Objective
Comparison of time and outcomes of National Institutes of Child Health and Human Development defined fetal heart rate acceleration criteria at ≤32 weeks (≥10 beats/min, ≥10 seconds) compared with standard criteria (≥15 beats/min, ≥15 seconds).
Study Design
Singleton high-risk pregnancies that were referred for nonstress testing at ≤32 weeks’ gestation were randomly assigned to 15 × 15 or 10 × 10 criteria. Data included nonstress test information, maternal data, and outcomes.
Results
One hundred forty-three women were randomly assigned to 15 × 15 (n = 71) or 10 × 10 (n = 72). The groups were similar in maternal and pregnancy characteristics. Median time to reactive nonstress testing was shorter in the 10 × 10 group (37.3 minutes) than the 15 × 15 group (41.3 minutes; P = .04). There were no serious adverse events.
Conclusion
The time to attain a reactive nonstress testing at ≤32 weeks’ gestation was 4 minutes shorter when the 10 × 10 criteria were used. There were no adverse events related to use of 10 × 10 nonstress testing criteria.
The National Institutes of Child Health and Human Development (NICHD) have defined fetal heart rate (FHR) accelerations at <32.0 weeks’ gestation as having a peak of ≥10 beats/min above baseline and a duration of ≥10 seconds. This difference between gestational ages of <32.0 weeks, compared with the historic definition of ≥15 beats/min for ≥15 seconds at ≥32.0 weeks’ gestation presumably reflects maturational processes in the fetal autonomic nervous system. Various types of reports support the role of maturation of the fetal autonomic nervous system to explain an increased amplitude of FHR accelerations with advancing gestational age and the inverse (ie, a reduced amplitude of FHR accelerations at earlier gestational ages). Research reports that have supported this mechanism include experimental animal studies, older reports in humans with the use of visual interpretation of standard FHR recordings, and reports that used computer analysis of FHR recordings. The NICHD definition for FHR acceleration (≥10 beats/min for ≥10 seconds) for gestational ages of <32.0 weeks is supported by the most recent American College of Obstetrics and Gynecology practice bulletin.
Despite the general acceptance that the amplitude of FHR accelerations at early gestational ages will be less than at later gestational ages, to date there have been no prospective, randomized, controlled trials that have examined whether the use of 10 × 10 criteria for accelerations at <32.0 weeks’ gestation impacts clinical staff time or can be used safely. Glantz and Bertoria, in a retrospective review of 751 nonstress testing (NST) in 488 inpatients at <32 weeks’ gestation, found no appreciable difference in significant perinatal morbidities between inpatients who met the 10 × 10 criteria vs the 15 × 15 criteria. The primary objective of our study was to perform a prospective, randomized trial to compare the outpatient antenatal testing time differences with the use of the standard NST criteria (≥15 beats/min for ≥15 seconds) with a group with the use of the NICHD suggested criteria (≥10 beats/min for ≥10 seconds) at <32.0 weeks’ gestation. A secondary aim was to review the clinical outcomes of women who started NST at <32 weeks’ gestation.
Materials and Methods
The study population consisted of pregnant women who had been referred to Sharp Mary Birch Hospital for Women and Newborns (SMBHWN) for antenatal testing from 24.0-32.0 weeks’ gestation. SMBHWN is a busy (>8000 deliveries/year) community-based, tertiary-care women and infants’ hospital. Subjects’ residences ranged from southern Riverside County to northern Baja California and east to the Arizona border. The ethnicity distribution was 48.3% white, 22.4% Hispanic, 7.7% African American, 13.3% Asian, 8.4% others. The 76 SMBHWN obstetric staff members range from solo providers to larger physician groups of >12 physicians who refer patients to the Antenatal Diagnostic Center (ADC). More than 40 outpatient NSTs are performed daily in the ADC on Monday, Tuesday, Thursday, and Friday. There are no medical students or residents available.
After approval by the institutional review board, we conducted a prospective, randomized clinical trial to compare resource use and outcomes that were related to 2 different NST methods. Inclusion criteria consisted of referral for antenatal testing, a gestational age of ≤32.0 weeks (as determined by the referring physician), singleton gestation, NST as the first type of antenatal test performed (not a contraction stress test or a biophysical profile [BPP]), informed consent, and the ability to follow the patient through serial antenatal testing and delivery at SMBHWN. Exclusion criteria included noncompliance in the following requirements: not to attend follow-up appointments, not to deliver at SMBHWN, and not to follow medical recommendations or to receive the chronic administration of medications (eg, narcotics) that are known to affect FHR. The indications for testing, as determined by the referring obstetrician, included maternal medical conditions (eg, hypertension, diabetes mellitus, advanced maternal age), a history of stillbirth, oligohydramnios or intrauterine growth restriction, umbilical cord conditions (velamentous, nuchal, other), decreased fetal movement, trauma/vaginal bleeding, fetal anomaly, and preterm labor.
