Fetal pulse oximetry: the end of a nice idea

Fetal pulse oximetry: the end of a nice idea

*Associate Professor and Director of Research
** Professor and Chair Department of Obstetrics, Gynecology and Reproductive Sciences, Division of Maternal-Fetal Medicine, University of Texas Houston Medical School, USA

Abstract

The goal of intrapartum fetal surveillance is to detect fetal hypoxemia before it progresses to severe acidemia, enabling the clinician to intervene before the stage of clinical impairment. One of the most recently proposed methods for fetal surveillance during labor was fetal pulse oximetry. The present review will provide a brief overview of the current status of fetal monitoring, and an outline of the evolution of fetal pulse oximetry from an international perspective. A large body of literature is available on the validity, reliability, and feasibility of fetal pulse oximetry. However, the ultimate end points in obstetrical care remain the neonatal and maternal outcomes. From this perspective, the aggregate interpretation of the 5 randomized clinical trials conducted so far does not demonstrate any medical value for fetal pulse oximetry in addition to the current standard of care. Keywords: fetal pulse oximetry, intrapartum fetal monitoring, fetal hypoxia, labor.

More frequently than not, labor is a physiological act perfectly tolerated by the fetus. For more than 100 years, intermittent auscultation of the fetal heart rate was all that the standard of care required. The practice of obstetrics however could not remain unaffected by the recent technological penetration in all the medical fields. Assessing the impact of modernization, or technical progress in obstetrics is a difficult task, primarily because in the large low risk population, major adverse outcomes such as intrapartum fetal death or neonatal encephalopathy are very rare (incidences of 6 per 1.000 livebirths and 1.6 per 10.000 births, respectively).1 Any outcome study having such endpoints would require a prohibitively large sample size to demonstrate even a minimal difference. By necessity, researchers have looked at related events, such as fetal/neonatal acidemia or neonatal depression, which are more frequent (roughly 1%), although not always indicative of intrapartum asphyxia, or a good predictor of long-term neurologic disability. The pathophysiology of acute intrapartum events is characterized by a fetal continuum including hypoxemia respiratory acidosis metabolic acidosis clinical impairment. The goal of intrapartum fetal surveillance is to detect fetal hypoxemia before it progresses to severe acidemia with subsequent perinatal mortality or long-term morbidity. It would be ideal if obstetricians could detect acidosis, as an indication of disturbed gas transfer, and intervene before the stage of clinical impairment. The last four decades of intrapartum fetal monitoring have witnessed various starts and stops of technologic advances that have alternatively looked promising or frustrating. One of the most recent methods touted for use during labor is fetal pulse oximetry, particularly appealing because it directly evaluates fetal oxygenation, the first step in the aforementioned pathophysiologic continuum. The present review will provide a brief overview of the current status of fetal monitoring, and an outline of the evolution of fetal pulse oximetry from an international perspective.

The assessment of intrapartum fetal well-being

Electronic fetal heart rate monitoring (EFM) was introduced with great optimism in the 1960s, and quickly adopted in the developed countries. Its widespread acceptance was not preceded by appropriate trials to determine its efficacy. Unfortunately, shortly after its generalization in practice, the reliability of subjective interpretations of fetal heart rate (FHR) tracings started to be questioned, and 30 years later, a meta- analysis of 12 randomized clinical trials involving 258,855 pregnant women cast doubt on the benefits of EFM. Besides mediocre benefits, it was suggested that EFM contributed to a significant increase in operative deliveries due to its high sensitivity and low specificity with respect to fetal hypoxemia and acidosis. Even the most ominous patterns on EFM are associated with low Apgar scores or acidosis in no more than 50% of cases. Because of the significant interpretation variability among practitioners, EFM was perceived to be in need of guidance, and the National Institute of Child Health and Human Development convened a panel of clinical experts in an attempt to standardize definitions for the visual interpretation of FHR tracings.But even after the publication of the standardized guidelines for FHR interpretation, the interobserver differences continued to be a common reality, rather than an exception.

Visual interpretation of FHR tracings is compromised not only by interobserver variation, but also by intraobserver variation, as demonstrated by Nielsen et al.The same clinician lacks consistency in her or his own interpretation of similar patterns. It was felt that computerized FHR analysis might help in eliminating subjectivity, increase reliability, and minimize diagnostic errors. However, the use of computer-based algorithms has not come even close to solving the problem. The application of such methods is not very effective because of the long-term non-stationary behavior of the FHR.FHR interpretation was obviously in need of assistance. It followed that improvement in the evaluation of intrapartum fetal well-being might come from the concomitant use of complementary methods of fetal assessment. Fetal scalp blood (FSB) sampling, proposed as early as 1962 by Saling,was viewed as such a technique able to provide a direct, objective fetal analyte, in contradistinction to the subjective FHR interpretation. It has been reported that the use of FSB sampling reduces the number of cesarean deliveries for fetal distress to approximately the same level as that observed in intrapartum women monitored by intermittent auscultation. However, the determination of the fetal acid-base status intrapartum, although a very accurate estimate of fetal well-being, is relatively cumbersome, and has not gained universal acceptance. Some authors even doubted its usefulness, making a point for its elimination from practice.

The technique is also only a snapshot in time, often requiring repeated measurements. A method providing a continuous measurement of fetal oxygenation or acid-base status, in addition to EFM, was obviously thought to be preferable, and even Erich Saling, the father of FSB sampling, at one point was a proponent of fetal pulse oximetry.Scalp stimulation and vibro-acoustic stimulation have also been proposed as methods to evaluate fetal well-being in conjuncture with EFM. These two methods were considered possible alternatives to FSB sampling. However, the initial enthusiasm was tampered by subsequent reports of a relatively high false-positive rate, and even high false- negative cases.More recently, fetal electrocardiogram ST-segment analysis has emerged as a promising intrapartum fetal monitoring method. Initially, it was shown to decrease the incidence of fetal acidemia and reduce the cesarean delivery rate in several European randomized controlled trials (RCTs). The STAN S21 device (Neoventa, Sweden) recently was approved by the Food and Drug Administration (FDA) in the United States. With the exception of a few feasibility studies, the method has not penetrated clinical practice in the United States yet. A disadvantage of the STAN S21 device is that it cannot be integrated into the available FHR monitors in the United States (Hewlett Packard or Corometrics). Furthermore, a recent Cochrane systematic review on the addition of fetal electrocardiogram monitoring reported no difference in the overall cesarean delivery rate when compared to EFM only.Ironically, after almost 4 decades of sophisticated attempts at modernization of intrapartum fetal monitoring, the clinical practice guidelines issued in 2002 by the Canadian Society of Obstetricians and Gynecologists specify that intermittent auscultation remains the preferred method of fetal surveillance in healthy pregnancies in spontaneous, non-augmented labor. For the remaining moderate or high-risk cases, EFM is warranted, with FSB sampling advisable when FHR tracing is uninterpretable or non-reassuring.

