Archives ofDisease in Childhood 1992; 67: 1237-1241 1237 Arch Dis Child: first published as 10.1136/adc.67.10_Spec_No.1237 on 1 October 1992. Downloaded from PERSONAL PRACTICE

The unreactive fetal heart rate

J S Smoleniec, D K James

The 'flat' fetal heart rate (FHR) trace both aged over 1/16th of a minute and the differ- before and during labour is viewed ominously ence between consecutive 1/16th of a minute by many obstetricians and midwives. The cur- intervals are used.5 rent medicolegal climate relating especially to Therefore it is long term FHR variation care in labour is such that much emphasis is which is most commonly referred to in clinical placed on an unreactive FHR. There is a practice. Long term variation is a visual tendency to link an unreactive FHR with an assessment of the irregular oscillations of the adverse outcome. However, it is only in a baseline FHR. The normal amplitude of vari- minority of cases that this is true. Furthermore, ation of the FHR baseline is commonly there is a wide variation in both definition and accepted as being between 5 and 15 bpm,6 interpretation. Like many aspects of though some maintain that 25 bpm is a more care, the unreactive FHR is an example of 'a realistic upper limit.7 The considerable inter- little knowledge being a bad thing'. observer and intraobserver variation in inter- pretation and quantification of FHR variation has led to the introduction of computerised Definition FHR analysis.8 There is no agreed definition of an unreactive Although theoretically computer analysis of FHR; there is more agreement about what FHR would appear to be the panacea for constitutes a reactive FHR. Therefore the problems of FHR interpretation, it is only as simplest definition of an unreactive FHR good as the program criteria. Such criteria might be one that fails to meet the criteria for need to be validated over tens of thousands of a reactive FHR. A reactive FHR is one which if they are to detect at has at least two accelerations of at least 15 high risk of intrauterine compromise. This beats per minute (bpm) amplitude (some point is well illustrated by a sinusoidal FHR would use 10 bpml), each lasting for at least with pathologically low FHR variability, being 15 seconds, during a set time period, on computed as having normal long term varia- no agreement, but it which could only be diagnosed as abnor- which there is general tion, http://adc.bmj.com/ varies from 10 to 120 minutes.2 3 From the mal by including a computation for short term results of fetal behavioural studies1 we use a variation.9 It is noteworthy that the computer 15 bpm amplitude of acceleration on at least criteria for long and short term variation two occasions in association with fetal move- (evolved by comparing the analysis of thou- ments within 40 minutes as our definition of sands of FHR recordings with the subsequent reactivity in the of 30 weeks' gestation or infant outcome) differ remarkably from those more, and consider 10 bpm to be more used clinically. Furthermore, although com- physiological at under 30 weeks (see below). puter analysis appears to detect antepartum on September 25, 2021 by guest. Protected copyright. The problem with defining an unreactive fetal hypoxia in small for dates fetuses, the FHR as one with no accelerations is that it same criteria were not found to be applicable does not take into account baseline FHR in labour. 10 Although promising, comput- variation. There are two types of variation: erised analysis needs further evaluation. short and long term. Short term variation is In clinical practice we suggest that an unre- the variation in successive beat-to-beat inter- active FHR should be defined as an FHR vals of the FHR. However, this variation can- which fails to demonstrate accelerations (at not be accurately measured using external least two periods of 15 seconds' duration with FHR monitoring via Doppler ultrasound with 15 bpm or more for pregnancies of more than autocorrelation. Even when true beat-to-beat 30 weeks, and 10 bpm for pregnancies less variation is accurately recorded using a fetal than 30 weeks) in association with fetal move- scalp electrode in labour, this cannot be visu- ments during 40 minutes' continuous record- ally quantified by someone observing the ing. If the FHR fails to fulfil these criteria University Department FHR. This