Subjects referred to the ADC unit at SMBHWN between March 2008 and June 2010 and who met the inclusion criteria were recruited for the study and gave informed consent. The consented patient was assigned randomly to the standard 15 × 15 or revised 10 × 10 criteria group. Information that was collected included demographics, medical history, indications of testing, results of antenatal testing, and clinical outcomes (antenatal, intrapartum, and neonatal). Maternal intervention data that were collected included the BPP, induction of labor, and cesarean section delivery. If the patient was admitted to the hospital for any complication, FHR information was not collected during hospitalization. Patients were discharged undelivered, returned for further testing in the antenatal testing unit, and continued to be study patients.
The NST was performed with the ADC’s standard policy and procedure, which follows national guidelines. The logistics of NST involved the positioning of the patient and the establishment of a technically satisfactory FHR recording. The fetus was monitored for 20 minutes. If the fetus was not reactive after 20 minutes, a single 3-second episode of vibroacoustic stimulation (VAS) was applied. Monitoring was continued. If after an additional 20 minutes, the tracing was still nonreactive by the assigned criteria, the on-call perinatal physician was alerted regarding the need to do a BPP. The FHR tracing was continued while awaiting the physician’s arrival. If, in that interval, the tracing met criteria for reactive, the BPP was not performed. The interval from the time of physician alert to physician arrival was variable. The interpretation of the NST was dependent on the group to which the subject was assigned randomly. In the 15 × 15 group, a reactive NST was defined as one that manifested ≥2 15 × 15 accelerations within a 20-minute interval. Subjects who were assigned randomly to the 10 × 10 group were treated in an identical fashion, except the criteria for reactive was ≥2 10 × 10 FHR accelerations in a 20-minute interval. Once the criteria for reactive was met, the patient was discharged from the ADC to return in 3-4 days for retesting. Amniotic fluid index was assessed at each visit by the sonography-certified ADC staff. Management was individualized if oligohydramnios or polyhydramnios was found. Once the patients reached >32.0 weeks’ gestation, NST data collection ended, and 15 × 15 criteria were used for subsequent antenatal testing.
Data were collected per protocol by study personnel that comprised 5 specially trained ADC registered nurses, a perinatal clinical nurse specialist, and a clinical research specialist. Given that the safety of the use of the NICHD criteria for interpretation of NSTs at <32 weeks’ gestation has not been established, the investigators believed it was imperative that this protocol be monitored carefully for adverse events. A safety monitoring committee was in place with members available to review any unanticipated or adverse event. If any consented study participant experienced an adverse event (nonreassuring fetal assessment [urgent or emergent unscheduled delivery temporally related to NST], a need for BPP evaluation that resulted in a score of <6, depression at birth, or a 5-minute Apgar score <4), the patient’s case was reviewed in its entirety to determine whether the antenatal testing process contributed to the adverse event. Study data and maternal and neonatal outcomes were evaluated and monitored on an ongoing basis, and unanticipated problems or adverse events were reported to the principal investigator immediately.
Subjects were assigned to groups in a 1:1 allocation ratio, with the use of a computer-generated sequence for blocks of 10 to maintain balance between study groups for interim safety and adverse event monitoring. Group assignment was placed in numerically ordered blinded opaque envelopes that could not be resealed and were stored in a secured location. Once the women were determined to be eligible for the study and had completed the informed consent process, the envelopes were used in numeric order. After receiving the group assignment, the nurse caregivers and physicians were aware of the study group assignment. The researcher who performed the interval safety and final statistical analyses was blinded to allocation; actual group assignment was not revealed until completion of the study.
Before the start of the study, sample size calculations were performed in the following manner: using 95% confidence level (1-alpha) and 80% power (1-beta) and an estimate of a 15% reduction in time to perform the NST, the total sample that was required to demonstrate a statistically significant difference between groups was estimated to be approximately 130 women (65 women in each group). A targeted total of 150 women was entered in the study to account for subjects who possibly would be lost to follow-up evaluation, attrition, a change in willingness to consent, or missing key data parameters.
Statistical analyses were performed with PASW Statistics (version 18.0; IBM Corporation, New York, NY). Baseline characteristics were compared with the use of the Student t tests, Wilcoxon rank sum test, χ 2 test, or Fisher exact test, as appropriate. Multivariable analyses included logistic regression, linear regression, and generalized estimating equations. The total number of minutes for all tests for each patient was adjusted for the number of NSTs to obtain a summary measure, which were minutes per subject for all NSTs.
