- Email: [email protected]
Fetal breathing movements and the nonstress test in high-risk pregnancies
Fetal breathing movements and the nonstress test in high-risk pregnancies FRANK
A.
LAWRENCE LOUISmE KIRK
MANNING, D.
PLATT,
SIPOS, A.
Los Azgeles,
M.D. M.D.
R.N.
KEEGAN,
JR.,
M.D.
Calijornia
Three hulndred ninety-eight observations of fetal breathing movements (FBM) and the nonstress test (NST) were made in 223 patients. The presence of FBM was noted in a significantly higher proportion of tests with a reactive NST (88%) than a nonreactive NST (67%). Conversely, a significantly higher proportion of tests were reactive when FBM were present (82.6%) than when FBM werje absent (49%). A significant relationship existed between either test and the outcome of pregnancy as judged by the five-minute Apgar score or the incidence of fetal distress in labor. Combining two normal tests did not improve the accuracy in predicting outcome; but the combinatoon of both tests, when abnormal, produced a significant improvement in predicting fetuses likely to heave an abnormal outcome. The combination of the normal with the abnormal test had a predictive accuracy similar to a single normal test in predicting a normal five-minute Apgar. This combination was associated with an incidence of fetal distress in labor intermediate between that seen with either the normal or abnormal test when alone. Neither a single test nor the combination of tests as helpful in identifying the small-for-gestational age (SGA) infants. These data indicate antepartuim fetal evaluation may be improved when more than one biophysical variable is used. (AM. J. O~BSTET. GYNECOL. 135:511, 1979.)
A N A L Y s I s of fetal biophysical variables holds promise as a means of antepartum fetal assessment. Both the association of acceleration of the fetal heart rate with fetal movements (the nonstress test or NST) and the presence of fetal breathing movements (FBM) have been shown in independent studies to be helpful in differentiating the normal from the compromised fetus.‘-” With either test the accuracy is greatest in predicting the normal fetus. Neither test is as accurate in identifying the compromised fetus. From the Depart.ment of Obstetriu and Gpecolog. University of Southern Calijo&a Sclwol of Medicine, and Women’s Hospital, Los Angeles CouxtplJniversity Southem California Medical Center. Supported by a grant from the National Foundation-March of Dimes awarded Manning, M.D. J Received for publication Revised
December
Accepted December
October
to Frank
of
A
13, 1978.
14, 1978.
Methods
28, 1978.
Rep% requests: Dr. Frank A. Manning, M.D., Women’s Hospital, .Room 5K-9, Los Angeles County -University of Southern California Medical Center, 1240 North Mission Rd., Los Angeles, California 90033. 0002.9378/79/200511+05!~00.50/0
Biophysical biorhythms have been demonstrated in animal fetuses and recently in the human fetus.j-’ Both the incidence and character of FBM are related to the sleep state of the fetus.6 Thus, whereas the presence of FBM appears to be a reliable index of fetal well-being, it is hot possible, in a single observation, to determine if the absence of FBM reflects a sleep state or compromise. The relationship of the NST to fetal sleep states is unknown. The similarities of the NST to FBM are striking in that a reactive pattern (normal) NST is an accurate index of Fetal well-being whereas a nonreactive pattern (abnormal) NST is a poor index of fetal compromise.* This study was designed to evaluate the relationship of two fetal biophysical variables, the NST and FBM, to compare the independent predictive accuracy of each test, and to determine if the combination of both tests can improve predictive accuracy.
0
1979
The
C. V. Mosby
Co.
Patients studied were selected at random from the high-risk pregnancies referred for antepartum fetal heart rate testing at Women’s Hospital, LACiUSC 511
Table
I. Srudy population
Post date (>42 Diabetic Class A (7) Class B (46)
weeks)
Class C (9) Class D (6) Class F (2) Siypertensive disease Suspected intrauterine growth retardation Kh isoimmunized pregnancy FJ,emoglobinopathies (SS, SC) Maternal medical disorders Total
76 70
26
18 10 5 18 m
SPedical C&W-. The NST M~S performed using a method previously described.” An NST was termed a reactive pattern (R) or normal if, during a 20 minute period, there were two or more fetal heart rate accelerations of 15 bpm or greater of at least 30 seconds duration associated with fetal movement. The NST was termed a nonreactive pattern (NR) or abnormal if there were less than two accelerations of the fetal heart Ivith fetal movement in a 40 minute period.” The resuits of the SST were used in clinical management. Fetal breathing movements were recorded by means of a real time B-mode ultrasound method (ADR Mode! 2130).” FBM were defined as convergent chest wail movements in conjunction with abdominal wall excursions. FBM were observed immediately before, during. or after the NST. The majority of observations were made immediately after the NST. Patients were scanned until either a period of FBM of 30 seconds or greater was observed or 30 minutes elapsed.” FBti were recorded as either present or absent. The results of FBM observation were not disclosed and hence did not influence clinical management. The results of the NST. FBM, and combined NSTi FBM were correlated with the outcome of pregnancy as reflected by the five-minute Apgar score, the incidence of fetal distress in labor, birth weight, and meconium staining of amniotic fluid. Fetal distress was defined as any abnormal intrapartum fetal heart rate pattern that precipitated emergency delivery of a fetus regardless of the condition of the fetus at delivery. Fetal heart rate patterns considered indicative of fetal distress included repetitive late decelerations, intermittent late decelerations with loss of heart rate variability, unremitting se\-ere variable decelerations, or persistent fetal brady-cardia. infants were considered small for dates if the birth weight was less than the tenth percentile for gestational age and sex.