is not surprising since the true after 40 minutes of continuous recording then of , mean (SE) beat-to-beat variation is 2 1 (0-2) it should be regarded as abnormal. This con- St Michael's Hospital, ms.4 In the clinical setting the only method of clusion would be reinforced by the demon- Southwell Street, Bristol BS2 8EG quantifying short term FHR variation is by stration of FHR baseline variation persistently J S Smoleniec using computerised FHR analysis. However, being less than 5 bpm during the 40 minutes. D K James even when this is available it appears that The presence of decelerations of the FHR (a 10 Correspondence to: beat-to-beat FHR variation is not as clinically fall of at least 15 bpm lasting for at least Dr Smoleniec. useful as when beat-to-beat intervals are aver- seconds) would be further evidence of patho- 1238 Smoleniec, _'ames Arch Dis Child: first published as 10.1136/adc.67.10_Spec_No.1237 on 1 October 1992. Downloaded from logical FHR pattern. However, it must be quiescence or fetal behavioural state 1F. 3 remembered that decelerations of the FHR At term a fetus can spend 30-40% of the time are a normal feature of the very immature in quiescence and the average (SD) con- FHRI and are therefore less likely to be tinuous period spent in quiescence is 21 (8 6) pathological before 28 weeks. minutes.14 Therefore we would recommend It is our view that this definition of the that an unreactive FHR should be investigated unreactive FHR with the caveats relating to if quiescence lasts longer than 40 minutes. baseline variation and decelerations is equally Another normal feature of advancing gesta- applicable to antepartum and intrapartum tion is the effect on baseline variation. Early in recordings of the FHR. the second trimester variation is very slight (less than 5 bpm) but gradually increases as pregnancy progresses so that it is usually Causes greater than 5 bpm after 26 weeks.1 The mean The normal FHR recording is determined by baseline variation is approximately 7 bpm at the dynamic interplay of the autonomic ner- 34 weeks. Thereafter the mean baseline varia- vous system, cerebral centres, and cardiovas- tion depends on the state of activity of the cular reflexes.11 The autonomic innervation of fetus: in quiescence, the mean variation falls the heart is the most important mechanism in to reach 4 bpm at 40 weeks, whereas in acti- controlling FHR variation. The parasympa- vity variation continues to rise to about 10 thetic muscarinic cardiodecelerator activity bpm at 40 weeks.1 Other fetal factors which (producing a decrease in FHR but increased are associated with an increase in FHR varia- variation) is of primary importance in influ- tion include fetal breathing and fetal encing beat-to-beat variation because of its mouthing movements.1 15 rapid effect on FHR. In contrast sympathetic Maternal physiological factors which influ- cardiac stimulation (producing an increase in ence FHR reactivity include diurnal variation FHR and decreased variation) leads to a (more fetal activity and hence reactivity in the much slower FHR response.12 evening'6) and glucose ingestion (increases The causes of unreactive FHR can be fetal breathing and baseline variation). broadly divided into physiological and patho- External stimulation (for example, with logical (table 1). vibratory or acoustic stimulators) of the fetus increases reactivity of the FHR and has been PHYSIOLOGICAL claimed to affect fetal behaviour generally.17 Gestation is the most important physiological influence on FHR. The FHR fails to manifest PATHOLOGICAL accelerations in association with movements A number of different pathologies can pro- and is thus normally unreactive before 24 duce an unreactive FHR by directly affecting weeks.1 From 24 weeks onwards, FHR accel- the fetal central nervous (CNS) or cardiovas- erations are seen for a greater proportion of cular systems (CVS). Examples of pathologies the time although these are not normally of 15 directly affecting the CNS are structural bpm or more until 30 weeks onwards.1 Over abnormalities such as anencephaly, infections

