- Email: [email protected]
Medical management of arrest disorders of labor: A current overview
Medical management of arrest disorders of labor: A current overview Sidney F. Bottoms, M.D., Victor J. Hirsch, B.S., and Robert J. Sokol, M.D. Cleveland, Ohio In order to evaluate the risks specifically related to arrest disorders of labor, obstetric management and perinatal outcome were studied in 5399 births selected to exclude risks clearly not caused by abnormal labor. Management was primarily medical-among patients with arrest disorders, the cesarean rate was 13% and the midforceps rate was 4%, and among those having cesarean or midforceps delivery, 96% received prior oxytocin stimulation. Perinatal outcome was assessed by the parameters of mortality, low Apgar scores, admission to the nursery intensive care unit, and neurological abnormality. With medical management, arrest disorders were not associated with an increased risk of the adverse perinatal outcome evaluated in this study. It appears possible to reduce operative rates for arrest disorders of labor safely by more extensive use of medical management. (AM J OBSTET GYNECOL 1987;156:935-9.)
Key words: Arrest of labor, oxytocin, cesarean section, neonatal prognosis There have been many changes in intrapartum management during the last 15 years. These changes include the advent of electronic fetal monitoring, declining midforceps use, and less reliance on radiopelvimetry. Concomitantly, there has been a threefold increase in cesarean birth rates. Approximately one third of this increase is attributable to operations performed for dystocia (i.e., abnormal labor progress).1. 2 The need for studies of alternative management of dystocia is apparent. Arrest disorders of labor represent severe dystocia. Their associations with increased frequencies of operative delivery, perinatal mortality, neonatal morbidity, and neurologic deficits have been well documented.' However, most of the studies reporting the perinatal risks associated with arrest disorders of labor predated the changes in intrapartum management outlined above. We have previously reported unexpectedly favorable perinatal outcomes with secondary arrest of dilatation, and noted that cesarean rates for dystocia at Cleveland Metropolitan General Hospital are less than half of those reported for national data.4 These findings stimulated a more detailed appraisal of the management of arrest disorders at our institution. We describe this From the Department of Obstetrics and Gynecology, Cleveland Metropolitan General Hospital, Case Western Reserve University. Supported in part by the Perinatal Clinical Research Unit, Cleveland Metropolitan General Hospital, Case Western Reserve University, grant No. MOJRR00210 from the Division ofResearch Resources, National Institutes of Health. Received for publication April I, 1986; revised October 20, 1986; accepted October 30, 1986. · Reprint requests: Sidney F. Bottoms, M.D., the Department of Obstetrics and Gynecology, Hutzel Hospital, Wayne State University, 4707 Saint Antoine Blvd., Detroit, Ml 48201.
management as medical since we rarely perform cesarean or midforceps delivery for arrest disorders unless medical management, including oxytocin stimulation, has failed. The goal of this study was to provide a current overview of the efficacy of medical management in reducing the risks associated with arrest disorders in a format allowing comparison with previous studies. In order to evaluate the risks specifically attributable to arrest disorders, patients with obstetric complications clearly not caused by arrest disorders were excluded from study. Obstetric management and perinatal outcomes were then compared for patients with and without an arrest disorder. It should be recognized at the outset that this study cannot hope to address every issue of interest related to arrest disorders and their management.
Material and methods Detailed antepartum, intrapartum, and neonatal information was routinely obtained and coded prospectively into a computerized patient information system at Cleveland Metropolitan General Hospital. This study is based on 7550 consecutive births between 1976 and 1979, excluding antepartum stillbirths. Exclusions, made to minimize confounding effects of risks other than arrest disorders, are detailed in Table I. Oxytocin was reserved for significant medical or obstetric complications, which, in turn, might affect perinatal outcome irrespective of labor pattern. Therefore all oxytocin induction was excluded, and oxytocin augmentation of spontaneous labor was excluded except for treatment of arrest disorders. Only the most obvious major congenital anomalies (those recognized in the delivery room) were excluded. Labor diagnoses were assigned prospectively by 935
936
Bottoms, Hirsch, and Sokol
April 1987 Am J Obstet Gynecol
Table I. Selection of the study group from 7550 consecutive births (excluding antepartum stillbirths) Exclusion
n
%
Oxytocin stimulation* Low birth weight (<2500 gm) Previous cesarean section Malpresentation Multiple pregnancy Incomplete data Obvious major anomaly Neonatal deatht Intrapartum fetal deatht
1866 1055 440 338 136 51 45 9 0
24.7 14.0 5.8 4.5 1.8 0.7 0.6 0.0 0.0
2977 4573
39.4 60.6
Total excluded Total studied
*All oxytocin inductions and all oxytocin augmentations except treatment of arrest disorders were excluded. tLimited to those without other exclusion criteria.
