OBJECTIVE: The majority of intrapartum fetal hypoxia occurs in uncomplicated pregnancies. Current intrapartum monitoring techniques have not resulted in a reduction in the incidence of cerebral palsy in term neonates. We report the development of a composite risk score to allow risk stratification of normal pregnancies before labor.
METHODS: Six hundred one women were recruited to this prospective observational study. All women underwent an ultrasound examination before active labor, during which fetal biometry and fetal Doppler flow resistance indices were measured. A composite risk score, amalgamating data from the umbilical artery, middle cerebral artery, and umbilical vein, was then developed and correlated with intrapartum outcomes.
RESULTS: In cases with the highest composite risk scores, the incidence of fetal compromise (the primary outcome) was 80.0% compared with just 15.3% in cases with the lowest risk scores (relative risk 5.2, 95% confidence interval 2.7–10.1). These cases were also at increased risk of cesarean delivery (53.3% compared with 3.4%, P<.001) and of developing a fetal heart rate pattern considered pathologic by National Institute for Health and Clinical Excellence criteria (P=.003). No significant variation in Apgar scores or umbilical artery pH was observed.
CONCLUSION: Intrapartum fetal compromise remains a significant global health issue. The composite risk score reported here can identify fetuses at both high risk and low risk of a subsequent diagnosis of intrapartum fetal compromise. This may enable more judicious use of current intrapartum fetal monitoring techniques, which are hampered by low specificity.
LEVEL OF EVIDENCE: II
Ultrasound assessment before active labor can identify fetuses at high risk and low risk of subsequent intrapartum fetal compromise.
Centre for Fetal Care, Queen Charlotte's and Chelsea Hospital, and the Institute for Reproductive and Developmental Biology, Imperial College London, London, United Kingdom; and the Mater Research Institute–University of Queensland, South Brisbane, Queensland, Australia.
Corresponding author: Professor Sailesh Kumar, Mater Research Institute/University of Queensland, Level 3, Aubigny Place, Raymond Terrace, South Brisbane, Queensland 4101, Australia; e-mail: firstname.lastname@example.org.
Financial Disclosure The authors did not report any potential conflicts of interest.
Current methods of intrapartum surveillance have failed to reduce the incidence of cerebral palsy in term neonates1 despite a significant increase in operative delivery rates.2 Although risk stratification is used in many areas of obstetric practice3,4 our ability to prospectively assess the risk of intrapartum fetal compromise is limited. Women without identified antenatal complications are currently considered “low risk,” but most cases of intrapartum fetal hypoxia occur within this normal cohort.5 These cases are also overrepresented in medicolegal litigation cases6 and represent a significant financial burden to health care providers.7
Intrapartum cardiotocography has been criticized for its high false-positive rate8 and is also associated with significant intra and interobserver disagreement.9 Fetal pulse oximetry10 and fetal ST segment analysis11 have been used to supplement cardiotocography, but all these techniques have minimal predictive value before labor.
Identification of fetuses at risk of compromise before labor would enable a more targeted approach to intrapartum care, potentially reducing the effect of diagnostic intrapartum tests with low specificity as well as empowering women to make more informed birthing choices.
Altered umbilical artery and middle cerebral artery (MCA) resistance indices and decreased umbilical vein flow are present in growth-restricted fetuses.12,13 Our group has recently demonstrated that Doppler indices of these vessels measured in early labor are predictive of subsequent intrapartum compromise.14,15
We report the development of a composite risk score, amalgamating data from the umbilical and middle cerebral arteries and umbilical vein, for use in risk stratification of normal, apparently low-risk pregnancies before labor.
