Antepartum electronic fetal monitoring interpretation: multicenter observational study

Abstract

Background

Antepartum EFM is the most widely used method worldwide to evaluate fetal well-being. However, neither NST nor other surveillance techniques can precisely determine the onset or degree of fetal hypoxemia and acidemia. The presence of accelerations seems to reflect fetal well-being; however, their absence does not predict fetal compromise. Evidence regarding the perinatal prognosis of normal NSTs with one or no accelerations compared with those classified as reactive remains insufficient.

Objective

To compare perinatal outcomes of the non-stress test (NST) interpreted by the usual reactive/non-reactive criterion versus the NST reported as “fetal well-being preserved” when there is one or no fetal heart rate (FHR) acceleration with moderate variability.

Methods

Multicenter, non-randomized, observational, comparative non-inferiority study evaluating two different NST interpretation criteria, conducted in Panama among high-risk pregnant patients with ≥ 34 weeks’ gestation. A reactive NST was defined as a tracing with baseline FHR 110–160 beats per minute, variability 6–25 [beats/min], no bradycardia, and at least two accelerations within 20 min. When one or no acceleration was present while the other criteria were maintained, the NST was defined as showing “fetal well-being preserved” (normal/reassuring).

Results

We analyzed 1,206 NSTs from 699 patients. There were no significant differences in gestational age at birth, time elapsed since the first NST, mode of delivery, 1- and 5-minute Apgar scores, or the presence of variable decelerations. No perinatal death was associated with follow-up by NST report. Twelve cases (0.99%) were reported as abnormal at the second NST performed 7 days after the initial NST. In those cases, no perinatal deaths occurred; two cases had a 5-minute Apgar < 7, one in each reporting group. Across the entire population, there were 6 deaths with no significant differences related to the initial NST reporting category.

Conclusion

This observational non-inferiority study reporting the NST as reactive (≥ 2 accelerations) or as fetal well-being preserved/reassuring (normal variability with ≤ 1 acceleration) yields no large differences in fetal well-being outcomes.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12884-026-09711-z.

Introduction

Electronic fetal monitoring (EFM), also known as cardiotocography or fetal electrocardiography, has been used to assess fetal well-being for more than 60 years [1] and is recognized as a fetal non-stress test (NST) aimed at preventing fetal death. Traditionally, baseline fetal heart rate (FHR), variability, and accelerations above baseline associated with fetal movements are assessed to infer fetal condition. Antepartum EFM is the most widely used method worldwide to evaluate fetal well-being. However, neither NST nor other surveillance techniques can precisely determine the onset or degree of fetal hypoxemia and acidemia [2, 3]. An antepartum EFM is reported as reactive (normal) when the baseline FHR is between 110 and 160 beats per minute with two or more accelerations within 20 min [2]; it is considered non-reactive (abnormal) if after 40 min two accelerations are not observed. An acceleration refers to an increase of the baseline FHR producing a peak ≥ 15 beats above baseline lasting > 15 s but < 2 min. Fetal movements usually trigger such accelerations. Studies show that fetal deaths decline after cases at risk are identified through antepartum EFM, but there is no conclusive evidence that antepartum monitoring significantly improves perinatal outcomes; therefore, indications for NST remain relative and it is mainly used in patients at high risk of perinatal death [2, 4]. It is a non-invasive, cost-effective, and easy-to-apply technique [5]. Previous studies evaluating NSTs considered non-reactive due to having only one acceleration found no differences in fetal well-being compared with tracings that had two accelerations [6]. Thus, the presence of accelerations seems to reflect fetal well-being; however, their absence does not predict fetal compromise; a possible explanation for false-positive rates exceeding 90% reported by some investigators [4]. Consequently, some authors have proposed a physiologic interpretation of a “normal/reassuring” tracing in which the NST shows a baseline FHR within the normal range, moderate baseline variability, absence of decelerations, and the presence of one or no accelerations [6–8]. Evidence regarding the perinatal prognosis of normal NSTs with one or no accelerations compared with those classified as reactive remains insufficient [9]. Indeed, very high false-positive NST rates have been reported [4]. International guidelines continue to use the classic dichotomous definition (reactive/non-reactive). In this context, it is relevant to compare the predictive value and clinical safety of conventional NST interpretation (reactive/non-reactive) versus NSTs considered “fetal well-being preserved/normal” (moderate variability, one or no acceleration, no pathologic decelerations).