The development of fetal oximetry Another idea emerging in the late 1980s was to develop technology similar to pulse oximetry in air-breathing individuals, in order to measure fetal arterial oxyhemoglobin saturation (FSpO2) during labor. The original technology for pulse oximetry was developed by allied biomedical engineers during World World II, as an application in military aviation. It was introduced in clinical medicine in the 1970s, and the first attempts at extending it to intrapartum fetal assessment were made in the United Kingdom. Pulse oximetry provides information about the percentage of oxygen bound to hemoglobin by using principles of optical spectrophotometry and plethysmography. Oxyhemoglobin (oxygenated hemoglobin) and deoxyhemoglobin (hemoglobin without oxygen) absorb red and infrared light differently: more red absorption by deoxyhemoglobin, and more infrared absorption by oxyhemoglobin. By measuring the relative absorption at each wavelength, the fraction of hemoglobin that carries oxygen can be determined, and the arterial oxygen saturation is expressed as a percentage. Although the physical principles remain the same, the sensors used for assessment of adult or neonatal arterial oxygen saturation were not appropriate for intrapartum fetal pulse oximetry and had to be modified.

The differences between the fetal and postnatal physiology had to be considered in the calibration of sensors and the software design for fetal application. In addition, the sensors had to be calibrated for the expected fetal range of biological values since the normal fetal oxygen saturation is much lower, fetal hemoglobin has a higher affinity for oxygen and higher concentration, there are more capillaries per unit of tissue in the fetus, and fetuses have a higher cardiac output and heart rate. The most challenging aspect so far has been the prerequisite to assure a good contact between sensor and fetal skin in order to avoid artifacts. Much research has gone into overcoming the above-mentioned logistical difficulties. In a transmission sensor (used in the adult or neonate), the light produced by two light emitting diodes (LED), one for red and the other for infrared light, will be picked up by the detector after traversing the interposed tissues. Interposition of tissue is not achievable in the fetus. The reflectance sensor developed for fetal oximetry has the LED and detector placed side by side, instead of being opposite to each other, and, as the name implies, the light to be analyzed is reflected from the tissues. This latter design adds variance in reading pending on the depth of light tissue penetration and device position.

changes. The Nellcor fetal oximetry sensor (Nellcor, Pleasanton, CA), largely used in clinical trials, is placed against the fetal temple, cheek, or forehead, being held in place by the uterine wall. The sensor usually descends and rotates with the fetal head, displacements are quite frequent, and adjustments in sensor placement may be needed. Several other models have attempted to stabilize the sensor against the fetal head employing either a suction device that attaches to the fetal scalp (as in the Zurich Reflectance Pulse Oximetry System18), or a spiral wire scalp electrode (as in the Respironics oximeter19). Another technical difference between the Zurich and the Nellcor system is that the first one records one measure per second, whereas the latter one averages values every 45 seconds. Respironics (Respironics Inc., Pittsburgh, PA) is a transmission oximeter incorporated into a scalp electrode. The LED and the detector are opposite each other within the hollow needle of the scalp electrode that is placed 2-3 mm under the skin, concomitantly interposing a fixed tissue segment. The readings obtained with the Zurich or Respironics system can be influenced by caput formation with venous congestion or by the plethysmographic changes during maternal pushing efforts.

The prerequisites for insertion of a fetal oximetry sensor are in general dilatation of at least 2 cm, ruptured membranes, no placenta previa, cephalic presentation, single fetus, and gestational age at least 36 weeks. The Nellcor N-400 system was commercially available in many European countries since 1995, and in Canada since 1998. The device was approved for use in the United States in 2000 as an adjunct to EFM in conditions of a nonreassuring FHR pattern. The FDA conditional approval was issued after the first randomized trial of fetal pulse oximetry performed in the United States by Garite and colleagues.20 The study was conducted between 1995 and 1998 in nine academic institutions, involving 1,010 women with pre-defined nonreassuring FHR patterns in labor. The study hypothesis was that improvement in fetal assessment secondary to the addition of fetal oximetry would lead to a reduction in cesarean delivery rate, without alterations in neonatal outcome. Indeed, in the group in which the oximetry device was used along with EFM, there was a significant reduction in the rate of cesarean delivery performed for a nonreassuring FHR tracing (4.5% vs 10.2%; p = 0.007). Nevertheless, the rate of cesarean delivery for dystocia was higher in the same group, offsetting any advantage in the overall cesarean delivery rate (29% in the intervention group vs 26% in the control group).

There were no significant differences in neonatal outcomes between the two groups. The unexpected result of increased cesarean deliveries for dystocia in the intervention group in the Garites study raises several possibilities: (1) given the unblinded study design, it is possible that clinicians, circumspect of the pulse oximetry information, continued to perform cesarean deliveries for nonreassuring FHR tracing, but labeled differently the indication for surgery. The validity of the dystocia diagnosis was discredited by a subsequent partogram analysis that showed that the rate of arrested labor was similar between the groups, (2) the nonreassuring FHR pattern in conditions of normal fetal oxygenation is predictive of dystocia. Previous randomized studies of EFM had suggested the same thing21, and (3) dystocia is the consequence of the device itself. Anecdotal observations have suggested a higher rate of persistent occiput posterior positions with fetal oximetry use. In a 2001 American College of Obstetricians and Gynecologists (ACOG) Committee Opinion, concerns were expressed about the potential increase in obstetrical care costs without any demonstrable improvement in outcome with the use of fetal oximetry.By 2002, only approximately 10% of the obstetrical units in the United States had acquired the device.