Patients who required VAS and/or BPP were identified and compared with those without VAS and/or BPP in the entire study population, between study groups, and within study groups. Conditional analyses were performed to examine differences in the duration of the first NST and minutes per patient for all NSTs of the anomalous infants vs the nonanomalous in the total group and in the 15 × 15 and 10 × 10 groups.
Results
A total of 166 patients were assessed for eligibility for the study; 16 women were excluded for various reasons ( Figure ), which resulted in 150 women who were assigned randomly to standard 15 × 15 (n = 75) or revised 10 × 10 (n = 75) groups. Four patients in the 15 × 15 group and 3 patients in the 10 × 10 group were excluded from the analyses because they did not deliver at SMBHWN and outcome data were not available. No consented subjects were excluded because of noncompliance ( Figure ). The final sample for the study was 15 × 15 (n = 71) and 10 × 10 (n = 72).
The 15 × 15 and 10 × 10 groups were similar in maternal and pregnancy characteristics that included maternal age, parity, prepregnancy body mass index, ethnicity, smoking status, marital status, gestational age at random assignment, and history of preterm delivery ( Table 1 ). Univariable and multivariable logistic regression models to compare the 15 × 15 group with the 10 × 10 group failed to reveal any statistically significant associations between groups in maternal age, parity, prepregnancy body mass index, or ethnicity in the full or reduced models (data not shown).
| Baseline characteristics | Standard 15 × 15 criteria (n = 71) | Revised 10 × 10 criteria (n = 72) | P value | Test |
|---|---|---|---|---|
| Visit 1 | ||||
| Maternal age, y a | 29.7 ± 5.5 | 31.6 ± 7.3 | .81 | Student t |
| Gestational age, wk a | 29.5 ± 1.6 | 29.7 ± 1.64 | .56 | Student t |
| Prepregnant body mass index, kg/m 2 a | 28.0 ± 8.0 | 27.8 ± 7.6 | .91 | Student t |
| Nulliparous, n (%) | 30 (42.3) | 30 (41.7) | .94 | χ 2 |
| Race, n (%) | .86 | χ 2 | ||
| White | 32 (45.1) | 37 (51.4) | ||
| Black | 5 (7.0) | 6 (8.3) | ||
| Hispanic | 17 (23.9) | 14 (19.4) | ||
| Asian | 10 (14.1) | 10 (13.9) | ||
| Other | 7 (9.9) | 5 (7.0) | ||
| History of preterm delivery, n (%) | 10 (14.1) | 14 (19.4) | .35 | χ 2 |
| Current smoker, n (%) | 1 (1.4) | 2 (2.8) | .57 | Fisher exact |
| Married, n (%) | 49 (69.0) | 49 (68.1) | .99 | χ 2 |
The indication categories for antenatal testing are shown in Table 2 . Subject assignment to 1 indication was not mutually exclusive. Maternal medical indications (eg, hypertension, preeclampsia, diabetes mellitus, thrombophilia) made up the largest group, followed by fetal risk conditions (eg, history of stillbirth, oligohydramnios, or intrauterine growth restriction). There were no statistically significant differences in indications for testing between the 15 × 15 and 10 × 10 groups.
| Indication a | Standard 15 × 15 criteria (n = 71) | Revised 10 × 10 criteria (n = 72) | P value | Test |
|---|---|---|---|---|
| Maternal medical, n (%) | 24 (33.8) | 23 (32.0) | .81 | χ 2 |
| Chronic hypertension, preeclampsia | 5 | 10 | ||
| Diabetes mellitus: gestational, types 1 and 2 | 11 | 12 | ||
| Advanced maternal age | 2 | 6 | ||
| Miscellaneous: fibroid tumors, cholestasis, low weight gain, others | 8 | 1 | ||
| History of stillbirth, oligohydramnios, intrauterine growth restriction, n (%) | 20 (28.2) | 16 (22.2) | .41 | χ 2 |
| Umbilical cord, n (%) | 6 (8.5) | 8 (11.1) | .59 | χ 2 |
| Velamentous | 2 | 1 | ||
| Nuchal | 5 | 7 | ||
| Miscellaneous | 2 | 7 | ||
| Trauma: vaginal bleeding, n (%) | 3 (4.2) | 1 (1.4) | .62 | Fisher exact |
| Fetal anomalies, n (%) | 5 (7.0) | 5 (6.9) | .98 | Fisher exact |
| Decreased fetal movement, n (%) | 3 (4.2) | 8 (11.1) | .12 | χ 2 |
| Preterm labor, incompetent cervix, cerclage, n (%) | 10 (14.1) | 11 (15.3) | .84 | χ 2 |
Considering the total standard 15 × 15 and revised 10 × 10 groups, median time in minutes to reach a reactive result for all NSTs was significantly shorter in the 10 × 10 group (37.3 minutes) vs the 15 × 15 group (41.3 minutes; P = .04; Table 3 ). This pattern of shorter intervals for the total 10 × 10 group to achieve a reactive test was also seen when we analyzed the proportion of patients with all NSTs reactive in ≤30, ≤40, or ≤60 minutes, but the differences were not statistically significant ( Table 3 ). The median duration of the first NST was shorter (34 minutes) in the 10 × 10 group compared with the 15 × 15 group (40 minutes), but the difference was not statistically significant ( P = .18).
