FETUS;PLACENTA, ANDNEWBORN Antepartum fetal heart rate testing I. Evolution LARRY
of the nonstress
On May 1, 1975, at Women’s Hospital, Los Angeles County-University of Southern California Medical Center, a new antepartum fetal heart rate (AFHRT) protocol was put into clinical use. This included the widely used contraction stress test (CST) and a new concept of nonstress testing (NST). The NST was based on FHR response associated with fetal movements and was categorized as reactive (normal) or nonreactive (abnormal). The nonreactive fetus was then evaluated with a CST if not contraindicated. During the 24 months, May 1,1975 to April 30,1977, a total of 2,422 NST’s were done in 1 ,169 patients, with 1,547 (64 per cent) reactive and 629 (35 per cent) nonreactive. CST was done 939 times, with 651 (90.6 per cent) negative, 29 (3 per cent) positive, 13 (1.4 per cent) equivocal, and 46 (5.0 per cent) unsatisfactory. There were ten (3.3 per cent) perinatal deaths within one week of a negative CST, five (1 .O per cent) within one week of a reactive NST, and two (6.7 per cent) with a positive CST. A reactive NST was as predictive of good outcome as was a negative CST. Analysis of the nonreactive NST showed that two or more accelerations were not associated with abnormal CST’s. Also, some nonreactive fetuses became reactive with oxytocin and had good outcome. These observations were utilized in the development of a newer, shorter NST which allows for fetal stimulation in an attempt to further define fetal well-being. (AM. J. OBSTET. GYNECOL. 133:29, 1979.)
From the Department of Ohstetrtis and Gynecolog?j,, U~~izvr.~ity of Southern California School of Medzclne Women’s Hotpztal, Los Angeles County-University Southern Cdiforr~ia Medicul Center.
Reprint 5K-22. 9003?.
repzrcsts: Dr. Richard H. Paul, Women’s Hospital 1240 N. Mkion Rd., Los Angeles. California
*Pre.wt addw~.v: Department of Obstetrin Gvwcolo~, United States Naval Regional Cknter. San Diego, California 92134. **Present address: Department of Obstetrics Gynecology, Hos ital Sk. Justine, Montreal. Carrada H3T I f :5. ***Present Gynecology, Califorwa
addw~s: Department Cedars-Sinaz Medicul 90048.
and Medical and Quebec,
of Obstetrics and Center, Los Angeles,
ANTEPARTUM fetal heart rate testing (AFHRT) has enjoyed widespread popularity in the United States since the report of Ray and associates’ on the use of the oxytocin challenge test (OCT) to follow pregnancies at risk for uteroplacental insufficiency. The rationale for this approach is that uterine contractions may provoke FHR patterns of late decelerations in the compromised fetus. Thus, the contractions being a repetitive stress, whether spontaneously occurring or oxytocin induced, form the basis of the contraction stress test (CST). The CST has the disadvantages that it is time consuming and not applicable to all patients (i.e., when oxytocin is contraindicated). Also, the “stress” applied may not accurately approximate the actual stress of labor which the test is attempting to imitate. Limitations of this approach are reflected in a false positive rate as high as 50 29
Ever&on et al.
per cent, in which a fetus predicted to be abnormal tolerates labor without signs of‘ tetal distress OI neonatal compromise. On the other hand, there is a low false negative rate (less than 1 per cent)’ suggesting that the normal fetus is readily identified by the test. Recently, reports have appeared in the [‘nited States literature on nonstressed antepartum FHR Iesting which relates the observations of’ FHR accelerations associated with fetal movement.‘i+‘i Early observations in this area weI-e made by Hammacher’ in the 19ciO’s. in which he explored the relationship of‘ FHR ffuctuations, oscillations, FHR accelerations, and perinatal outcome in the absence of. exogenous strexs f’;lctor>. Kubli and colleagues” employed a similar appl-oath. but added ohseryations of WR response to uterine contractions and other indicators of tctal evaluation. such as estriols, ultrasound. and amnioscop) in ;I more comprehensive approach. Studies b) Lee, Trienvriler,:’ and Fox!’ and their associates in the Cnited States correlated strcsh test and perinatal outcome ltith FHR acceleration occurrence, but did no1 clearly define precise criteria for these ohservations. Rochard and Schifrin used the fetal movement-associated FHR acceleration response in a nonstress test (NS’I‘) approach and introduced the terms used in this study. i.e.. reactive. nonreactive, and sinusoidal FHR patterns. Schifi-in was instrumental in the development of‘ the protocol used in this study. which defined the reartihe and nonreactive patterns precisely and explored the use of this approach in an attempt to replace the time-comsuming (ST with 21 shorter, simpler test of’ fetal well-being. In setting up the protocol, variability, vAich is ditficult to IIIWSUI‘C~ with external (primarily ultrasound) FHR s!‘stems, was not used in the tetal evaluation. ‘I‘he purpose of this communication is to detail :I two-\-car esper-ienc-c using the NST as the primary- approach in an~eparrum ;ISsessment of fetal lvell-being.
