Fetal scalp blood sampling in labor – a review

Authors

  • Jan S. Jørgensen,

    Corresponding author
    1. Department of Gynecology and Obstetrics, Odense University Hospital, University of Southern Denmark, Institute of Clinical Research, Perinatal Research Unit, Odense, Denmark
    • Correspondence

      Jan S. Jørgensen, Department of Gynecology and Obstetrics, Odense University Hospital, Sdr. Boulevard 29, 5000 Odense C, Denmark.

      E-mail: jan.stener.joergensen@rsyd.dk

    Search for more papers by this author
  • Tom Weber

    1. Department of Gynecology and Obstetrics, Hvidovre Hospital, University of Copenhagen, Hvidovre, Denmark
    Search for more papers by this author

  • The authors have stated explicitly that there are no conflicts of interest in connection with this article; however, both authors hold a strong conviction that FBS is a beneficial supplement to CTG.

Abstract

During the 1970s and 1980s, electronic fetal monitoring and fetal scalp blood sampling were introduced without robust evidence. With a methodical review of the published literature, and using one randomized controlled trial, seven controlled studies, nine randomized studies of various surveillance methods and data from the Danish National Birth Registry, we have assessed the usefulness of fetal scalp blood sampling as a complementary tool to improve the specificity and sensitivity of electronic cardiotocography. Based on heterogeneous studies of modest quality with somewhat inconsistent results, we conclude that fetal scalp blood sampling in conjunction with cardiotocography can reduce the risk of operative delivery. Fetal scalp blood sampling can provide additional information on fetal wellbeing and fetal reserves at a time before decisions are made concerning the need for and timing of operative delivery and the choice of anesthesia, and be an adjunct in the interpretation of cardiotocography patterns.

Abbreviations
CI

confidence interval

CS

cesarean section

CTG

cardiotocography

FBS

fetal scalp blood sampling

RCT

randomized controlled trial

Key Message

As a supplement to cardiotocography, fetal blood sampling can:

  • Reduce the rate of operative delivery.
  • Provide more accurate information on the fetal status and reserves before decisions on operative delivery and choice of anesthesia are made.

Introduction

Following the initial development of electronic fetal motoring by Hon and Petrie [1], during the 1970s and 1980s this surveillance modality was rapidly introduced into obstetric practice, spreading to most high- and middle-income countries. The introduction was made on the assumption that electronic fetal motoring would help to reduce cerebral palsy (CP) caused by intrapartum asphyxia [2]. What actually happened was that the incidence of severe intrapartum asphyxia was only slightly reduced and CP rates among infants born at term remained unchanged. What changed dramatically was the cesarean section (CS) rate, which rose four- to fivefold between 1970 and 1990 [3]. One reason for this was that whereas a normal cardiotocography (CTG) tracing is a strong indicator of the absence of intrapartum asphyxia, the interpretation of CTGs carries significant inter- and intra-observer variation and a non-reassuring or abnormal CTG may not be associated with acidosis and subsequent asphyxia. False positives from high sensitivity but low specificity may result in unnecessary interventions such as operative vaginal delivery and emergency CS [4]. A supplementary test was needed to increase the specificity of CTG and to reduce the number of unnecessary interventions. Fetal scalp blood sampling (FBS) to measure scalp pH and lactate was introduced as a tool to improve the performance of electronic fetal motoring. This review summarizes the development and evaluates the effect and usefulness of FBS as a supplement to CTG.

History of biochemical fetal monitoring during labor

In Berlin in 1962, Erich Saling first reported the use of FBS during labor. He described how with an amnioscope, a light source and a knife (originally a razor blade), he could collect by mouth suction, drops of fetal scalp blood into a heparinized tube. These were then transferred to an acid-base laboratory for rapid analysis and reporting of the fetal acid-base balance [5]. Almost 20 years later, he reported on the number of fetal deaths over a 30-year period among 50 000 deliveries during the change in fetal monitoring from auscultation alone (1955–60) to auscultation in combination with FBS (1961–67) to CTG and FBS (1968–79). Intrapartum deaths fell in these three time periods from 0.8 to 0.32%, and then to 0.15%. He credited FBS as the main reason for this decline [6]. FBS was introduced as an additional clinical test in many obstetric units during the 1970s and 1980s and remains in daily use [7].