| Outcome measure | Standard 15 × 15 criteria (n = 71) | Revised 10 × 10 criteria (n = 72) | P value | Test |
|---|---|---|---|---|
| Duration of the first nonstress testing, min a | 40 (28, 54) | 34 (27, 45) | .18 | Mann-Whitney U |
| Median number of nonstress tests, n a | 3 (2.0, 6.0) | 3 (2.0, 5.0) | .66 | Mann-Whitney U |
| Minutes per subject to reactive nonstress testing a | 41.3 (32, 55) | 37.3 (30, 43) | .04 | Mann-Whitney U |
| All nonstress testing reactive, n (%) | ||||
| At ≤30 min | 9 (12.7) | 17 (23.6) | .09 | χ 2 |
| At ≤40 min | 17 (23.9) | 24 (33.3) | .21 | χ 2 |
| At ≤60 min | 41 (57.7) | 49 (68) | .20 | χ 2 |
VAS was used in 18 of the women in the 15 × 15 group and 17 of the women in the 10 × 10 group. Although the number of women who used VAS was similar in the 2 groups, total number of VAS episodes was greater in the 15 × 15 group (n = 43) than in the 10 × 10 group (n = 21). This difference in total number of NSTs that used VAS was due to an increased number of the 15 × 15 subjects who used VAS in multiple NSTs. In the 15 × 15 group, 6 women used VAS in 1 NST; 2 women used VAS in 2 NSTs; 8 women used VAS in 3 NSTs; and in 1 subject, VAS was used in 4 NSTs. In contrast in the 10 × 10 group, VAS was used in 1 NST by 15 women; 2 of the women in the 10 × 10 group used VAS during 3 NSTs.
Because of the difference in the use of VAS between the 15 × 15 and 10 × 10 groups, the subjects were subcategorized into those with or without VAS. The median duration of the first in the 10 × 10 group and the minutes per subjects for all NSTs for both the 15 × 15 and 10 × 10 groups were significantly longer when VAS was used ( Table 4 ). The between group (15 × 15 or 10 × 10) comparisons of with or without VAS were statistically significant, with the exception of the 15 × 15 groups durations of the first NST ( Table 4 ). Similar to the total between group comparisons of the median minutes per subjects for all NSTs ( Table 3 ; P = .04), the subjects who did not require VAS achieved a reactive pattern with a 4-minute difference in the median minutes per subjects for all NSTs ( Table 4 ; P = .03).
| Variable | Standard 15 × 15 criteria (n = 71) | Revised 10 × 10 criteria (n = 72) | P value |
|---|---|---|---|
| VAS | |||
| Duration of the first nonstress testing | |||
| Without VAS | 39 (27, 53) a | 33 (25, 44) b | .17 |
| With VAS | 43 (31, 98) a | 40 (29, 58) b | .71 |
| Minutes per subject/all nonstress tests | |||
| Without VAS | 40 (30, 52) c | 36 (29, 40) d | .03 |
| With VAS | 52 (37, 60) c | 45 (22, 71) d | .75 |
| BPP | |||
| Duration of the first nonstress testing | |||
| Without BPP | 40 (28, 53) e | 34 (27, 45) | .13 |
| With BPP | 55 (31, 107) e | 104 (98,–) d | .39 |
| Minutes per subject: all nonstress testing | |||
| Without BPP | 40 (32, 53) d | 38 (30, 42) f | .06 |
| With BPP | 60 (51, 74) d | 45 (29, 45) f | .30 |
| Anomaly | |||
| Duration of the first nonstress testing | |||
| Without anomaly | 40 (28, 55) g | 34 (27, 45) h | .14 |
| With anomaly | 32 (30, 55) g | 43 (26, 57) h | .99 |
| Minutes per subject: all nonstress testing | |||
| Without anomaly | 41 (32, 53) i | 37 (30, 44) j | .052 |
| With anomaly | 44 (33, 62) i | 39 (28, 40) j | .39 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
