Archives of Disease in Childhood 1994; 70: F195-F200 F195

Cerebral palsy and neonatal encephalopathy Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.70.3.F195 on 1 May 1994. Downloaded from

Geraldine Gaffney, Valerie Flavell, Ann Johnson, Marian Squier, Susan Sellers

Abstract obstetrically preventable.2 In some term A retrospective cohort study was carried infants who later develop cerebral palsy, how- out to test the hypothesis that children ever, there are pre-existing genetic or develop- born at term with cerebral palsy with signs mental factors which could have resulted in of neurological dysfunction preceded by damage to the developing brain in the depression at birth (termed neonatal antenatal period, or which may have altered encephalopathy) differ from those without the vulnerability of the fetus to the normal such signs in the frequency of antenatal stress of labour.3 4 and perinatal factors, and in the severity The challenge is to identify those infants and characteristics of their impairment who show the effects of an acute intrapartum and disability. hypoxic-ischaemic injury as opposed to an The study was carried out in the area earlier antenatal insult. Neuropathological covered by Oxford Regional Health study of infants who die may help to dis- Authority. Antenatal, intrapartum, neo- tinguish the two groups by identifying the age natal factors, and the later clinical status and characteristics of the lesions in the brain. ofthe two groups of children were used as In term infants acute asphyxia typically the main outcome measures. produces damage in the hippocampus, Although most maternal and antenatal thalamus, brain stem, and anterior horn cells characteristics were similar in the two of the spinal cord.5-9 Earlier antenatal groups, the mothers of children with a ischaemic injury affecting the less mature brain history of neonatal encephalopathy were tends to produce more severe damage in the more likely to be primigravidae (odds white matter of the cerebral hemispheres with ratio (OR) 2-0; 95% confidence interval relative sparing of the cerebral cortex.'0 11 (CI) 1P0 to 4-3) and to have a pregnancy of In life, this distinction may not be made so greater than 41 weeks' gestation (OR 3 5; clearly. Neuroimaging, such as ultrasound, 95% CI 1*0 to 12-1). Intrapartum compli- may not detect mild diffuse lesions of antenatal cations were more frequent in the neo- origin and may also be unreliable soon after an natal encephalopathy group: meconium acute hypoxic injury.'2 Newer techniques staining ofthe amniotic fluid (OR 3*5; 95% such as magnetic resonance imaging'3 14 CI 1-5 to 7.8), an ominous first stage and computed tomography,'5-17 though more cardiotocograph (OR 10-2; 95% CI 2-9 to discriminating, have not been available for 36.4), with a longer median duration of routine use until recently. abnormality (200 v 48 minutes). At 5 years We therefore have to turn to clinical obser- ofage those with neonatal encephalopathy vations to distinguish when brain injury might http://fn.bmj.com/ were more likely to have developed spastic have occurred. For many years it has been quadriplegia (OR 4*8; 95% CI 2-2 to 10.5), assumed that intrapartum clinical signs of to have visual impairment (OR 3-0; 95% fetal distress such as changes on a cardiotoco- CI 1-0 to 8.6), and to be non-walking (OR graph or meconium staining of amniotic fluid 4-0; 95% CI 1-8 to 8.8) than those without indicate acute hypoxic stress. It has become

neonatal encephalopathy. clear, however, that these signs may also on September 25, 2021 by guest. Protected copyright. Children with cerebral palsy who were reflect earlier antenatal ischaemic damage and born at term and have neonatal encepha- this non-specificity limits their usefulness as lopathy are more likely to have had signs markers of recent hypoxic events.'8 It has of intrapartum asphyxia and are more been suggested that signs of neonatal neuro- likely to have a more severe form of logical abnormality in the first hours after National Perinatal cerebral palsy than those without a history birth are the most reliable indicator of a recent Epidemiology Unit, ofneonatal encephalopathy. Although this period of hypoxia.'920 This view has been Radcliffe Infirmary, Oxford OX2 6HE group represents only one in 10 ofall cases adopted to the extent that the term 'hypoxic- G GafEney of cerebral palsy, some of these may be ischaemic encephalopathy' has been used to V Flavell obstetrically preventable. describe infants with abnormal neonatal A Johnson (Arch Dis Child 1994; 70: F195-F200) neurological signs. Over the last few years a John Radcliffe consensus has evolved that, to attribute Maternity Hospital, cerebral palsy to intrapartum hypoxia, the Oxford S Sellers Cerebral palsy is used to refer to a spectrum of infant must be both depressed at birth and motor disorders of differing aetiology and have signs of neonatal neurological abnor- Department of clinical manifestation. A number of predispos- mality.2' It is likely that not all infants with Neuropathology, Radcliffe Infirmary, ing antepartum, intrapartum, and postnatal neurological signs have had intrapartum Oxford factors have been identified.' In children with hypoxia and a more general term such as M V Squier cerebral palsy born after 37 weeks' gestation neonatal encephalopathy is preferable.22 In Correspondence to: and who have no congenital anomaly, the this study we have used the term neonatal Dr Johnson. neurological deficit is often attributed to an encephalopathy to describe infants who Accepted 11 February 1994 acute intrapartum hypoxic episode, possibly have neonatal neurological dysfunction and F196 Gaffney, Flavell, Johnson, Squier, Sellers