Normally, pH is measured in the fetal scalp blood. Usually 15–50 μL is required (≈ three to four drops of blood) for measurement in an acid-base machine/laboratory. A pH value >7.25 is regarded as normal, demonstrating fetal wellbeing and normal oxygenation. Values between 7.25 and 7.20 are regarded as sub-normal and require extra vigilance and repeat sampling within 20–30 min. Values of pH < 7.20 (or <7.15 in the second stage of labor) are early warnings of fetal hypoxia requiring intervention such as intrauterine resuscitation or operative delivery [8].

Lactate concentration in fetal scalp blood samples has also been investigated and tested [9] and found to be reliable, although normal values need to be derived for every different lactate measurement system. Normal values have been described as values of <4.2 mmol/L. Values between 4.2 and 4.8 are considered intermediate, and values >4.8 mmol require intervention. Fetal monitoring with scalp blood lactate has been compared with pH monitoring, and umbilical cord blood lactate has been shown to correlate well with corresponding pH and base excess values [10-12].

During the 1980s and 1990s, several electrodes and systems were investigated in the hope of finding a biochemical method that could be used to monitor fetal wellbeing continuously, such as fetal tissue pH [13, 14], fetal transcutaneous pO2 [15] and pCO2 [16], fetal base excess [17] and fetal pulse oxymetry [18]. None of these methods are currently used in clinical practice.

Material and methods

A computerized search of the literature using Medline, Embase and Cochrane databases was conducted in September 2013. We used the MeSH terms: “Fetal scalp blood sampling and asphyxia” with 72 different combinations of key words related to asphyxia and eight to fetal scalp blood sampling. No language restrictions were employed. One of the authors (J.S.J.) conducted the search assisted by a research librarian. The study selection and the data collection process were deliberately conducted with a special focus on the risk of selection bias and are presented in a flow chart (Figure 1).

Figure 1.

Study selection flow chart.

We included randomized controlled trials (RCT) comparing fetal monitoring (auscultation and/or electronic fetal motoring) with and without FBS, and non-randomized prospective studies with interventions and controls. We also included retrospective controlled studies, studies of outcome before and after introduction or elimination of FBS in one or more institutions and RCTs of other types of fetal monitoring systems including FBS. Studies on sampling time and success as well as reports on FBS complications, and comparisons between FBS-lactate and FBS-pH were included. Finally, reviews were searched for relevant research studies.

Data were extracted from each study included with special focus on: (i) reduction in operative deliveries, fetal outcome, sampling success rates and complications; (ii) external validity (numbers and characteristics of patients, country and year of origin); (iii) characteristics of studies (RCTs, prospective or retrospective, inter- or intra-institutional, intervention and control group). The characteristics of the included studies were summarized, organized and presented in 2 × 2 contingency tables and text. We performed a power calculation for a future RCT with material large enough to prove that FBS might reduce unwanted obstetrical outcomes [19].

Data from the Danish National Birth Registry were retrieved for information on the use of FBS in relation to fetal outcome and operative deliveries. All women (99.8%) giving birth in Denmark are reported to the Danish Medical Birth Registry (= 443 281 between 2005 and 2011). The population selected for the present study was women with term and post-term deliveries (= 413 423). The data were analyzed with the software package BASE-SAS 9-2 (SAS Institute, Cary, NC, USA). Selected outcomes were analyzed for changes in trend over time in the percentage of deliveries using the Cochran-Armitage test (Z-value) with a one-sided p-value of <0.05 regarded as significant for the observed trend.

Results

The database search provided a total of 548 citations. After removal of duplicates, 440 remained. A total of 38 studies were identified for inclusion in the final review (Figure 1).

Only one prospective RCT compared auscultation with CTG and CTG plus FBS [20]. This was a study of 695 high-risk deliveries where 232 were monitored by auscultation alone, 233 by CTG alone, and 230 by CTG and FBS. The three groups were comparable and there were no significant differences in fetal outcome at delivery or at 9 months' follow-up [21]. When comparing CTG with CTG plus FBS, there was a reduction of CS (18% vs. 11%), the relative risk for a vaginal delivery in the CTG plus FBS group was 1.07 [95% confidence interval (CI) 1.00–1.15] and the number needed to treat to prevent one CS was 15.9 (95% CI 7.9–1192.9). The difference was not significant but was supported by several non-randomized retrospective studies showing that the use of FBS was paralleled by a reduction in the rate of CS of 6–53% [22-26] (Table 1).