Validity (or accuracy) and reliability (or precision) are measures of test performance that provide the foundation of all forms of scientific measurement. Validity is a concept that refers to how well an instrument measures what it purports to measure. Reliability implies the instruments capacity to give the same result - whether correct or not - on repeat testing (test-retest correlation or reproducibility). Reliability indicates that the instrument is free of measurement error, or random influence. A reliable test may be valid or invalid. An unreliable test is very unlikely to be valid. In medicine, validity (or accuracy) of a test underlies how successful the test is to distinguish between those with, and without an independently verified condition. The degree of correct categorization is expressed as measures such as sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The predictive values most directly and practically indicate the applicability of a test. In evaluating the accuracy of fetal pulse oximetry, the first question to ask may be: What does fetal pulse oximetry measure? Pulse oximetry is a noninvasive means of assessing the adequacy of oxygenation.

However, it gives only indirect information regarding the partial pressure of oxygen in the blood, and provides no data concerning perfusion or acid-base status. In other clinical applications, oxygen saturation is not an acceptable substitute for arterial blood gas analysis. The pulse oximeter is not a hemoximeter. Only the latter directly and reliably determines blood oxygen saturation by spectrophotometry.24 Even the calculated oxygen saturation values provided automatically by modern blood gas analysers are not very accurate.25 Luttkus et al, from Germany, have compared fetal oxygen saturation by hemoximetry in a FSB sample and fetal oxygen saturation by fetal pulse oximetry (FSpO2) immediately before the blood sampling.26 The FSpO2 medians were always higher than the hemoximetry saturation in the FSB sample, leading to false negative results in hypoxic babies. Other studies have looked at the correlation between pulse oximetry measurements and fetal arterial oxygen saturation. Although correlation cannot be perceived as a measure of accuracy, because it only gives information on the agreement of two instruments, the strength of the reported relations remains informative. In animal studies pulse oximetry measurements correlated well with the simultaneously measured arterial oxygen saturation (r = 0.98, p = 0.01).27 Human studies though, reached inconsistent results. While McNamara et al reported a good correlation between FSpO2 measurements and umbilical arterial blood oxygen saturation at birth (r = 0.59, p < 0.001),28 Langer et al could not find any relationship between FSpO2 levels determined during pushing efforts and oxygen saturation in umbilical venous blood at birth.29 Such ambiguous conclusions may be accounted for by differences in practice (use of umbilical venous or arterial blood, recording during pushing or between pushing efforts), different intervals from FSpO2 reading to umbilical blood sampling, or study groups that are not comparable (as an example, monitoring with fetal pulse oximetry all women in labor, or only those with FHR tracing abnormalities).

There are several limitations that should be taken into account when assessing the assumption that fetal pulse oximetry values correlate with fetal arterial oxygen saturation. Fetal pulse oximetry measures arterial oxygen saturation during the systolic pulse wave in the skin microcirculation at head level. This specific circulation sector in the fetus is part of the preductal circulation, with oxygen saturation levels somewhere between umbilical arterial and umbilical venous blood oxygen saturation. At least theoretically, FSpO2 should be closer to FSB sampling than umbilical blood sampling results. Although FSB samples consist of capillary blood, which is not exactly central arterial blood, we know that at least in the neonate, the differences are small.30 In the intrapartum fetus, several intervening variables possibly contribute to weaker relationships including different intervals between the last pulse oximetry signal and blood sampling after delivery, differences in local tissue perfusion status such as congestion with accumulation of nonpulsatile arterial or venous blood,31 and perfusion changes during fetal compromise, as the fetus will centralize its blood flow, with vasoconstriction in the skin circulation. From the very beginning in fetal pulse oximetry research, it was important to establish the critical threshold for fetal oxygen desaturation. Studies in catheterized fetal sheep model had suggested that the level below which metabolic acidosis can be anticipated was a FSpO2 of approximately 30%. Subsequent human studies looking at the distribution of FSpO2 values, indicated that a FSpO2 of 33% represents approximately the 10th percentile on the normal distribution, and a FSpO2 of 29-30 % would represent the 3rd - 5th percentile of values during normal outcome labor.

The threshold of 30% is also supported by prospective human data obtained in a German multicenter study.34 According to those data, a FSpO2 <30 % has 100% sensitivity in predicting a fetal scalp blood pH < 7.20. FSpO2 < 30% also correlated with lack of variability on the FHR tracing.35 It is important to understand that the FSpO2 cutoff value of 30% is not an indication of fetal distress, but rather a threshold at which increasing fetal acidosis will be encountered. Oxygen saturation is a dynamic biologic parameter with a relatively broad range of variation. Furthermore, the normal fetus seems to have a remarkable capacity to compensate for transient episodes of desaturation. A single reading cannot accurately reflect fetal condition; the trend in FSpO2 has to be taken into account. Transient, less than 2 minute periods of desaturation are very common, and only FSpO2 < 30% persisting over 2 minutes,36 or over 10 minutes37 are likely to be associated with development of intrapartum acidosis. More recently, Gorenberg et al, in a retrospective analysis, correlated FSpO2 with umbilical artery blood pH and found that the 30% threshold by itself, or even the duration of the longest episode of FSpO2 < 30% do not correlate well with fetal acidemia (umbilical artery blood pH < 7.20).38 Rather, the repetitive nature of such episodes is more predictive. The study was underpowered to detect a significant difference in acidemia, and the study design did not allow for sufficient observation time in order to detect the natural progression of hypoxia to metabolic acidosis, a better indicator of fetal compromise. The authors final impression was that more than 10 episodes of FSpO2 < 30% would overcome the ability of the fetus to compensate.