depression at birth. We assume that this group The children with cerebral palsy were Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.70.3.F195 on 1 May 1994. Downloaded from of infants includes those most likely to have divided into those with signs of neonatal had a recent hypoxic episode. encephalopathy and those without. This was If this is so, children with cerebral palsy based on information recorded in the neonatal without neonatal encephalopathy might be case notes. We defined neonatal encephalo- expected to differ in a number of ways from pathy as depression at birth, based on a one those with encephalopathy. For example, those minute Apgar score of less than or equal to six, without neonatal encephalopathy would have: followed by evidence of neonatal neurological (a) a higher frequency of adverse maternal and abnormality such as lethargy, coma, impaired antenatal factors (as they are more likely to respiration, seizures, and/or tone changes. have cerebral palsy of antenatal origin); (b) a Infants with transient jitteriness were lower frequency of acute events in labour such excluded. We did not attempt to grade or as haemorrhage and less severe signs of fetal allocate a level of encephalopathy. Not all distress (as measured by the frequency and infants with seizures were allocated to the neo- duration of abnormalities on a cardiotoco- natal encephalopathy group. Infants who had graph); and (c) a different pattern and distri- seizures but who appeared neurologically bution of motor and associated sensory and normal between seizures (that is not lethargic intellectual deficits. Those without neonatal or hypotonic) and who were not depressed at encephalopathy (cerebral palsy assumed to be birth were not included in the neonatal of antenatal origin) are more likely to have encephalopathy group. signs reflecting white matter damage, which The Oxford regional register uses a standard predominantly affects motor tracts. Those with system for describing children with central neonatal encephalopathy (cerebral palsy motor deficit.23 On the basis ofthis, the clinical assumed to be of acute intrapartum origin) characteristics of the children in the study were would be more likely to have evidence of described in terms of the distribution of tone extensive deep grey matter damage in addition changes, as walking or non-walking, and with to the white matter being affected, manifest as or without intellectual deficit, vision loss, severe motor deficits with associated intellec- seizures, involuntary movement, or bulbar tual and sensory impairment. signs such as difficulty in swallowing. We tested these hypotheses by comparing the antenatal histories, intrapartum events, the duration of stress in the fetus, and the clinical INFORMATION ON ANTENATAL AND status of the surviving children between two INTRAPARTUM EVENTS groups derived from a total birth population. The obstetric notes of mothers included in the These two groups are: (1) children with study were obtained. All information about the cerebral palsy born of a singleton pregnancy outcome of the infant was masked by a after 37 completed weeks' gestation without researcher who did not participate in the congenital anomaly and who did not have review of the notes. Information was neonatal encephalopathy (no neonatal abstracted about the antenatal period and encephalopathy group); and (2) children with events during labour, delivery, and immedi- cerebral palsy born of a singleton pregnancy ately after birth. Information was collected after 37 completed weeks' gestation without about the antenatal and intrapartum character- http://fn.bmj.com/ congenital anomaly and who did have neonatal istics blind to the condition of the neonate and encephalopathy (neonatal encephalopathy later development. group). Gestation was estimated from accurate menstrual data or by ultrasound assessment before 20 weeks' gestation. If there was a dis-

Methods crepancy of greater than 14 days between the on September 25, 2021 by guest. Protected copyright. CHILDREN WITH CEREBRAL PALSY two, the ultrasound estimate was used. Children with cerebral palsy born between 1984 and 1987 who were singleton deliveries after 37 completed weeks' gestation were FETAL HEART RATE MONITORING identified from the Oxford regional register of At the time of birth of the subjects, continuous early childhood impairment. This register electronic fetal heart rate monitoring was widely includes children of mothers who are resident used in all 10 obstetric units in the region. Some within the Oxford health region at the time of of the mothers in the study who did not have delivery. Multiple sources of ascertainment are continuous electronic fetal heart rate monitor- used to compile the register and the status of ing had an admission cardiotocograTh with children is determined at 3 and 5 years. We intermittent monitoring during labour; the excluded children in whom a major congenital remainder had intermittent auscultation. It was anomaly was diagnosed. This comprised a accepted practice that ifthere were signs offetal mixed aetiological group; some had well heart rate abnormality on intermittent ausculta- recognised syndromes such as the 'prune belly' tion, continuous electronic fetal heart rate syndrome, others had evidence of intrauterine monitoring was started. infection such as congenital varicella, and The original traces were reviewed for the some a genetically determined disorder such study and the presence and duration of as X linked spastic paraplegia. In addition, ominous changes were noted. The terminology children in whom there was a definite posmatal used to describe and classify the cardioto- cause for cerebral palsy, such as neonatal cograph was that used in the Dublin trial of meningitis or trauma, were excluded. continuous fetal heart rate monitoring.24 Cerebral palsy and neonatal encephalopathy F197