Table 1. Studies with historical and geographical controls showing that FBS paralleled reductions of obstetric interventions (Pro FBS) and studies showing that FBS did not improve obstetrical outcomes (Con FBS)
Author(s)Country, yearResults/conclusionPopulation, characteristics and study period
  1. FBS, fetal blood sampling; CS, cesarean section.

Pro FBS
Zalar and Quilligan [25]USA, 1979156 of 298 (53%) avoided CS for fetal distressAnalysis of 298 deliveries with FBS in 1976. Historical controls from 1970, 1974 and 1977, = 33 000
Young et al. [26]Canada, 198010% reduction in CS for fetal distress

232 FBS in a 9 month period

Retrospective review of CTG traces and outcome

Total number or rates of deliveries and CS = not stated

Ayromlooi and Garfinkel [27]USA, 1980FBS decreases the rate of cesarean section as a means of dealing with fetal distress from 78.0 to 57.5% without a significant change in the Apgar score

Total number of deliveries before and after:

1974: 3039; 1977: 3542

Total number or rates of CS = not stated

Irvine and Shaw [28]UK, 198932% reduction in CS rate for fetal distress

Periods compared. (I) 1/11/1985–31/10/86 with (II) 1/11/1986–31/10/87

Number of CS for fetal distress: Period: (I) 82 – (II) 55

Total number of deliveries: (I/II) 2000/1956 and CS: 314/294

Reif et al. [29]Germany, 2011Operative delivery could be avoided in 6.4% of study population in spite of the non-reassuring CTG

Retrospective study of 669 women with FBS performed in 2008–2009

Total number of deliveries in the period: 4437/All CS: 733 (16.5%); of these, 230 CS (31.4%) were for fetal distress

Con FBS
Perkins [29, 49]USA, early 1980sAbsence of FBS did not result in extraordinary rates of stillbirth, operative intervention, or neonatal compromise

Three years of perinatal results in a predominantly high-risk population of 7100 deliveries compared with similar populations managed in other institutions, with and without FBS

Total numbers of rates of deliveries, CS and FBS from other institutions = not stated

Goodwin et al. [31]USA, 1986–1992Fetal scalp blood sampling has been virtually eliminated without an increase in the CS rate for fetal distress, or an increase in perinatal asphyxiaFBS, CS for fetal distress, Apgar scores, asphyxia and meconium aspiration 1986–92 (= 112 000) FBS rate the first 3 years 1.76% to a low of 0.03% in 1992

In a meta-analysis of nine randomized trials, auscultation was compared with CTG with and without the use of FBS [27]. The auscultation group had 6474 monitored deliveries and the CTG plus FBS group 6490. The odds ratio for CS was 4.14 (95% CI 2.29–7.15) in the studies not including FBS and 1.98 (95% CI 1.33–2.94) in the studies including FBS, which indicated that the use of FBS may be associated with a non-significant reduction in CS. The neonatal outcome did not differ significantly between the groups apart from the risk of neonatal convulsions, which was halved in the CTG plus FBS group (odds ratio 0.49; 95% CI 0.29–0.82).

In the 1980s in the USA, former spokesmen for the use of FBS published statements of retrospective data from their units comparing obstetric outcomes before and after the use of FBS or between units using FBS or not [28, 29]. They did not find any benefit from FBS. However, despite dealing with high-risk pregnancies, they reported very low frequencies in the use of FBS of <1.5–2% (Table 1).

Fetal monitoring by stimulation of the fetal scalp and acoustic stimulation [30-32] as well as fetal ECG (PR-interval analysis) supplementary to CTG monitoring [33] were also investigated. The studies described a reduction in the need for FBS, albeit with limitations (Table 2). A prospective study of 100 FBS reported a success rate of 89% and a median duration of “decision to result” of 18 min (inter-quartile range 12–25), but with 9% lasting more than 30 min [34].