Given the study limitations, such a conclusion remains to be confirmed by other investigations. Many of the aforementioned studies were conducted under the assumption that fetal pulse oximetry values should correlate with fetal acid-base condition. It has been reported that whenever the oxygen saturation in the umbilical artery is  30%, acidosis (pH < 7.13) in the same blood is very rare, only 1%.39 The correlation between fetal pulse oximetry values and fetal acid-base status is much poorer.26 Leszczynska- Gorzelak et al, from Poland, could not demonstrate any relationship between FSpO2 levels in either 1st or 2nd stage of labor and pH, or partial pressure of oxygen in umbilical venous blood at delivery.40 Other authors are of the same opinion, considering that intrapartum FSpO2 is of limited use as a test for predicting acidosis at birth, regardless of the FSpO2 cutoff value used.41 More recent reports from the Netherlands42 and the Czech Republic43 conclude that fetal pulse oximetry is a poor diagnostic tool for fetal acidosis. Rijnders et al found no significant correlation between fetal scalp blood or umbilical artery blood pH and mean FSpO2 measurements for the last 30 minutes before the sampling (r = 0.02, p = 0.9).42 Even the lowest FSpO2 level did not correlate with umbilical artery blood pH (r = 0.04, p = 0.84). None of the three cases of umbilical artery blood pH < 7.05 in the Dutch study would have been detected using the mean FSpO2 value before delivery, and only one would have been detected using the lowest FSpO2.

In the Czech study, only 25% of cases with umbilical artery blood pH < 7.20 had FSpO2  30%, whereas 50% of the acidotic cases were associated with pathological fetal electrocardiogram (STAN21) recordings.43 In a French multicenter study conducted between 1994 and 1995 with the Nellcor pulse oximetry system in 164 cases with abnormal FHR tracing, a good correlation between fetal pulse oximetry and FSB sampling (r = 0.29, p < 0.01) was noted in the 1st stage of labor, but FSpO2 readings in the 2nd stage of labor did not correlate with either oxygen saturation, partial pressure of oxygen, pH, or bicarbonate level in the umbilical artery at birth.44 In an observational series of 128 fetuses with nonreassuring FHR pattern reported by Schmidt et al from Germany, only 2 of the 11 cases with umbilical artery blood pH < 7.20 were detected by pulse oximetry recordings < 30% during the last 30 minutes of the 2nd stage of labor, and out of 5 cases with hypoxic readings during the 2nd stage only 2 fetuses were acidotic at birth.

The calculated sensitivity was 18%, specificity 92%, PPV 40%, and NPV 80%. The authors also mentioned that a low Apgar score was never predicted by fetal pulse oximetry, and concluded that fetal distress is insufficiently identified by means of fetal pulse oximetry. Others, using the same fetal pulse oximetry Nellcor system over the final 30 minutes of labor, and a cutoff for umbilical blood acidemia of pH < 7.13, reported similar numbers: sensitivity 28%, specificity 94%, PPV 40%, and NPV 80%.41 Vitoratos et al from Greece, analyzing FSpO2 readings in active labor (not limited to the last 30 minutes before delivery), obtained somewhat better values: sensitivity 72%, specificity 93%, PPV 61.5%, and NPV 95.8% for an umbilical artery blood pH < 7.15.46 The impression that the validity of fetal pulse oximetry is higher in earlier labor than during the 2nd stage is also supported by data from Switzerland. Stiller et al, employing the Zurich pulse oximetry system at different times in labor (1st and 2nd stage), with a cutoff for hypoxemia of 33% and a cutoff for acidemia of umbilical artery blood pH  7.14 reported a sensitivity of 67 80%, and a specificity of 62 90%.47 The authors noted a decrease in PPV with imminence of birth and suggested that identification of fetal acidosis by means of fetal pulse oximetry becomes increasingly unreliable as birth approaches.

They considered that the reason might be loss of signal precision with increasing movement, pressure, and congestion. It is debatable whether there is any change in mean FSpO2 with advancing labor. Leszczynska-Gorzelak et al48 found a significant decrease in mean FSpO2 from the 1st stage to the 2nd stage of normal labor (51.9 % vs 43.8 %, p < 0.001), and Dildy et al33 noted a similar difference upon analyzing 160 normal labors (59% vs 53%). Other studies, in Bulgaria49 and France44, could not verify such differences. The Bulgarian study assessed only cases of normal labor and recorded mean FSpO2 values of 48.8 % in the 1st stage, not significantly different from 46% in the 2nd stage of labor.49 The aforementioned measures of validity (sensitivity, specificity, PPV, NPV) have been incorrectly presented as characteristics of fetal pulse oximetry. In all the studies, fetal pulse oximetry was not used independently, but rather as an adjunct to EFM. EFM has a sensitivity for fetal acidosis of 93%, specificity of 29%, PPV of 2.6%, and NPV of 99.5%.50 From a statistical point of view, whenever two evaluation methods that have the same end point (fetal acidosis) are combined, sensitivity decreases while specificity increases, theoretically resulting in less unnecessary intervention.

It was hoped that fetal oximetry success and applicability might not depend so much on the prediction of acidosis, but on the identification of the well-oxygenated fetus so as to allow labor to safely continue in the presence of a concerning, but not ominous FHR tracing.51 Another common deficiency that limits the acceptability of the available observational studies on validity of fetal pulse oximetry is the unrealistic cutoff for defining pathologic fetal acidemia (umbilical artery blood pH < 7.13 7.20). It is widely accepted, both in the United States52 and Europe,53 that pathologic fetal acidemia should be defined as an umbilical artery blood pH of < 7.00. Even in this group, 2/3 of the neonates will be unaffected by morbidity. It is also accepted that the presence of a metabolic component to acidemia may be as important, if not more important, than a single pH cutoff.52 Beginning at a pH < 7.13, the transition from respiratory (reversible) to metabolic acidosis (possibly consequential) takes place.39 Only a minority of human studies of fetal pulse oximetry have distinguished between respiratory and metabolic fetal acidemia, and when they did, intrapartum pulse oximetry monitoring was unable to predict umbilical artery base excess.41,44 In the first randomized study of fetal pulse oximetry published by Garite et al, there were 7 neonates with an umbilical artery blood pH < 7.00 (3 in the intervention group, and 4 in the control group).