means of a computer program developed in our laboratory. Validation of this program with visual graphic analysis of labor has been reported previously. 5 · 6 The diagnoses classified as arrest disorders were secondary arrest of dilatation (no progression in dilatation for ~2 hours during active phase of labor), prolonged deceleration phase (deceleration phase duration ~3 hours for nulliparas or :;;,, I hour for multiparas), and arrest of descent (no progression in descent :;;,, I hour for nullipara or ;;.0.5 hour for multiparas during deceleration phase or second stage). 7 Although individualized, management of arrest disorders generally followed a common approach. Ambulation was employed selectively, particularly in multiparous patients with poor contraction patterns. Typically, when abatement was used, the arrest disorder appeared to be related to administration of analgesics. When oxytocin stimulation of labor was planned, our policy was to use internal electronic fetal monitoring for both fetal heart rate and uterine contractions. If the fetal membranes were intact, external electronic fetal monitoring was temporarily employed until the vertex was well applied to a cervix sufficiently dilated to permit safe amniotomy. Many arrest disorders resolved during the interval when monitoring was initiated and/or amniotomy was performed; oxytocin was not needed in these cases. Oyxtocin was administered intravenously as a dilute solution with a constant infusion pump, at an initial dosage of I to 2 mU of oxytocin per minute. Fetal scalp sampling was utilized frequently when the diagnosis of fetal distress was made. Apgar scores were assigned by nurse or pediatrician. Neurological abnormality was assessed by the pediatric staff as part of the routine neonatal physical examination performed at time of discharge from the hospital. Any deviation from the normal neurological eval-
uation, however minor, was classified as a neurological abnormality for purposes of this study. Results were analyzed on the bases of approximate relative risks and 95% confidence intervals. 8 Results
The study group remaining after the exclusions outlined in Table I was at low risk for operative intervention and adverse perinatal outcomes. For example, 73% of the cesarean births and 79% of infants with 5-minute Apgar scores <7 were excluded. Among intrapartum and neonatal deaths, 93% had additional exclusion criteria. However, 86% of all patients with arrest disorders were in the study group. Of patients in the study group, 13% had one or more arrest disorders; among those excluded, the frequency was 3%; overall, the frequency was 9%. lntrapartum management is outlined in Fig. 1. Radiopelvimetry was rarely performed (I. 7% of patients with an arrest disorder). Operative intervention with midforceps or cesarean delivery was rarely employed without first attempting oxytocin stimulation of labor. Of the 96% of patients with arrest disorders whose initial management was medical, 50% had resolution of the arrest disorder without the use of oxytocin or midforceps or cesarean delivery. The remaining 50% of patients in this category received oxytocin stimulation. Of those who received oxytocin, 71 % had a spontaneous or low-forceps delivery. Statistical analysis of intrapartum management is reported in Table II. Electronic monitoring and all methods of operative delivery were increased with arrest disorders. Cesarean birth was five times more likely among patients with an arrest disorder than among patients without this complication. During the study period, all intrapartum fetal deaths and all but nine neonatal deaths met other exclusion criteria; none of these nine was associated with an arrest disorder of labor. Analysis of neonatal morbidity related to the effects of arrest disorders is detailed in Table Ill. Among cesarean births, those preceded by arrest disorders were less likely to result in admission of the infant to the nursery intensive care unit. No other significant difference in neonatal morbidity was attributable to arrest disorders. Neonatal morbidity among pregnancies complicated by an arrest disorder was also analyzed according to the use of oxytocin. Admission to the nursery intensive care unit was more common after stimulation with oxytocin (Table IV). However, admission to the nursery intensive care unit was also more common after cesarean birth (Table V). This risk was significantly greater than the risk associated with oxytocin stimulation and was significantly less for those with arrest disorders than
Medical management of arrest of labor 937
Volume 156 Number 4
ARREST DISORDER 593111%)
MIDFORCEPS WITHOUT OXYTOCIN 511%)
MEDICAL MANAGEMENT 569(96%)
CESAREAN WITHOUT OXYTOC IN 19(3%)
OXYTOCIN 285 (50"fo)
NO OXYTOCIN 284 (50%)
I SPONTANEOUS 258191%)
LOW FORCEPS 26 (9%)
SPONTANEOUS 161 (56%)
LOW FORCEPS 43115%)
MIDFORCEPS 2117%)
CESAREAN 60121%)
Fig. I. Management of arrest disorders in the study group. Numbers of births are indicated below each category of management. Percentages refer to the proportion of the larger category immediately above it in the figure.