The general objective of this study is to compare perinatal outcomes and predictive capacity of NSTs interpreted as reactive/non-reactive versus NSTs considered normal or “fetal well-being preserved (reassuring)” in patients deemed by their clinicians to be high-risk pregnancies with ≥ 34 weeks’ gestation.

Methodology

Study design

Observational, multicenter and comparative non-inferiority study of pregnant patients with ≥ 34 weeks’ gestation considered high risk by their treating clinicians and undergoing antepartum EFM.

Study population

Conducted in three hospitals in the Republic of Panama: Complejo Hospitalario Dr. Arnulfo Arias Madrid (Panama City), Complejo Hospitalario Dr. Manuel Amador Guerrero (Colón), and Hospital Rafael Estévez (Coclé). We included high-risk pregnancies ≥ 34 weeks’ gestation referred by their treating clinicians for EFM to monitor fetal health. Study period: August 2023 to November 2024. Exclusions: age < 18 years, twin pregnancy, known fetal malformations, labor.

Ethics

Approved by the Research Bioethics Committee of the Complejo Hospitalario Dr. Arnulfo Arias Madrid (CSS), reference CBI-CHDrAAM-056-2023. Verbal and written consent to participate was necessary in all women.

Procedure, interpretation, and reporting

Eligible patients were verbally informed about the study; those who agreed received written informed consent and were given as much time as needed to read and sign. This was done when EFM was reported as reactive or as fetal well-being preserved (normal). NSTs were performed with the patient comfortably supine, without dietary restrictions, in the morning. The study was conducted with patients seen in the high-risk clinic (outpatients) and with patients hospitalized for some pathology and under conservative management. All these patients receive an EFM every week, independent of this study and the EFM interpretation is not accompanied by an ultrasound. The EFM classification or interpretation used in this study is already used by clinicians, however it has not been investigated.

For this study, the EFM was interpreted by the treating physician in the consultation or in the hospital; if there was doubt in the interpretation, one of the authors of the study was consulted so that together they could make the final report.

A tracing/NST was defined as reactive when baseline FHR was 110–160 beats per minute, variability 6–25 [beats/min], no bradycardia, and at least two accelerations within 20 min (Fig. 1A). The presence of variable decelerations was not used to categorize tracings as non-reactive or to exclude cases. An NST with fetal well-being preserved (normal, reassuring) was defined as baseline FHR 110–160 beats per minute, variability 6–25 [beats/min], no bradycardia, and one or no acceleration within 20 min (Fig. 1 B1, B2, B3). If the clinician (attending physician) considered the trace to be abnormal or required more time to analyze it (> 20 min) the case was not included in the study. The presence of variable decelerations was not used to define “well-being not preserved” or to exclude cases. For this study, the duration of EFM was 20 min.

Fig. 1.

A. Reactive non-stress test (NST). Baseline FHR: 120 beats per minute; ≥ 2 FHR accelerations; normal baseline variability; no FHR decelerations. B1. Fetal well-being preserved (normal) NST. Baseline FHR: 150 beats per minute; < 2 FHR accelerations; normal baseline variability; no FHR decelerations. B2. Fetal well-being preserved (normal) NST. Baseline FHR: 122 beats per minute; < 2 FHR accelerations; normal baseline variability with 1 acceleration; no FHR decelerations. B3. Fetal well-being preserved (normal) NST with variable deceleration. Baseline FHR: 140 beats per minute; > 2 FHR accelerations; normal baseline variability with variable deceleration

Baseline FHR

Refers to the average FHR maintained for a minimum of 2 min with moderate variability during a 10-minute window, in the absence of accelerations or decelerations. Normal baseline FHR is 110–160 beats per minute.

