Transcript of "Fetal pain implications for research and practice"
British Journal of Obstetrics and GynaecologySeptember 1999, Vol106, pp. 881-886RE VIEW Fetal pain: implications for research and practicePain is a subjective experience. The fetus cannot tell us sive and ascending process, with the cerebral cortex thewhat it is feeling, and there is no objective method for last region to develop.the direct measurement of pain. To address the question Connections from the periphery to the spinal cord areof pain in the fetus, one must use indirect evidence from formed early, at about eight weeks; C fibres begin toa variety of sources, and then make an informed guess. grow into the human fetal spinal cord at about 10This approach is similar to that which we use with ani- weeks4. The substantia gelatinosa in the dorsal horn ismals. We cannot ask animals how they feel, but infer the spinal cord region of interneurones which is thoughtfrom a variety of indirect approaches including study of to play a major part in the modulation of noxious inputs;their behaviour, anatomy, and physiology. by 30 weeks of gestation it has most features of the adult4. The cerebral cortex starts to form at 10 weeks, although at that stage it is isolated from the rest of theDoes the fetus feel pain? brain5. Cortical development involves the structural dif- ferentiation and maturation of cortical neurones, fibres,Consciousness glia and blood vessels, and this starts only at about 17To feel pain, or suffer discomfort, one needs to be con- weeks of gestation with layers VI and V, but continuesscious, to be aware. We do not know when, if at all, con- until long after birth. From 15 weeks, the cortex issciousness starts in the fetus. The biological basis of underlain by the subplate zone, a layer of neuronesconsciousness is little understood although at least in below the cortex that is specific to the fetus. Synapsesadult humans, the evidence suggests that it is in some appear within the cortical plate from mid-gestation. Theway associated with electrical activity in the cerebral subplate zone expands considerably between 17 and 20cortex. Crick has suggested that one is conscious of weeks, while from about 17 weeks, there is a shiftingsomething when there is electrical activity in specific population of connections from the thalamus to thislarge neural cortical networks, particularly in layers IV region6. Thalamic fibres penetrate the cortical plateto VI of the cerebral cortex2. from 24 to 28 weeks~~, at this stage the full anatom- and Greenfield has emphasised that one should not think ical pathways necessary for nociception are in place.of consciousness as an all or none phenomenon, rather I. Kostovic, who has been involved in many of thethat it may come on like a dimmer switch. This concept fundamental human fetal neuroanatomical studiesG9,of evolving consciousnesscould apply to the developing has written in a personal communication: Between 22fetus, in whom experience is most unlikely to be the and 26 weeks of gestation the subplate zone contains ansame as that of an adult. Furthermore, the fetus may not abundant mixture of cholinergic, thalamo-cortical andhave the same physical basis for conscious experience corticocortical waiting neurones, and there are tran-as the older human. Frogs, for example, do not have a sient fetal synaptic circuits between the subplate anddeveloped cerebral cortex, lacking layers IV to VI. If cortical plate neurones. It seems probable that extrinsicthey are conscious at all, their experience must be asso- influences (via the thalamocortical pathways at least)ciated with activity in a less complex neuronal network, could change the activity of the neocortical alange atpossibly more analogous to the fetal subplate zone3. that stage, and possibly even earlier e.g. in a 20 week fetus when thalamocortical and cholinergic afSerents already have synapses with upper subplate neurones,Anatomical evidence and these neurones very probably send axons into theThe most important evidence is anatomical. For the cortical plate. At least from mid-gestation onwards itfetus to feel pain, it is necessary for the requisite noci- seems that extrinsic