Table II. lntrapartum management according to the occurrence of arrest disorders
Management
Arrest disorder (N = 593) (n as% of N)
No arrest disorder (N = 3980) (n as% of N)
Relative risk*
Electronic monitor Low-forceps Mid forceps Cesarean section
398 (67.1) 69 (11.6) 26 (4.4) 79 (13.3)
1,085 (27.3) 293 (7.4) 35 (0.9) 120 (3.0)
5.4 (4.5-6.6)t 1.7 (l.2-2.2)t 5.2 (3.l-8.7)t 4.9 (3.6-6.7)t
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
Table III. Neonatal morbidity according to the occurrence of arrest disorders, stratified by mode of delivery Arrest disorder (n as% of N)
Cesarean birth (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality Vaginal birth (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality Total (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality
79 I 16 3 514 8 51 12 593 9 67 15
(1.3) (20.3) (3.8) (1.6) (9.9) (2.3) (1.5) (11.3) (2.5)
No arrest disorder (n as% of N)
120 8 48 3 3860 29 290 57 3980 37 338 60
Relative risk*
(6.7) (40.0) (2.5)
0.2 (0.0-1.5) 0.4 (0.2-0.7)t 1.5 (0.3-8. I)
(0.8) (7.5) (1.5)
2.1 1.4 1.6
(0.9) (8.5) (1.5)
1.6 (0.8-3.5) 1.4 (l.0-1.8) 1.7 (0.9-3.0)
(0.9-4.7) ( 1.0-1.9) (0.8-3.0)
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
for those without arrest disorders. Low Apgar scores were more frequent only among cesarean births not preceded by arrest disorders.
Comment The increasing use of cesarean delivery has pointed to the need for studies of contemporary management of dystocia. Our findings suggest that medical management is a safe and effective alternative to immediate
operative intervention in almost all cases of arrest disorders of labor meeting the criteria of this study. Details of the basis for this interpretation follow, after consideration is given to the advantages and limitations of a study of this design. The methods used in this study were chosen, in part, to facilitate comparison with Friedman's'· 9 analysis of the prognosis of arrest disorders. The validity of this comparison depends on similarity of the patients stud-
938
Bottoms, Hirsch, and Sokol
April 1987 Am J Obstet Gynecol
Table IV. Neonatal morbidity among those with arrest disorders according to the use of oxytocin stimulation Oxytocin stimulation (N = 285) (n as% of N)
5 min Apgar score <7 Nursery intensive care unit Neurological abnormality
(0.7) (15.1) 7 (2.5)
2 43
No oxytocin stimulation (N = 308) (n as% of N)
7 24 8
(2.3) (7.8) (2.6)
Relative risk*
0.3 2.1 0.9
(0.1-1.5) (l.2-3.6)t (0.3-2.7)
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
Table V. Neonatal morbidity according to mode of delivery, stratified by the occurrence of arrest disorders Cesarean birth (n as% of N)
Arrest disorder (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality No arrest disorder (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality Total (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality
79 I 16 3 120 8 48 3 199 9 64 6
( 1.3) (20.3) (3.8) (6.7) (40.0) (2.5) (4.5) (32.2) (3.0)
Vaginal birth (n as% of N)
514 8 (1.6) 51 (9.9) 12 (2.3) 3860 29 (0.8) 290 (7.5) 57 (1.5) 4374 37 (0.8) 341 (7.8) 69 (l.6)
Relative risk*
0.8 2.3 1.7
(0.1-6.9) ( l.2-4.3)t (0.4-6.1)
9.4 8.2 1.7
(4. l-21.5)t (5.5-12. I )t (0.5-5.7)
5.6 5.6 1.9
(2.6-l l.8)t (4.l-7.8)t (0.8-4.6)
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
ied, which in turn depends on exclusions, diagnostic criteria, and grouping. Exclusion criteria are similar in the two studies; Friedman limited analysis to singleton, vertex, term births of infants with birth weights >2500 gm and corrected perinatal mortality rates for major congenital malformations. However, cesarean birth, excluded from Friedman's study, was included in an attempt to include patients similar to those managed with midforceps delivery in the earlier study. Diagnostic criteria and grouping are identical in the two studies. The comparability of methodology is supported by similar frequencies of resolution without oxytocin or operative intervention; Friedman9 reported this occurred in about half of arrest disorders, a finding confirmed in this study. The advantage of comparability with previous work should be accompanied, however, by an understanding of limitations inherent in the design of both studies. The use of relatively rare events as outcome measures requires relatively large numbers of patients and makes it impractical to address many issues of interest. For example, since more than one type of arrest disorder can occur in a single labor, arrest disorders represent up to seven distinct subgroups (three with a single type of arrest disorder and four with various
combinations). Meaningful statistical analysis for each subgroup is precluded by the limitations of sample size in both studies. In addition to statistical considerations, there are practical limitations of collecting and analyzing detailed data in large studies. Neither of these studies, nor any other single study of which we are aware, has attempted detailed analysis of oxytocin administration (dosage and duration) or other medical alternatives (amniotomy, ambulation, abatement, and expectant management) as well as labor pattern, method of delivery, and perinatal outcome. Also, it should be understood that despite efforts to assure comparability, the results of alternative managements can be evaluated more accurately when it is possible to perform a prospective randomized trial. Friedman's'· 9 earlier study detected significant increases in perinatal mortality and depressed Apgar scores in association with arrest disorders of labor. This finding was not confirmed in the current study, despite the larger sample size and inclusion of cesarean births. This apparent change in the perinatal prognosis of arrest disorders represents an improvement in both relative and absolute terms. Explanations for the overall improvement in perinatal prognosis are speculative but might include changes in general health, prenatal care,