Materials and methods Test indications in order of decreasing frcquenq included:. prolonged pregnancy. diabetes mellitus. hypertensive disorders. abnormal estriols. suspected intrauterine growth retardation. 2nd history of a previous stillbirttl. Testing routinely started at a gestational age (approximately 34 weeks’) at which time intervention for an abnormal Lest would be considered. Patients undenzent testing in a room used only for antepartum FHR testing and tests were performed by a team of’ specially trained nurses with consistent physician supervisors. External FHR monitoring techniques were employed with the initial attempt made b\ means
of a phonocardiograph microphone* and ii standard fetal nionit0r.f It‘ an inadequate FHR +V:IS obtained. ultrasonic methods \verr used. The ti-cqurnc\ a11d duration of‘ uterine contractions were evaluated h riiealls of’an external tocodvnalnc,nleter.. A sec~ntl observation ivirh the rocodyllal~iomrter I\ as I hc rapid spiking deflections macle bv f’cral movements illld confirmed h! the mother and/or the examining nurst’. Careful documentation of these f’etal movements and the ashociatecl 1;HR c haracteristics (i.e., accelcralion) formed the basis usecl in the interpretation of‘thc nollstrehs test (NST). Thr patient was placed in the semi-l;owlcr position and blood pressure raken al IO minute itltervals to avoid supine hypotension as a possible cause for :1n abnormal test. Baseline FHR and maternal pulse wrre taken aI the beginning and end of’thc rest. BeCause 110 specific criteria fi)t- interpretation oft he NSI‘ rxistetl, criteria were \omewhar arbitrarily defined, based on clinical
fetal rest-activit! state would be observed. based on the existent information regarding
This was neonatal rest-activir! c1.c les ofapJ”-(~“i,narely 40 minutes.” During a 5’0 minute time period, five accelerations associ-
of’ ar least
mal” rest. If’ the FHR accelerations wt’rc les\ than 15 b.p.m. or f&cr than 5 in an!. 20 minute period. the test ~a considered nonreacti\,e and the patient underwent ;I ET.
with a w;1$
pet. \$ceE, antI
\L hich was coupled
fi)llowecl witlt both NST tfemotwratcd of this
‘1‘1~ only of
the last two if’the
Results During :$O,
the t!zo-year- period,
May 1. 1975. to April underwent 2.422 tests: 1 ..?A7 (64 per cent) were reactive, 829 ;I total
“Modified C:orometrics FMS 10 I A with ‘l‘oiutsu microphone and manual gain control. Corometrics Medical Systems, Wallingt;n-d. Connecticut. *Standard Coromrtrics FMS 101 B. Coromrtrics Medical Systems. Wallingford. Connecticut, and Hewlett-Pat kard HPS 030A, Hewletr-Packard, Waltham, Massachusetts.
(34 per cent) were nonreactive, 43 (2 per cent) were unsatisfactory, and three tests were described as sinusoidal. The reason for unsatisfactory nonstress tests was an inability to adequately record FHR due to extreme obesity most often, and on occasion, due to excessive fetal activity which interfered with continuous FHR counting. The FHR tracing was obtained utilizing the abdominal wall FECG in 5 per cent, the phonocardiographic technique in 42 per cent, and ultrasound was used in 53 per cent of patients. A CST was performed 939 times with 851 (90.6 per cent) negative, 29 (3 per cent) positive (27 patients, one patient had three positive tests), 13 (1.4 per cent) equivocal, and 46 (5.0 per cent) unsatisfactory. The reasons for the 46 unsatisfactory tests included 30 in which the FHR was inadequate and 16 in whom uterine contractions sufficient to be called a CST were not achieveable. Perinatal death. Eight hundred eleven patients were delivered within seven days of the last test. The AFHRT result was related to perinatal death as an indicator of predictability. When only the last test within one week of delivery is considered, there were 493 reactive tests with five perinatal deaths (rate 10 per 1,000) and 302 nonreactive with ten deaths (rate 33 per 100) which is statistically significant (p = 0.05). The perinatal mortality rate associated with a nonreactive last test coupled with a negative CST was 27 per 1,000 and a nonreactive NST with a positive CST was 87 per 1,000. There was no statistical difference between the rates in the “normal tests” (a reactive pattern, 10 per 1,000) and nonreactive CST (27 per 1,000). Twenty-six of the 27 patients who had a positive CST delivered within 24 hours of the test. The outcome in this group of patients is detailed in a separate report. There were no perinatal deaths in patients with equivocal CST’s. There was one death in the unsatisfactory CST or NST group. This occurred in an infant with severe cardiac anomalies and a complete heart block. A sinusoidal pattern occurred three times, in one pre-eclamptic patient and in two with postdates. All tolerated labor and delivered normal infants. Accelerations. In order to evaluate the validity of using five accelerations in 20 minutes as the criteria for the NST. all 302 nonreactive tests with delivery 57 days were reviewed and classified by the number of movement-associated accelerations seen in the first 20 minutes of the test. Breakdown relative to accelerations was: no accelerations-181 patients (59 per cent), one acceleration-56 patients (19 per cent), two accelerations-35 patients (12 per cent), three accelerations25 patients (9 per cent), and four accelerations-4 patients (1 per cent).