influences (via thalamo-corticalceptive pathways to be developed. This involves neural pathways) can act through demonstrable synapses,connections between peripheral receptors and the spinal which, ifphysiologically active, may be involved in thecord, upward transmission via the spinal cord to the tha- modulation of the activity of thefetal neocortex.lamus, and from there to the outer cerebral layers. The Assuming that activity in the cerebral cortex or sub-development of the human nervous system is a progres- plate zone is necessary for consciousness, then for the0 RCOG 1999 British Jounzal of Obstetrics and Gynaecology 881
882 REVIEWfetus to be conscious of an external experience, these sound, over the next few weeksi6. It can respond toregions need to be connected with incoming nervous sound from 20 weeks and discriminate between differ-activity. Most incoming pathways, including nocicep- ent tones from 28 weeks17.tive ones, are routed through the thalamus and, as stated With the preterm baby, who now can be kept aliveabove, penetrate the subplate zone from about 17 weeks. from 23 weeks, one can observe behavioural responsesHowever, the earliest cortical links with the external to various clinical interventions. Such babies show aworld are formed even earlier than this; these comprise distinct pattern of behaviour to painful stimuli, such as athe catecholamine pathways of noradrenergic and heel prick. This includes a wide range of facial expres-dopaminergic neurones, and do not pass through the tha- sions and behaviours, including screwing up the eyes,lamus. These monoamine fibres start to invade the sub- opening the mouth, as well as clenching the hands andplate zone at 13 weeks and reach the cortex at about 16 limb withdrawal, which an older baby would also show,weeks6,o..This puts an early limit on when it is likely if in paini8. Most nurses and mothers looking afterthat the fetus might be aware of anything that is going preterm babies are convinced that they are both sentienton in its body or elsewhere. and feel pain. This type of evidence is similar to that The last pathways in the nociceptive system to be from animals. Most people believe that their cat feelsformed are the inhibitory descending serotonin neu- pain if someone treads on its tail.rones, which can block the ascending pathways. Thesedo not form until after birth4,raising the possibility thatthe fetus may actually be more sensitive to noxious Theoretical considerationsstimuli than the older child, and may explain why the Does one need previous experience to feel pain?newborn shows exaggerated behavioural responses tosensory provocation4. It has been argued that the fetus cannot feel pain, because pain is a complex phenomenon affected by previous e x p e r i e n ~ e ~It ~is . generally agreed that - ~Physiological evidence stimulation of a particular nociceptive pathway in theThere is some evidence for a primitive electroen- adult can be associated with various types of consciouscephalogram from 19 to 20 weeks, and sustained elec- experience, even in the same individual. Such varia-troencephalogram from 22 weeks; these have been tion may depend on previous experience, or on otherobtained from very preterm infantsy.*. Electroen- simultaneous occurrences. It is well known, for exam-cephalic patterns are clearly measurable in older ple, that a soldier wounded in battle often feels nothingpreterm babies and have been well characterised from at the time. It is also possible to sensitise the experi-28 weeks to termg. Studies of evoked responses in ence: people who are depressed often feel more painpreterm babies show that both visual and somatosensory than at other times. This complexity of the experiencepotentials can be elicited from as early as 24 weeks and of pain in adults is not controversial. However, the factare well developed by 27 weeks". The fact that a primi- that the suffering associated with nociceptive stimula-tive somatosensory potential can be evoked at 24 weeks tion in adults can be affected by activity in other partssuggests that the nociceptive pathways from the periph- of the brain, does not prove that in a ndve being, suchery to the cortex are