Volume 156 Number 4
and neonatal care. However, these general changes provide little explanation for the improvement in perinatal prognosis of pregnancies with arrest disorders relative to pregnancies without this complication; changes in intrapartum management would seem to be a more likely explanation for this relative improvement. These changes include technical advances such as electronic fetal monitoring, fetal scalp sampling, and continuous infusion pumps for oxytocin, as well as the medical management approach used in this study. Our study cannot determine the relative importance of each of these factors. Thus we do not imply that previous studies were in error, that the risks of arrest disorders without appropriate obstetric care have been reduced, that all potential risks of arrest disorders (e.g., longterm neonatal neurological morbidity) have been eliminated, or that all of the improvements in perinatal prognosis can be attributed to medical management. Our findings, however, do suggest that medical management, as used in this study, is associated with a high degree of fetal safety as compared with older treatment protocols. The apparent advantage of medical management is the reduced need for operative intervention. Friedman' previously estimated the need for cesarean birth in arrest disorders to be approximately 50%. Only 18% of patients with arrest disorders in this study had midforceps or cesarean delivery. Perhaps the most important factor in the reduction of the operative delivery rate was the exclusion of radiopelvimetry from the management protocol. Older protocols that relied on radiopelvimetry resulted in as many as 37% of patients with arrest disorders having immediate cesarean delivery for the indication of cephalopelvic disproportion.' Several recent studies have suggested that radiopelvimetry has little predictive value and is rarely indicated in the management of labors with vertex presentation.'°· 11 In the current study, those patients who might have been diagnosed as having cephalopelvic disproportion on the basis of radiopelvimetry and treated with immediate cesarean section were managed medically. As might be expected from recent studies of radiopelvimetry, the subsequent course of labor for all patients receiving medical management in the current study was no different than that Friedman' described for a group of patients with an adequate pelvis by radiopelvimetry. Fifty percent of arrest disorders resolved without use of oxytocin, and 71 % of arrest disorders in which oxytocin stimulation was employed resulted in spontaneous or low forceps delivery. Groups led by O'Driscoll, Studd, Philpott, Barber, Sokol, Bottoms, Fine, and others•·'· 11 • 15 have previously espoused a medical approach for the management of
Medical management of arrest of labor 939
dystocia. Variation in terminology, treatment criteria, cultural background, ethnicity, and socioeconomic status as well as questions about the safety of oxytocin stimulation of labor may have led some to question the general applicability of these studies to clinical practice. However, in the current study, even when cesarean delivery was reserved for only the most severe or otherwise complicated cases of arrest disorders, the neonates fared as well as those of patients without arrest disorders and significantly better than neonates delivered by cesarean section for other indications. In summary, these findings suggest that with few exceptions, medical management, including oxytocin stimulation, is indicated before operative intervention with midforceps or cesarean delivery in otherwise uncomplicated arrest disorders of labor. REFERENCES I. Bottoms SF, Rosen MG, Sokol RJ. The increase in cesarean birth rate. N Engl J Med l 980;302:559. 2. Cesarean childbirth. Washington, DC: US Government Printing Office, 1982; publication no. 1982-0-522025/5510; NIH publication no. 82-2067. 3. Friedman EA. Labor: clinical evaluation and management. 2nd ed. New York: Appleton-Century-Crofts, 1978. 4. Bottoms SF, Rosen MG, Sokol RJ. Short arrest of cervical dilatation: a risk for maternal/fetal/infant morbidity. AM j 0BSTET GYNECOL 1981; 140: 108. 5. Sokol RJ, Stojkov BS, Chik L, Rosen MG. Normal and abnormal labor progress: I. A quantitative assessment and survey of the literature. J Reprod Med 1977;18:47. 6. Sokol RJ, Nussbaum RS, Chik L, Rosen MG. Computer diagnosis of labor progression: II. Application of an online interactive program in 45 high-risk labors. J Reprod Med 1973;11:154. 7. Sokol Rj, Nussbaum RS, Chik L, Rosen MG. Computer diagnosis of labor progression: V. Reliability of a subroutine for evaluating station and descent of the fetal presenting part. J Reprod Med 1974; 13: 183. 8. Armitage P. Statistical methods in medical research. New York: Wiley, 1971. 9. Friedman EA. Patterns of labor as indicators of risk. Clin Obstet Gynecol 1973;16:172. 10. O'Brien WF, Cefalo RC. Evaluation of x-ray pelvimetry and abnormal labor. Clin Obstet Gynecol 1982;25: 157. 11. Fine EA, Bracken M, Berkowitz RI. An evaluation of the usefulness of x-ray pelvimetry: comparison of the Thoms and modified Ball methods with manual pelvimetry. AM j 0BSTET GYNECOL 1980;137:15. 12. O'Driscoll K,Jackson RJA, Gallagher JT. Active management of labour and cephalopelvic disproportion. J Obstet Gynaecol Br Commonw 1970;77:385. 13. Studd J, Clegg DR, Sanders RR, Hughes AO. Identification of high risk labours by labour nomogram. Br Med J 1975;2:545. 14. Philpott RH. Cervicographs in the management oflabour in primigravidae. II. The action line and treatment of abnormal labour.] Obstet Gynaecol Br Commonw 1972; 79:599. 15. Barber HRK, Graber EA, Orlando A. Augmented labor. Obstet Gynecol l 972;39:933. 16. O'Driscoll K, Foley M. Correlation of decrease in perinatal mortality and increase in cesarean section rates. Obstet Gynecol 1983;6 l: I.