Table 2. Fetal sclap blood sampling (FBS) compared with other tests of fetal wellbeing
Author(s) and titleCountry, yearResults/conclusion

Ingemarsson and Arulkumaran [32]

“Reactive fetal heart rate response to vibroacoustic stimulation in fetuses with low scalp blood pH”

Sweden/Singapore, 1989Mean FBS pH values significantly higher in fetuses that showed fetal heart rate acceleration compared with those who had no response or a deceleration after vibro-acoustic stimulation (pH 7.30 and 7.22, respectively). Acidotic scalp blood pH values (7.16 and 7.18) seen in two fetuses with reactive responses to vibro-acoustic stimulation and pain stimuli by FBS

Elimian et al. [31]

“Intrapartum assessment of fetal well-being: a comparison of scalp stimulation with scalp blood pH sampling”

USA, 1997Fetal scalp stimulation reduced the need of FBS by 50%; minor risk of false negative tests (2/21 with non-reactive CTG but normal variabilty had scalp-pH < 7.15).

Van Wijngaarden et al. [33]

“Improved intrapartum surveillance with PR interval analysis of the fetal electrocardiogram: a randomized trial showing a reduction in fetal blood sampling

USA, 1996Fetal ECG analysis (PR-analysis) + electronic fetal monitoring reduced need for FBS and increased its efficacy without increasing adverse outcomes

Twelve articles reported a total of 37 cases of severe complications following FBS [26, 35-46]. These were principally scalp hemorrhage (n = 21) including fetal collapse and death (n = 2) and lacerations of the fetal scalp with the need for suturing (n = 1). In addition, severe infections including scalp abscess (n = 1) and neonatal sepsis (n = 3) and retained scalpel fragments (n = 6) were reported. In Odense and Copenhagen between 2003 and 2013, with the use of FBS procedures in more than 25 000 patients, no such severe complications have been observed. Two articles reported conflicting results with respect to transmission of HIV when using FBS. One article reported a higher risk of transmission [47], whereas in another, the authors disagreed [48].

The Danish Birth Registry started to register FBS from 2005. From 2005 to 2011 the use of FBS rose from 3.8 to 6% of all term and post-term deliveries (Figure 2a–d). The emergency CS rate in the same cohort declined slightly from 8.2% to 7.9%. The vacuum extraction rate declined significantly from 8.5% to 7.9%. These figures did not disclose whether signs of asphyxia or failure to progress in labor were the indications for emergency CS and vacuum extraction. In the same period, the frequency of umbilical artery pH < 7.05 fell significantly from 0.74% to 0.34%, but the incomplete and increasing rate of registration of umbilical artery pH among normal parturients implied that these figures could not be interpreted. The registration of the Apgar score at 5 min, however, has been constant and valid, and the frequency of Apgar scores below 7 has remained constant around 0.5%.

Figure 2.

National data from the Danish Medical Birth Registry. (a) Emergency caesarean section (CS); (b) Vacuum extraction; (c) Umbilical cord artery pH < 7.05 (UCA pH < 7.05); and (d) Apgar score at 5 min <7 in relation to rates of fetal scalp blood sampling (FBS). The Z-value represents the trend. The p-values the statistical significance of the trend. See text for further interpretation. GA, gestational age.

A power calculation has been made for future RCTs powered to address the hypothesis that FBS might reduce adverse obstetric outcomes such as (i) intrapartum asphyxia (measured by umbilical cord artery pH < 7.05 and/or Apgar score at 5 min <7); (ii) emergency CS and (iii) vacuum extraction. Based on historical Danish national figures from the time when the use of FBS was not yet introduced systematically, and assuming a possible reduction of asphyxia by 25% (from an incidence of 1% in controls to 0.75% in the intervention group), of emergency CS by 25% (from and incidence of 10% in controls to 7.5% in the intervention group), and vacuum extraction by 50% (from and incidence of 10% in controls to 5% in the intervention group), the numbers needed in each arm would be 21 000, 2004 and 434, respectively.