the last trimester fetal rest/activity cycles such as varicella, acute hypoxia (secondary to http://adc.bmj.com/ become manifest and these evolve into sophis- uterine contractions in labour, or placental ticated fetal behavioural states from 35 abruption) and a major non-fatal hypoxic or weeks.'3 During the last trimester the com- anoxic insult'8 causing brain death such as monest reason for an unreactive FHR is fetal placental abruption or maternal cardiorespira- tory collapse.'9 The pathophysiological effect of chronic placental insufficiency as seen with Table 1 Causes of an unreactive FHR severe pre-eclampsia or recurrent placental on September 25, 2021 by guest. Protected copyright. Physiological Pathological bleeding is to cause hypoxia which affects In cases when decom- Fetal Immaturity Central nervous system: FHR variation.20 such Anomalies (for example anencephaly) pensation occurs there will be a resultant pro- Infection (for example varicella) Hypoxia: gressive reduction in FHR variation which will Rest/activity states, Chronic (for example nearly always be accompanied by recurrent IF behavioural state,* placental insufficiency, pre-eclampsia) variable 22 This effect or quiescence Acute (for example intrapartum late or deceleration.2' asphyxia, placental abruption or has been shown by cordocentesis to be associ- maternal cardiorespiratory collapse resulting in brain death) ated with fetal hypoxia.23 If the decompensa- Cardiovascular system: tion continues, hypoxic acidosis will result Arrhythmias Anomalies which is characterised by loss of accelerations, Cardiac failure (various causes) decreased baseline FHR variation (<3 bpm) Maternal Diurnal variation Accidents (for example and shallow decelerations in response to External stimuli cardiorespiratory arrest) 25 Drugs: Braxton Hicks' contractions.24 Opiates The commonest CVS cause for an un- Anaesthetics (for example epidural) Parasympatholytic reactive FHR to cardiac arrhythmia. This can Sympathomimetic be primary idiopathic (supraventricular tachy- Magnesium sulphate cardia, atrial flutter), or secondary to a struc- *Four categories of behavioural states are used to describe the fetus and these are numbered 1 tural cardiac abnormality. Cardiac failure in to 4 and the letter 'F' is added to differentiate fetal states from neonatal states. Behavioural to abnormal states are determined by heart rate, fetal gross body movements, and eye movements. For the fetus (secondary anaemia, example, state 1F, which is synonymous with fetal quiescence, is characterised by absent eye karyotype, cardiac anomaly or arrhythmia) movements, a stable fetal heart rate with low baseline variation, which can be interrupted by FHR but this is isolated heart rate accelerations which are strictly associated with brief gross body movements, also results in a unreactive mostly startles. probably due to the resulting hypoxia. The unreactive fetal heart rate 1239 Arch Dis Child: first published as 10.1136/adc.67.10_Spec_No.1237 on 1 October 1992. Downloaded from Maternal 'pathological' causes of an unre- assessment is mandatory. A biophysical pro- active FHR are primarily confined to the file should be performed28 which if normal is effects of drugs taken by the mother, which reassuring (which can be especially helpful in cross the and affect the fetal CNS interpreting unreactive FHR below 28 weeks) and CVS. Those that predominantly have but if it is abnormal it is associated with an their effect on the CNS26 include opioids such increased risk of fetal compromise. In addi- as pethidine, hypnotics such as diazepam, and tion, during the profile procedure more criti- anaesthetics both general and regional (epidu- cal observation of the relationship between the ral, spinal). However, the effect of an epidural FHR and fetal activity may prove useful in or spinal anaesthetic in producing an unreac- diagnosing fetal pathology. For example, in tive FHR may be more commonly due to an brain death there is an unreactive FHR and effect on uterine perfusion. Drugs with pre- absent fetal movement, whereas dissociation dominant effects on the fetal CVS include of fetal activity with an unreactive FHR is parasympatholytic agents, such as atropine found with anencephaly and congenital and tricyclic antidepressants, sympath- abnormalities.29 omimetic agents such as ritodrine and salbu- A detailed anomaly scan will help identify tamol. Other drugs include magnesium sul- fetuses with obvious CNS, CVS, and karyo- phate and even paracetamol.27 typic abnormalities. Cardiac arrhythmias A rise in maternal temperature will result in should be categorised