Key words: Arrest of labor, oxytocin, cesarean section, neonatal prognosis There have been many changes in intrapartum management during the last 15 years. These changes include the advent of electronic fetal monitoring, declining midforceps use, and less reliance on radiopelvimetry. Concomitantly, there has been a threefold increase in cesarean birth rates. Approximately one third of this increase is attributable to operations performed for dystocia (i.e., abnormal labor progress).1. 2 The need for studies of alternative management of dystocia is apparent. Arrest disorders of labor represent severe dystocia. Their associations with increased frequencies of operative delivery, perinatal mortality, neonatal morbidity, and neurologic deficits have been well documented.' However, most of the studies reporting the perinatal risks associated with arrest disorders of labor predated the changes in intrapartum management outlined above. We have previously reported unexpectedly favorable perinatal outcomes with secondary arrest of dilatation, and noted that cesarean rates for dystocia at Cleveland Metropolitan General Hospital are less than half of those reported for national data.4 These findings stimulated a more detailed appraisal of the management of arrest disorders at our institution. We describe this From the Department of Obstetrics and Gynecology, Cleveland Metropolitan General Hospital, Case Western Reserve University. Supported in part by the Perinatal Clinical Research Unit, Cleveland Metropolitan General Hospital, Case Western Reserve University, grant No. MOJRR00210 from the Division ofResearch Resources, National Institutes of Health. Received for publication April I, 1986; revised October 20, 1986; accepted October 30, 1986. · Reprint requests: Sidney F. Bottoms, M.D., the Department of Obstetrics and Gynecology, Hutzel Hospital, Wayne State University, 4707 Saint Antoine Blvd., Detroit, Ml 48201.
management as medical since we rarely perform cesarean or midforceps delivery for arrest disorders unless medical management, including oxytocin stimulation, has failed. The goal of this study was to provide a current overview of the efficacy of medical management in reducing the risks associated with arrest disorders in a format allowing comparison with previous studies. In order to evaluate the risks specifically attributable to arrest disorders, patients with obstetric complications clearly not caused by arrest disorders were excluded from study. Obstetric management and perinatal outcomes were then compared for patients with and without an arrest disorder. It should be recognized at the outset that this study cannot hope to address every issue of interest related to arrest disorders and their management.
Material and methods Detailed antepartum, intrapartum, and neonatal information was routinely obtained and coded prospectively into a computerized patient information system at Cleveland Metropolitan General Hospital. This study is based on 7550 consecutive births between 1976 and 1979, excluding antepartum stillbirths. Exclusions, made to minimize confounding effects of risks other than arrest disorders, are detailed in Table I. Oxytocin was reserved for significant medical or obstetric complications, which, in turn, might affect perinatal outcome irrespective of labor pattern. Therefore all oxytocin induction was excluded, and oxytocin augmentation of spontaneous labor was excluded except for treatment of arrest disorders. Only the most obvious major congenital anomalies (those recognized in the delivery room) were excluded. Labor diagnoses were assigned prospectively by 935
936
Bottoms, Hirsch, and Sokol
April 1987 Am J Obstet Gynecol
Table I. Selection of the study group from 7550 consecutive births (excluding antepartum stillbirths) Exclusion
n
%
Oxytocin stimulation* Low birth weight (<2500 gm) Previous cesarean section Malpresentation Multiple pregnancy Incomplete data Obvious major anomaly Neonatal deatht Intrapartum fetal deatht
1866 1055 440 338 136 51 45 9 0
24.7 14.0 5.8 4.5 1.8 0.7 0.6 0.0 0.0
2977 4573
39.4 60.6
Total excluded Total studied
*All oxytocin inductions and all oxytocin augmentations except treatment of arrest disorders were excluded. tLimited to those without other exclusion criteria.