MATERIALS AND METHODS
Six hundred one women were recruited to this study over a 24-month period between March 2011 and March 2013. All women were recruited from the delivery suite, antenatal ward, and day assessment unit at Queen Charlotte's and Chelsea Hospital. This is a tertiary referral maternity unit in London, United Kingdom. Patients were recruited at term (37–42 weeks of gestation) in early labor (4 cm or less cervical dilatation). Women with both spontaneous and induced onset of labor were considered eligible for inclusion. Those undergoing induction of labor were only considered eligible if the indication for this was postdates pregnancy (before 42 weeks of gestation) or if induction had been planned for social reasons. Women undergoing induction as a result of fetal or maternal complications were not included. Exclusion criteria included multiple pregnancy or known fetal anomaly, known or suspected fetal growth restriction (estimated fetal weight less than 10th percentile with umbilical artery pulsatility index greater than the 95th percentile) or placental dysfunction (small for gestational age or reduced growth velocity) identified during routine antenatal care, maternal hypertension or preeclampsia, cervical dilatation greater than 4 cm (indicating active labor), ruptured membranes with meconium stained liquor, or evidence of intrauterine infection. The aim was to recruit women with apparently uncomplicated, low-risk pregnancies, with appropriately grown fetuses, who would deliver within 72 hours of recruitment to this study. All women received written information about the study and gave informed written consent before their inclusion. Ethical approval for this project was granted by the London Research Ethics Committee (Reference No. REC 10/H0718/26).
After recruitment, an ultrasound scan was performed to measure fetal biometry as well as Doppler resistance indices in the umbilical artery, umbilical vein, and MCA. All ultrasound scans were performed by a single trained practitioner using a Voluson e ultrasound machine with a 4- to 8-MHz curvilinear transabdominal transducer. In a subcohort of 50 women, measurements were repeated by the same operator after a 30-minute interval to establish intraobserver variability and by a different operator to establish interobserver variability. To avoid caval compression, women were positioned in a slightly left tilt supine position with the head of the bed elevated. Resistance indices were measured using pulse wave Doppler with triplicate readings taken and the mean value used for data analysis. Doppler waveforms were not recorded during periods of fetal breathing movements or during contractions. The angle of insonation was kept as close to 0° as possible, and always less than 30°, to ensure the accuracy of recorded waveforms. The umbilical vein was also imaged in gray scale, in longitudinal section, and its internal diameter recorded. Umbilical venous flow rate was then calculated (assuming laminar flow of a newtonian fluid in a cylindrical vessel) using the following formula:
The use of this formula for calculation of umbilical venous flow has previously been reported.15,16
Clinical staff managing the labor were blinded to the ultrasound results and retained responsibility for all decisions regarding intrapartum care. At our institution, all decisions for cesarean delivery are made by the attending physician.
All participants were followed up within 48 hours of delivery. Medical records were reviewed and intrapartum and neonatal outcome details recorded. The indication for delivery was noted as recorded by the obstetric team. Participants were then divided quantitatively into three groups (less than 10th percentile, 10th–90th percentile, and greater than 90th percentile) based on the pulsatility index of each of the fetal vessels (and flow rate in the case of the umbilical vein). Points (0–2) were awarded to each case for each Doppler parameter (Table 1). The minimum number of points each participant could accumulate was zero, whereas the maximum number of points was eight. Participants were then subclassified according to the scores from all fetal vessels and intrapartum and neonatal outcomes compared. The incidence of diagnoses of fetal compromise was compared between the risk score groups using the χ2 test and maternal demographic characteristics compared using a combination of χ2 and analysis of variance and Kruskal-Wallis tests.
The main outcomes measures for this study were mode of delivery and the presence or absence of a diagnosis of intrapartum fetal compromise. The diagnosis of fetal compromise, made by the obstetric team, was based on abnormal fetal heart rate monitoring (interpreted according to National Institute for Health and Clinical Excellence criteria, which are similar to those of the American College of Obstetricians and Gynecologists; see the Appendix17), abnormal fetal blood sampling, or both. In all cases, to ensure consistency of interpretation, intrapartum monitoring was reviewed by two senior members of the obstetric team before a decision for intrapartum cesarean delivery for presumed fetal compromise was performed. Other outcome measures included the presence or absence of meconium-stained liquor and fetal heart rate abnormalities as well as neonatal outcomes including Apgar scores and umbilical artery pH at delivery.