functional from that timei4. The as the fetus, there can be no experience of pain. Theflexor reflex, a measure of nociceptive function in the fact that the sensation of pain can be affected by previ-central nervous system, is also present in preterm infants ous experience, does not entail the conclusion that pre-tested from 26 weeks4.This evidence thus suggests that vious experience is necessary to feel pain. Such anthe nociceptive system is functional from at least 24-26 argument would suggest that a newborn baby could notweeks, but gives little information concerning earlier feel pain either. The view that to experience pain it isgestations. necessary to have experienced pain previously is self defeating: there could never be a first experience of pain.Behavioural responsesOne has to be cautious about interpreting behavioural Is self consciousness needed?responses in terms of conscious experience, for some, atleast, could be purely reflex. It is well known that decor- It has also been suggested that consciousness impliesticate experimental animals show a wide range of self consciousness, and as the fetus is not self consciousbehavioural responses to noxious stimuli. it cannot be conscious eithe9. However, consciousness The fetus starts to make movements in response to does not necessarily imply self consciousness in thebeing touched from eight weeks15, and more complex adult sense. All that is needed for the fetus to feel pain, ismovements build up, as detected by real-time ultra- that it has a simple awareness of what is going on in 0 RCOG 1999 Br J Obstet Gynaecol 106,881-886
itself. It does not need the more complex understanding ovine, fetus is capable of an acute brain sparingthat it itself is different from the outside world. response. What then is the use of measuring stress responses? In considering stress responses in relation to the questionStress responses of fetal pain, the ‘null hypothesis’ is of relevance: ifRecent research has concentrated on the stress responses there were no change in stress hormone level, it wouldof the fetus to various interventions, just as neonatal be very unlikely that the fetus was experiencing pain.research did in the previous It is important to Stress responses can also be used to give some sort ofclarify the relevance of this work to a discussion of pain. index, though imperfect, of the degree of traumaStress responses, defined as an activation of specific involved, and further determine the effect of analgesiahormonal and neurotransmitter systems, do not provide or anaesthesia. Finally, high levels of stress hormonesa direct index of pain. Although stress hormones are may have long term consequences, as discussedusually increased when a subject is experiencing pain, below”.there are many other situations which are not painful, It was the demonstration of stress responses in thesuch as exercise, which also can increase their levels. newborn during surgery, that precipitated the change inFurthermore, production and release of stress hormones attitude in the medical and nursing care of newbornsuch as cortisol can be mediated by the hypothalamus, infants. It must be emphasised, however, with both thewithout involvement of the cortex or other higher brain fetus and the newborn, that a stress response in itselfregions involved in sentience. does not tell us directly what the baby is feeling. There is now evidence that the human fetus canmount substantial stress responsesz4.These have been Long term implicationsshown both by examining stress hormone levels in theblood before and after invasive procedure^^^^*^ and by There is evidence to suggest that a single early painfulexamining the redistribution of blood flow within the or stressful experience can have long term effects, andfetus27.We have shown that after intrauterine transfu- sensitise the child to pain and stress later. A recentsions carried out at the placental cord insertion, which is prospective study showed that male babies who hadnot innervated, there is little or no change in any of these been circumcised at five days or less, four to six monthsvariables. However, after procedures through the intra- later cried more, and showed