using M mode ultra- a rise in the FHR baseline with a reduction in sound. Fetal size assessment is important in variation, but accelerations tend to be pre- diagnosing intrauterine growth retardation served. which is a common feature of many of the It should be stressed that not all these pathologies associated with an unreactive pathological causes of an unreactive FHR are FHR. Similarly, abnormal Doppler ultra- confined to labour. Indeed the majority pre- sound studies of the umbilical artery date labour onset. This observation leads to (increased systolic to diastolic ratio or absent the speculation that at least some of the unre- end diastolic frequencies) are found in many active FHR records which are identified in conditions which can result in an unreactive labour may reflect some longstanding problem FHR. Indeed the combination of intrauterine which has not been identified because no growth retardation, abnormal Doppler record- FHR recording was undertaken before labour ings, and an unreactive FHR is associated (for example undiagnosed intrauterine growth with a poor prognosis for the fetus.30 retardation). Fetal stimulation has been used as an adjunct to FHR monitoring in order to assess fetal wellbeing. The oxytocin stress test is Management of an unreactive FHR commonly used in North America if the FHR ANTEPARTUM is unreactive. However, it is not without risks The first principle of management when an such as aggravating fetal compromise and unreactive FHR is reported or suspected is to preterm labour. An alternative stimulation test ensure that the strict criteria for definition are is vibroacoustic stimulation which has been

fulfilled (see above) since the commonest found useful antenatally in shortening the http://adc.bmj.com/ cause will be fetal quiescence. The next prin- time to achieve a reactive FHR.31 However, to ciple of management is to try and ascertain date whereas a positive response is reported as the cause (table 2). A clinical assessment reassuring (associated with a non-acidotic should be undertaken to exclude conditions fetus), it is less useful in identifying fetal dis- such as pre-eclampsia, antepartum haemor- tress.32 Furthermore, there have been sugges- rhage, drug taking, and a small for dates fetus. tions that vibrocoustic stimulation may be to an compro- After this clinical evaluation, ultrasound fetal potentially hazardous already on September 25, 2021 by guest. Protected copyright. mised fetus.33 If the cause is still not certain then invasive Table 2 Management ofan unreactive FHR testing may be necessary. In the antepartum Ensure that the criteria for diagnosis are fulfilled to exclude period the diagnosis of chromosomal abnor- physiological causes mality is the commonest indication for inva- Unreactive FHR-need to exclude a pathological cause sive testing, especially if ultrasound stigmata History and examination Pre-eclampsia of anomalies are identified. While fetal blood Antepartum haemorrhage will provide a relatively quick Drugs sampling Reduced/absent fetal activity answer, within two to three days, allow Small for dates fetus haemoglobin and blood gas and acid-base Ultrasound assessment Biometry (fetal size) Anatomy CNS/cardiac estimation (in case of fetal hydrops), and sero- Cardiac rhythm (M mode) logical screening for viral infections, it is not Biophysical assessment Coincidence in fetal behaviour without significant risk to the fetus especially Fetal Doppler studies in the retarded fetus.34 Chorionic Stress/stimulation tests growth Invasive testing villus sampling may prove to be safer and Antepartum Karyotype and the results as reliable. The Viral serology quicker Haemoglobin relative merits of the two approaches have not Blood gases yet been tested in controlled studies. Ante- Intrapartum Fetal scalp sampling Treatment partum definitive treatment will depend upon Antepartum Fetal specific cause (for example the cause. If investigations conclude that the antiarrhythmic drug or blood transfusion) fetus does not have a lethal irreversible cause Delivery for the unreactive FHR or one which can be 1240 Smoleniec, James Arch Dis Child: first published as 10.1136/adc.67.10_Spec_No.1237 on 1 October 1992. Downloaded from safely treated in utero (for example anti- For the future, research should first concen- arrhythmic therapy35 or fetal blood trans- trate