This study was powered (power of 0.80 and significance level of <.05) to demonstrate a difference of 60% in the incidence of a diagnosis of fetal compromise between pregnancies with the highest and lowest risk score groups. Data from a pilot study of 100 participants was used to estimate effect size and the likely distribution of cases among the risk score groups. The study was not powered to investigate rare adverse neonatal outcomes such as neonatal encephalopathy.
Six hundred one women were recruited to the study over a 24-month period. Maternal demographics were recorded for all participants (Table 2). The mean gestational age at delivery was 40.5 weeks (range 37.0–42.0 weeks of gestation), and the mean birth weight was 3,531 g (range 1,780–5,026 g). Within the study population, 40 neonates were identified to have a birth weight less than the 10th percentile for gestation. Intrapartum outcomes are presented with these cases included and excluded (Table 3).
The mean umbilical artery pulsatility index for the cohort was 0.80 (range 0.46–1.53). The 10th percentile value for the umbilical artery pulsatility index was 0.64 and the 90th percentile value was 0.98. Sixty-three participants (10.5%) had pulsatility indices less than the 10th percentile, 475 participants (79.0%) had pulsatility indices between the 10th and 90th percentile, and 63 participants (10.5%) had pulsatility indices greater than the 90th percentile (see Table 1 for points allocation).
The mean MCA pulsatility index for the cohort was 1.37 (range 0.68–2.32). The 10th percentile value for the MCA pulsatility index was 1.04, and the 90th percentile value was 1.71. There were 61 cases (10.1%) with pulsatility indices less than the 10th percentile, 480 cases (79.9%) had pulsatility indices between the 10th and 90th percentiles, and 60 cases (10.0%) had pulsatility indices greater than the 90th percentile (see Table 1 for points allocation).
The mean cerebral–umbilical ratio was 1.77 (range 0.56–3.15). The 10th percentile value for the cerebral–umbilical ratio was 1.26, and the 90th percentile value was 2.32. There were 66 participants (11.0%) with a cerebral–umbilical ratio less than the 10th percentile, 475 participants (79.0%) had cerebral–umbilical ratios between the 10th and 90th percentile, and 60 participants (10.0%) had a cerebral–umbilical ratio greater than the 90th percentile (see Table 1 for points allocation).
The mean umbilical venous flow was 60.4 mL/min/kg. The 10th percentile value for umbilical venous flow was 41.6 mL/min/kg and the 90th percentile value was 79.1 mL/min/kg. Sixty participants (10.0%) had an umbilical venous flow rate of less than the 10th percentile, 481 participants (79.6%) had an umbilical venous flow rate between the 10th and 90th percentile, and 60 participants (10.0%) had an umbilical venous flow rate greater than the 90th percentile (see Table 1 for points allocation). Using the mean flow velocity, and mean umbilical vein diameter, the Reynolds number for umbilical venous flow was calculated as 1,030. This supports our assumption of laminar flow in this vessel.
The interobserver and intraobserver variability of these measurements was satisfactory with correlation coefficients 0.94 or greater in all cases.
The mean accumulated score was then compared between neonates born by different modes of delivery (Fig. 1). Neonates born by emergency caesarean delivery for presumed fetal compromise had the highest mean score at 4.6 (median 4, range 2–8), whereas neonates born by instrumental delivery for a prolonged second stage had the lowest mean score at 3.8 (median 4, range 0–6). Neonates born by normal vaginal delivery had a mean score of 3.9 (median 4, range 1–7) and those born by instrumental delivery for presumed fetal compromise had a mean score of 4.2 (median 4, range 1–7). The distribution of scores was significantly different between the mode of delivery groups (Kruskal-Wallis P<.001). Other than a higher percentage of primiparous women undergoing operative delivery, no significant differences in maternal demographics were observed between the different mode of delivery groups (data not presented).