means of a computer program developed in our laboratory. Validation of this program with visual graphic analysis of labor has been reported previously. 5 · 6 The diagnoses classified as arrest disorders were secondary arrest of dilatation (no progression in dilatation for ~2 hours during active phase of labor), prolonged deceleration phase (deceleration phase duration ~3 hours for nulliparas or :;;,, I hour for multiparas), and arrest of descent (no progression in descent :;;,, I hour for nullipara or ;;.0.5 hour for multiparas during deceleration phase or second stage). 7 Although individualized, management of arrest disorders generally followed a common approach. Ambulation was employed selectively, particularly in multiparous patients with poor contraction patterns. Typically, when abatement was used, the arrest disorder appeared to be related to administration of analgesics. When oxytocin stimulation of labor was planned, our policy was to use internal electronic fetal monitoring for both fetal heart rate and uterine contractions. If the fetal membranes were intact, external electronic fetal monitoring was temporarily employed until the vertex was well applied to a cervix sufficiently dilated to permit safe amniotomy. Many arrest disorders resolved during the interval when monitoring was initiated and/or amniotomy was performed; oxytocin was not needed in these cases. Oyxtocin was administered intravenously as a dilute solution with a constant infusion pump, at an initial dosage of I to 2 mU of oxytocin per minute. Fetal scalp sampling was utilized frequently when the diagnosis of fetal distress was made. Apgar scores were assigned by nurse or pediatrician. Neurological abnormality was assessed by the pediatric staff as part of the routine neonatal physical examination performed at time of discharge from the hospital. Any deviation from the normal neurological eval-
uation, however minor, was classified as a neurological abnormality for purposes of this study. Results were analyzed on the bases of approximate relative risks and 95% confidence intervals. 8 Results
The study group remaining after the exclusions outlined in Table I was at low risk for operative intervention and adverse perinatal outcomes. For example, 73% of the cesarean births and 79% of infants with 5-minute Apgar scores <7 were excluded. Among intrapartum and neonatal deaths, 93% had additional exclusion criteria. However, 86% of all patients with arrest disorders were in the study group. Of patients in the study group, 13% had one or more arrest disorders; among those excluded, the frequency was 3%; overall, the frequency was 9%. lntrapartum management is outlined in Fig. 1. Radiopelvimetry was rarely performed (I. 7% of patients with an arrest disorder). Operative intervention with midforceps or cesarean delivery was rarely employed without first attempting oxytocin stimulation of labor. Of the 96% of patients with arrest disorders whose initial management was medical, 50% had resolution of the arrest disorder without the use of oxytocin or midforceps or cesarean delivery. The remaining 50% of patients in this category received oxytocin stimulation. Of those who received oxytocin, 71 % had a spontaneous or low-forceps delivery. Statistical analysis of intrapartum management is reported in Table II. Electronic monitoring and all methods of operative delivery were increased with arrest disorders. Cesarean birth was five times more likely among patients with an arrest disorder than among patients without this complication. During the study period, all intrapartum fetal deaths and all but nine neonatal deaths met other exclusion criteria; none of these nine was associated with an arrest disorder of labor. Analysis of neonatal morbidity related to the effects of arrest disorders is detailed in Table Ill. Among cesarean births, those preceded by arrest disorders were less likely to result in admission of the infant to the nursery intensive care unit. No other significant difference in neonatal morbidity was attributable to arrest disorders. Neonatal morbidity among pregnancies complicated by an arrest disorder was also analyzed according to the use of oxytocin. Admission to the nursery intensive care unit was more common after stimulation with oxytocin (Table IV). However, admission to the nursery intensive care unit was also more common after cesarean birth (Table V). This risk was significantly greater than the risk associated with oxytocin stimulation and was significantly less for those with arrest disorders than
Medical management of arrest of labor 937
Volume 156 Number 4
ARREST DISORDER 593111%)
MIDFORCEPS WITHOUT OXYTOCIN 511%)
MEDICAL MANAGEMENT 569(96%)
CESAREAN WITHOUT OXYTOC IN 19(3%)
OXYTOCIN 285 (50"fo)
NO OXYTOCIN 284 (50%)
I SPONTANEOUS 258191%)
LOW FORCEPS 26 (9%)
SPONTANEOUS 161 (56%)
LOW FORCEPS 43115%)
MIDFORCEPS 2117%)
CESAREAN 60121%)
Fig. I. Management of arrest disorders in the study group. Numbers of births are indicated below each category of management. Percentages refer to the proportion of the larger category immediately above it in the figure.
Table II. lntrapartum management according to the occurrence of arrest disorders
Management
Arrest disorder (N = 593) (n as% of N)
No arrest disorder (N = 3980) (n as% of N)
Relative risk*
Electronic monitor Low-forceps Mid forceps Cesarean section
398 (67.1) 69 (11.6) 26 (4.4) 79 (13.3)
1,085 (27.3) 293 (7.4) 35 (0.9) 120 (3.0)
5.4 (4.5-6.6)t 1.7 (l.2-2.2)t 5.2 (3.l-8.7)t 4.9 (3.6-6.7)t
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
Table III. Neonatal morbidity according to the occurrence of arrest disorders, stratified by mode of delivery Arrest disorder (n as% of N)
Cesarean birth (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality Vaginal birth (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality Total (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality
79 I 16 3 514 8 51 12 593 9 67 15
(1.3) (20.3) (3.8) (1.6) (9.9) (2.3) (1.5) (11.3) (2.5)
No arrest disorder (n as% of N)
120 8 48 3 3860 29 290 57 3980 37 338 60
Relative risk*
(6.7) (40.0) (2.5)
0.2 (0.0-1.5) 0.4 (0.2-0.7)t 1.5 (0.3-8. I)
(0.8) (7.5) (1.5)
2.1 1.4 1.6
(0.9) (8.5) (1.5)
1.6 (0.8-3.5) 1.4 (l.0-1.8) 1.7 (0.9-3.0)
(0.9-4.7) ( 1.0-1.9) (0.8-3.0)
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
for those without arrest disorders. Low Apgar scores were more frequent only among cesarean births not preceded by arrest disorders.