on improving the understanding of the fusion) then delivery is the usual treatment, pathophysiological processes which ultimately often by . lead to an unreactive FHR both antepartum and intrapartum. Secondly, efforts should be directed towards extending the range of inves- INTRAPARTUM tigations which are available, especially in In labour while clinical assessment in order to labour, to obtain a more comprehensive pic- determine the cause is still possible and ture of the fetal condition. should be undertaken, the full range of inves- 1 Pillai M, James DK. The development of fetal heart rate tigations may be neither available nor feasible. patterns during normal pregnancy. Obstet Gynecol In practice the commonest investigation em- 1990;76:812-5. 2 Paul RH. The evaluation of antepartum fetal well-being ployed is fetal blood pH estimation obtained using the . In: Petrie RH, ed. Fetal monitoring. by fetal scalp sampling (which has an added Clinics in perinatology. London: Saunders, 1982;9:254. 3 Brown R, Patrick J. The nonstress test: how long is advantage of acting as a stimulation test). enough? Am J Obstet Gynecol 1981;141:646. However, there are limitations to this 4 Lawson GW, Dawes GS, Redman CWG. A comparison of two fetal heart rate ultrasound detector systems. Am Jf approach. Unless detailed fetal assessment has Obstet Gynecol 1982;143:840-2. been undertaken before the onset of labour 5 Lawson GW, Dawes GS, Redman CWG. Analysis of fetal heart rate on-line at 32 weeks gestation. Br J Obstet (for example as part of the admission assess- Gynaecol 1984;91:542-50. ment), management in labour is going to be 6 Steer PJ, Eigbe MB, Lissauer JJ, Beard RJ. Inter- relationships among abnormal cardiotocograms in labour, largely on the basis of fetal acidaemia so that meconium staining of the , arterial cord fetal normality before labour has to be blood pH and Apgar scores. Obstet Gynecol 1989; 74:715-21. assumed. This assumption may be proved to 7 FIGO News-Guidelines for the use of fetal monitoring. Int be incorrect once the infant is delivered. J Gynaecol Obstet 1987;25:159-67. 8 Dawes GS, Visser GHA, Goodman JDS, Redman CWG. Furthermore, a normal scalp pH in the Numerical analysis of the human fetal heart rate: the presence of an unreactive FHR does not quality of ultrasound records. Am J Obstet Gynecol 1981 ;141:43-52. exclude pathology.36 9 Street P, Dawes GS, Moulden M, Redman CWG. Short- In labour, general fetal resuscitational mea- term variation in abnormal antenatal fetal heart rate records. AmJ Obstet Gynecol 1991;165:515-22. sures should be employed even when the scalp 10 Dawes GS, Rosevear SK, Pello LC, Moulden M, Redman pH is low since the cause of the unreactive CWG. Computerised analysis of episodic changes in fetal heart rate in early labour. Am J Gynecol 1991;165: FHR and decelerations maybe secondary to 618-24. reversible causes such as cord compression or 11 Renou P, Newman W, Wood C. Autonomic control of fetal heart rate. Am J Obstet Gynecol 1969;105:949. hypertonic uterine contractions or uterine 12 Warner HR, Cox A. A mathematical model of heart rate hypoperfusion after epidural anaesthesia. The control by sympathetic and vagus efferent information. J Appl Physiol 1982;17:349. advantages are that in utero resuscitation is 13 Nijhuis JG, Prechtl HFR, Martin CBJr, Bots RSGM. Are more efficient than neonatal resuscitation and there behavioural states in the human fetus? Early Hum Dev 1982;6:177. an operative delivery may be avoided.37 These 14 Pillai M, James DK. Behavioural states in normal mature measures may include stopping an oxytocin human fetuses. Arch Dis Child 1990;65:39-43. 15 Wheeler T, Gennser G, Lindvall R, Murrills AJ. Changes infusion and/or administering tolcolytics, fluid in the fetal heart rate associated with fetal breathing infusion, changing the mother's position to and fetal movement. Br J7 Obstet Gynecol 1980;87: 1068-79. avoid caval compression, and giving facial 16 Patrick J, Campbell K, Carmichael L, Natale R, http://adc.bmj.com/ oxygen. If these measures are unsuccessful, Richardson B. Patterns of gross fetal body movements over 24-hour observation intervals in the last 10 weeks of is not imminent and the scalp pregnancy. Am 7 Obstet Gynecol 1982;142:363-71. pH is <7-2, delivery by caesarean section 17 Gagnon R, Hunse C, Carmichael L, et al. Effects of vibra- tory acoustic stimulation on human fetal breathing and should be performed. gross fetal body movements near tern. Am J Obstet Gynecol 1986;155:1227. 