Table 1 Characteristics ofmothers ofchildren with cerebral palsy with and without postnatal cause for their disability and 123 Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.70.3.F195 on 1 May 1994. Downloaded from neonatal encephalopathy (NE) (36%) were born before 37 completed weeks' No (%) No (%) gestation. Of the remainder, 42 (22%) had evi- without NE with NE Odds ratio dence of major congenital anomaly and two Maternal characteristic (n= 100) (n=41) (95% CI) were term multiple births. The remaining 145 Unmarried 11 (11) 4 (10) 0 9 (0-3 to 29) children were eligible for inclusion in the Maternal disease 12 (12) 3 (7) 0-6 (0-2 to 2 2) Primigravida 34 (34) 21 (51) 2-0 (1-0 to 4 3) study. There was insufficient information in Recurrent abortion 2 (2) 1 (2) 1-2 (0-1 to 13-9) the case notes of four, three of these being born Poor obstetric history 5 (5) 3 (7) 1-5 (0-3 to 6-6) Previous preterm labour 2 (2) 2 (5) 2-5 (0 4 to 18-5) outside the region, and these were excluded. Maternal smoking 25 (25) 7 (17) 0-6 (0-2 to 1-6) Of the 141 children with cerebral palsy in Mean age (years) 26-5 26-5 0 (-1-83 to 1.83)* Mean length of menstrual cycle (days) 30-8 29-1 1-8 (-1 1 to 4-7)* the study, 100 (7 1%) had no sign of neonatal encephalopathy and 41 (29%) had evidence of *Difference of means (95% confidence intervals). neonatal encephalopathy.

Table 2 Antenatalfactors in mothers ofchildren with cerebral palsy with and without neonatal encephalopathy (NE) MATERNAL AND ANTENATAL CHARACTERISTICS Existing maternal disease occurred more often No (Oo) No (%/6) without NE with NE Odds ratio in the group without neonatal encephalopathy Antenatalfactor (n= 100) (n=41) (95% CI) and previous preterm labour more often in the Antenatal infection 4 (4) 2 (5) 1-2 (0-2 to 7 0) neonatal encephalopathy group; numbers were Premature rupture of membranes 1 (1) 2 (5) 5-1 (0-5 to 57 6) however, and confidence intervals wide Pre-eclampsia 33 (33) 13 (32) 0 9 (0 4 to 2-1) small, Severe pre-eclampsia 8 (8) 3 (7) 0 9 (0-2 to 3 6) (table 1). The only maternal factor which dif- Antepartum haemorrhage 3 (3) 1 (2) 0-8 (0-1 to 8 0) at the 5% level between the two groups Previous infertility 10 (10) 3 (7) 0-7 (0-2 to 2 7) fered Induced conception 4 (4) 2 (5) 1-2 (0-2 to 6-9) was primigravidity, which was more frequent Raised maternal serum a fetoprotein 3/54 (6) 1/23 (4) 0-8 (0 1 to 8 0) group (odds Polyhydramnios 2 (2) 1 (2) 1-2 (0 1 to 13-9) in the neonatal encephalopathy Oligohydramnios 3 (3) 1 (2) 0-8 (0 1 to 8 0) ratio (OR) 2-0; 95% confidence interval (CI) Reduced fetal movement 11 (11) 4 (10) 0 9 (0-3 to 2-9) 10 to 43). Complicated antenatal course 51 (51) 20 (49) 0 9 (0-4 to 1-9) Overall, half of the mothers of infants with- out neonatal encephalopathy (51/100) and Table 3 Intrapartum factors in mothers ofchildren with and without neonatal mothers of infants with neonatal encephalo- encephalopathy (NE) pathy (20/41) had one or more complicating antenatal factors (table 2). The frequency of No (%) No (%/6) without NE with NE Odds ratio other individual complicating antenatal factors Intrapartum factor (n= 100) (n=41) (95% CI) did not differ at the 5% level between the two Breech presentation 4 (4) 1 (2) 0-6 (0 1 to 5-5) groups. Pregnancy duration ¢42 weeks' gestation 5 (5) 6 (15) 3-5 (1 0 to 12-1) Induction of labour 24 (24) 17 (42) 2-2 (1 0 to 4 9) Induction of labour m primigravidae 9/34 (27) 12/21 (57) 3-7 (1-2 to 11-7) Augmentation of labour 13 (13) 6 (15) 1-2 (0 4 to 3-3) INTRAPARTUM CHARACTERISTICS First stage >12 hours 13/84 (16) 7/35 (20) 1-4 (0 5 to 3-8) Second stage >2 hours 6/82 (7) 8/31 (26) 4-4 (1-4 to 14-0) Pregnancy which continued after 41 com- Meconium stained amniotic fluid 17 (17) 17 (42) 3-5 (1-5 to 7-8) weeks' was more frequent in the neo- Haemorrhage in labour 2 (2) 1 (2) 1 2 (0 1 to 13-9) pleted Spontaneous vaginal delivery 66 (66) 10 (24) 0-2 (0 1 to 0 4) natal encephalopathy group (OR 3-5; 95% CI Forceps delivery 16 (16) 16 (39) 3-4 (1-5 to 7 7) to 12@1) (table 3), particularly in primi- http://fn.bmj.com/ Breech delivery 1 (1) 1 (2) 2-5 (0-2 to 40 5) IP0 All caesarean sections 16 (16) 14 (34) 2-7 (1-2 to 6-3) gravidae (OR II-0; 95% CI 1.2 to 102-5). All emergency caesarean sections 9 (9) 14 (34) 5-2 (2-1 to 13-4) Induction of labour (OR 2-2; 95% CI 1-0 to All emergency caesarean sections in labour 4 (4) 10 (24) 7-7 (2-3 to 26-4) 4.9), particularly in primigravidae (OR 3-7; 95% CI 1 2 to 11.7), was more frequent in the ETHICS APPROVAL neonatal encephalopathy group. A second