Comment The increasing use of cesarean delivery has pointed to the need for studies of contemporary management of dystocia. Our findings suggest that medical management is a safe and effective alternative to immediate
operative intervention in almost all cases of arrest disorders of labor meeting the criteria of this study. Details of the basis for this interpretation follow, after consideration is given to the advantages and limitations of a study of this design. The methods used in this study were chosen, in part, to facilitate comparison with Friedman's'· 9 analysis of the prognosis of arrest disorders. The validity of this comparison depends on similarity of the patients stud-
938
Bottoms, Hirsch, and Sokol
April 1987 Am J Obstet Gynecol
Table IV. Neonatal morbidity among those with arrest disorders according to the use of oxytocin stimulation Oxytocin stimulation (N = 285) (n as% of N)
5 min Apgar score <7 Nursery intensive care unit Neurological abnormality
(0.7) (15.1) 7 (2.5)
2 43
No oxytocin stimulation (N = 308) (n as% of N)
7 24 8
(2.3) (7.8) (2.6)
Relative risk*
0.3 2.1 0.9
(0.1-1.5) (l.2-3.6)t (0.3-2.7)
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
Table V. Neonatal morbidity according to mode of delivery, stratified by the occurrence of arrest disorders Cesarean birth (n as% of N)
Arrest disorder (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality No arrest disorder (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality Total (N) 5 min Apgar score <7 Nursery intensive care unit Neurological abnormality
79 I 16 3 120 8 48 3 199 9 64 6
( 1.3) (20.3) (3.8) (6.7) (40.0) (2.5) (4.5) (32.2) (3.0)
Vaginal birth (n as% of N)
514 8 (1.6) 51 (9.9) 12 (2.3) 3860 29 (0.8) 290 (7.5) 57 (1.5) 4374 37 (0.8) 341 (7.8) 69 (l.6)
Relative risk*
0.8 2.3 1.7
(0.1-6.9) ( l.2-4.3)t (0.4-6.1)
9.4 8.2 1.7
(4. l-21.5)t (5.5-12. I )t (0.5-5.7)
5.6 5.6 1.9
(2.6-l l.8)t (4.l-7.8)t (0.8-4.6)
*Approximate relative risks are presented with 95% confidence intervals indicated in parentheses. tp < 0.05 since 95% confidence interval excludes 1.0.
ied, which in turn depends on exclusions, diagnostic criteria, and grouping. Exclusion criteria are similar in the two studies; Friedman limited analysis to singleton, vertex, term births of infants with birth weights >2500 gm and corrected perinatal mortality rates for major congenital malformations. However, cesarean birth, excluded from Friedman's study, was included in an attempt to include patients similar to those managed with midforceps delivery in the earlier study. Diagnostic criteria and grouping are identical in the two studies. The comparability of methodology is supported by similar frequencies of resolution without oxytocin or operative intervention; Friedman9 reported this occurred in about half of arrest disorders, a finding confirmed in this study. The advantage of comparability with previous work should be accompanied, however, by an understanding of limitations inherent in the design of both studies. The use of relatively rare events as outcome measures requires relatively large numbers of patients and makes it impractical to address many issues of interest. For example, since more than one type of arrest disorder can occur in a single labor, arrest disorders represent up to seven distinct subgroups (three with a single type of arrest disorder and four with various
combinations). Meaningful statistical analysis for each subgroup is precluded by the limitations of sample size in both studies. In addition to statistical considerations, there are practical limitations of collecting and analyzing detailed data in large studies. Neither of these studies, nor any other single study of which we are aware, has attempted detailed analysis of oxytocin administration (dosage and duration) or other medical alternatives (amniotomy, ambulation, abatement, and expectant management) as well as labor pattern, method of delivery, and perinatal outcome. Also, it should be understood that despite efforts to assure comparability, the results of alternative managements can be evaluated more accurately when it is possible to perform a prospective randomized trial. Friedman's'· 9 earlier study detected significant increases in perinatal mortality and depressed Apgar scores in association with arrest disorders of labor. This finding was not confirmed in the current study, despite the larger sample size and inclusion of cesarean births. This apparent change in the perinatal prognosis of arrest disorders represents an improvement in both relative and absolute terms. Explanations for the overall improvement in perinatal prognosis are speculative but might include changes in general health, prenatal care,
Volume 156 Number 4