All participants were then categorized according to their accumulated score and delivery outcomes compared between the different groups (Table 3). Fetuses with the highest scores (7 or 8) had the highest incidence of emergency cesarean delivery for presumed fetal compromise with an incidence of 53.3% seen in this group. In comparison, fetuses with the lowest scores (0–2) had an incidence of cesarean delivery for presumed fetal compromise of just 3.4%. When compared with fetuses with the lowest scores, those with the highest scores were at increased risk of cesarean delivery for presumed fetal compromise with a relative risk of more than 15. These fetuses were also more likely to be diagnosed with fetal compromise during the labor and were less likely to be born vaginally (either instrumental or spontaneous vaginal delivery). Maternal demographic parameters were compared among the four different risk score groups (Table 4). Significant differences were found only in the mean gestation at assessment, which was lower (40.1 weeks of gestation) in the lowest scoring group than in the other groups (40.5, 40.7, and 40.5 weeks of gestation).
A risk score of 7–8 gave a positive likelihood ratio for fetal compromise during labor of 9.7 (assuming a score of 7–8 is a positive test and a score of less than 7 is a negative test), whereas a score of 0–2 gave a negative likelihood ratio of 0.4 (assuming a score of 0–2 indicates a negative test and a score of greater than 2 a positive test). A score of 7–8 had a positive predictive value for a diagnosis of fetal compromise during labor of 80.0%, a sensitivity of 6.8%, and a specificity of 99.3%. A score of 0–2 had a negative predictive value for a diagnosis of fetal compromise during labor of 84.8%, a sensitivity of 94.9%, and a specificity of 11.8%.
Intrapartum outcomes were also evaluated (Table 5). Fetuses with higher risk scores were significantly more likely to subsequently be diagnosed with a pathologic cardiotocography (P=.003). Fetuses with the highest risk scores were more than six times more likely to subsequently have a pathologic fetal heart rate pattern during labor than those with the lowest risk scores. These fetuses were also found to have a significantly lower umbilical artery oxygen partial pressure at delivery (P=.003) (Table 6). No significant differences in the incidence of meconium-stained liquor or cardiotocographies classified as suspicious were observed among the different risk score groups. Fetuses in the highest risk score groups were significantly smaller (birth weight- and gestation-matched birth weight percentile) than those in the lowest risk score groups (P=.01 and <.001, respectively). Excluding the difference in umbilical artery oxygen partial pressure at delivery, no significant differences were observed in any of the neonatal outcome variables (Table 6) among the different risk score groups.
In this article, we report the development of a composite risk score for the prediction of intrapartum fetal compromise, in appropriately grown term neonates, using simple Doppler measurements from three fetal vessels. Fetuses with the highest scores were at significantly increased risk of being diagnosed with fetal compromise during labor and of requiring emergency cesarean delivery. They were also significantly less likely to be born vaginally. Although multivessel fetal Doppler is used in the management of fetal growth restriction, its value in the appropriately grown fetus has not been proven.
Our group has previously demonstrated that the umbilical artery pulsatility index, MCA pulsatility index, cerebral–umbilical ratio,14 and umbilical venous flow rate,15 when measured in early labor, vary significantly between fetuses subsequently born by different modes of delivery. Although all had good negative predictive value, particularly the cerebral–umbilical ratio (negative predictive value 99%), their positive predictive value was relatively low (36%). By combining these parameters into a composite risk score, we now demonstrate that the positive predictive value can be improved to greater than 50% for the likelihood of intrapartum compromise for neonates with the highest risk scores while maintaining good negative predictive value for those with the lowest risk scores. In this study, fetuses with the highest risk scores were almost 16 times more likely to require cesarean delivery for fetal compromise when compared with those with the lowest risk scores.
We did not observe any differences in neonatal outcomes between the neonates in the different risk score groups. However, this study was not powered to detect differences in neonatal outcomes. The neonatal outcome measures assessed in this study represent the condition of the fetus at the time of delivery rather than the time a decision for delivery was made. The fetal condition at delivery may be influenced by management instigated between these time points such as cessation of Syntocinon, potentially improving the fetal condition before caesarean delivery for fetal compromise or, conversely, a difficult instrumental delivery potentially worsening the fetal condition. Bloom et al18 compared neonatal outcomes in neonates delivered by emergency cesarean delivery for fetal compromise both within 30 minutes of the decision to deliver and after 30 minutes. Neonatal outcomes were worse in the group of neonates delivered fastest. The authors suggest this relates to the greater urgency with which the more clinically compromised neonates were delivered. However, it may also relate to a shorter period of “fetal resuscitation” before delivery.
