Objective
To examine the association between electronic fetal heart rate monitoring and neonatal and infant mortality, as well as neonatal morbidity.
Study Design
We used the United States 2004 linked birth and infant death data. Multivariable log-binomial regression models were fitted to estimate risk ratio for association between electronic fetal heart rate monitoring and mortality, while adjusting for potential confounders.
Results
In 2004, 89% of singleton pregnancies had electronic fetal heart rate monitoring. Electronic fetal heart rate monitoring was associated with significantly lower infant mortality (adjusted relative risk, 0.75); this was mainly driven by the lower risk of early neonatal mortality (adjusted relative risk, 0.50). In low-risk pregnancies, electronic fetal heart rate monitoring was associated with decreased risk for Apgar scores <4 at 5 minutes (relative risk, 0.54); in high-risk pregnancies, with decreased risk of neonatal seizures (relative risk, 0.65).
Conclusion
In the United States, the use of electronic fetal heart rate monitoring was associated with a substantial decrease in early neonatal mortality and morbidity that lowered infant mortality.
During labor electronic fetal monitoring (EFM) is used to assure well-being because the inexplicable interplay of antenatal complications, inadequate placental perfusion, and intrapartum events can lead to adverse outcomes. Even uncomplicated pregnancies are monitored for asphyxial injury and intrapartum death. Indeed, EFM during labor is the most common obstetric procedure in the United States. From 1997 to 2003 in the United States, EFM was used in 84% of the over 27 million births.
See related editorial, page 455
For Editors’ Commentary, see Table of Contents
Despite the ubiquitous use, there are concerns about the efficacy of EFM. As noted by the American College of Obstetricians and Gynecologists (ACOG) practice bulletin, the efficacy of monitoring is adjudicated by comparing the neonatal morbidity, including seizure and cerebral palsy, or mortality averted vs the unnecessary interventions (operative vaginal or cesarean delivery) undertaken. Because all the randomized clinical trials (RCTs) with EFM compare it with intermittent auscultation (IA), the efficacy is determined by calculating the relative risk (RR) of interventions, neonatal seizure, cerebral palsy, or death. Compared with IA, EFM is associated with a significantly increased likelihood of operative vaginal delivery, overall cesarean delivery, as well as with nonreassuring fetal heart rate tracing or fetal acidosis. Though the use of EFM and intrapartum interventions significantly decreases the rate of neonatal seizures, its use is not associated with a significantly lower rate of cerebral palsy or of neonatal death. A recent Cochrane review by Alfirevic et al reported that EFM was associated with 1 additional cesarean delivery for every 58 women monitored continuously and 661 women would have to have EFM during labor to prevent 1 neonatal seizure.
Although the efficacy of EFM is debatable, it is noteworthy that there are some concerns regarding the 12 RCTs, which sampled 37,000 women. Only 2 of these trials are of high quality, and only 3 trials reported data in low-risk women. The risk of cerebral palsy was ascertained in one trial, which randomized newborn infants at <32 weeks and risk of hypoxic ischemic encephalopathy by another. The combined sample size of 12 RCTs is insufficient to determine whether EFM can significantly lower neonatal mortality. Alfirevic et al noted that to test the hypothesis that continuous monitoring can prevent 1 death in 1000 births, more than 50,000 women need randomization. In addition, there are concerns that a metaanalysis that combines results of RCTs published before the introduction of the CONSORT (Consolidated Standards of Reporting Trials) guidelines or with inadequate study sample size may not reflect the outcomes in actual practice. Thus, we sought to determine the efficacy of EFM by comparing the outcomes among women who were vs were not monitored electronically during labor.
The primary objective of this study was to examine the association between EFM during labor and corrected neonatal and infant mortality in the United States. The secondary objectives were to assess the relative risk of operative vaginal delivery or primary cesarean delivery as well as neonatal morbidity (Apgar score <4 at 5 minutes or neonatal seizures) by EFM status.
Materials and Methods
We used the US 2004 birth cohort linked birth/infant death dataset assembled by the National Center for Health Statistics. Because this deidentified data are publicly available, our study was not considered as “human subjects research,” and did not require approval from the institutional review board.
The study population consisted of singleton live births documented using the 1989 revision of the standard certificate of live birth, used by 41 states and the District of Columbia. We restricted the study to US residents, and pregnancies that ended between 24 and 44 weeks in gestation. We excluded newborn infants with congenital anomaly or implausible birthweight-gestational age combinations based on an algorithm developed by Alexander et al. We also excluded births delivered by repeat cesarean sections. The primary outcomes included neonatal and infant mortality (death within the first year). We further categorized neonatal deaths as early neonatal (deaths within the first 6 days), late neonatal (deaths between 7-27 days), and postneonatal mortality (deaths between 28-364 days).