Three hundred rrinety-eight obser%,c:ions 3f FBM and the NST were made in 223 patients (Table i)~ Most patients delivered within one week of the last test (86%) and all delivered within three weeks. Overall, FBM were present in 83.7% of observations and the NST was reactive in 78.4% (Table II). FBM were present in 88% of observations when the NST was reactive as compared to 67% when the NST wa6 nonreactive (p < 0.05). Conversely: the NST was rcaciive in 82.6% of observations when FBM were present and reactive in 56.9% of observations when FBM were absent (p < 0.05). FBM,
NST,
and
five-minute
Apgar
(,TabXe
III).
A
normal five-minute A.pgar score (~7’) was observed in 179 of 187 patients (95.7%) with FBM present on the last observation before delivery and in 27 of 36 patients (75%) with FBM absent (p < 0.05). One hundred sixty-one of 169 patients (95.2%) with a reactive NST on the last obsenJation delivered an infant with a normai five-minute Apgar score as corAlpared to 15 of 54 (83.3%) with a nonreactive NST (p < 0.03). The predictive accuracy of the NST and FBM was not significantly different. A normal five-minute Apgar score was observed in l46 of 151 patients (96.7%) when both the NST was reactive and FBM were present in the :ast observation before delivery. The predictive accuracy of two normal tests offered no significant improvement over that of either normal test alone. Thirty-five of 36 patients (97.2%) with a nonreactive NST but with FBM present on the last observation delivered an infant with a normal five-minute Apgar score. The predictive accuracy of this combination was similar to that of the presence of FBM alone (96%) but significantlv different from the predictive accuracy of the nonreaciive NST (83.3%) when considered alone (p < 0.05). Seventeen of 18 patients (94.4%) with a reactive NST but with FEM absent on the last observation delivered an infant with a normal five-minute Apgar score. The predictive accuracy of this combination was similar to that of a reactive NST alone (95.2%) but significantly different from the predictive accuracy of the absence of FBM (75%) (p < 0.05). Finally, 8 of 18 patients (44.3%) with both a nonreactive NST and absent FBM on the last observation delivered infants with abnormal five-minute Apgar scores (<7). This combination was associated with a significantly higher incidence of abnormal five-minute Apgar scores than with any single test or combination of tests (p < 0.01). FBM, (Table
NST, IV).
and
the
Spontaneous
incidence
of
or induced
iabor occurred
fetal
distress
in
Volume Number
135 4
Table
!I.
FBM
Observations
333 (83.7%)
Absent
65 (16.3%)
Total
Table
R NR R NR
275 58 37 28
NST
(82.6%) (17.4%) (56.9%) (43.1%)
No. oft&s
Reactive
NSTIFBM
Nonreactive
and five-minute
Apgar
FBM
test*
FBM FBM FBM FBM
86 (21.6%)
513
+ + -
275 37 58 28
(88%) (12%) (67%) (33%)
398
score 5 Minute
Antepartum
pregnancies
FBM
3 12 (78.4%)
398
III.
in high-risk
on FBM and NST NST
Present
and NST
No. of patients
Apgar , qIIIpr-
27
alone
FBM present FBM absent Total NST alone NST - R. NST - MR
Total NSTIFBM combinsd NST-RIFBM-P NST-RIFBM-A.
NST-NRIFBM-P NST-NRIFBM-A
Total
187 36 223
179 (95.7%) 27 (75%)
8 (4.3%) 9 (25%)
co.05
169 54 223
161 (95.2%) 45 (83.3%)
8 (4.8%) 9 (1617%)
co.05
151 1‘8 36 18 223
146 17 35 10
5 1 1 8
(96.7%) (94.4%) (97.2%) (55.5%)
(3.3%) (5.6%) (2.8%) (45.5%)
NS P < 0.051 p
P<0.05?
*Last test before delivery. tAs compared to abnormal test. 184 of 223 patients studied (82.5%). The remaining 39 patients were delivered by cesarean section before labor. Fetal distress in labor was observed in 11 of 153 patients (7.2%) with FBIM present on the last test before delivery and in 11 of 31 patients (35.5%) with FBM absent (p < 0.05). C:omparatively, fetal distress was present in 7 of 137 patients (5.1%) when the last NST was reactive and in 15 of 47 patients (31.9%) when the last NST was nonreactive (p < 0.05). There was no significant difference in the predictive accuracy of fetal distress in labor between the two tests. Fetal distress in labbr occurred in 5 of 121 patients (4.1%) in whom both a reactive NST and the presence of FBM were observed in the last test before delivery. This combination offerred no significant improvement in predictive accuracy of fetal distress.when compared to the reactive ?JST or the presence of FBM alone. Six of 32 patients (18.7%) with a last test that revealed a nonreactive KSIT but with FBM present developed fetal distress in labor. The frequency of fetal distress with this combination was significantly greater than the frequency observed with the presence qf FBM alone (p < 0.05). Two of 16 patients (12.5%) with a reactive NST but with FBM absent developed fetal distress in
Table
IV.