Antepartum FHR testing. I 31
Fig. 1. Antepartum July, 1977.
DELIVERY OR FURTHER EVALUATION
fetal heart rate testing at LAC/USC in
I. CST outcome relative to number of accelerations in the first 20 minutes with delivery 17 days
56 36 25
20 2 0 0 0
: 3 4
1 0 0 0
Table II. Positive, equivocal CST reactive versus nonreactive with CST Positive
Reactive 87 Nonre112 active 199
*False positive. The groups exhibiting various numbers of’ accelerations per 20 minutes were compared to possible outcome measures which included: (1) the subsequent CST, (2) fetal death, (3) Apgar scores, and (4) neonatal death. The only positive result was found in the relationship of the number of accelerations and an abnormal CST outcome. All the equivocal and positive CST’s occurred in the groups who had zero and one accelera-
32 Evertson et al.
tions. There were no abnormal CST’s when two OI more accelerations were present. Stimuli. In the group of 302 patients who had nonreactive patterns, there were 199 who underwent oxytocin infusion (OCT) to achieve a CST. In the remainder, there was sufficient uterine contraction to qualify as a CST or oxytocin was contraindicated. During the process of starting an intravenous, signing consent forms, or oxytocin infusion, fetal movements and associated FHR accelerations sufficient to qualify as a reactive pattern often occurred. Thus, of these 199 patients, 87 became reactive and 112 remained nonreactive during the CST. Subsequent perinatal losses occurred only in the group which remained nonreactive during the CST. Low 5 minute Apgar scores were more than three times as frequent in the group remaining nonreactive during CST’s. Furthermore, all of the positive and equivocal CST’s were in the group which remained nonreactive except one. That single abnormal CST was subsequently regarded as false positive, since on induction the fetus tolerated labor without evidence of late deceleration and was born in excellent condition (Table II).
Comment As one attempts to judge the efficacy of any anteparturn method which assesses fetal condition, a great dilemma is encountered since outcome endpoints may be unreliable and by circumstances are often remote in time. Thus, one is faced with making comparisons between antepartum observations and measures such as perinatal death, which by necessity can be adversely affected by critical problems such as the process of labor. This is a major weakness of this and similar studies, which is fully recognized. From the soft data herein presented, it would appear that perinatal outcome is significantly different when a reactive pattern group is compared to those exhibiting one which is nonreactive. However, one must realize that the overwhelming number of patients demonstrating a nonreactive pattern as defined had normal outcomes. There was no difference in perinatal outcome when the NST reactive pattern was compared with the negative CST that had been preceded by a nonreactive pattern. This observation is a significant one in that these “normal tests” appear to have equal predictability. Even the worst combination of AFHRT observations, a nonreactive pattern followed by a positive CST (87 per l,OOO), was most often associated with “good” outcome. In evaluating the criteria for a nonreactive pattern, the current data suggest that abnormal CST’s occurred only in patients who had zero or one acceleration in 20 minutes. Redefinition using this finding would havr
reduced the nonreactive group from 34 per cent to 26 per cent of the overall group, thus requiring fewer time-consuming CST’s. The observation that an “induced reactive” pattern during oxytocin infusion was infrequently associated with an abnormal CST was of great interest. It would appear that stimuli, as proposed years ago. might be useful in fetal assessment schemes. The conclusions that were drawn from this study can be summarized as follows: 1. ‘4bnormal CST’s occurred only in patients with zero or one acceleration. 