In the regression analyses, the primary exposure variable was EFM status during labor (yes vs no). Covariates that were considered as potential confounders included maternal age (categorized as <20, 20-34, and ≥35 years), maternal race/ethnicity (white/non-Hispanic whites, African-American/non-Hispanic blacks, Hispanic, and other races), maternal educational attainment (categorized into less than high school, completed high school, beyond high school), marital status (currently married, not currently married), self-reported tobacco use or alcohol use during pregnancy (yes vs no), and infant’s sex (male vs female). We limited our potential confounders because of concerns for overadjustment for variables in the causal pathway between the EFM and perinatal outcomes.
We calculated the risk of corrected early neonatal, late neonatal, postneonatal, and infant mortality rate for EFM and non-EFM groups. Multivariable log-binomial regression models were fitted to evaluate the association between EFM and mortality with adjustments for confounders. Results are presented as adjusted risk ratio (aRR) and 95% confidence intervals. We also calculated the “number needed to treat” (NNT) to avoid 1 adverse birth outcome ; that is, the number of women who need to have an EFM to prevent 1 death.
Additional secondary analyses were conducted. We compared the distribution of having operative vaginal or primary cesarean delivery based on EFM status. We estimated the relative risk of the neonatal morbidities, which include Apgar score at 5 minutes (<4 vs ≥4), neonatal seizure (yes vs no), or either 1, for all women, or those with gestational ages ≥37, <37, ≥34, <34 weeks, as well as in high- and low-risk women. Low risk was defined as all women excluding those with cardiac, acute or chronic lung disease, diabetes, hydramnios or oligohydramnios, hemoglobinopathy, chronic hypertension, pregnancy-induced hypertension, eclampsia, renal disease, Rh sensitization, uterine bleeding, and preterm delivery (gestational age <37 weeks). High risk was defined as gestational age <37 weeks or presence of any of the above high-risk factors.
The proportion of data with missing values for each covariate was <2%. All analyses were performed using SAS 9.2 (SAS Institute, Inc, Cary, NC.)
Results
In 2004, there were 4,118,956 live births in the United States. We excluded 6903 births to foreign resident mothers, 139,494 births with multiple gestation, 80,374 newborn infants delivered before 24 weeks or after 44 weeks, 13,847 infants with implausible birthweights, 422,396 infants delivered by repeat cesarean sections, 38,013 anomalous neonates, 793,571 births using the 2003 revision of the standard certificate, 888,774 infants from areas where variables used in the analyses were not reported, and 3373 infants with unknown use of EFM. After these exclusions, there were 1,732,211 singleton live births (42% of all live births) that comprised our study population, of whom 89% used EFM during labor. Table 1 presents the sample characteristics.
| Mortality | |||||
|---|---|---|---|---|---|
| Characteristics | Total live births (n = 1,732,211) | Early neonatal (n = 1568) | Late neonatal (n = 919) | Post-neonatal (n = 2927) | Infant (n = 5414) |
| Mother’s age, y | |||||
| <20 | 11% (196,225) | 17% (266) | 20% (186) | 21% (624) | 20% (1076) |
| 20–34 | 76% (1,322,142) | 71% (1120) | 70% (644) | 72% (2097) | 71% (3861) |
| ≥35 | 12% (213,844) | 12% (182) | 10% (89) | 7% (206) | 9% (477) |
| Mother’s race/ethnicity | |||||
| White/non-Hispanic | 63% (1,090,445) | 49% (755) | 50% (457) | 53% (1551) | 51% (2763) |
| Black/non-Hispanic | 16% (281,740) | 34% (523) | 33% (303) | 31% (915) | 32% (1741) |
| Hispanic | 14% (247,710) | 13% (195) | 13% (121) | 10% (298) | 11% (614) |
| Other/non-Hispanic | 6% (99,235) | 5% (80) | 4% (33) | 5% (146) | 5% (259) |
| Mother’s marital status | |||||
| Nonmarried | 37% (635,672) | 53% (836) | 58% (532) | 63% (1832) | 59% (3200) |
| Married | 63% (1,096,539) | 47% (732) | 42% (387) | 37% (1095) | 41% (2214) |
| Mother’s education | |||||
| Less than high school | 19% (332,758) | 26% (392) | 29% (255) | 35% (1022) | 32% (1669) |
| High school completed | 30% (515,557) | 36% (545) | 37% (326) | 36% (1051) | 36% (1922) |
| Above high school | 50% (863,027) | 37% (561) | 35% (309) | 28% (821) | 32% (1691) |
| Use of tobacco | |||||
| No | 89% (1,528,545) | 85% (1313) | 76% (696) | 72% (2089) | 77% (4098) |
| Yes | 11% (194,339) | 15% (233) | 24% (218) | 28% (806) | 23% (1257) |
| Use of alcohol | |||||
| No | 99% (1,710,120) | 98% (1519) | 98% (897) | 98% (2855) | 98% (5721) |
| Yes | 1% (12,457) | 2% (25) | 2% (18) | 2% (46) | 2% (89) |
| Infant sex | |||||
| Female | 49% (847,007) | 41% (636) | 40% (371) | 41% (1188) | 41% (2195) |
| Male | 51% (885,204) | 59% (932) | 60% (548) | 59% (1739) | 59% (3219) |
| Electronic fetal monitoring | |||||
| No | 11% (195,940) | 21% (324) | 12% (112) | 12% (349) | 14% (785) |
| Yes | 89% (1,536,271) | 79% (1244) | 88% (807) | 88% (2578) | 86% (4629) |
The corrected early neonatal, late neonatal, postneonatal, and infant mortality rates for all subjects were 0.9, 0.5, 1.7, and 3.1 per 1000 births, respectively. Figure 1 presents the mortality rate by EFM groups. The risk of corrected mortality rate was different between those with vs without EFM during the early neonatal period (0.8 vs 1.7 per 1000 births, respectively; P < .001), but not in late (0.5 vs 0.6; P = .402) or postneonatal periods (1.7 vs 1.8; P = .296). The corrected infant mortality rate was 3.0 per 1000 births for those with EFM and was significantly lower than 4.0 per 1000 births for those without EFM ( P < .001).