FBM/NST
Antepartum
test*
and mode of delivery
/ ;;e:!
K
FBM alone FBM-P FBM-A
153 31
Total
184
NST alone NST-R NST-NR
137 47
Total Combined
142 (92.8%) 20 (64.5%) 162 (88.0%)
184
7 (5.1%) 15 (31.9%)t 22 (12.0%)
13p (94.9%) 32 (68.1%) 162 (S+O%)
121
5 (4.1%)
NSTIFBM
NST-RIFBM-P NST-RIFBM-A
NST-NRIFBM-P NST-NRIFBM-A
Total
11 (7.2%) 11 (35.5%)t 22 (12.0%)
16 32 15
184
2 6 9 22
(12.5%)$ (18.8%)$ (60.0%)$ (12.0%)
116 (95.9%) 14 26 6 162
(87.5%) (81.2%) (40.0%) (88.0%)
*Last test before delivery. tp 0.05. $p 0.05 as compared to the normal test alone. labor. This combination had a significantly higher incidence of fetal distress than with a reactive NST alone (p < 0.05) Fisher exact test). Finally. 9 of 15 patients (60%) with both a nonreactive NST and the absence of FBM on the last test developed fetal distress. The incidence of fetal distress with this combination was sig-
Table
Y. PGS~TiFBM and fetal weight
and meconium
FBt%I crlon~
Present Absent XST ulonr
Reactive Nonreactive FB.\II,VST combined YST-RIFBM-P NST-RIFBM-A NST-KRIFBM-P NST-KRIFBM-A
187 36
i83 (98.0%)
4 (2.0%)
30 (83.3%)
6 (16.6%)
175 (93.6%) 33 (91.6%)
12 (8.4%) 3 (8.4%)
169 54
167 (Y9.W) 48 (88.9%)
4 (2.3%) 6 (ll.i%)
159 (Y4.0%) 49 (91.0%)
10 (6.0%) .5 (Y.O%)
I51 18 36 18
149 16 34 14
2 2 2 4
143 16 32 17
.+&A = Appropriate for gestational age, %A “Last test betixe delivery.
(98.6%) (88.8%) (94.4%) (77.7%)
(1.4%) (11.2%) (5.6%) (22.3%)
(Y4.7R) (88.8%) (88.8%) (94.4%)
8 (L3%j 2 (I 1.2%) 4 (11.2%) 1 (5.6rCj
= small for gestationai age.
nilicantly greater than with any single iest aione or combination of tests (p < 0.0 I). ha, NST, and birth weight (Table Y). Te11 of 223 fetuses studied (4.5%) weighed less than the tenth percentile for gestational age. FBM were present on the East test in four of these SGA infants and absent in six. Similarly, in four of these patients the last SST was reactive and in six it was nonreactive. Combined KST and FRM results were not helpful in specifically identifying the SGA infant since in two fetuses both tests Lvere normal, in two the NST was reactive but with FBM absent, in two the N’ST Evasnonreactive with FBM present, an in four both tests were abnormal. Nonetheless. in 8 of 10 SGA fetuses (80%) at least one test was abnormal. Further, in both SGA fetuses that r+ere delxessed at delivery both tests were abnormal. FBM, NST, and meconium-stained amniotic fluid (Table V). Meconium was present in 15 of 223 patients studied (6.7%). No apparent relationship between the incidence of meconium and either test or combination of tests was observed.
In separate studies both the NST and FBM have been sho\vn to be useful methods for antepartum fetal evaluation.1-4 In this studv where we have measured bot variables in the same patient we note remarkable similarities between the predictive accuracy of either test. When considered independently, the accuracy of’ a rlortnal test in predicting a normal outcome was statistically better than Teas the abnormal test in predicting an abnormal outcome. Further, the combination of two normal tests did not improve the predictive accuracy as compared to either normal test alone. Conversely, the combination of tv,w abnormal tests produced a high]) significant improvement in the predictive accuracy of an abnormal outcome as compared to a single abnormai tes!. The relationship of a normal and abnormal
combination of tes& w outcome is i~rigrui~lg. I‘he predictive accuracy of a normal five-iniltute Apgai score with this combination was similar io the accuracy of the normal test alone, but significai:tiy different from that of the abnormal test alone. Howe\-er, the incidence of fetal distress in labor with the combination of a normal and abnormal test was intermediate, that is. significantly less than that observed with a single abnorma! test alone but significantIy greater than that observed with a single normal test alone. It is apparent that neither tear is helpful in speciticallv identifying the SGA infant. In experimental models chronic placental insufficieq of sufficient magnitude to impair fetal growth is uswily associated with normal arterial pH and a small fall in arrerial pO,.’ A similar mechanism may help :o explain the relatively high frequency of at least one abnormal biophysical test in the SCA fetuses studied, The relationship betlveen meconium passage and fetal distress is weak as the majority of‘i‘etuses that pass meconium are not acidemic.” Hence it is not surprising rhat no relationship hetween either antepartum test and the incidence of meconium was noted. Overall, these data appear to indicate rhat the normal biophysical test is a more reliable index of fetal health than is the abnormal test of fetai compromise. The explanation for this observation ma)- he related to the mechanisms responsible for the control of fetal biophysical variables. Fetal breathing movements, for example, are not random but rather reflect complex integrated neurologic controls. In the normal fetus. breathing movements are intermittent \t-ith periods of apnea of up to 108 minutes.” The periodic+ of FBM is at least in part due to sleep-wake cycles of the fetus.” Depression of the fetal centrai nervous system b> hypoxemia, acidemia, hypoglycemia, or dl,ugs can reduce or abolish FB?JI.~. lo Thus, whereas the preser;ce of FBM suggests an intact fetal central nervous system,
VolumeNumber
135 -1
FBM
the absence of FBM may reflect either normal periodicity or central nervous system depression. Continuous or intermittent FBM have been reported in the asphyxiated fetal lamb before death. However, to date, we have not observed either pattern before death in the asphyxiated primate’” or in the asphyxiated or dying human fetus.r5 C’ntil recently, biophysical monitoring of the fetus has been largely restricted to heart rate analysis. The development of real time B-mode ultrasound methods
and NST
in high-risk
pregnancies
515
now enables the physician to assess additional fetal biophysical variables. These data indicate that anteparturn fetal evaluation may be improved when two biophysical variables are considered. The authors wish to thank Marcie Smith, R.N., Paula Broussard, R.N., and Dorothy McCart, R.N., for technical support. In addition we wish to thank Drs. E. J. Quilligan and R. H. Paul for their critical review of and comments on this manuscript.
REFERENCES
1. Rochard, F., Schifrin, B. S., Sureau, C., et al.: Non-stress cardiotachometry ftor antepartum fetal evaluation, Obstet. Gynecol. 451433. 1975: 2. Martin. C, B.. and Schifrin, B. S.: In, Aladiem, S., and Brown, A. K., editors: Prenatal Fetal Monitoring in Perinatal Intensive Care, St. Louis, Missouri, 1977, The C. V. Mosby Comlany, p. 155. 3. Manning, F. A.: Fetal breathing movements as a reflection of fetal status, Postgrad. Med. 61:116, 1977. 4. Platt. L. D., Manning, F. A., Lemay, M., and Sipos, L.: Fetal breathing movements and outcome of pregnancy, AM.
J. OBSTET‘:GYMECOL.
132:542,
1978.
_
9.
a
10.
11.
1
5. Patrick. 1.. Fetherson. W.. Vick, H., and Voeeelin. R.: Human Jfe;al breathing movements and gross f&l body movements at weeks 34 to 35 of gestation, AM. J. OBSTET. GYNECOL. 130:693, 1978. 6. Boddv, K.. and Dawes, G. S.: Fetal breathing, Br. Med. Bull. 31:3,’ 1975. 7. Boddy, K., Dawes, G. S., and Robinson, J. S.: A 24 hour rhvthm in the foetus. in Comline. R. S.. Cross, K., Dawes, G./S., and Nathan&z. P., editors: Foetal and Neonatal Physiology, Proceedings of Sir Joseph Barcroft Centenary Symposium, Cambridge, 1973, Cambridge University Press, p. 63. 8. Creasy, R. K., DeSwiet, M., Kahanpaa, K. V., Young, W. P., and Rudolph, A. M.: Pathophysiological changes in the foetal lamb with growth retardation, in Comline, R. S., Cross, K., Dawes,2. S., and Nathanielsz, P., editors:
Fetal and Neonatal Physiology, Proceedings of Sir Joseph Barcroft Centenarv Svmoosium. Cambridge. 1973, Cambridge University Press, ‘p, 398.’ ” Miller. F. C., Sacks, D. A., Yeh, S., Paul, R. H., Schifrin, B. S,, Martin, C. B., and Hon, E. H.: Significance of meconium in labor. AM. 1. OBSTET. GYNECOL. 122:573, 1975. Manning, F. A., Platt, L. D., and Lemay, M.: Maternal hypoxemia and fetal breathing movements, Obstet. Gynecol. In press, 1978. Evertson, L., Gauthi’er, R., Keegan, K., Paul, R. H., Schifrin, B. S., and Martin, C. B.: Evolution of the nonstress test, AM. J. OBSTET. GYNECOL. 133:29, 1979. Patrick, J. E., Dalton, K. J., and Dawes, G. S.: Breathing movements before death in fetal lambs, AM. J. OBSTET. GYNECOL. 125:73, 1976. Chapman, R. L. K., Dawes, G. S., Rurak, D. W., and Wilds, P. L.: Intermittent breathing before death in fetal lambs, AM. J. OBSTET. GYNECOL. 131:894, 1978. Manning, F. A., Martin, C. B., Jr., Murata, Y., Miyaki, K., and Danzler, G.: Breathing movements before death in the primate fetus (IZlacaca mulatta), AM. J. OBSTET.
12.
13.
14.
GYNECOL.
135:71,
1979.
F. A., and Platt, L. D.: Fetal breathing move15. Manning, ments: Antepartum monitoring of fetal conditions, in Quilligan, E. J., editor: Clinics in Obstetrics and Gynecology, vol. 6, London, 1979, W. B. Saunders Company, Ltd., p. 335.