18 Nijhuis JG, Kruyt N, van Wijack JAM. Fetal brain death: two case reports. BrJ Obstet Gynaecol 1988;195:198. Conclusion 19 van der Moer PE, Gerretsen G, Visser GHA. Fixed fetal heart rate pattern after intrauterine accidental decerebra- on September 25, 2021 by guest. Protected copyright. When faced with an unreactive FHR it is tion. Obstet Gynecol 1985;65:215-7. important that the criteria for diagnosis are 20 Henson GL, Dawes GS, Redman CWG. Antenatal fetal heart rate variation relation to fetal acid-base status at fulfilled. The majority of so called 'unreactive caesarean section. BrJ Obstet Gynecol 1983;90:516. FHR traces' are only such because FHR 21 Devoe LD, McKenzie J, Searle NS, et al. Clinical sequelae of the extended nonstress test. Am J Obstet Gynecol 1985; monitoring has not been continued for long 151:1074. enough. In the antepartum period an un- 22 Visser GHA, Bekedam DJ, Ribbert LSM. Changes in antepartum heart rate patterns with progressive deteriora- reactive FHR should be investigated as tion of the fetal condition. Int J Biomed Comput quickly as possible. In labour the options are 1990;25:239-46. 23 Visser GHA, Sadovsky G, Nicolaides KH. Ante- more limited and testing for fetal acidaemia is partum heart rate patterns in small for commonly the only investigation possible. third trimester fetuses: correlations with blood gases obtained at cordocentesis. Am J Obstet Gynecol 1990; However, it must be remembered that an 182:698-703. intrapartum unreactive FHR may be caused 24 Visser GHA, Redman CWG, Huisjes HJ, Turnbull AC. Nonstressed antepartum heart rate monitoring: implica- by antepartum pathology. tions of decelerations after spontaneous contractions. Am Prompt delivery is the usual therapeutic J Obstet Gynecol 1980;138:429. 25 Krebs HB, Petres RE, Dunn U. Intrapartum fetal heart option if the cause of the unreactive FHR is rate monitoring. Obstet Gynecol 1983;145:297-305. thought to be pathological, the effects on the 26 Petrie RH, Yeh SY, Murata Y, et al. The effects of drugs on fetal heart rate variation. Am J Obstet Gynecol fetus reversible and non-lethal. 1978,130:294-99. At present there is no objective assessment 27 Rosevear SK, Hope PL. Favourable neonatal outcome fol- lowing maternal paracetamol overdose and severe fetal of FHR other than by computer analysis distress. Case report. Br J Obstet Gynaecol 1989;96: which if used serially may enable management 491-3. 28 Manning FA, Platt LD, Sipos L. Antepartum fetal evalua- of an unreactive FHR in the antepartum tion: development of a fetal biophysical score. Am J period to be improved. Obstet Gynecol 1980;136:787-95. The unreactive fetal heart rate 1241 Arch Dis Child: first published as 10.1136/adc.67.10_Spec_No.1237 on 1 October 1992. Downloaded from 29 Pillai M, Garrett C, James D. Bizarre fetal behaviour as- HFR. Vibro-acoustic stimulation of the human fetus: sociated with lethal congenital anomalies: a case report. effect of behavioural state organization. Early Hum Dev EurJ Obstet Gynecol Reprod Biol 1991 ;39:215-8. 1989;19:285-96. 30 Maulik D, Yarlagadda P, Downing G. Doppler velocimetry 34 Maxwell DJ, Johnson P, Hurley P, Neales K. Allan L, in obstetrics. In: Morrison JC, ed. Antepartalfetal surveil- Knott P. Fetal blood sampling and pregnancy loss in lance. Obstet Gynecol Clin North Am. London: Saunders, relation to indication. Br Jf Obstet Gynecol 1991 ;98: 1990;17: 181. 892-7. 31 Smith CV, Phelan JP, Broussard P, et al. Fetal acoustic 35 Sykes GS, Molloy PM, Johnson P, Stirrat GM, Tumbull stimulation testing III. Predictive value of a reactive test. AC. Fetal distress and the condition of the newbom J Reprod Med 1988;33:217. infants. BMJ 1983;287:943-5. 32 Shaw KJ, Paul RH. Fetal responses to external stimuli. In: 36 Smoleniec JS, James DK. Intermittent fetal tachycardia Morrison JC, ed. Antepartal fetal surveiUlance. Obstet and hydrops fetalis. Arch Dis Child 1991;66:1160-1. Gynecol Clin North Am. London: Saunders, 1990;17: 37 Caldeyro-Barcia R. Intrauterine fetal reanimation in 235-46. acute intrapartum distress. Jf Perinat Med 1991;19: 33 Visser GHA, Mulder HH, Wit HP, Mulder EJH, Prechtl (2): 11. http://adc.bmj.com/ on September 25, 2021 by guest. Protected copyright.