Variability

An important indicator of FHR functionality: it is the amplitude between the maximum and minimum FHR within one minute (i.e., variability is a number of beats representing the amplitude). Variability is absent if no fluctuation is present; minimal if ≤ 5; moderate (normal) when the fluctuation amplitude is 6–25 beats; and marked if > 25.

Accelerations

Increases in FHR above baseline by ≥ 15 beats lasting ≥ 15 s. Vibroacoustic stimulation was not used in the patients who entered the study.

Variable decelerations

Occur abruptly, number 1–3, and last < 30 s per 20-minute tracing. Their presence does not influence the final categorization.

Termination of pregnancy

At the time of pregnancy termination, clinicians were unaware whether the interpretation of the NST was part of a study. They knew of the existence of a normal reported NST of at least 7 days prior but did not know under which of the two criteria it had been reported. The tracing (paper) was not attached to the file at the time of birth, only the normal report, since the study was based on normal reports with one or the other method. This allowed clinicians to make decisions based on their own judgment, independent of the NST report. The plan was to recruit most of the patients in the high-risk obstetric outpatient clinic.

Data collected

For each patient, a data sheet recorded: NST reported as normal (reactive vs. fetal well-being preserved/reassuring), maternal/fetal high-risk indication, gestational age, maternal age, maternal weight, parity, medication use, second NST at 7 days, and perinatal outcomes (gestational age at birth, 1- and 5-minute Apgar, birthweight, NICU admission, perinatal death).

Statistical analysis

Based on a possible 7 stillbirths per 1,000 NSTs reported as reactive within 7 days of EFM [10], we planned to evaluate 1,200 EFM tests—600 in the “reactive EFM” group and 600 in the “fetal well-being preserved (reassuring)” group—to assess differences in perinatal mortality from the time of NST through 7 days, considering perinatal mortality as the result of a comparative non-inferiority study. Continuous variables are reported as mean and standard deviation (SD); categorical/binary variables as counts and percentages. Categorical comparisons used Pearson’s χ² with continuity correction or Fisher’s exact test, as appropriate. Student’s t test compared continuous variables between groups. No correction was applied. Two-sided p < 0.05 was considered statistically significant.

Results

We analyzed 1,206 antepartum EFM NSTs from 699 high-risk pregnancies ≥ 34 weeks’ gestation across the 3 hospitals over 16 months; 42% (507) of NSTs were performed 7 days after the initial NST. Iron was the most frequently used medication among these patients; hypertension was the most common diagnosis overall (Table 1). There were no significant differences between groups in gestational age at birth, time from the first NST to delivery, mode of delivery, 1- or 5-minute Apgar scores, or the presence of variable decelerations (Table 2). No perinatal death was associated with follow-up according to NST report among the 1,206 NSTs studied. Twelve cases (0.99%) were reported as abnormal at the second NST performed 7 days after the initial one. In those 12 cases, one case per group had reported variable decelerations, no perinatal deaths occurred; two cases had a 5-minute Apgar < 7, one per reporting group. Regarding the study’s primary outcome (perinatal mortality), there were 6 deaths with no significant differences related to how the initial NST was reported (Table 2). Among the 6 deaths in the entire cohort, no altered fetal monitoring had been reported. Two deaths occurred between the initial NST and 7 days later (one per group), and the other 4 occurred after the second NST. These deaths were associated with maternal conditions (two placental abruptions—one in chronic hypertension with diabetes, the other with severe preeclampsia), two cases with fetal growth restriction, one due to intrapartum complications (fetal distress), and one due to subsequent course (neonatal sepsis). NICU admissions did not differ significantly between reporting groups.

Table 1.