Gestational age-specific early neonatal mortality differed significantly among those with vs without EFM ( Figure 2 ). Table 2 presents the adjusted RR of mortality and neonatal morbidity, according to gestational age at birth and EFM use. Having EFM during labor was associated with a lower risk of early neonatal deaths (RR, 0.50) and a reduced risk of having Apgar score <4, regardless of gestational age. EFM was also significantly associated with a lower risk of infant mortality (RR, 0.75), particularly for births within 24-27, 28-31, 32-33, and ≥37 gestational weeks.
| Gestational ages, wk | Mortality | Morbidity | |||||
|---|---|---|---|---|---|---|---|
| Early neonatal | Late neonatal | Postneonatal | Infant | Apgar score <4 | Neonatal seizures | Apgar score <4 or neonatal seizures | |
| (0-6 d) | (7-27 d) | (28-364 d) | (0-364 d) | ||||
| All | 0.50 (0.44–0.57) b | 0.91 (0.74–1.12) | 0.91 (0.81–1.02) | 0.75 (0.69–0.81) b | 0.54 (0.49–0.59) b | 0.85 (0.70–1.03) | 0.60 (0.55–0.65) b |
| 24-27 | 0.63 (0.54–0.72) b | 0.97 (0.72–1.31) | 1.06 (0.77–1.45) | 0.78 (0.70–0.87) b | 0.58 (0.50–0.67) b | IC a | 0.61 (0.53–0.72) b |
| 28-31 | 0.51 (0.37–0.70) b | 1.25 (0.67–2.33) | 0.98 (0.65–1.49) | 0.75 (0.60–0.95) b | 0.55 (0.41–0.74) b | 0.49 (0.16–1.49) | 0.57 (0.43–0.76) b |
| 32-33 | 0.39 (0.21–0.75) b | 0.88 (0.31–2.52) | 0.83 (0.47–1.49) | 0.66 (0.44–0.98) b | 0.49 (0.31–0.78) b | IC a | 0.55 (0.34–0.86) b |
| 34-36 | 0.41 (0.24–0.71) b | 1.08 (0.54–2.15) | 1.03 (0.73–1.44) | 0.87 (0.67–1.13) | 0.48 (0.36–0.64) b | 0.80 (0.45–1.43) | 0.54 (0.42–0.70) b |
| >37 | 0.65 (0.47–0.90) b | 0.90 (0.63–1.27) | 0.89 (0.78–1.03) | 0.86 (0.76–0.97) b | 0.62 (0.54–0.71) b | 0.86 (0.70–1.07) | 0.68 (0.61–0.77) b |
a Incalculable because there were 10 or less neonatal seizures;
Table 3 presents the number needed to treat (NNT) (the number of EFMs that need to be performed) to prevent 1 death. There is a positive correlation between gestational age and NNT to prevent early neonatal and infant death. For infant mortality, the NNT ranged from as low as 1:15 for gestations 24-27 weeks to as high as 1:4078 for term gestations.
| Gestational ages, wk | Early neonatal mortality | Infant mortality | ||
|---|---|---|---|---|
| NNT | 95% CI | NNT | 95% CI | |
| All | 1266 | 1073–1542 | 1012 | 798–1352 |
| 24-27 | 12 | 9–17 | 15 | 10–28 |
| 28-31 | 71 | 48–131 | 82 | 46–424 |
| 32-33 | 291 | 174–887 | 217 | 112–3472 |
| 34-36 | 1449 | 912–3529 | 1753 | NNTB 611–infine–NNTH 2016 |
| ≥37 | 10,949 | 6275–42,921 | 4078 | 2232–23,646 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