All 4 cases in the control group were allowed to undergo vaginal delivery. In the control group there were also 6 cases of severe base excess (< -16 mEq/L). No such cases were recorded in the intervention group, and the 3 cases with acidemia at birth were recognized antepartum, leading to cesarean delivery. The authors suggested an improved sensitivity and specificity for metabolic acidemia in the intervention group, a very promising observation in contrast with previously reported observational data. Unfortunately, the study design did not guarantee that patient management was based exclusively on EFM with or without fetal pulse oximetry. In the study, vibro-acoustic stimulation or FSB sampling was required before proceeding to cesarean delivery in both groups. In the situations where FSpO2 was <30% for the entire interval between two contractions, or FSpO2 was unobtainable, the clinicians were supposed to revert to EFM interpretation, and in case of the latter one being persistently nonreassuring, the option existed for scalp stimulation or FSB sampling. Therefore the study could not answer the question whether clinical decisions could be made based exclusively on fetal pulse oximetry. Schmidt et al considered that such an exclusive application of fetal pulse oximetry might actually jeopardize fetal health.45 Is the application of fetal pulse oximetry capable to improve various neonatal outcomes?

Neonatal outcome is undoubtedly the ultimate end point in obstetrical care. In Garites study, there was no difference in neonatal outcome between the groups using, or not using fetal pulse oximetry.20 According to Chua et al, FSpO2 levels, measured even 10 minutes before delivery, have no relation with neonatal outcome.54 Leszczynska- Gorzelak et al are of the opinion that FSpO2 is more predictive of neonatal outcome in the 1st stage rather than in the 2nd stage of labor, except for the Apgar score which had no relationship with FSpO2 readings obtained in either the 1st or 2nd stage of labor.40 Butterwegge reported 6 cases of FSpO2 < 30% for more than 30 minutes, all with good neonatal outcome,55 and Alshimmiri et al pointed out that only normal FSpO2 correlates with fetal well-being.41 The reliability (precision) of fetal pulse oximetry has been evaluated at both the technical and practical level. At the technical level, dual sensor studies compared agreement using two devices concomitantly. Davies and Greene placed two FS14 sensors (Nellcor, Pleasanton, CA) on one fetal head and noted that above a FSpO2 of 30% the differences between the two sensors were approximately 1%, but below 30% more scattering was present, sometimes with differences exceeding 15%.56 These data suggested a good agreement in documenting fetal well-being, but decreased precision in indicating fetal jeopardy.

It has also been calculated that the precision of reflectance pulse oximeters at the low saturation levels characteristic for intrauterine life is approximately 4.7%.27 Thus, the actual value when a reading of 30% is posted may range from 25.3% to 34.7%. At the practical level, reliability of fetal pulse oximetry was evaluated as part of a series of feasibility studies. A very important aspect in the evaluation of fetal pulse oximetry was the so-called posting time: the amount of reliable saturation readings during the total time that the sensor is in position. In adult or neonatal transmission pulse oximeters, the posting time is 100%. The fetal Nellcor system was designed so that it monitored the quality of the measurement, and no value was displayed if the signal received did not have the characteristics of a fetal arterial plethysmographic curve, or if the contact between sensor and skin was insufficient. As a result of fetal movements and other artifacts, the posting time is always less than 100%. In the French multicenter study the mean reliable signal time in the 1st stage of labor was only 64.7%, and even less in the 2nd stage of labor (54%).44 A low reliable signal time of 68% was also reported by Schmidt et al.45 Other authors have encountered a much lower rate of problematic recordings of about 19-20%,40,42 and even lass than 10%.20,57 With the Zurich oximeter system, which is presumably less frequently dislodged, invalid signals have still been recorded 35% of the time.47 But what is more concerning, according to Luttkus et al, 58 is that the posting time tends to decrease with acidemia, rendering oximetry more questionable as a diagnostic tool for fetal jeopardy.

Many artifacts may impede signal acquisition and impact the readings reliability including the position of the sensor relative to fetal head ( the difference in FSpO2 readings between forehead and occiput may be as much as 13.4%), incomplete sensor-to-skin contact, marked caput formation, the increased intrauterine pressure with contraction, especially with low presentation station, below +2 (FSpO2 monitoring requires detection of fetal pulses, which may decrease or be undetectable when the surrounding pressure is high, resulting in loss of signal), the interposition of vernix or fetal hair, and the presence of meconium which acts in a similar manner to a red light filter, altering the ratio red/infrared light, and resulting in artificially low values.60 The latter theoretical concern is rejected by Yam et al who did not observe any effect of meconium on FSpO2 values.59 In conditions of meconium-stained amniotic fluid, Carbonne et al have shown that fetal pulse oximetry is a better predictor of meconium aspiration syndrome (MAS) than FSB sampling.61 More recently, in a Hungarian retrospective analysis of 118 cases, fetal pulse oximetry combined with amnioinfusion reportedly prevented MAS and reduced the cesarean delivery rate ( from 31% to 22%).62 The conclusions are questionable though, because the study did not have the required power to detect a difference in such a rare outcome as MAS (incidence of 1-3%), and even the difference in cesarean delivery rate was not statistically significant. The data regarding the influence of meconium on FSpO2 readings are still contradictory. All the above impairments in precision may contribute to the poor sensitivity of fetal pulse oximetry. Feasibility has been assessed from other perspectives as well. An Australian survey aimed to determine clinicians perceptions during placement of the oximetry sensor.63 Ease of placement was rated as good or excellent in 71% of cases, and the patients comfort was rated as good or excellent in 90% of cases. Chua et al reported a mean insertion time of 90 seconds, and that a reliable signal could be obtained within 5 minutes in 87% of placements.64 The French multicenter study bore the conclusion that the procedure is satisfactory as feasibility, and easier to use than FSB sampling.44 The device itself was reported to be harmless to the mother and fetus.

Fetal pulse oximetry after year 2000

A small randomized controlled trial of 146 women in labor at term with an abnormal FHR pattern was reported from Germany in 2004.65 The findings were positive, consisting of a significant decrease in both overall cesarean delivery rate and cesarean delivery rate for nonreassuring fetal status when fetal pulse oximetry was added to EFM, compared to EFM without fetal oximetry. There was no increase in the rate of cesarean delivery performed for dystocia. The impact on the cesarean delivery rate was achieved without a concomitant increase in adverse neonatal outcome. Another observation was that the use of FSB sampling was reduced by 50% in the group additionally monitored with fetal oximetry. The authors considered that fetal oximetry improved the sensitivity and specificity of surgical delivery for nonreassuring fetal status. However, the study was conducted heavily relying on FSB sampling. At least one FSB sampling was performed on all patients in both groups.