A.
LAWRENCE LOUISmE KIRK
MANNING, D.
PLATT,
SIPOS, A.
Los Azgeles,
M.D. M.D.
R.N.
KEEGAN,
JR.,
M.D.
Calijornia
Three hulndred ninety-eight observations of fetal breathing movements (FBM) and the nonstress test (NST) were made in 223 patients. The presence of FBM was noted in a significantly higher proportion of tests with a reactive NST (88%) than a nonreactive NST (67%). Conversely, a significantly higher proportion of tests were reactive when FBM were present (82.6%) than when FBM werje absent (49%). A significant relationship existed between either test and the outcome of pregnancy as judged by the five-minute Apgar score or the incidence of fetal distress in labor. Combining two normal tests did not improve the accuracy in predicting outcome; but the combinatoon of both tests, when abnormal, produced a significant improvement in predicting fetuses likely to heave an abnormal outcome. The combination of the normal with the abnormal test had a predictive accuracy similar to a single normal test in predicting a normal five-minute Apgar. This combination was associated with an incidence of fetal distress in labor intermediate between that seen with either the normal or abnormal test when alone. Neither a single test nor the combination of tests as helpful in identifying the small-for-gestational age (SGA) infants. These data indicate antepartuim fetal evaluation may be improved when more than one biophysical variable is used. (AM. J. O~BSTET. GYNECOL. 135:511, 1979.)
A N A L Y s I s of fetal biophysical variables holds promise as a means of antepartum fetal assessment. Both the association of acceleration of the fetal heart rate with fetal movements (the nonstress test or NST) and the presence of fetal breathing movements (FBM) have been shown in independent studies to be helpful in differentiating the normal from the compromised fetus.‘-” With either test the accuracy is greatest in predicting the normal fetus. Neither test is as accurate in identifying the compromised fetus. From the Depart.ment of Obstetriu and Gpecolog. University of Southern Calijo&a Sclwol of Medicine, and Women’s Hospital, Los Angeles CouxtplJniversity Southem California Medical Center. Supported by a grant from the National Foundation-March of Dimes awarded Manning, M.D. J Received for publication Revised
December
Accepted December
October
to Frank
of
A
13, 1978.
14, 1978.
Methods
28, 1978.
Rep% requests: Dr. Frank A. Manning, M.D., Women’s Hospital, .Room 5K-9, Los Angeles County -University of Southern California Medical Center, 1240 North Mission Rd., Los Angeles, California 90033. 0002.9378/79/200511+05!~00.50/0
Biophysical biorhythms have been demonstrated in animal fetuses and recently in the human fetus.j-’ Both the incidence and character of FBM are related to the sleep state of the fetus.6 Thus, whereas the presence of FBM appears to be a reliable index of fetal well-being, it is hot possible, in a single observation, to determine if the absence of FBM reflects a sleep state or compromise. The relationship of the NST to fetal sleep states is unknown. The similarities of the NST to FBM are striking in that a reactive pattern (normal) NST is an accurate index of Fetal well-being whereas a nonreactive pattern (abnormal) NST is a poor index of fetal compromise.* This study was designed to evaluate the relationship of two fetal biophysical variables, the NST and FBM, to compare the independent predictive accuracy of each test, and to determine if the combination of both tests can improve predictive accuracy.
0
1979
The
C. V. Mosby
Co.
Patients studied were selected at random from the high-risk pregnancies referred for antepartum fetal heart rate testing at Women’s Hospital, LACiUSC 511
Table
I. Srudy population
Post date (>42 Diabetic Class A (7) Class B (46)
weeks)
Class C (9) Class D (6) Class F (2) Siypertensive disease Suspected intrauterine growth retardation Kh isoimmunized pregnancy FJ,emoglobinopathies (SS, SC) Maternal medical disorders Total
76 70
26
18 10 5 18 m
SPedical C&W-. The NST M~S performed using a method previously described.” An NST was termed a reactive pattern (R) or normal if, during a 20 minute period, there were two or more fetal heart rate accelerations of 15 bpm or greater of at least 30 seconds duration associated with fetal movement. The NST was termed a nonreactive pattern (NR) or abnormal if there were less than two accelerations of the fetal heart Ivith fetal movement in a 40 minute period.” The resuits of the SST were used in clinical management. Fetal breathing movements were recorded by means of a real time B-mode ultrasound method (ADR Mode! 2130).” FBM were defined as convergent chest wail movements in conjunction with abdominal wall excursions. FBM were observed immediately before, during. or after the NST. The majority of observations were made immediately after the NST. Patients were scanned until either a period of FBM of 30 seconds or greater was observed or 30 minutes elapsed.” FBti were recorded as either present or absent. The results of FBM observation were not disclosed and hence did not influence clinical management. The results of the NST. FBM, and combined NSTi FBM were correlated with the outcome of pregnancy as reflected by the five-minute Apgar score, the incidence of fetal distress in labor, birth weight, and meconium staining of amniotic fluid. Fetal distress was defined as any abnormal intrapartum fetal heart rate pattern that precipitated emergency delivery of a fetus regardless of the condition of the fetus at delivery. Fetal heart rate patterns considered indicative of fetal distress included repetitive late decelerations, intermittent late decelerations with loss of heart rate variability, unremitting se\-ere variable decelerations, or persistent fetal brady-cardia. infants were considered small for dates if the birth weight was less than the tenth percentile for gestational age and sex.