Approval for the study was obtained from the stage of labour exceeding two hours (OR 4-4; on September 25, 2021 by guest. Protected copyright. ethics committees of all eight health districts in 95% CI 1-4 to 14.0) and meconium staining of the Oxford region. the amniotic fluid were all more frequent in the neonatal encephalopathy group (OR 3-5; 95%/o CI 1.5 to 7 8). STATISTICAL ANALYSIS Seven infants without neonatal encephalo- Analysis of categorical data was performed by pathy were delivered by elective caesarean comparison of characteristics of those with section compared with none in the neonatal neonatal encephalopathy and those without encephalopathy group; the indication to neonatal encephalopathy expressed as odds perform five of these seven was a previous ratios and their 95%/o confidence intervals. caesarean section. Delivery by forceps or Continuous data were analysed comparing the emergency caesarean section was more difference between means and 95% confidence frequent in the neonatal encephalopathy group intervals for data with a normal distribution (OR 3-4; 950/o CI 1.5 to 7 7), (OR 7-7; 950/o CI and the Mann-Whitney test for non-normally 2-3 to 26-4) (table 3). distributed data. A total of 48% of mothers of infants without neonatal encephalopathy and 66% of mothers of infants with neonatal encephalopathy had a Results cardiotocograph available for analysis (table There were 339 children with cerebral palsy on 4). Eight per cent of the group without neo- the regional register who were born between natal encephalopathy and 48% of the neonatal 1984 and 1987 to mothers resident in the encephalopathy group had signs of fetal dis- Oxford region. Of these 27 (8%) had a tress with an increased likelihood of ominous F198 Gaffney, Flavell, Johnson, Squier, Sellers