Although the largest neonates were found in the group with the lowest risk scores, no significant differences in maternal demographic characteristics were observed among the different risk score groups. The greater size of neonates in the lowest risk score groups may suggest that placental function is better in these pregnancies, resulting in a lower incidence of intrapartum compromise and emergency delivery.
Potential limitations of this study include the presumptive nature of the diagnosis of fetal compromise, which was based on clinical interpretation of fetal heart rate patterns supplemented by the clinical scenario given that there is no gold standard test for fetal compromise. A second potential limitation is the inclusion of women undergoing induction of labor in the study. However, grouped linear regression analysis confirmed that the Doppler parameters were not significantly different between women undergoing induction and those in spontaneous labor. Although significant differences in gestation and birth weight were observed between the different composite risk score groups, the absolute differences were small (218 g and 0.6 weeks) and are unlikely to be of clinical significance or to have had a confounding influence. We also acknowledge that this study was powered to demonstrate a difference in the diagnosis of fetal compromise only and not in the other delivery outcomes.
Continuous intrapartum fetal heart rate monitoring is associated with high false-positive rates.8 As a result, current U.K. guidance issued by the National Institute for Health and Clinical Excellence advises against its use in “low-risk” women.17 However, as demonstrated by our results, some of those assumed to be “low risk” based on their antenatal history are actually at considerable risk of intrapartum fetal compromise. Ultrasound assessment at term, although not being a perfect test, does provide a much more accurate assessment of risk than antenatal history alone and as such could be used to better target intrapartum monitoring to truly “high-risk” groups.
The recent Birthplace in England and Wales study19 observed reduced intervention rates in all nonobstetric birth place settings, but also high transfer rates to an obstetric unit in nulliparous women. It is possible that transfer rates could be reduced by using a risk stratification tool such as that presented here. Risk stratification before labor could be used to better identify those pregnancies that warrant closer intrapartum monitoring as well as those that do not, potentially reducing unnecessary intervention and improving patient satisfaction rates.
Normal—a cardiotocography trace in which all four features are classified as reassuring.
Suspicious—a cardiotocography trace with one feature classified as nonreassuring and all others as reassuring.
Pathologic—a cardiotocography trace with two or more features classified as nonreassuring or one or more classified as abnormal.
National Institute for Health and Clinical Excellence Criteria for Cardiotocography Classification (Clinical Guideline 55: Intrapartum care, National Institute for Health and Clinical Excellence, October 2012) Classification of cardiotocography features
○ Nonreassuring—100–109, 161–180
○ Abnormal—less than 100, greater than 180, sinusoidal pattern for greater than 10 min
○ Reassuring—greater than 5
○ Nonreassuring—less than 5 for 40–90 min
○ Abnormal—less than 5 for greater than 90 min
○ Nonreassuring—typical variable deceleration with more than 50% of contractions for more than 90 min or single prolonged deceleration for up to 3 min
○ Abnormal—atypical decelerations with more than 50% of contractions or late decelerations for more than 30 min or a single prolonged deceleration for greater than 3 min
○ The absence of accelerations with an otherwise normal cardiotocography trace is of uncertain significance
Modified from National Institute for Health and Clinical Excellence. Intrapartum care: Care of healthy women and their babies during childbirth. Available at: http://www.nice.org.uk/nicemedia/live/11837/36280/36280.pdf. Retrieved November 1, 2013. The content is accurate at the time of going to press. Cited Here...
1. Clark SL, Hankins GDV. Temporal and demographic trends in cerebral palsy—fact and fiction. Am J Obstet Gynecol 2003;188:628–33.