Three hundred rrinety-eight obser%,c:ions 3f FBM and the NST were made in 223 patients (Table i)~ Most patients delivered within one week of the last test (86%) and all delivered within three weeks. Overall, FBM were present in 83.7% of observations and the NST was reactive in 78.4% (Table II). FBM were present in 88% of observations when the NST was reactive as compared to 67% when the NST wa6 nonreactive (p < 0.05). Conversely: the NST was rcaciive in 82.6% of observations when FBM were present and reactive in 56.9% of observations when FBM were absent (p < 0.05). FBM,
NST,
and
five-minute
Apgar
(,TabXe
III).
A
normal five-minute A.pgar score (~7’) was observed in 179 of 187 patients (95.7%) with FBM present on the last observation before delivery and in 27 of 36 patients (75%) with FBM absent (p < 0.05). One hundred sixty-one of 169 patients (95.2%) with a reactive NST on the last obsenJation delivered an infant with a normai five-minute Apgar score as corAlpared to 15 of 54 (83.3%) with a nonreactive NST (p < 0.03). The predictive accuracy of the NST and FBM was not significantly different. A normal five-minute Apgar score was observed in l46 of 151 patients (96.7%) when both the NST was reactive and FBM were present in the :ast observation before delivery. The predictive accuracy of two normal tests offered no significant improvement over that of either normal test alone. Thirty-five of 36 patients (97.2%) with a nonreactive NST but with FBM present on the last observation delivered an infant with a normal five-minute Apgar score. The predictive accuracy of this combination was similar to that of the presence of FBM alone (96%) but significantlv different from the predictive accuracy of the nonreaciive NST (83.3%) when considered alone (p < 0.05). Seventeen of 18 patients (94.4%) with a reactive NST but with FEM absent on the last observation delivered an infant with a normal five-minute Apgar score. The predictive accuracy of this combination was similar to that of a reactive NST alone (95.2%) but significantly different from the predictive accuracy of the absence of FBM (75%) (p < 0.05). Finally, 8 of 18 patients (44.3%) with both a nonreactive NST and absent FBM on the last observation delivered infants with abnormal five-minute Apgar scores (<7). This combination was associated with a significantly higher incidence of abnormal five-minute Apgar scores than with any single test or combination of tests (p < 0.01). FBM, (Table
NST, IV).
and
the
Spontaneous
incidence
of
or induced
iabor occurred
fetal
distress
in
Volume Number
135 4
Table
!I.
FBM
Observations
333 (83.7%)
Absent
65 (16.3%)
Total
Table
R NR R NR
275 58 37 28
NST
(82.6%) (17.4%) (56.9%) (43.1%)
No. oft&s
Reactive
NSTIFBM
Nonreactive
and five-minute
Apgar
FBM
test*
FBM FBM FBM FBM
86 (21.6%)
513
+ + -
275 37 58 28
(88%) (12%) (67%) (33%)
398
score 5 Minute
Antepartum
pregnancies
FBM
3 12 (78.4%)
398
III.
in high-risk
on FBM and NST NST
Present
and NST
No. of patients
Apgar , qIIIpr-
27
alone
FBM present FBM absent Total NST alone NST - R. NST - MR
Total NSTIFBM combinsd NST-RIFBM-P NST-RIFBM-A.
NST-NRIFBM-P NST-NRIFBM-A
Total
187 36 223
179 (95.7%) 27 (75%)
8 (4.3%) 9 (25%)
co.05
169 54 223
161 (95.2%) 45 (83.3%)
8 (4.8%) 9 (1617%)
co.05
151 1‘8 36 18 223
146 17 35 10
5 1 1 8
(96.7%) (94.4%) (97.2%) (55.5%)
(3.3%) (5.6%) (2.8%) (45.5%)
NS P < 0.051 p
P<0.05?
*Last test before delivery. tAs compared to abnormal test. 184 of 223 patients studied (82.5%). The remaining 39 patients were delivered by cesarean section before labor. Fetal distress in labor was observed in 11 of 153 patients (7.2%) with FBIM present on the last test before delivery and in 11 of 31 patients (35.5%) with FBM absent (p < 0.05). C:omparatively, fetal distress was present in 7 of 137 patients (5.1%) when the last NST was reactive and in 15 of 47 patients (31.9%) when the last NST was nonreactive (p < 0.05). There was no significant difference in the predictive accuracy of fetal distress in labor between the two tests. Fetal distress in labbr occurred in 5 of 121 patients (4.1%) in whom both a reactive NST and the presence of FBM were observed in the last test before delivery. This combination offerred no significant improvement in predictive accuracy of fetal distress.when compared to the reactive ?JST or the presence of FBM alone. Six of 32 patients (18.7%) with a last test that revealed a nonreactive KSIT but with FBM present developed fetal distress in labor. The frequency of fetal distress with this combination was significantly greater than the frequency observed with the presence qf FBM alone (p < 0.05). Two of 16 patients (12.5%) with a reactive NST but with FBM absent developed fetal distress in
Table
IV.