more often other forms ofhepatic vein, that involve piercing the fetal abdomen, pain behaviour in response to their vaccination jab, thanwhich is innervated, there are major changes. Cortisol, those who had not been circumcised”. It has also beenj3-endorphin and noradrenaline rise substantially after shown that repeated insults, such as heel lancing, mayblood transfusions, a slow procedure that takes at least induce a state of hypersenitivity in the response to10 minutes. With shorter interventions, such as fetal pain32.Reynolds and F i t ~ g e r a l d ~ ~ shown in a rat haveblood sampling without transfusion, cortisol and j3- model that in the neonatal period there can be long last-endorphin remained constant after procedures at either ing sensory nerve sprouting at the site of a wound; thissite. There is, however, evidence of a more rapid could be one mechanism for long term hypersensitivityincrease in noradrenaline, with procedures that involve to pain.piercing the fetal abdomen, from at least 18 weeks of Animal studies suggest that brief fetal or neonatalgestation, although this response seems variablez6. stress can have long term effects. The developing ner- With Doppler ultrasound, our group has also shown vous system appears to be at a very plastic stage andthat with invasive procedures there is a significant fall in vulnerable to insult. In rats born at a stage equivalent tothe pulsatility index in the middle cerebral artery, con- the late fetus in man early postnatal handling perma-sistent with a redistribution of blood flow to the brain nently increases both the density of glucocorticoid(‘brain sparing’). It has been found after procedures that receptors in the hippocampu~~~, the behavioural andinvolve piercing the fetal trunk, and occurs rapidly, as responses to stress throughout life. Handled ratsearly as 70 s after the These include fetal secreted less corticosterone and showed a faster returnblood sampling, tissue and urine sampling, body cavity to basal levels in a stressful situation. Handling in theaspirations, and insertion of feto-amniotic shunts. These first postpartum week had greater long term effects, thanresponses have been found at all gestations studied from handling in the subsequent two weeks35.as early as 16 weeks. It has been well established that Fuji et u Z . ~showed that exposure of pregnant rats to ~this cerebral redistribution response occurs in the human hydrocortisone for only three days, affected the longfetus during the chronic stress associated with intrauter- term development and behaviour of the offspring. A sin-ine growth restriction and hypoxaemia, and in response gle dose of dexamethasone administered to the rhesusto the acute stresses of haemorrhage or hypoxaemia in macaque was sufficient to damage fetal hippocampalanimal modelsz9.It thus appears that the human, like the formation, especially the CA3 region37.Prenatal stressQ RCOG 1999 Br J Obstet Gynaecol 106,881-886
884 REVIEWof pregnant rhesus monkeys has also been shown to 20 and 24 weeks and only a very few after that. Late ter-have long term effects on the offspring, especially aug- minations may cause pain to the fetus if they involve anmenting their hormonal and behavioural responses to invasive procedure, such as surgical dismemberment.new stress or^^^.^^. This may, at least in part, be due to the Modification of the technique, such as preparatorydirect transfer of some maternal cortisol across the pla- occlusion of the umbilical cord, may be ap~ropriate~~.centaN. Whether potassium-induced termination of pregnancy Whether fetal experience of pain or the activation of at a viable gestation or the hypoxaemia caused by uter-major stress responses, either in utero or at birth, has ine contractions in terminations induced by prosta-any long term effects in humans is not known. These glandins cause pain or discomfort is hard to assess.animal experiments suggest that it may be possible andthat this is an important area for future research. Childbirth The experience of the baby during birth is not usuallyClinical implications considered. It is generally assumed that as birth is a nat-The fetus is currently treated as though it feels