidence of uterine rupture was low, the event is most often seen as an acute emergency. Prevention should be directed toward timely diagnosis and prompt management of labor dystocias. Staff and facilities for safe management of a uterine scar separation are a requisite for the conduct of a vaginal birth after previous cesarean section. (AM J OBSTET GYNECOL 1991 ;165:996-1001.)
Key words: Uterine rupture, cesarean section, dehiscence The incidence of cesarean section m the United States remains at a high level, approximately 25%.1 Financial and medical considerations and patient preferences have initiated pressures to decrease the number of cesarean sections by promoting vaginal birth after a
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Southern California, and Los Angeles County-University of Southern California Medical Center, Women's Hospital. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetricians, San Francisco, California, January 28-February 2, 1991. Reprint requests: Arnold L. Medearis, MD, Department of Obstetrics and Gynecology, University of Southern California, LAC 1USC Medical Center, Women's Hospital, Room 5K40, 1240 N. Mission Road, Los Angeles, CA 90033. 616131204
996
previous cesarean section. Counter balancing this trend is concern about a possible catastrophic scar separation during labor. Previous studies have estimated the incidence of uterine scar separation to be between 0.2% and 3.5%.2.6 However, interpretation of incidence figures has been complicated by a lack of agreement on definitions. Uterine rupture has not been uniformly distinguished from uterine scar dehiscence. Bloodless uterine scar dehiscence has not been shown to have serious consequences for the mother or infant, whereas the effect of uterine rupture can be devastating to both. Previous studies from this department and others 7. lo have demonstrated the relative safety of a trial of labor after a cesarean section. The low incidence and small numbers of uterine ruptures, however, have precluded a comprehensive examination of this complication.
Uterine rupture in patients undergoing trial of labor 997
Volume 165 Number 4, Part 1
Table I. Occurrence of scar separation in patients with previous cesarean section Dehiscence
All patients with previous cesarean section Elective repeat cesarean section All patients with trial of labor Successful trial of labor Failed trial of labor (cesarean section)
Rupture
n
%
No.
%
No.
11,041
0.6
67
0.6
70
3,443
0.4
13
0.3
9
7,598
0.7
54
0.8
61
6,021 1,577
0.5 1.7
28 26
0.1 3.4
7 54
This study was undertaken to examine the Los Angeles County-University of Southern California Women's Hospital experience with uterine rupture in patients who underwent a trial of labor after previous cesarean section. An attempt was made to define maternal risk factors associated with a uterine rupture. Material and methods
The study population consisted of patients who were delivered at the Los Angeles County-University of Southern California Women's Hospital, the teaching hospital of the University of Southern California School of Medicine. Patients with a previous cesarean section who had a scar separation were identified during the period from January 1983 through December 1989 from coded medical records and department morbidity and mortality review documents. The year 1983 marked the beginning of the systematic introduction of vaginal birth after previous cesarean section. The records of 137 patients with a scar separation were retrospectively reviewed. The patients who underwent a trial of labor had either a low-transverse, low-vertical, or unknown uterine incision. Exclusion criteria from 1983 through 1985 were an unknown uterine scar, a previous classic incision, or transverse lie that failed version. After that time, patients with unknown scars were included. Patients with multiple gestation or breech presentation were not excluded. If the patient refused a trial of labor, a repeat cesarean section was planned. The fetal heart rate was monitored with a fetal scalp electrode or external Doppler technique. Uterine activity was monitored with an intrauterine pressure catheter when cervical dilatation allowed or with an external tocodynamometer. In each case blood was available and intravenous access established. Oxytocin and an epidural anesthetic were administered for obstetric indications. A uterine exploration was performed after each successful vaginal delivery to establish scar status. Patient management was conducted by resident physicians with staff supervision.
Uterine rupture was defined for our purposes by a modification of Donnelly and Franzoni. l l Uterine dehiscence was defined as a uterine wall defect with no emergent laparotomy, no evidence of fetal distress, and no excessive bleeding. A uterine rupture was defined as a uterine wall defect and emergency laparotomy, acute fetal distress necessitating operative intervention, or acute maternal bleeding with an estimated blood loss of> 1 L. Statistical analysis was conducted with the Crunch statistical software package (Crunch Software Corp., Oakland, Calif.). A t test was used to compare groups of patients, with p < 0.05 considered significant. Timetrend analysis was conducted with the BMDP package (BMDP Statistical Software Inc., Los Angeles) statistical software with a Cochran-Mantel-Haenszel analysis. Results
During the period from January 1983 through December 1989 there were 119,395 deliveries at Los Angeles County-University of Southern California Women's Hospital, including 11,041 (9.3%) women with a history of a cesarean section. Of these, 7598 patients (69%) underwent a trial of labor, with 6021 (79.2%) delivered vaginally. The uterine rupture rate with a previous cesarean section was 0.6% (70 cases), whereas the uterine dehiscence rate was 0.6% (67 cases). The distribution of patients into trial of labor and elective repeat cesarean section groups is shown in Table I. In those patients who underwent a trial of labor, the incidence of uterine rupture and dehiscence was 0.8% (61 cases) and 0.7% (54 cases), respectively. The uterine rupture and dehiscence rates over the years studied are shown in Table II for patients who underwent a trial of labor. Fetal distress was the primary indication for cesarean section in patients with a uterine rupture (85.2%, 46/54 cases), with arrest disorders second (13.0%, 7/54 cases). In patients with a uterine dehiscence, arrest disorders were the most common indication for cesarean section (69.2%, 19/26 cases), with fetal distress sec-
998
Farmer et al.