and neonatal care. However, these general changes provide little explanation for the improvement in perinatal prognosis of pregnancies with arrest disorders relative to pregnancies without this complication; changes in intrapartum management would seem to be a more likely explanation for this relative improvement. These changes include technical advances such as electronic fetal monitoring, fetal scalp sampling, and continuous infusion pumps for oxytocin, as well as the medical management approach used in this study. Our study cannot determine the relative importance of each of these factors. Thus we do not imply that previous studies were in error, that the risks of arrest disorders without appropriate obstetric care have been reduced, that all potential risks of arrest disorders (e.g., longterm neonatal neurological morbidity) have been eliminated, or that all of the improvements in perinatal prognosis can be attributed to medical management. Our findings, however, do suggest that medical management, as used in this study, is associated with a high degree of fetal safety as compared with older treatment protocols. The apparent advantage of medical management is the reduced need for operative intervention. Friedman' previously estimated the need for cesarean birth in arrest disorders to be approximately 50%. Only 18% of patients with arrest disorders in this study had midforceps or cesarean delivery. Perhaps the most important factor in the reduction of the operative delivery rate was the exclusion of radiopelvimetry from the management protocol. Older protocols that relied on radiopelvimetry resulted in as many as 37% of patients with arrest disorders having immediate cesarean delivery for the indication of cephalopelvic disproportion.' Several recent studies have suggested that radiopelvimetry has little predictive value and is rarely indicated in the management of labors with vertex presentation.'°· 11 In the current study, those patients who might have been diagnosed as having cephalopelvic disproportion on the basis of radiopelvimetry and treated with immediate cesarean section were managed medically. As might be expected from recent studies of radiopelvimetry, the subsequent course of labor for all patients receiving medical management in the current study was no different than that Friedman' described for a group of patients with an adequate pelvis by radiopelvimetry. Fifty percent of arrest disorders resolved without use of oxytocin, and 71 % of arrest disorders in which oxytocin stimulation was employed resulted in spontaneous or low forceps delivery. Groups led by O'Driscoll, Studd, Philpott, Barber, Sokol, Bottoms, Fine, and others•·'· 11 • 15 have previously espoused a medical approach for the management of
Medical management of arrest of labor 939
dystocia. Variation in terminology, treatment criteria, cultural background, ethnicity, and socioeconomic status as well as questions about the safety of oxytocin stimulation of labor may have led some to question the general applicability of these studies to clinical practice. However, in the current study, even when cesarean delivery was reserved for only the most severe or otherwise complicated cases of arrest disorders, the neonates fared as well as those of patients without arrest disorders and significantly better than neonates delivered by cesarean section for other indications. In summary, these findings suggest that with few exceptions, medical management, including oxytocin stimulation, is indicated before operative intervention with midforceps or cesarean delivery in otherwise uncomplicated arrest disorders of labor. REFERENCES I. Bottoms SF, Rosen MG, Sokol RJ. The increase in cesarean birth rate. N Engl J Med l 980;302:559. 2. Cesarean childbirth. Washington, DC: US Government Printing Office, 1982; publication no. 1982-0-522025/5510; NIH publication no. 82-2067. 3. Friedman EA. Labor: clinical evaluation and management. 2nd ed. New York: Appleton-Century-Crofts, 1978. 4. Bottoms SF, Rosen MG, Sokol RJ. Short arrest of cervical dilatation: a risk for maternal/fetal/infant morbidity. AM j 0BSTET GYNECOL 1981; 140: 108. 5. Sokol RJ, Stojkov BS, Chik L, Rosen MG. Normal and abnormal labor progress: I. A quantitative assessment and survey of the literature. J Reprod Med 1977;18:47. 6. Sokol RJ, Nussbaum RS, Chik L, Rosen MG. Computer diagnosis of labor progression: II. Application of an online interactive program in 45 high-risk labors. J Reprod Med 1973;11:154. 7. Sokol Rj, Nussbaum RS, Chik L, Rosen MG. Computer diagnosis of labor progression: V. Reliability of a subroutine for evaluating station and descent of the fetal presenting part. J Reprod Med 1974; 13: 183. 8. Armitage P. Statistical methods in medical research. New York: Wiley, 1971. 9. Friedman EA. Patterns of labor as indicators of risk. Clin Obstet Gynecol 1973;16:172. 10. O'Brien WF, Cefalo RC. Evaluation of x-ray pelvimetry and abnormal labor. Clin Obstet Gynecol 1982;25: 157. 11. Fine EA, Bracken M, Berkowitz RI. An evaluation of the usefulness of x-ray pelvimetry: comparison of the Thoms and modified Ball methods with manual pelvimetry. AM j 0BSTET GYNECOL 1980;137:15. 12. O'Driscoll K,Jackson RJA, Gallagher JT. Active management of labour and cephalopelvic disproportion. J Obstet Gynaecol Br Commonw 1970;77:385. 13. Studd J, Clegg DR, Sanders RR, Hughes AO. Identification of high risk labours by labour nomogram. Br Med J 1975;2:545. 14. Philpott RH. Cervicographs in the management oflabour in primigravidae. II. The action line and treatment of abnormal labour.] Obstet Gynaecol Br Commonw 1972; 79:599. 15. Barber HRK, Graber EA, Orlando A. Augmented labor. Obstet Gynecol l 972;39:933. 16. O'Driscoll K, Foley M. Correlation of decrease in perinatal mortality and increase in cesarean section rates. Obstet Gynecol 1983;6 l: I.