nothing, ural phenomenon, undergone for thousands of yearsand is given no analgesia or anaesthesia for potentially without pain relief, that it is painless for the baby. Thispainful interventions.This is similar to the way in which may not be the case.newborn babies used to be treated, until the major Noradrenaline levels in umbilical cord blood afterchange of practice which arose out of the work of Ayns- spontaneous vaginal delivery are 10 to 20 times thoseley Green, Anand and colleague^^^*^^*^. They compared before labouP, and several times higher than contem-newborn infants undergoing cardiac surgery who poraneous maternal levels. Babies born by elective cae-received deep anaesthesia with sufentanil with those sarean section have smaller rises in cortisol,given a lighter regimen of halothane and morphine. The noradrenaline, met-enkephalin and P-endorphin thansufentanil group, in whom the responses of cortisol and those born by vaginal delivery4749.Vaginal deliverynoradrenaline were reduced to the baseline, had a much seems advantageous for transitional respiratory adapta-better post-operative outcome in terms of sepsis and tion at birth, and catecholamines are known to facilitatemortality, than the latter, in whom the stress hormone resorption of pulmonary fluidSo. Thus a degree of stressresponse was not reduced to the same extent4*.Now at birth assists the babys adaptation to the externalsuch pain relief is routinely given to babies not just for world. The rise in stress hormones, however, can beopen surgery, but for more minor procedure^^^. How- considerably greater in assisted than in normal vaginalever, it is possible that opiate drugs may themselves deliveries". Elevation of fetal catecholamines in thehave long term adverse effects, and research is needed umbilical cord are associated with p02 and pH levelsto determine their risk-benefit ratio for different inter- suggestive of h y p ~ x i aIt ~ . likely that the mechanical ~ isventions, both in the fetus and in the neonate. effects of instrumental delivery may add to this stress, and it may be that in the future obstetricians will con- sider giving analgesia before, or immediately after, suchZnvasive procedures deliveries.There are several areas where it is appropriate to con-sider pain relief or anaesthesia in the fetusu. Firstly, Analgesia for the fetustherapeutic interventions, such as shunt insertions orblood transfusions which are carried out for the benefit Failure to provide adequate analgesia for preterm babiesof the fetus. Many of these procedures are carried out in is now considered substandard and unethical practice52.the third trimester at gestational ages comparable to There have been similar calls for fetuses to be givenpreterm infants, who would be given analgesia. Sec- analgesia during invasive procedure^^^*^^, even thoughondly, there are diagnostic sampling or aspiration proce- there is no current evidence that analgesia blunts noci-dures in the fetus which are only of concern in this ceptive responses in utero, or how analgesia may becontext if the procedure transgresses the fetal body; this safely and effectively administered.will not be the case with needle insertion of the placenta, The opioid agonists, such as fentanyl, are the drugsamniotic fluid or umbilical cord. most widely used for sedation and analgesia in neona- tology. Fentanyl has been given directly intravenously to fetuses before open fetal surgery, without apparentTermination of pregnancy adverse effect, but also without any proof that it workP.The majority of terminations of pregnancy are camed Direct administration requires intravenous injection toout before 13 weeks of gestation, but 10% are per- the fetus, and this is known to have risks. These could beformed between 13and 19 weeks, less than 1%between reduced but not abolished by direct fetal intramuscular 0 RCOG 1999 Br J Obstet Gynaecol 106,881486