Discussion

In the 1970s and 1980s, both CTG and FBS were introduced into obstetric practice around the world without proper validation. No large-scale RCTs were conducted beforehand which were powered to demonstrate measurable clinical benefits with respect to improvement of fetal or maternal outcome. A search of the literature yielded only one RCT comparing CTG monitoring with and without FBS [20]. This single 35-year-old study was not sufficiently powered to demonstrate any statistically significant beneficial effect but it did demonstrate a clear trend towards a reduction in the number of unnecessary CS. There were no differences in short-term or long-term outcomes for babies born in the study population or at follow-up [21]. In reviewing nine RCTs comparing auscultation compared with CTG with or without FBS [30], there was a dramatic reduction in CS of more than 50% with the use of FBS, with the same fetal outcome, apart from a halving in the rate of neonatal convulsions in those who were monitored with CTG plus FBS.

Several retrospective observational studies, primarily from the USA, lend support to these findings [22-26]. A few retrospective studies from USA researchers, initially in favor of FBS, claimed that abandoning this procedure in their large units did not negatively affect the neonatal outcome [29, 49]. We speculate that these findings were due to a very low frequency of the use of FBS. A common feature of the reviewed studies is that they were conducted many years ago, and the literature search which addressed the risk of selection and measurement bias was undermined by the very low number of relevant studies. A considerable heterogeneity also exists. Nevertheless, these studies indirectly indicate that the use of FBS in conjunction with CTG reduces the risk of CS.

Although scalp blood lactate measurement requires only a very small amount of blood, this method has not been shown to improve the clinical usefulness compared with scalp blood pH monitoring [11]. Lactate changes correlate well with changes in base excess and it would be interesting to compare the performance of these parameters. However, there are no data evaluating full acid base balance on fetal scalp blood with suggested cut-off values for intervention [10].

Tested against other fetal monitoring systems such as fetal scalp stimulation and fetal ECG (PR-interval) analysis, FBS performs similarly. The studies are few and the numbers are small [30-33]. Only one systematic but small study on the FBS success rate and sampling time was found [34] and this reported a clinically recognizable success rate of nearly 90%, but a very long median sampling duration of 18 min. This is not in keeping with our experience, where most FBS are taken with a sufficient result within five to 10 min depending on dilation of the cervix and descent of the fetal presenting part.

Despite the large number of FBS taken worldwide over the last three decades, only a few serious complications of FBS have been reported [23, 35-46]. Most of these reports date back to the introduction of FBS with the uncontrolled use of scalpel knife-blades and the need for a substantial volume of blood (>125 μL) for a blood-gas analysis. In our own substantial units in the last decade, we have not seen any serious complications. To put our findings into perspective and for further research purposes, we conducted a power calculation for a hypothetical RCT to reduce unwanted obstetric outcomes using FBS. A 25% reduction of intrapartum asphyxia and of emergency CS would require 21 000 and 2004 cases in each arm, respectively, whereas a halving of vacuum extractions would only require 434 women in each arm. Thus a study which could include STAN (ST analysis) would be feasible – especially with combined maternal and fetal outcomes.

The figures from the RCT and the indirect comparison studies are supplemented by the data from the Danish Birth Registry indicating that in a recent 7-year period with over 400 000 term deliveries in a population with an already low incidence of intrapartum asphyxia, the frequency of low Apgar scores remained unchanged. However, low Apgar scores may be caused by many other conditions than asphyxia, such as infection, malformation and metabolic diseases. Therefore, a slight reduction in the incidence of asphyxia is not necessarily followed by a significant reduction in the rate of low Apgar scores In the same time period, when the registered rate of FBS increased, the rates of both emergency CS and vacuum extraction were reduced slightly, by 0.3 and 0.6%, respectively.

Conclusion

From a single old RCT and heterogeneous and sometimes inconsistent studies of modest quality it seems that FBS reduces operative delivery and possibly even neonatal asphyxia. In Denmark, an increasing use of FBS in conjunction with CTG seems to be associated with a reduction in operative delivery. However, our experience and expert opinion, from more than 25 years of frequent and daily use of FBS, is that it has prevented unnecessary emergency CS despite non-reassuring CTG traces. For doctors in training, FBS is a useful tool to provide a deeper understanding of the underlying physiology and interpretation of CTG traces. Furthermore, FBS provides more accurate information on fetal wellbeing and fetal metabolic reserves in addition to CTG before decisions are made with respect to urgency of operative delivery and type of anesthesia.

Acknowledgements

The authors would like to thank Tove Faber Frandsen, chief librarian at “Videncentret,” Odense University Hospital, Denmark, for her hard work and help in designing and conducting the systematic literature search.

Ancillary