2. Induced “reactivity” is equivalent to spontaneous “reacti\itv.” 3. A reactive NST has a predictive reliability equivalent to a negative CST. The need to simplify AFHRT and yet retain or improve the predictive precision is evident to those who use such testing methods. As a result of this study, a revision of the AFHRT protocol was accomplished. This was done on the assumption that: (1) a reactive NST was as adequate a predictor of “normal” as a negative CST, (2) a reactive pattern could be defined as two or more accelerations in a period of 20 minutes or less, and (3) a stimulus which evoked a reactive pattern was equivalent to that seen spontaneously. The present testing approach allows AFHRT in both inpatient and outpatient facilities. Patients are selected Ii-om the high-risk clinics and antepartum ward and are tested using an external FHR and uterine activitv system in the semi-Fowler position, as previously done. Minimum observation time (NST) is 10 minutes with a maximum of .40 minutes. Two or more accelerations ot 15 b.p.m. lasting at least 15 seconds during any 20 minutes in the observation period is considered a reactive pattern and the patient is retested in seven davs. If zero or one acceleration is encountered in the first 20 minutes. the teSt is nonreactive and stimulation in the form of palpation and manipulation of the fetus is performed. A reactive pattern persisting live minutes after the stimulus is considered “reactive” and the patient is I-etcsted in one week. If the test remains nonreactive. oxytocin via ini’usion pump is given to yield three contractions in ten minutes for CS7‘ interpretation, as under the previous protocol. The latter is done only in a hospital setting. If the fetus exhibits a reactive pattern during this procedure. it is considered “normal” and will be retested in one week. even though the CST was uot completed. A nonreactive test during the CST Ivill be reported as nonl.eactive/neg~~:ative CS’I‘ (retest one week). Ilonl-e;lctive/ecluivocal (retest nest day ), “I‘ nonreactive/positive (evaluate for delivery). ,4 sporitaneous (X7‘ lvill be reported similarly (Fig. 1),
Preliminary experience with this protocol suggests that the need for a CST may be reduced to less than 10 per cent. The value of this approach as it relates to fetal stimulation, repeated NST usage as opposed to im,mediate CST, and its predictive value is the subject of current investigation. It appears that by further decreasing the number of patients requiring a CST, more patients can be screened with a simplified NST. Hopefully, through the use of fetal stimulation, be it sound, manipulation of fetus and/or mother, or yet other stimuli, a simple,
REFERENCES 1. Ray, M., Freeman, R., Pine, S., and Hesselgesser, R.: Clinical experience with the oxytocin challenge test, AM. J. OBSTET. GYNECOL. 114: 1, 1972. 2. Ever&on, L. R., Gauthier, R. J., and Collea, J. V.: Fetal demise following negative contraction stress tests, Obstet. Gynecol. 51: 671, 1978. 3. Trierweiler, M. W., Freeman, R. K., and James, J.: Baseline fetal heart rate characteristics as an indicator of fetal status during the antepartum period, AM. J. OBSTET. GYNECOL. 125: 618, 1976. 4. Lee. C. Y., DiLoreto, P. C., and Logrand, B.: Fetal activity acceleration determination for the evaluation of fetal reserve, Obstet. Gynecol. 48: 19, 1976. 5. Farahani, G., and Fenton, A. N.: Fetal heart rate acceleration in relation to the oxytocin challenge test, Obstet. Gynecol. 49: 163, 1977. 6. Rochard, F., Schifrin, B. S., Goupil, F., LeGrad, H., Blottiere, J., and Sureau, C.: Nonstressed fetal heart rate monitoring in the antepartum period, AM. J. OBSTE~T. GYNEC:OL. 126: 699, 1976.
Antepartum FHR testing. I 33
rapid, reliable, specific screening test using FHR can be developed. Thus, the high-risk fetus can be identified and
by the antepartum nurse specialists: Marci Smith, R.N., Paula Broussard, R.N., and Dorothy McCart, R.N. Their attention, advice, and assistance
7. Hammacher, K.: The clinical significance of cardiotocography, in Huntingford, P. J., Huter, E. A., and Saling, E., editors: Perinatal Medicine, New York, 1969, Academic Press, Inc., p. 80. 8. Kubli, F. W., Kaeser, O., and Kinselmann, M.: Diagnostic management of chronic placental insufficiency, in Pecile, A., and Finzi, C., editors: The Foeto-Placental Unit, Amsterdam, 1969, Excerpta Medica Foundation, p. 323. 9. Fox, H. E., Steinbrecher, M., and Ripton, B.: Antepartum fetal heart rate and uterine activity studies, AM. J. OBSTET. GYNECOL. 126: 61, 1977. 10. Sterman, M. B., and Hoppenbrouwers, T.: The development of sleep-waking and rest-activity patterns from fetus to adult in man, in Sterman, M. B., McGinty, D. M., and Adinolfi, A. M., editors: Brain Development and Behavior, New York, 1971, Academic Press, Inc., pp. 203227. 11. Martin, C. B., Jr., and Schifrin, B. S.: Prenatal fetal monitoring, in Aladjem, S., and Brown, A. K., editors: Perinatal Intensive Care, Chap. 1, St. Louis, 1976, The C. V. Mosby, Company.