REVIEW 885administration, the stress of which is likely to be rela- available anatomical and physiological evidence. Thetively minor, but would prolong the procedure in order physical system for nociception is present and func-to await absorption. Intra-amniotic injections of lipid tional by 26 weeks and it seems likely that the fetus issoluble opiods results in sub-therapeutic levels in the capable of feeling pain from this stage. The first neu-fetuss6. rones to link the cortex with the rest of the brain are Intravenous administration of fentanyl in the mother monoamine pathways, and reach the cortex from aboutis unsatisfactory since the rate of transfer across the pla- 16 weeks of gestation. Their activation could be associ-centa is slow: 10 minutes after administration of 1 pgkg ated with unpleasant conscious experience, even if notthe average fetal:maternal ratio was 0.31”. Larger doses pain. Thalamic fibres first penetrate the subplate zone atmay cause respiratory depression in the mother. Intra- about 17 weeks of gestation, and the cortex at 20 weeks.venous benzodiazepines cross freely into the fetal circu- These anatomical and physiological considerations arelation, with fetomaternal equilibrium occurring within important, not only because of immediate suffering, but5-10 minute^^^.^^. However, not only do they cause also because of possible long term adverse effects of thissedation, they may also have adverse behavioural early experience. Research in these areas is urgentlyeffects if delivery soon follows, impairing fetal respon- required.siveness. General anesthesia has significant risks in The eighteenth century philosopher, Jeremy Ben-pregnancy. The potential benefits of analgesia in the tham, wrote of animals The question is not Can they rea-fetus need to be balanced against the risk of additional son?, noc Can they talk?, but Can they suffer?. Thisprocedures, and the potential for adverse long term drug caused a change in attitude towards animals and theireffects.Administration of safe and effective analgesia to treatment that is continuing to day, such that in the UK,the fetus, without adverse effects in the mother, is a con- even frogs and fishes are required by Act of Parliamentsiderable challenge. to be protected by anaesthesia from possible suffering due to invasive procedures. Why not human beings?Medical versus scientific caution AcknowledgementsThere is clearly is not enough evidence to be certain ifand when the fetus starts to feel pain. By 26 weeks the The authors would like to thank Dr M. Marin-Padillafull anatomical system for nociception has been formed, and Dr I. Kostovic for helpful information on fetal neu-the electroencephalogramshows activity in the cerebral roanatomy. Our work in this area is funded by WellBe-cortex, and the preterm baby of the same gestational ing, The Henry Smith Charity, and the Childrenage, if delivered, shows a complex range of pain Nationwide Medical Research Fund.behaviour. Some have concluded that it is not possiblefor the fetus to be aware of events before 26 weeks of Vivette Glover, Reader & Nicholas M. Fisk, Professorgestation20.60, not to feel pain until considerably later and Division of Paediatrics, Obstetrics and Gynaecology,than that6’. This seems unduly certain, given the avail- Imperial College School of Medicine, Queenable evidence. Before 26 weeks, too little is known Charlotte’s and Chelsea Hospital, Londonabout the physical basis of consciousness in the fetus,the function of the subplate zone, the function of tran- Referencessient connections to the cortical plate and subplate zone, 1 Crick F. The Astonishing Hypothesis. London: Simon and Schusterand the role of the monoamine innervation, to be sure Ltd, 1994.that the fetus has no awareness. Given the anatomical 2 Greenfield SA. Journeys to the Centres of the Mind. Towards a Sci- ence of Consciousness. New York: WH Freeman, 1995.evidence, it is possible that the fetus can feel pain from 3 Marin-Padilla M. Ontogenesis of the pyramidal cell of the mam-20 weeks and is caused distress by interventions from as malian neocortex and developmental cytoarchitectonics: a unifyingearly as 15 or 16 weeks. This sets a limit to the earliest theory. J Comp Neuroll992; 321: 223-240. 4 Fitzgerald M. Development of pain pathways and mechanisms. In:stage that analgesia might be considered. Anand KIS & McGrath PJ, editors. Pain Research and Clinical Man- agement. Vol5. Pain in Neonates. Amsterdam: Elsevier. 