2. Alfirevic Z, Devane D, Gyte G. Continuous cardiotocography (CTG) as a form of electronic fetal monitoring (EFM) for fetal assessment during labour. The Cochrane Database of Systematic Reviews 2013, Issue 5. Art. No.: CD006066. DOI: 10.1002/14651858.CD006066.pub2.
3. Nicolaides KH. Screening for fetal aneuploidies at 11 to 13 weeks. Prenat Diagn 2011;31:7–15.
4. DeFranco EA, Lewis DF, Odibo AO. Improving the screening accuracy for preterm labor: is the combination of fetal fibronectin and cervical length in symptomatic patients a useful predictor of preterm birth? A systematic review. Am J Obstet Gynecol 2013;208:233.e1–6.
5. Low JA, Pickersgill H, Killen H, Derrick EJ. The prediction and prevention of intrapartum fetal asphyxia in term pregnancies. Am J Obstet Gynecol 2001;184:724–30.
6. Johnson SL, Blair E, Stanley FJ. Obstetric malpractice litigation and cerebral palsy in term infants. J Forensic Leg Med 2011;18:97–100.
7. NHS Litigation Authority. Ten years of maternity claims: an analysis of NHS litigation authority data. London (UK): NHS Litigation Authority; 2012.
8. Nelson KB, Dambrosia JM, Ting TY, Grether JK. Uncertain value of electronic fetal monitoring in predicting cerebral palsy. N Engl J Med 1996;334:613–8.
9. Bernardes J, Costa-Pereira A, Ayres-de-Campos D, Geijn HP, Pereira-Leite L. Evaluation of interobserver agreement of cardiotocograms. Int J Gynecol Obstet 1997;57:33–7.
10. Gardosi J, Carter M, Becket T. Continuous intrapartum monitoring of fetal oxygen saturation. Lancet 1989;2:692–3.
11. Westgate J, Harris M, Curnow JSH, Greene KR. Plymouth randomized trial of cardiotocogram only versus ST waveform plus cardiotocogram for intrapartum monitoring in 2400 cases. Am J Obstet Gynecol 1993;169:1151–60.
12. Harman C, Baschat A. Comprehensive assessment of fetal wellbeing: which Doppler tests should be performed? Curr Opin Obstet Gynecol 2003;15:147–57.
13. Rigano S, Bozzo M, Ferrazzi E, Bellotti M, Battaglia FC, Galan HL. Early and persistent reduction in umbilical vein blood flow in the growth-restricted fetus: a longitudinal study. Am J Obstet Gynecol 2001;185:834–8.
14. Prior T, Mullins E, Bennett P, Kumar S. Prediction of intrapartum fetal compromise using the cerebroumbilical ratio: a prospective observational study. Am J Obstet Gynecol 2013;208:124.e1–6.
15. Prior T, Mullins E, Bennett P, Kumar S. Umbilical venous flow rate in term fetuses: can variations in flow predict intrapartum compromise? Am J Obstet Gynecol 2014;210:61.e1–8.
16. Acharya G, Wilsgaard T, Rosvold Berntsen GK, Maltau JM, Kiserud T. Reference ranges for umbilical vein blood flow in the second half of pregnancy based on longitudinal data. Prenat Diagn 2005;25:99–111.
17. National Institute for Health and Care Excellence. Clinical Guideline 55: Intrapartum care. London (UK): National Institute for Health and Care Excellence; 2007.
18. Bloom SL, Leveno KJ, Spong CY, Gilbert S, Hauth JC, Landon MB, et al.. Decision-to-incision times and maternal and infant outcomes. Obstet Gynecol 2006;108:6–11.
© 2014 by The American College of Obstetricians and Gynecologists.
19. Birthplace in England Collaborative Group, Brocklehurst P, Hardy P, Hollowell J, Linsell L, Macfarlane A, et al.. Perinatal and maternal outcomes by planned place of birth for healthy women with low risk pregnancies: the Birthplace in England national prospective cohort study. BMJ 2011;343:d7400.