The extensive use of FSB sampling, not unusual in Germany, creates difficulties in interpretation, at least relative to the independent effect modification that can be attributed to fetal pulse oximetry. Still another RCT assessed the use of fetal pulse oximetry in addition to EFM in women with nonreassuring FHR pattern in labor, extending the evaluation to a population not studied previously: preterm deliveries between 28 and 36 weeks gestation, deliveries complicated by chorioamnionitis, and high (more than -2) fetal head station.66 In this American study of 327 deliveries, the authors found no difference in the overall cesarean delivery rate or the rate of cesarean delivery performed out of concern for fetal status between the group adding fetal oximetry to EFM and the group with EFM alone. In 2006, the results of two multicenter studies from Australia and the United States were published. The Australian RCT (the FOREMOST trial),67 conducted in four hospitals included 600 deliveries and essentially replicated the findings in Garites study.

The investigators compared the operative delivery rate (cesarean or instrumental delivery) for nonreassuring fetal status with and without the addition of fetal oximetry to EFM. They found that the addition of fetal pulse oximetry to EFM resulted in a significant decrease in operative intervention for nonreassuring fetal status, without modification in neonatal outcomes. Nevertheless, the overall operative delivery rates were not different between the groups because of a concomitant increase in instrumental deliveries for failure to progress in the 2nd stage of labor. The largest to date RCT of fetal pulse oximetry was conducted by the National Institutes of Child Health and Human Development Maternal- Fetal Medicine Units Network in the United States and included 5341 women delivering at 14 university hospitals (the FOX trial).68 The participants, in spontaneous labor at term with a singleton fetus in vertex presentation, were randomized to either open or masked fetal pulse oximetry in addition to EFM. In the open group the fetal oxygen saturation values were displayed to the clinician, whereas in the masked group the values were recorded by the computer, but hidden from the clinician. Besides this original study design intended to ensure strict comparability between the groups, in contrast to other studies, a nonreassuring FHR tracing was not a condition for inclusion and randomization.

The knowledge of intrapartum fetal oxygen saturation had no effect on the rates of cesarean delivery overall or specifically for the indication of nonreassuring FHR tracing or dystocia. In addition, knowledge of fetal oxygen saturation did not affect infant outcomes. The authors conclusion was that fetal oximetry is of no benefit as an adjunct for the interpretation of EFM. As we have seen, RCTs conducted in different obstetrical environments may have conflicting results. The aggregate interpretation of their findings in a meta-analysis may improve power and precision of effect estimates. A 2007 Cochrane meta-analysis concluded that the available data provide limited support for the use of fetal pulse oximetry to reduce the cesarean delivery rate.69 It is still possible that fetal pulse oximetry may be an effective tool in other clinical scenarios not yet studied, such as fetal arrhythmias with uninterpretable FHR tracing, including fetal bradycardia caused by a complete heart block and fetal tachycardia associated with maternal fever, thyrotoxicosis, or fetal supraventricular tachycardia, when distinguishing other contributions to tachycardia may be difficult. In such situations, complementary evaluations may help, and for improved sensitivity and specificity, it has been proposed to combine fetal pulse oximetry with fetal electrocardiogram.65 However, one wonders how much more technology is justifiable in the labor and delivery room.

Conclusion

Fetal pulse oximetry may be accurate, precise, and reliable in doing what it claims that it does recording of fetal oxygen saturation but unfortunately, it did not consistently or convincingly demonstrate medical value in addition to the current standard of care. To conclude on an optimistic note, rigorous research of fetal pulse oximetry prevented the acceptance in practice of still another piece of technology of equivocal scientific merit. In this case, clinical practice followed the science, and not the other way around, as too frequently seen in the past, including the adoption of EFM. The commercial availability of the Nellcor fetal pulse oximetry system used in the five RCTs20,65-68 was discontinued in 2006. Other systems remain available, but have not been subject to rigorous trials.

References:

Badawi N, Kurinczuk JJ, Keogh JM, et al. Intrapartum risk factors for newborn encephalopathy: the Western Australian case-control study. BMJ 1998;317:1554-1558 Hon EH, Quilligan EJ. The classification of fetal heart rate. II. A revised working classification. Conn Medicine 1967;31:779-784 Banta HD, Thacker SB. Assessing the costs and benefits of electronic fetal monitoring. Obstet Gynecol Surv 1979;34:627-640 Thacker SB, Stroup DF, Peterson HB. Efficacy and safety of intrapartum electronic fetal monitoring: an update. Obstet Gynecol 1995;86:613-620 National Institute of Child Health and Human Development Research Planning Workshop. Electronic fetal heart rate monitoring: research guidelines for interpretation. Am J Obstet Gynecol 1997;177:1385- 1390 Nielsen PV, Stigsby B, Nickelsen C, Nim J. Intra- and interobserver variability in the assessment of intrapartum cardiotocograms. Acta Obstet Gynecol Scand 1987;66:421-424 Chaffin DG, Goldberg CC, Reed KL. The dimension of chaos in the fatal heart rate. Am J Obstet Gynecol 1991;165:1425-1429 Saling E. Neues vorgehen zur untersuchung des kindes unter der geburt. Arch Gynakol 1962;197:108-122 Neilson JP. Fetal blood sampling as adjunct to heart rate monitoring. Oxford (UK): Cochrane Database of Systematic Reviews;1995.Issue 2 Goodwin TM, Milner-Masterson L, Paul RH. Elimination of fetal scalp blood sampling on a large clinical service. Obstet Gynecol 1994;83:971-974 Saling E. Comments on past and present situation of intensive monitoring of the fetus during labor. J Perinat Med 1996;24:7-13 Irion O, Stuckelberger P, Moutquin JM, Morabia A, Extermann P, Beguin F. Is intrapartum vibratory acoustic stimulation a valid alternative to fetal scalp pH determination? Br J Obstet Gynaecol 1996;103:642-647 Salamalekis E, Vitoratos N, Lochis C, et al. Evaluation of non-reassuring fetal heart rate patterns with fetal pulse oximetry combined with vibratory acoustic stimulation. J Matern Fetal Neonatal Med 2003;13:110-114 Amer-Wahlin I, Hellsten C, Noren H, et al. Cardiotocography only versus cardiotocography plus ST segment analysis of fetal electrocardiogram for intrapartum fetal monitoring: a Swedish randomized controlled trial. Lancet 2001;358:534-538 Westgate J, Harris M, Curnow JS, Greene KR. Plymouth randomized trial of cardiotocogram only versus ST waveform plus cardiotocogram for intrapartum monitoring in 2400 cases. Am J Obstet Gynecol 1993;169:1151-1160 Neilson JP. Fetal electrocardiogram (ECG) for fetal monitoring during labour. Cochrane Database of Systematic Reviews, 3, 2006 Liston R, Crane J, Hamilton E, et al. Fetal health surveillance in labour. J Obstet Gynaecol Can 2002;24:250-276 Knig V, Huch R, Huch A. Reflectance pulse oximetry: principles and obstetric application in the Zurich system. J Clin Monit Comput 1998;14:403-412 Nijland MJM, Shankar V, Iyer V, Ross MG. Assessment of fetal scalp oxygen saturation determination in the sheep by transmission pulse oximetry. Am J Obstet Gynecol 2001;183:1549-1553 Garite TJ, Dildy GA, McNamara H, et al. A multicenter controlled trial of fetal pulse oximetry in the intrapartum management of nonreassuring fetal heart rate patterns. Am J Obstet Gynecol 2000;183:1049- 1058 Haverkamp AD, Orleans M, Langendoerfer S, et al. A controlled trial of the differential effects of continuous fetal heart rate monitoring in labor: a randomized trial. Am J Obstet Gynecol 1979;134:399-412 ACOG Committee Opinion. Number 258, September 2001. Fetal pulse oximetry. Obstet Gynecol 2001;98:523-524 Dildy GA. A guest editorial: fetal pulse oximetry. Obstet Gynecol Surv 2003;58:225-226 24. Zijlstra WG, Buursma A, Zwart A. Performance of an automated six-wavelength photometer (radiometer OSM3) for routine measurement of hemoglobin derivatives. Clin Chem 1988;34:149-152 Porath M, Sinah P, Dudenhausen JW, Luttkus AK. Systematic instrumental errors between oxygen saturation analysers in fetal blood during deep hypoxemia. Clin Chim Acta 2001;307:151-157 Luttkus AK, Lbke M, Bscher U, Porath M, Dudenhausen JW. Accuracy of fetal pulse oximetry. Acta Obstet Gynecol Scand 2002;81:417-423 Carter AM, Stiller R, Knig V, Jorgensen JS, Svendsen P, Huch R. Calibration of a reflectance pulse oximeter in fetal lambs for arterial oxygen saturations below 70%. J Soc Gynecol Invest 1998;5:255-259 McNamara H, Cung CD, Lilford R, Johnson N. Do fetal pulse oximetry readings at delivery correlate with cord blood oxygenation and acidemia? Br J Obstet Gynaecol 1992;99:735-738 Langer B, Boudier E, Haddad J, Pain L, Schlaeder G. Fetal pulse oximetry during labor of 62 patients. Fetal Diagn Ther 1996;11:37-45 Harrison AM, Lynch JM, Dean JM, Witte MK. Comparison of simultaneously obtained arterial and capillary blood gases in pediatric intensive care unit patients. Crit Care Med 1997;25:1904-1908 Nijland R, Jongsma HW, Nijhuis JG, Oeseburg B. Accuracy of fetal pulse oximetry and pitfalls in measurements. Eur J Obstet Gynecol Reprod Biol 1997;72:S21-27 Nijland R, Jongsma H, Nijhuis J, van den Berg P, Oeseburg B. Arterial saturation in relation to metabolic acidosis in fetal lambs. Am J Obstet Gynecol 1995;172:810-819 Dildy GA, van den Berg PP, Katz M, et al. Intrapartum fetal pulse oximetry: fetal oxygen saturation trends during labor and relation to delivery outcome. Am J Obstet Gynecol 1994;171:679-684 Khnert M, Seelbach-Gbel B, Butterwegge M. Predictive agreement between fetal arterial oxygen saturation and fetal scalp pH: results of the German multicenter study. Am J Obstet Gynecol 1998;178:330-335 Salamalekis E, Thomopoulos P, Giannaris D, et al. Computerised intrapartum diagnosis of fetal hypoxia based on heart rate monitoring and fetal pulse oximetry recordings utilizing wavelet analysis and neural networks. Br J Obstet Gynaecol 2002;109:1137-1142 Bloom SL, Swindle RG, McIntire DD, Leveno KJ. Fetal