1993: 1%38. 5 Marin-Padilla M. Structuralorganisation of the human cerebral cortexConclusion prior to the appearance of the cortical plate. Anat Embryo1 1983; 168: 21-40.It is not possible to measure pain directly in the fetus. 6 Kostovic I, Rakic P. Developmental history of the transient subplateStudies of stress responses can be used to give an index zone in the visual and somatosensory cortex of the macaque monkey and human brain. J Comp Neurol1990; 297: 441470.of the degree of trauma induced by different interven- 7 Mrzljak L, Uylings HBM, Kostovic I. van Eden CG. Prenatal devel-tions, and also the response to analgesia or anaesthesia, opment of neurones in the human prefrontal cortex. J Comp Neurolbut they do not indicate what the fetus actually experi- 1988;271: 355-386. 8 Molliver M, Kostovic I, van der Loos H. The development ofences. The assessment of whether or when the fetus is synapses in cerebral cortex of human fetus. Brain Res 1973; 50:likely to feel pain has to be based on an evaluation of the 403-407.0 RCOG 1999 Br J Obstet Gynaecol 106,8231-886
886 REVIEW 9 Kostovic I, Knezevic S, Wisniewski HM, Spilich GJ, editors. Neu- adrenal stress response. Cell Mol Neumbioll993; 13: 321-347. rodevelopment,Aging and Cognition. Boston: Birkhauser, 1992. 36 Fuji T, Horinaka M, Hata M. Functional effects of glucocorticoid10 Berger B, Verney C, Golman-Rakic PS. Prenatal monoamine innerva- exposure during fetal life. Prog Neumpsychopharmacol Biol Psychiat tion of the cerebral cortex: differences between rodents and primates. 1993; 17: 279-293. In: Kostovic I, Knezevic S, Wisniewski HM, Spilich GJ, editors. Neu- 37 Uno H, Lohmiller L, Thieme C et al. Brain damage induced by prena- rodevelopment, Ageing and Cognition. Berlin: Birkhauser: 1992: tal exposure to dexamethasone in fetal rhesus macaques. 1. H i p 18-36. pocampus. Dev Brain Res 1990,53: 157-167.11 Zecevic N, Verney C. Development of the catecholamineneurons in 38 Clarke AS, Wittwer DJ, Abbott DH, Schneider ML. Long term effects human embryos and fetuses, with special emphasis on the innervation of prenatal stress on HPA reactivity in juvenile rhesus monkeys. Dev of the cerebral cortex. J Comp Neuroll995; 351: 509-535. Psychobioll994;27: 257-269.12 Flower MJ. Neuromaturation of the human fetus. JMed Philos 1985; 39 Schneider ML, Coe CL, Lubach GR. Endocrine activation mimics the 1 0 237-25 1 . adverse effects of prenatal stress on the neufomotor development of13 Klimach VJ, Cook RWI. Maturation of the neonatal somatosensory the infant primate Dev Psychobioll992;25: 427439. evoked response in preterm infants. Dev Med Child Neumll988; 30: 40 Gitau R, Cameron A, Fisk NM,Glover V. Fetal exposure to m a t e d 208-214. cortisol. Lancet. 1998; 352 707-708.14 Hrbek A, Karlberg P, Olsson T. Development of visual and 41 Anand KW, Sippell WG, Schofield NM, Aynsley-Green A. Does somatosensory evoked responses in pre-term newborn infants. Elec- halothane anaesthesia decrease the metabolic and endocrine stress tmenceph Clin Neumphysioll973;3 :225-232. 4 responses of newborn infants undergoing operation? BMJ 1988; 2 %15 Prechtl HF. Ultrasound studies of human fetal behaviour. Early Hum 668-672. Dev 1985; 12:91-98. 42 Anand KJS, Hickey PR. Halothanemorphine compared with high16 De Vries J, Vissier G, Prechtl H. The emergence of fetal behaviour. dose sufenntanil for aneastheisa and postoperative analgesia in neona- Early Hum Dev 1982; 12: 301-322. tal cardiac surgery. NEngl JMed 1992; 326 1-9.17 Hepper PG, Shahidullah BS.The development of fetal hearing. Fetal 43 de Lima J, Lloyd-Thomas AR, Howard RF, Sumner E, Quinn TM. Mat MedRev 1994;6: 167-179. Infant and neonatal pain: anaesthetists’ perceptions and prescribing18 Grunau RVE, Craig KD. Pain expression in neonates: facial action patterns. BMJ 1996; 313: 787. and cry. Pain 1987; 2 8 395410. 44 Glover V, Fisk N. Do fetuses feel pain? We don’t know; better to err19 Derbyshire SWG. Fetal stress responses. Lancet 1994; 344:615. on the safe side from mid-gestation. BMJ 1996; 313: 796.20 Derbyshire SWG, Furedi A. ‘Fetal pain’ is a misnomer. BMJ 1996; 45 Bennett P Fetal stress responses. Lancet 1994; 344: 615. . 313 795. 