Table 4 Findings on cardiotocograph (CTG) in mothers ofchildren with cerebralpalsy encephalopathy had a birth weight less than or Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.70.3.F195 on 1 May 1994. Downloaded from with and without neonatal encephalopathy (NE) equal to the 10th centile (OR 1 -1; 95% CI 0 4 No (%) No (%) to 2 8) and 19 (19%) ofthose without neonatal without NE with NE Odds ratio encephalopathy and nine (22%) with neonatal CTGfinding (n= 100) (n=41) (95% CI) encephalopathy had a head circumference less CTG performed 63 (63) 31 (76) 1-8 (0-8 to 4-1) than or equal to the 10th centile (OR 1-2; 95% CTG missing 15 (24) 4 (13) 0-6 (0-2 to 2-0) Ominous first stage CTG 4/48 (8) 13/27 (48) 10-2 (2-9 to 36-4) CI 0 5 to 2 9). The proportion of infants Ominous second stage CTG 19/45 (42) 21/25 (84) 7-2 (2-1 to 24-4) whose birth weight or head circumference was Median (range) duration of first stage 48-5 200-0 0.3t abnormality (min) (38-287) (15-480) less than or equal to the 10th centile did not Median (range) duration of abnormal 38-0 100-0 0-003t differ between the two groups. total CTG (min)* (8-287) (12-480) Among the infants without neonatal *Based on ominous CTG in either or both the first and second stages of labour. encephalopathy were 13 who had neonatal tMann-Whitney (p value). seizures but no other signs of encephalopathy. Two had ominous cardiotocograph changes Table 5 Outcome at 5 yearsfor those with and without neonatal encephalopathy (NE) but all had an Apgar score at five minutes between eight and 10. Two of these infants No (%/6) No (%o) without NE with NE Odds ratio were delivered by elective caesarean section. Outcome at S years (n= 100) (n=41) (95% CI) Quadriplegia 23 (23) 24 (58) 4-8 (2-2 to 10-5) Hemiplegia 39 (39) 7 (17) 0 3 (0-1 to 0 8) TYPES OF CEREBRAL PALSY Diplegia 9 (9) 3 (7) 0-8 (0-2 to 3-1) Hypotonia/varying tone 12 (12) 3 (7) 0-6 (0-2 to 2 2) At 5 years of age children with neonatal Other tone change 17 (17) 4 (10) 0-5 (0-2 to 1-7) encephalopathy were more likely than those No independent walking 23 (24) 21 (55) 4 0 (1-8 to 8 8) without neonatal encephalopathy to have No usefulvision 8 (9) 8 (22) 3 0 (1-0 to 8 6) Severe intellectual delay 22 (23) 15 (43) 2-5 (1-1 to 5 7) quadriplegia (OR 4-8; 95°/O CI 2.2 to 10-5) and Involuntary movements 17 (17) 11 (27) 1-8 (0-8 to 4 3) were less likely to have hemiplegia (OR 0 3; Swallowing difficulty 20 (20) 18 (44) 3-1 (1-4 to 6-9) 95% CI 041 to 0 8) (table 5). The motor disability seemed more severe in the neonatal encephalopathy group with a higher propor- changes on the cardiotocograph in both the tion of children unable to walk independently first (OR 10-2; 95% CI 2-9 to 36 4) and at 5 years (OR 4-0; 95% CI 1.8 to 8-8). There second (OR 7-2; 95% CI 2 1 to 24 4) stage on was an increased frequency among those with cardiotocograph in the neonatal encephalo- neonatal encephalopathy of associated impair- pathy group. The duration of ominous changes ments; severe visual impairment (OR 3 0; for the total labour was longer in the neonatal 95% CI 1-0 to 8&6), severe intellectual delay encephalopathy group than in the group with- (OR 2-5; 95% CI 1 1 to 5.7), and swallowing out neonatal encephalopathy (median duration difficulty (OR 341; 95% CI 1V4 to 6-9). without neonatal encephalopathy 38 minutes, Although the frequency of involuntary with neonatal encephalopathy 100 minutes; movement was increased in the neonatal Mann-Whitney, p=0 003). encephalopathy group this was not statistically significant at the 5% level (OR 1 8; 95% CI 0-8 to 4.3). NEONATAL CONDITION http://fn.bmj.com/ By definition, all infants who developed neo- natal encephalopathy had a one minute Apgar Discussion score of less than six, but among those without In recent years there has been a considerable neonatal encephalopathy 23 (23%) had a one change in our understanding of the origin of minute Apgar score of less than six but did not cerebral palsy. The earlier view that most

go on to demonstrate neonatal neurological cerebral palsy was attributable to intrapartum on September 25, 2021 by guest. Protected copyright. abnormality. The median Apgar scores at one hypoxia, much of which was obstetrically and five minutes for those with neonatal preventable, has been supplanted by a counter- encephalopathy were two and five respectively, view that 'birth asphyxia is a rare cause of whereas for those without neonatal encephalo- cerebral palsy'.25 pathy the median Apgar scores at one and five Although it is likely that in some infants who minutes were nine and 10 respectively (Mann- develop cerebral palsy, injury has been Whitney; p<0-0001). acquired around the time ofbirth, in others the Only 15 infants in the study had umbilical insult occurs antenatally and in many the brain cord gas analysis; all of the 10 with neonatal abnormality may have evolved over a long encephalopathy had an arterial pH of 7-2 or period oftime. These concepts ofthe origins of less, as did three of the five without neonatal fetal and neonatal brain injury are based on encephalopathy; and arterial base deficit of neuropathological,5-11 26 radiological,27-30 and - 12-0 was found in all 10 of those with neo- epidemiological3l-33 evidence. It would be natal encephalopathy and two of four without helpful to have a clearer understanding, neonatal encephalopathy. Among the infants according to the time of hypoxic insult, of the with neonatal encephalopathy, 29% had differences in the clinical characteristics of evidence ofrenal failure and 10% had evidence infants who later develop cerebral palsy. This of myocardial ischaemia. None of the group would not only help to identify preventable without neonatal encephalopathy had either factors, but also in deciding which infants may renal failure or myocardial ischaemia. Eighteen benefit from treatment which may reduce the (1 8%) of the infants with neonatal encephalo- secondary cell damage after acute intrapartum pathy and eight (20%) of those with neonatal hypoxia.34 Cerebral palsy and neonatal encephalopathy F199