Baseline characteristics

Characteristic*	Reactive NST	Reassuring NST
Initial NST reported reactive, n(%)	350(50.0)	349(50.0)
Hospitals
CHM Hospital	213(60.9)	203(58.2)
Aguadulce Hospital	119(34.0)	128(36.7)
Colón Hospital	18(5.1)	18(5.1)
NST 7 days later	249(71.1)	258(73.9)
Most used medications**
Iron	115(32.9)	110(31.5)
Vitamins	82(23.4)	89(25.5)
Amlodipine	67(19.1)	64(18.3)
Folic acid	55(15.7)	58(16.6)
Aspirin	20(5.7)	24(6.9)
Insulin	13(3.7)	12(3.4)
Metformin	13(3.7)	10(2.9)
Most common diagnoses†
Hypertension	207(59.1)	205(57.1)
Maternal age ≥ 35 years	100(28.6)	109(31.2)
Diabetes	99(28.2)	102(28.4)
Obesity	37(10.6)	30((8.6)
Fetal growth restriction (FGR)	10(2.9)	7(2.0)
Maternal toxoplasmosis	7(2.0)	4(1.1)
Thrombophilias	4(1.1)	5(1.4)

*No significant difference between groups. NST = antepartum electronic fetal monitoring non-stress test 

**Some patients used more than one medication. † Some patients had more than one diagnosis

Table 2.

Outcomes comparing reactive NST vs. fetal well-being preserved (reassuring) NST

Variable	Reactive NST	Reassuring NST	p
Patients (1st NST)	350	349	–
Total NSTs (1,206), n (%)	597 (49.5)	609 (50.5)	–
Vaginal delivery, n (%)	224 (64)	201 (58)	0.08
Apgar 1, mean (SD)	8.5 (1.0)	8.5 (1.0)	0.52
-Apgar < 5, n (%)	1(0.2)	1(0.2)	1.00
Apgar 5, mean (SD)	8.9 (0.5)	8.8 (0.8)	0.46
-Apgar < 7, n (%)	1(0.2)	1(0.2)	1.00
Initial GA, weeks (SD)	36.9 (1.8)	36.5 (1.9)	0.08
Days from initial NST to birth, mean (SD)	14.1 (10.6)	15.0 (10.8)	0.26
Parity, mean (SD)	2.4 (1.4)	2.2 (1.2)	0.08
Maternal weight, kg (SD)	82.6 (17.7)	84.0 (16.6)	0.27
Birthweight, g (SD)	3,150 (573)	3,240 (490)	0.02
1–3 variable decelerations, n (%)	89 (25)	69 (20)	0.09
NST abnormal ≥ 7 days after initial, n (%)*	7 (2.0)	5 (1.4)	0.31
Perinatal mortality**, n (%)
Neonatal death, n (%)	3 (0.9)	1 (0.3)	0.31
In utero death, n (%)	1 (0.3)	1 (0.3)	1.00
NICU admission, n (%)	17 (4.9)	19 (5.4)	0.60

* None with perinatal death, ** None with an abnormal NST

Discussion

In this observational non-inferiority study of 1,206 NSTs, reporting monitoring as reactive (≥ 2 accelerations) or as fetal well-being preserved/reassuring (normal variability with ≤ 1 acceleration) led to no large differences in fetal well-being outcomes. That is, if moderate (normal) FHR variability is present with one or no acceleration, outcomes are similar to those with two accelerations. In addition, among high-risk pregnancies ≥ 34 weeks, repeating the NST at ≥ 7 days appears safe for fetal surveillance.

NST was introduced in the 1960s to continuously monitor FHR and uterine contractions with the goal of reducing perinatal deaths [11]. Unfortunately, despite decades of use, uncertainties persist regarding interpretation and outcomes [2, 7]. The pathophysiology by which hypoxia causes absent, or diminished FHR variability is unknown. However, reduced or absent variability may reflect hypoxia, acidemia, or central nervous system dysfunction due to placental, umbilical cord, or fetal factors. Fetal hypoxemia leads to acidemia, which is associated with adverse neonatal outcomes including neonatal encephalopathy, cerebral palsy, and death [12].