October 1991 Am J Obstet Gynecol
Table II. Incidence of scar separation in patients who underwent trial of labor after previous cesarean section, by year Dehiscence
Rupture
Year
n
%
No.
%
No.
1983 1984 1985 1986 1987 1988 1989
908 1153 1367 1131 730 999 1310
0.1 0.8 0.4 1.3 0.8 0.6 0.8
I 9 6 15 6 6 II
0.4 0.4 0.6 1.0 1.2 0.4 1.5
4 5 7 II 10 4 20
TOTAL
7595
0.7
(54)
0.8
(61)
Cochran-Mantel-Haenszel trend analysis; statistically significant increase in rate of uterine rupture (p nificant increase in rate of uterine dehiscence.
=
0.003) with no sig-
Table III. Occurrence of uterine scar separation as function of type of previous uterine incision in patients who underwent trend of labor Dehiscence (n = 54) Type of previous uterine incision
No.
Low-transverse Unknown Classic Vertical
37 16 0 I
I
ond (26.9%, 7/26 cases). The seven cases of fetal distress in the uterine dehiscence group were not believed to be related to the scar separation, because the dehiscences were small «2 em). The type of previous uterine incision in patients who underwent a trial of labor, either low-transverse, lowvertical, classic, or unknown, is shown in Table III. The type of previous scar was identified from operative reports or by intraoperative identification of the scar site. Eleven of 115 patients (9.5%) who underwent a trial of labor had uterine rupture that was distant and separate from the previous scar. The average number of previous cesarean sections in those patients who underwent a trial of labor was 1.4, with only four patients having more than two previous cesarean sections. Patients were not excluded from a trial of labor on the basis of the number of previous cesarean sections, nor was the indication for previous cesarean section a cause for exclusion, with the exception of a previous scar separation. In such patients a repeat elective cesarean section was recommended. The number of successful vaginal deliveries after a cesarean section is shown in Table I for the trialof-labor population. Oxytocin was used during the trial of labor in 73.8% of the patients with a uterine rupture and in 74.1 % of the patients with a uterine dehiscence. Overall use of oxytocin during a trial of labor was 74.0% of patients
Rupture (n = 61)
%
No.
68.5 29.6 0 1.9
29 25 5 2
I
% 47.5 41.0 8.2 3.3
with a scar separation. An epidural anesthetic was administered during labor to 27.9% and 35.2% of the uterine rupture and dehiscence populations, respectively. The concurrent administration of epidural anesthesia and oxytocin was seen in 24.6% of the uterine rupture and 33.3% of the uterine dehiscence groups. Average labor duration in those patients with a uterine rupture was 14.8 ± 10.6 hours, whereas the average duration was 15.0 ± 8.2 hours in patients with a uterine dehiscence. The maternal and infant variables of age, gravidity, parity, estimated blood loss, Apgar scores, birth weight, and gestational age are shown in Table IV for patients with a scar separation. Maternal morbidity and mortality as evidenced by maternal death, transfusion, urologic injury, and hysterectomy are seen in Table V. Thirteen patients were transfused, with an average of 2.7 units per patient. No patient with uterine dehiscence received a transfusion. The 15 urologic injuries consisted predominantly of intentional cystotomy or bladder laceration, with no ureteral injuries reported. Cystotomy with or without the passage of stents was most often done for the determination of ureter position or patency. Fourteen hysterectomies were performed in the uterine rupture group. Neonatal morbidity and mortality, as evidenced by neonatal death, fetal death, and neonatal intensive care unit admission, are also seen in Table V. There were no late infant
Volume 165 Number 4, Part I
Uterine rupture in patients undergoing trial of labor 999
Table IV. Patient characteristics Dehiscence
Age (yr) Gravidity Parity Estimated blood loss (ml) Apgar score at 1 min