We used the clinical finding of depression at was an increased frequency of induction of Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.70.3.F195 on 1 May 1994. Downloaded from birth followed by neonatal neurological abnor- labour and of pregnancies of greater than 41 mality as the most reliable marker of an acute weeks' gestation among the mothers of infants episode of intrapartum hypoxia. This finding who went on to develop neonatal encephalo- was used as the basis for differentiating those pathy. In addition, prolonged labour and signs children with a predominantly antenatal origin of prolonged severe distress were more fre- of cerebral palsy and those with a primarily quent in the neonatal encephalopathy group. intrapartum origin. We recognised, however, This cluster of adverse intrapartum events that this division may be rather crude. It is were also described by Minchom et al in infants possible, for example, that within the group of who developed early neonatal seizures.37 Such children with cerebral palsy with neonatal associations do not, of course, imply a causal encephalopathy there might be some whose link. encephalopathic signs reflected a longstanding This group (neonatal encephalopathy), neurological disorder. Further, it is possible however, includes those cases of cerebral palsy that a more critical evaluation of neurological which are most likely to be obstetrically states in the first few days of life might help to preventable. For example, it could be argued identify different aetiological subgroups. For that reduction in the duration of fetal distress example, we were aware of a group of infants by expediting delivery might have altered the who had neonatal seizures but did not show outcome. The greater frequency of fetal any other signs of encephalopathy and who distress among those with neonatal encephalo- were not depressed at birth. This group of pathy, however, could equally reflect long- infants appeared to be aetiologically distinct standing neurological dysfunction which is from those with changes in consciousness and not, at present, obstetrically preventable. muscle tone between seizures. It was also of The challenge is to find reliable ways of interest that in nine of the 13 infants in this identifying the infant who is neurologically group seizures began before the age of 48 abnormal before or during labour. Several hours. In the past it has been suggested that potential methods are being examined; for seizures which start in the first 48 hours of life example, the use of near infrared spectroscopy are those most likely to reflect intrapartum in labour and antenatal biophysical assess- hypoxia.35 A large systematic study of infants ment.3842 Using the latter, abnormal in utero with clinical signs of neurological dysfunction behavioural states can be shown in some in the neonatal period together with infants who later show abnormal neonatal neuroimaging would clarify these issues. neurological function. Neonatal evidence of Many infants who later have signs of damage to other organs may help to identify a cerebral palsy do not appear to have abnormal recent acute hypoxic event. Ischaemic damage neurological signs in the neonatal period. If to the kidneys or myocardium in an infant with neonatal encephalopathy is considered to be a neonatal encephalopathy who subsequently condition of term infants, then only 12% has cerebral palsy provides supporting (41/339) of all children with cerebral palsy evidence of an intrapartum origin to the later born to residents of the area in the four birth neurological deficit.43-45 years studied had evidence of neonatal It has been previously suggested that clinical encephalopathy. This once again emphasises subgroups of cerebral palsy may have different http://fn.bmj.com/ the need to look at aetiological factors which aetiologies.46 Our hypothesis that children may influence earlier intrauterine cerebral with cerebral palsy who have a history of development, not just events related to labour, neonatal encephalopathy have more extensive in the majority of children with cerebral palsy. motor impairment and associated sensory and The search for clinical markers of adverse intellectual disability was sustained. Further

antenatal factors in populations of children exploration of the relation between the neuro- on September 25, 2021 by guest. Protected copyright. with cerebral palsy, however, is fairly pathology of the injury, the later clinical mani- unrewarding. Although previous work has festations, and the appearance of the brain on shown an increased risk of pre-eclampsia and neuroimaging may be useful in understanding intrauterine growth retardation among the time of injury and hence in identifying mothers of term infants with cerebral palsy those insults which are most likely to be compared with a control population,36 this risk preventable. does not appear to be greater in the group In conclusion, we consider that children without neonatal encephalopathy as we had with cerebral palsy and a history of neonatal hypothesised. Apart from a lower frequency of encephalopathy (depression at birth followed primigravidae, we were unable to find ante- by neonatal neurological dysfunction) are natal characteristics in the group without more likely than those without neonatal neonatal encephalopathy which differentiated encephalopathy to have: (a) evidence of these infants from the group with neonatal prolonged fetal distress as measured by the encephalopathy. It is likely that more sophisti- duration of cardiotocograph changes; (b) cated fetal neuroimaging techniques are evidence of other organ damage; and (c) a needed to detect signs of antenatal ischaemic severe and extensive form of motor deficit with damage, whether focal lesions or the more dif- intellectual and sensory involvement. fuse white matter lesions described with We suggest that this cluster of clinical signs intrauterine hypoxia.'0 indicates that the aetiology is likely to be an It was clear, however, that the clinical intra- acute intrapartum insult, preventable or non- partum course was different in the infants with preventable. It follows that in the absence of and without neonatal encephalopathy. There this cluster of signs, intrapartum asphyxia is F200 Gaffney, Flavell, _rohnson, Squier, Sellers