FBM/NST
Antepartum
test*
and mode of delivery
/ ;;e:!
K
FBM alone FBM-P FBM-A
153 31
Total
184
NST alone NST-R NST-NR
137 47
Total Combined
142 (92.8%) 20 (64.5%) 162 (88.0%)
184
7 (5.1%) 15 (31.9%)t 22 (12.0%)
13p (94.9%) 32 (68.1%) 162 (S+O%)
121
5 (4.1%)
NSTIFBM
NST-RIFBM-P NST-RIFBM-A
NST-NRIFBM-P NST-NRIFBM-A
Total
11 (7.2%) 11 (35.5%)t 22 (12.0%)
16 32 15
184
2 6 9 22
(12.5%)$ (18.8%)$ (60.0%)$ (12.0%)
116 (95.9%) 14 26 6 162
(87.5%) (81.2%) (40.0%) (88.0%)
*Last test before delivery. tp 0.05. $p 0.05 as compared to the normal test alone. labor. This combination had a significantly higher incidence of fetal distress than with a reactive NST alone (p < 0.05) Fisher exact test). Finally. 9 of 15 patients (60%) with both a nonreactive NST and the absence of FBM on the last test developed fetal distress. The incidence of fetal distress with this combination was sig-
Table
Y. PGS~TiFBM and fetal weight
and meconium
FBt%I crlon~
Present Absent XST ulonr
Reactive Nonreactive FB.\II,VST combined YST-RIFBM-P NST-RIFBM-A NST-KRIFBM-P NST-KRIFBM-A
187 36
i83 (98.0%)
4 (2.0%)
30 (83.3%)
6 (16.6%)
175 (93.6%) 33 (91.6%)
12 (8.4%) 3 (8.4%)
169 54
167 (Y9.W) 48 (88.9%)
4 (2.3%) 6 (ll.i%)
159 (Y4.0%) 49 (91.0%)
10 (6.0%) .5 (Y.O%)
I51 18 36 18
149 16 34 14
2 2 2 4
143 16 32 17
.+&A = Appropriate for gestational age, %A “Last test betixe delivery.
(98.6%) (88.8%) (94.4%) (77.7%)
(1.4%) (11.2%) (5.6%) (22.3%)
(Y4.7R) (88.8%) (88.8%) (94.4%)
8 (L3%j 2 (I 1.2%) 4 (11.2%) 1 (5.6rCj
= small for gestationai age.
nilicantly greater than with any single iest aione or combination of tests (p < 0.0 I). ha, NST, and birth weight (Table Y). Te11 of 223 fetuses studied (4.5%) weighed less than the tenth percentile for gestational age. FBM were present on the East test in four of these SGA infants and absent in six. Similarly, in four of these patients the last SST was reactive and in six it was nonreactive. Combined KST and FRM results were not helpful in specifically identifying the SGA infant since in two fetuses both tests Lvere normal, in two the NST was reactive but with FBM absent, in two the N’ST Evasnonreactive with FBM present, an in four both tests were abnormal. Nonetheless. in 8 of 10 SGA fetuses (80%) at least one test was abnormal. Further, in both SGA fetuses that r+ere delxessed at delivery both tests were abnormal. FBM, NST, and meconium-stained amniotic fluid (Table V). Meconium was present in 15 of 223 patients studied (6.7%). No apparent relationship between the incidence of meconium and either test or combination of tests was observed.
In separate studies both the NST and FBM have been sho\vn to be useful methods for antepartum fetal evaluation.1-4 In this studv where we have measured bot variables in the same patient we note remarkable similarities between the predictive accuracy of either test. When considered independently, the accuracy of’ a rlortnal test in predicting a normal outcome was statistically better than Teas the abnormal test in predicting an abnormal outcome. Further, the combination of two normal tests did not improve the predictive accuracy as compared to either normal test alone. Conversely, the combination of tv,w abnormal tests produced a high]) significant improvement in the predictive accuracy of an abnormal outcome as compared to a single abnormai tes!. The relationship of a normal and abnormal
combination of tes& w outcome is i~rigrui~lg. I‘he predictive accuracy of a normal five-iniltute Apgai score with this combination was similar io the accuracy of the normal test alone, but significai:tiy different from that of the abnormal test alone. Howe\-er, the incidence of fetal distress in labor with the combination of a normal and abnormal test was intermediate, that is. significantly less than that observed with a single abnorma! test alone but significantIy greater than that observed with a single normal test alone. It is apparent that neither tear is helpful in speciticallv identifying the SGA infant. In experimental models chronic placental insufficieq of sufficient magnitude to impair fetal growth is uswily associated with normal arterial pH and a small fall in arrerial pO,.’ A similar mechanism may help :o explain the relatively high frequency of at least one abnormal biophysical test in the SCA fetuses studied, The relationship betlveen meconium passage and fetal distress is weak as the majority of‘i‘etuses that pass meconium are not acidemic.” Hence it is not surprising rhat no relationship hetween either antepartum test and the incidence of meconium was noted. Overall, these data appear to indicate rhat the normal biophysical test is a more reliable index of fetal health than is the abnormal test of fetai compromise. The explanation for this observation ma)- he related to the mechanisms responsible for the control of fetal biophysical variables. Fetal breathing movements, for example, are not random but rather reflect complex integrated neurologic controls. In the normal fetus. breathing movements are intermittent \t-ith periods of apnea of up to 108 minutes.” The periodic+ of FBM is at least in part due to sleep-wake cycles of the fetus.” Depression of the fetal centrai nervous system b> hypoxemia, acidemia, hypoglycemia, or dl,ugs can reduce or abolish FB?JI.~. lo Thus, whereas the preser;ce of FBM suggests an intact fetal central nervous system,
VolumeNumber
135 -1
FBM
the absence of FBM may reflect either normal periodicity or central nervous system depression. Continuous or intermittent FBM have been reported in the asphyxiated fetal lamb before death. However, to date, we have not observed either pattern before death in the asphyxiated primate’” or in the asphyxiated or dying human fetus.r5 C’ntil recently, biophysical monitoring of the fetus has been largely restricted to heart rate analysis. The development of real time B-mode ultrasound methods
and NST
in high-risk
pregnancies
515
now enables the physician to assess additional fetal biophysical variables. These data indicate that anteparturn fetal evaluation may be improved when two biophysical variables are considered. The authors wish to thank Marcie Smith, R.N., Paula Broussard, R.N., and Dorothy McCart, R.N., for technical support. In addition we wish to thank Drs. E. J. Quilligan and R. H. Paul for their critical review of and comments on this manuscript.