46 GulmezogluAM, Mahomed K, Hofmeyr GJ, Nikodem VC, KramerT.21 SzawarskiZ. Probably no pain in the absence of self. BMJ 1996; 313: Fetal and maternal catecholamine levels at delivery. J Perinat Med 796-797. 1996; 24: 687691.22 Anand KJS, Brown MJ, Causon RC, Christofides ND, Bloom SR, 47 Procianoy RS, Cecin SKG. The influence of labor and delivery on Aynsley-Green A. Can the human neonate. mount an endocrine and preterm fetal adrenal function. Acta Paediat Scand 1985; 7 4 metabolic response to surgery? J Paediatr Surg 1985; 2 0 4 1 4 8 . 4m04.23 Anand KJS, Hickey PR. Pain and its effects in the human neonate and 48 Tropper PJ, Warren WB, Jozak SM, Conwell IM, Stark RI, Goland fetus. NEngl J Med 1987; 317: 1321-1329. RS. Corticotropinreleasing hormone concentrations in umbilical cord24 Glover V, Giannakoulopoulos X. Stress and pain in the fetus. In: blood of preterm fetuses. J Dev Physioll992; 1 8 81-85. Aynsley-Green A, Platt W, Lloyd-ThomasAR. editors. Balliere’sC i ln 49 Hawdon JM, Ward Platt MP, Aynsley-Green A. Patterns of metabolic Paediat. 1995:495-510. adaptation for preterm and term infants in the f i st postnatal week25 Giannakoulopoulos X, Sepulveda W, Kourtis P, Glover V, Fisk N. Arch Dis Child 1992;67: 357-365. Fetal plasma cortisol and b-endorphin response to intrauterine 50 Walters DV, Walters RE. The role of catecholamines in lung liquid needling. Lancer 1994; 344:77-81. absorption at birth. Ped Res 1978; 12: 239.26 GiannakoulopoulosX, Teixeira J, Fisk N, Glover V. Human fetal and 51 Rothenberg SJ, Chicz-DeMet A, Schnaas L, Karchmer S, Salinas V, maternal noradrenaline responses to invasive procedures. Ped Res Guzman LA. Umbilical cord b-endorphin and early childhood motor 1999; 45: 494499. development. Ear Hum Dev 1996; 46:83-95.27 Teixeira J, Fogliani R, Giannakoulopoulos X,Glover V,Fisk N. Fetal 52 Walco GA, Cassidy RC, Schechter NL.Pain, hurt and harm. N Eng J haemodynamic stress response to invasive procedures. Lancet 1996; Med 1994; 331: 541-544. 347: 624. 53 Furness M. Diagnostic potential of fetal renal biopsy. Prenar Diagn28 Teixeira J, Glover V, Fisk NM. Acute cerebral redistribution in 1994; 14:415. response to invasive precedures in the human fetus. Am J Obster 54 Commission of Enquiry into Fetal Sentience. Human sentience before Gynecoll999. In press. birth. London: Care T u t 1996. rs,29 Nathanielsz PW. The role of basic science in preventing low birth 55 Adzick NS, Harrison MR. Fetal surgical therapy. Lancer 1994; 343: weight. Future Child 1995; 5: 57-70. 897-902.30 Sapolsky RM. Why stress is bad for your brain. Science 1996; 273: 56 Sxto HH, Mann LI, Bhakthavathsalan A, Liu M, Inturrisi CE. 749-750. Meperidine pharmacokinetics in the maternal-fetal unit. J Phannacol3 1 Taddio A, Katz J, nersich AL, Koren G. Effect of neonatal circumci- Exp Ther 1978; 206: 448459. sion on pain response during subsequent routine examination. Lancer 57 Eisele J, We A. Newborn and maternal fentanyl levels at caesarean 1997; 349 599-603. section. Anesth Analg 1982; 61: 179-180.32 Fitzgerald M, Millard C, Mchtosh N. Cutaneous hypersenstivity fol- 58 Bakke 0, Haam K. Time course of transplacental passage of lowing peripheral tissue damage in newborn infants and its reversal diazepam: influence of injection-delivery interval on neonatal drug with topical anaesthesia. Pain 1989;3 9 31-36. concentrations. Clin Pharmacokinet 1982; 7: 353-362.33 Reynolds ML, Fitzgerald M. Long-term sensory hyperinnervation fol- 59 Guerre-Millo M, Rey E, Challier J, Turquais J. d’Athis P, GO. Trans- lowing neonatal skin wounds. J Comp Neuroll995; 358: 487498. fer in vitro of three benzodiazepines across the human placenta. Eur J34 Meaney M, Aitken DH. The effects of early postnatal handling on hip- Clin Pharmacoll979; 15: 171-173. pocampal glucocorticoid receptor concentrations: temporal parame- 60 Wise. J. Fetuses cannot feel pain before 26 weeks. BMJ 1997; 315: ters. Dev Brain Res 1985; 22: 301-304. ll-00 ll00.35 Meaney M, Bhatnagar S, Diorio J et al. Molecular basis for the devel- 61 Derbyshire S. Locating the beginnings of pain. Bioethics 1999; 13: opment of individual differences in the hypothalamic-pituitary- 1-31. 0 RCOG 1999 Br J Obstet Gynaecol 106,881-886