unlikely to be the cause of a subsequent neuro- 23 Evans P, Johnson A, Mutch L, Alberman E. A standard Arch Dis Child Fetal Neonatal Ed: first published as 10.1136/fn.70.3.F195 on 1 May 1994. Downloaded from form for recording clinical findings in children with a logical deficit. motor deficit of central origin. Dev Med Child Neruol 1989; 31: 119-27. We are grateful to Rosemary King, administrative coordinator 24 Macdonald D, Grant A, Sheridan-Pereira M, Boylan P, of the Oxford regional register of early childhood impairment; Chalmers I. The Dublin randomized controlled trial of to the regional obstetricians and midwives for their assistance; intrapartum fetal heart rate monitoring. Aml 7 Obstet to Georgina Berridge for computing support; and to colleagues Gynecol 1985; 152: 524-39. at the National Perinatal Epidemiology Unit for commenting 25 Bryce R, Stanley F, Blair E. The effects of intrapartum care on drafts of the paper. The Oxford regional register of early on the risk of impairments in childhood. In: Chalmers I, childhood impairment is funded by Oxford Regional Health Enkin M, Keirse MSNC, Eds. Effective care in pregnancy Authority. Geraldine Gaffney was funded by the Medical and childbirth. Oxford: Oxford University Press, 1989: Research Council. 1313-21. 26 Myers RE. Two patterns of perinatal brain damage and their conditions of occurrence. A]n Obstet Gynecol 1972; 1 Paneth N. The causes of cerebral palsy. Recent evidence. 116: 246-76. Clin Invest Med 1993; 16: 95-102. 27 Bejar P, Wozniak P, Allard M, et al. Antenatal origin of neu- 2 Paneth N, Fox HE. The relationship of Apgar score to rologic damage in newborn infants. Am.Y Obstet Gynecol neurological handicap: a survey of clinicians. Obstet 1988; 159: 357-63. Gynecol 1983; 61: 547-50. 28 Scher MS, Belfar H, Martin J, Painter MJ. Destructive 3 Stanley F, Blair E. Why have we failed to reduce the brain lesions of presumed fetal onset: antepartum causes frequency of cerebral palsy? Med3Aust 1991; 154: 623-6. of cerebral palsy. Pediatrics 1991; 88: 898-906. 4 Hill A. Current concepts of hypoxic-ischemic injury in the 29 Johnson MA, Pennock JM, Bydder GM, Dubowitz LMS, term newborn. Pediatr Neurol 1991; 7: 317-25. Thomas DJ, Young IR. Serial MR imaging in neonatal 5 Pasternak JF, Predey TA, Mikhael MA. Neonatal asphyxia: cerebral injury. AmjiYRadiol 1987; 8: 83-92. vulnerability of basal ganglia, thalamus, and brainstem. 30 Rorke LB, Zimmerman RA. Prematurity, postmaturity and Pediatr Neurol 1991; 7:147-9. destructive lesions in utero. Am]Y Neuroradiol 1992; 13: 6 Roland EH, Hill A. Selective brainstem injury in an 517-36. asphyxiated newborn. Ann Neurol 1988; 23: 89-92. 31 Nelson KB, Ellenberg JH. Antecedents of cerebral palsy. N 7 Leech RW, Alvord EC. Anoxic-ischaemic encephalopathy EnglY Med 1986; 8: 81-6. in the human neonatal period. The significance of brain 32 Blair E, Stanley FJ. Intrapartum asphyxia: a rare cause of stem involvement. Arch Neurol 1977; 34: 109-13. cerebral palsy. JPediatr 1988; 112: 515-9. 8 Dambska M, Laure-Kamionowska M, Liebhart M. 33 Nelson KB. Relationship of intrapartum and delivery room Brainstem lesions in the course of chronic fetal asphyxia. events to longterm neurologic outcome. Clin Perinatol ClinNeuropathol1987;6: 110-5. 1989; 16: 995-1007. 9 Clancy RR, SladkyJT, Rorke LB. Hypoxic-ischaemic injury 34 Levene MI. Role of excitatory amino acid antagonists in the following perinatal asphyxia. Ann Neurol 1989; 25: 185-9. management of birth asphyxia. Biol Neonate 1992; 62: 10 Squier MV, KeelingJW. The incidence of prenatal brain 248-51. injury. NeuropatholAppl Neurobiol 1991; 17: 29-38. 35 Dennis J, Chalmers 1. Very early neonatal seizure rate: a 11 Schuman HM, Selednik LJ. Periventricular leukomalacia. possible epidemiological indicator of the quality of peri- Arch Neurol 1980; 37: 231-5. natal care. Br] Obstet Gyniaecol 1982; 89: 418-26. 12 Hope PL, Gould SJ, Howard S, Hamilton PA, Costello 36 Gaffney G, SellersS, Flavell V, Squier M, Johnson A. Case- AMdelL, Reynolds EOR. Precision of ultrasound diagnosis control study of intrapartum care, cerebral palsy, and of pathologically verified lesions in the brains of very death. BM.7 1994; 308: 743-50. preterm infants. Dev Med Child Neurol 1988; 30: 457-71. 37 Minchom P, Niswander K, Chalmers I, et al. Antecedents 13 BarkovitchJ. MR and CT evaluation of profound neonatal and outcome of very early neonatal seizures in infants and infantile asphyxia. Am .7 Neuroradiol 1992; 13: born at or after term. Br 7 Obstet Gvnaecol 1987; 94: 959-72. 431-9. 14 Krageloh-MannI, Hagberg B, Petersen D, Riethmuller J, 38 Wyatt JS. Near infrared spectroscopy. Investigation and Gut E, Michaelis R. Bilateral spastic cerebral palsy- patho- assessment of perinatal brain injury. Biol Neonate 1992; genetic aspects from MRI. Neuropediatrics 1992; 23: 46-8. 62: 290-4. 15 Fitzhardinge PM, Flodmark 0, Fitz CR, Ashby S. The 39 Peebles DM, Edwards AD, Wyatt JS, et al. Changes in prognostic value of computed tomography as an adjunct human fetal cerebral haemoglobin concentration and to assessment of the term infant with postasphyxial oxygenation during labor measured by near-infrared encephalopathy. Pediatrics 1981; 99: 777-81. spectroscopy. AmY Obstet Gynecol 1992; 166: 1369-73. 16 Adsett DB, Fitz CR, Hill A. Hypoxic-ischaemic cerebral 40 Nijhuis JG, Prechtl HFR, Martin CB, Bots RSGM. Are injury in the term newborn: correlation of CT findings there behavioural states in the human fetus? Early Hum with neurological outcome. Dev Med Child Neurol 1985; Dev 1982; 6:177-95. 27: 155-60. 41 Manning FA, Baskett TF, Morrison I, Lange IR. Fetal 17 Lipp-Zwahlen AE, Deonna T, Micheli JL, Calame A, biophysical profile scoring: a prospective study in 1184 Chrzanowski R, Cetre E. Prognostic value of neonatal CT high-risk patients. Ami _7 Obstet Gynecol 1981; 140:

scans in asphyxiated term babies: low density score 289-94. http://fn.bmj.com/ compared with neonatal neurological signs. 42 Horimoto N, Koyanagi T, Macda H, et al. Can brain Neuropediatncs 1985; 16: 209-17. impairment be detected by in-utero behavioural patterns. 18 Gaffney G, Squier M, Johnson A, Flavell V, Sellers S. Arch Dis Child 1993; 69: 3-8. Clinical associations of prenatal ischaemic white matter 43 Perlman J, Tack E, Martin T, Shackleford G, Amon E. injury. Arch Dis Child 1994; 70: F101-6. Acute systemic organ injury in term infants after asphyxia. 19 Sarnat HB, Sarnat MS. Neonatal encephalopathy following Amn Dis Child 1989; 143: 617-20. fetal distress. Arch Neurol 1976; 33: 696-705. 44 Dauber I, Krauss A, Symchych P, Auld P. Renal failure 20 Levene MI, Kornber J, Williams THC. The incidence and following perinatal anoxia. 7 Pediatr 1976; 88: 851-5. severity of post-asphyxial encephalopathy in full-term 45 Ruth VJ. Prognostic value of creatine kinase-BB isoenzyme infants. Early Hum Dev 1985; 11: 21-6. in high risk newborn infants. Arch Dis Child 1989; 64: on September 25, 2021 by guest. Protected copyright. 21 Hall DMB. Birth asphyxia and cerebral palsy. BM7 1989; 563-8. 299: 279-82. 46 Stanley FJ, Blair E, Hockey A, Petterson B, Watson L. 22 Nelson K, Leviton A. How much of neonatal encephalo- Spastic quadriplegia in Western Australia: a genetic pathy is due to birth asphyxia? Am 7DisChild 1991; 145: epidemiological study. 1: Case population and perinatal 1325-31. risk factors. Dev Med Child Neuirol 1993; 35: 191-201.