Antenatal fetal surveillance tests have high false-positive rates [4, 9] (non-reactive results in non-compromised fetuses) and low positive predictive value (low risk of fetal death after an abnormal result). Nevertheless, many clinicians act upon such reports; over a quarter (27.3%) of primary cesarean deliveries in the United States have been attributed to management based on non-reactive NST reports [13].

Overall, the best NST parameter to predict adverse perinatal outcome appears to be decreased or absent variability [7, 14, 15]. Indeed, such variability is directly associated with absent accelerations, whereas with moderate (normal) variability, accelerations are frequently observed. Loss or reduction of variability plus decelerations is frequently associated with fetal acidemia [4, 5, 7, 11, 16, 17]. This association has also been reported during intrapartum monitoring [18].

Our findings highlight that when baseline FHR is 110–160 with moderate variability (6–25 [beats/min]) and one or no acceleration during 20 min, there is no fetal compromise; this is equivalent to having two or more accelerations. Interpreting NSTs this way could avoid extending tests beyond 20 min.

Up to 50% of variable decelerations can be observed in otherwise normal NSTs [19, 20]; therefore, their mere presence is not a sign of intrauterine fetal compromise. In our study, 1–3 variable decelerations per 20-minute NST were seen in 22.6% of the population, without significant differences between groups. These findings support the limited importance of such decelerations in the NST context. In the presence of oligohydramnios, variable decelerations require different interpretation/analysis [9].

Regarding optimal timing to repeat an NST after a normal report, evidence-based recommendations are scarce due to the challenges of conducting prospective trials in pregnancies complicated by stillbirth and the heterogeneity of high-risk conditions. Therefore, antenatal surveillance recommendations largely rely on expert consensus and observational studies [2, 7]. A false-negative risk of approximately 1.9 per 1,000 has been suggested after a non-stress test [21]. No large randomized trials guide recommended frequency; thus, the optimal interval remains unknown. In our study, after a reactive NST or an NST with fetal well-being preserved, there was one perinatal death per group at 7 days, both associated with maternal intrapartum complications and therefore excluded as NST false negatives. Only 0.99% of cases with initial NSTs had an abnormal NST 7 days later, with no significant differences by initial reporting group. Thus, in high-risk pregnancies where FHR variability is normal, the presence or absence of accelerations indicates preserved/reassuring fetal well-being in 99% of cases over the next 7 days.

Similar studies with larger sample sizes, possibly randomized designs, and varying intervals for repeating the NST are needed before generalizing these findings.

This study presents strengths such as comparative design, strategy that shortens study time without compromising patient safety, large number of tests, multicenter and addresses an evergreen topic. Some limitations as non-random group allocation, absence of perinatal deaths directly detected by NST raises the possible underpower due to low perinatal mortality.

In summary, In this observational non-inferiority study, among high-risk pregnancies ≥ 34 weeks with an NST showing moderate variability, the presence of 0–1 accelerations was not associated with worse short-term perinatal outcomes compared to ≥ 2 accelerations. However, the study was underpowered for rare events such as perinatal death, and the 7-day repeat interval requires further validation in larger, randomized trials.

Abbreviations

EFM

Electronic fetal monitoring

NST

Non-stress test

FHR

Fetal heart rate

CSS

Caja de Seguro Social

NICU

Neonatal intensive carae unit

SD

Standard deviation

Authors' contributions

PVDG conceived the study, designed the methodology, and supervised data analysis. CRP, CCA, KG, GS, AG, and RM contributed to study design refinement, data interpretation, and manuscript drafting. PVDG performed statistical analysis and prepared the initial manuscript draft. All authors critically reviewed and approved the final version of the manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Approved by the Research Bioethics Committee of the Complejo Hospitalario Dr. Arnulfo Arias Madrid (CSS), reference CBI-CHDrAAM-056-2023. Verbal and written consent to participate was necessary in all women.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.