pulse oximetry: duration of desaturation and intrapartum outcome. Obstet Gynecol 1999;93:1036-1040 Seelbach-Gbel B, Heupel M, Khnert M, Butterwegge M. The prediction of fetal acidosis by means of intrapartum fetal pulse oximetry. Am J Obstet Gynecol 1999;180:73-81 Gorenberg DM, Patillo C, Hendi P, Rumney PJ, Garite TJ, Fetal pulse oximetry: correlation between oxygen desaturation, duration, and frequency and neonatal outcomes. Am J Obstet Gynecol 2003;189:136-138 Dildy GA, Thorp JA, Yeast JD, Clark SL. The relationship between oxygen saturation and pH in umbilical blood: implications for intrapartum fetal oxygen saturation monitoring. Am J Obstet Gynecol 1996;175:682-687 Leszczynska-Gorzelak B, Poniedzialek-Czajkowska E, Oleszczuk J. Fetal blood saturation during the 1st and 2nd stage of labor and its relation to the neonatal outcome. Gynecol Obstet Invest 2002;54:159-163. Alshimmiri M, Bocking AD, Gagnon R, Natale R, Richardson BS. Prediction of umbilical artery base excess by intrapartum fetal oxygen saturation monitoring. Am J Obstet Gynecol 1997;177:775-779 Rijnders RJ, Mol BW, Reuwer PJ, Drogtrop AP, Vernooij MM, Visser GH. Is the correlation between fetal oxygen saturation and blood pH sufficient for the use of fetal pulse oximetry? J Matern Fetal Neonatal Med 2002;11:80-83 Hajek Z, Vrablik J, Haddad R, Kratochvil B, Parizek A. The fetal ECG-ST analysis in the diagnosis of fetal hypoxia. Ceska Gynekol 2002;67:16-19 Goffinet F, Langer B, Carbonne B, et al. Multicenter study on the clinical value of fetal pulse oximetry. I. Methodologic evaluation. The French Study Group on Fetal Pulse Oximetry. Am J Obstet Gynecol 1997;177:1238-1246 Schmidt S, Koslowski S, Sierra F, Meyer-Wittkopf M, Heller G. Clinical usefulness of pulse oximetry in the fetus with non-reassuring heart rate pattern? J Perinat Med 2000;28:298-305 Vitoratos N, Salamalekis E, Saloum J, Makrakis E, Creatsas G. Abnormal fetal heart rate patterns during the active phase of labor and the value of fetal oxygen saturation. J Matern Fetal Neonatal Med 2002;11:46-49 Stiller R, von Mering R, Knig V, Huch A, Huch R. How well does reflectance pulse oximetry reflect intrapartum fetal acidosis? Am J Obstet Gynecol 2002;186:1351-1357 Leszczynska-Gorzelak B, Poniedzialek-Czajkowska E, Zych I, Grzechnik M, Oleszczuk J. The usefulness of the intrapartum fetal pulse oximetry in anticipating the neonatal outcome. Ginekol Pol 2001;72:1183-1188 Nikolov A, Dimitrov A, Vakrilkova L, Iarkova N. Fetal oxygen saturation during normal delivery. Akush Ginekol 2000;40:3-6 Low JA, Victory R, Derrick EJ. Predictive value of electronic fetal monitoring for intra-partum fetal asphyxia and metabolic acidosis. Obstet Gynecol 1999;93:285-291 Vidaeff AC, Ramin SM. Fetal pulse oximetry: 8 vital questions. OBG Management 2004;16:28-44 Goldaber KG, Gilstrap LC III, Leveno K, Dax JS, McIntire DD. Pathologic fetal acidemia. Obstet Gynecol 1991;78:1103-1107 Boog G. La souffrance foetale aigu. J Gynecol Obstet Biol Reprod 2001;30:393-432 Chua S, Yeong SM, Razvi K, Arulkumaran S. Fetal oxygen saturation during labour. Br J Obstet Gynaecol 1997;104:1080-1083 Butterwegge M. Fetal pulse oximetry and non-reassuring heart rate. Eur J Obstet Gynecol Reprod Biol 1997;72:S63-66 56. Davies MG, Greene KR. Fetal pulse oximetry a preliminary report on sensor precision determined by dual sensor studies. Eur J Obstet gynecol Reprod Biol 1997;72:S35-41 57. Butterwegge M, Seelbach-Gbel B, Khnert M. ber den einsatz der fetalen pulsoxymetrie bei risikogeburten in deutschen kliniken. Z Geburtshilfe Neonatol 2002;206:83-87 Luttkus AK, Friedmann W, Thomas S, Dimer JA, Dudenhausen JW. The safety of fetal pulse oximetry in parturients requiring fetal scalp blood sample. Obstet Gynecol 1997;90:533-537 Johnson N, Johnson VA, Bannister J, McNamara H. The effect of meconium on neonatal and fetal reflectance pulse oximetry. J Perinat Med 1990;18:351-355 Yam J, Chua S, Arulkumaran S. Intrapartum pulse oximetry: Part 1: Principles and technical issues. Obstet Gynecol Surv 2000;55:163-172 Carbonne B, Cudeville C, Sivan H, Cabrol D, Papiernik E. Fetal oxygen saturation measured by pulse oximetry during labor with clear and meconium-stained amniotic fluid. Eur J Obstet Gynecol Reprod Biol 1997;72:S51-55 62. Halvax L, Szabo I, Vizer M, Csermely T, Ertl T. Simultaneous use of intrapartum fetal pulse oximetry and amnioinfusion in meconium stained amniotic fluid. Eur J Obstet Gynecol Reprod Biol 2002;104:105-108 East CE, Colditz PB. Clinicians perceptions of placing a fetal oximetry sensor. J Qual Clin Pract 2000;20:161-163 Chua S, Yam J, Razvi K, Yeong SM, Arulkumaran S. Intrapartum fetal oxygen saturation monitoring in a busy labour ward. Eur J Obstet Gynecol Reprod Biol 1999;82:185-189 Kuhnert M, Schmidt S. Intrapartum management of nonreassuring fetal heart rate patterns: a randomized controlled trial of fetal pulse oximetry. Am J Obstet Gynecol 2004;191:1989-1995 Klauser CK, Christensen EE, Chauhan SP, et al. Use of fetal pulse oximetry among high-risk women in labor: a randomized clinical trial. Amer J Obstet Gynecol 2005;192:1810-9 East CE, Brennecke SP, King JF, et al. The effect of intrapartum fetal pulse oximetry, in the presence of a nonreassuring fetal heart rate pattern, on operative delivery rates: A multicenter randomized, controlled trial (the FOREMOST trial). Amer J Obstet Gynecol 2006;194:606.e1-606.e16 Bloom SL, Spong CY, Thom E, et al. Fetal pulse oximetry and cesarean delivery. N Engl J Med 2006;355:2195-202 East CE, Chan FY, Colditz PB, et al. Fetal pulse oximetry for fetal assessment in labour. Cochrane Database Syst Rev. 2007 Apr 18;(2):CD004075 Luttkus AK, Stupin JH, Callsen TA, et al. Feasibility of simultaneous application of fetal electrocardiography and fetal pulse oximetry. Acta Obstet Gynecol Scand 2003;82:443-448