REFERENCES
1. Rochard, F., Schifrin, B. S., Sureau, C., et al.: Non-stress cardiotachometry ftor antepartum fetal evaluation, Obstet. Gynecol. 451433. 1975: 2. Martin. C, B.. and Schifrin, B. S.: In, Aladiem, S., and Brown, A. K., editors: Prenatal Fetal Monitoring in Perinatal Intensive Care, St. Louis, Missouri, 1977, The C. V. Mosby Comlany, p. 155. 3. Manning, F. A.: Fetal breathing movements as a reflection of fetal status, Postgrad. Med. 61:116, 1977. 4. Platt. L. D., Manning, F. A., Lemay, M., and Sipos, L.: Fetal breathing movements and outcome of pregnancy, AM.
J. OBSTET‘:GYMECOL.
132:542,
1978.
_
9.
a
10.
11.
1
5. Patrick. 1.. Fetherson. W.. Vick, H., and Voeeelin. R.: Human Jfe;al breathing movements and gross f&l body movements at weeks 34 to 35 of gestation, AM. J. OBSTET. GYNECOL. 130:693, 1978. 6. Boddv, K.. and Dawes, G. S.: Fetal breathing, Br. Med. Bull. 31:3,’ 1975. 7. Boddy, K., Dawes, G. S., and Robinson, J. S.: A 24 hour rhvthm in the foetus. in Comline. R. S.. Cross, K., Dawes, G./S., and Nathan&z. P., editors: Foetal and Neonatal Physiology, Proceedings of Sir Joseph Barcroft Centenary Symposium, Cambridge, 1973, Cambridge University Press, p. 63. 8. Creasy, R. K., DeSwiet, M., Kahanpaa, K. V., Young, W. P., and Rudolph, A. M.: Pathophysiological changes in the foetal lamb with growth retardation, in Comline, R. S., Cross, K., Dawes,2. S., and Nathanielsz, P., editors:
Fetal and Neonatal Physiology, Proceedings of Sir Joseph Barcroft Centenarv Svmoosium. Cambridge. 1973, Cambridge University Press, ‘p, 398.’ ” Miller. F. C., Sacks, D. A., Yeh, S., Paul, R. H., Schifrin, B. S,, Martin, C. B., and Hon, E. H.: Significance of meconium in labor. AM. 1. OBSTET. GYNECOL. 122:573, 1975. Manning, F. A., Platt, L. D., and Lemay, M.: Maternal hypoxemia and fetal breathing movements, Obstet. Gynecol. In press, 1978. Evertson, L., Gauthi’er, R., Keegan, K., Paul, R. H., Schifrin, B. S., and Martin, C. B.: Evolution of the nonstress test, AM. J. OBSTET. GYNECOL. 133:29, 1979. Patrick, J. E., Dalton, K. J., and Dawes, G. S.: Breathing movements before death in fetal lambs, AM. J. OBSTET. GYNECOL. 125:73, 1976. Chapman, R. L. K., Dawes, G. S., Rurak, D. W., and Wilds, P. L.: Intermittent breathing before death in fetal lambs, AM. J. OBSTET. GYNECOL. 131:894, 1978. Manning, F. A., Martin, C. B., Jr., Murata, Y., Miyaki, K., and Danzler, G.: Breathing movements before death in the primate fetus (IZlacaca mulatta), AM. J. OBSTET.
12.
13.
14.
GYNECOL.
135:71,
1979.
F. A., and Platt, L. D.: Fetal breathing move15. Manning, ments: Antepartum monitoring of fetal conditions, in Quilligan, E. J., editor: Clinics in Obstetrics and Gynecology, vol. 6, London, 1979, W. B. Saunders Company, Ltd., p. 335.