Garcia v East Lancashire Hospitals NHS Trust

This is a claim for damages for injury and loss said to have been caused by the Defendant in the management of the Claimant’s mother’s pregnancy and labour and of the Claimant’s birth on 8 August 1998. By an open letter dated 21 December 2001, the Defendant admitted that it was in breach of duty in failing to deliver the Claimant some 26¼ hours before she was actually delivered. However, the Defendant denies that its breach of duty was causative of any damage. It contends that the fetus had already suffered the relevant injury before the breach of duty occurred, so that an earlier delivery would not have avoided the injury. In May 2005, an order was made for a split trial of the issues of liability (including causation) and quantum. This judgment is concerned with the issue of causation. I have to decide whether, if the Claimant had been delivered at the time she should have been, her injury would have been avoided.

It had been contended on behalf of the Claimant that the Defendant was in breach of duty by reason of Mr Mahady’s failure to act on the CTG trace beginning at 1330 hours (which the experts agreed was abnormal) and to deliver the Claimant by 1600 hours on 7 August at the latest. In the event, this allegation was not pursued.

At 1615 hours, Mrs Garcia’s blood pressure was raised at 160/95. By 2000 hours, it had dropped to 130/80 and she was feeling ‘much more relaxed’. At 2000 hours, urinalysis showed “++glucose no protein”, i.e. no sign of pre-eclampsia. At 1957 hours, a CTG was commenced. It was reviewed at 2030 hours by Dr Nasah, registrar, who noted at 2047 hours:

“Trace repeated; base line 170 bpm, base line variability about 5 bpm, no decelerations or accelerations seen. Trace faxed to IWM (Mr Mahady). Instructions re: action awaited.”

The notes record that, at 2100 hours, a telephone message was received from Mr Mahady, indicating that he was “happy with CTG”. No further action was to be taken that night but the CTG was to be repeated the following morning and he was to review it. A note of the nursing handover, timed at 2130 hours, records that pre-eclampsia bloods were normal and that fetal movements had been noted. The Defendant admitted that, as a result of Mr Mahady’s failure to act on the CTG traces of the evening of 7 August and to deliver the Claimant by 2130 hours, it was in breach of duty.

It is clear from the notes that the midwives attending Mrs Garcia were concerned at Mr Mahady’s response to the CTG. A note recorded at 1530 hours indicates that the Head of Midwifery Services, Miss Margaret Pickles, was aware of the situation and was to see the CTG. Later, the concerns were reported to Ms Connor, supervisor, by the senior midwife on duty, Sister Smith. At 0230 hours on 8 August, a note records a telephone call to the ward from Sister Smith. Ms Connor had instructed her that Mrs Garcia was to have a CTG when she wakened and the procedure was then to continue as normal, i.e. if the midwives were concerned, they should contact the registrar on call and the on call consultant if necessary. The midwives were instructed to keep Sister Smith informed of events.

At 0335 hours, Mrs Garcia was awake. Her blood pressure was 145/100 (i.e. slightly raised). She was asymptomatic. A CTG was commenced and Sister Smith was informed. At 0405 hours, the CTG was noted to be “unreactive” and Dr Nasah was informed. He attended and viewed the trace. He noted that the CTG was “same as before”. Mrs Garcia was reassured and was to await review by Mr Mahady later that morning provided that her blood pressure remained stable and she was asymptomatic. Sister Smith was informed of the position by the midwives caring for Mrs Garcia and she was to contact Ms Connor.

The CTG was recommenced at 0907 hours on 8 August. At 1000 hours, Mr Mahady reviewed Mrs Garcia and noted that she was now contracting every 4 to 5 minutes. The fetus was said to be “active” and the CTG “reactive”. A vaginal examination revealed that the cervix was partly effaced and the os was 2cms, with the vertex still 3cms above the ischial spines. Since he judged that Mrs Garcia was now in spontaneous labour, Mr Mahady directed that she should be transferred to the delivery suite. He said that he would re-assess her at 1400 hours. The CTG was discontinued at 1010 hours. Mrs Garcia was then transferred to the delivery suite where a partogram was set up at 1100 hours. At that time, a fetal heart rate of 160 bpm was noted, together with tightenings every 4 minutes. By midday, she was having mild short-lasting contractions every 5 or 10 minutes.

The next CTG was recorded between 1230 and 1300 hours. At 1345 hours, Mr Mahady attended. He noted that contractions were occurring and the CTG was “normal”. On vaginal examination, the cervix was partly effaced, the os was 2–3cms and the vertex 3cms above the ischial spines. Mr Mahady noted that he would reassess Mrs Garcia at 1600 hours.

In the event, it appears that Mr Mahady returned rather later than planned. The next CTG was commenced at 1630 hours and lasted for approximately 50 minutes. The midwifery notes record that, at 1630 hours, Mrs Garcia was having “weak, irregular contractions” and the baby was “active”. The CTG was noted as “having no variability 160 bpm. Small deceleration with contraction”. Mrs Garcia was awaiting review by Mr Mahady and her blood pressure was 140/90.

At 1830 hours, Mr Mahady attended again and saw the CTG. He performed a vaginal examination and artificially ruptured the membranes. Meconium liquor was present. A further CTG was commenced at 1835 hours and Mrs Garcia was prepared for an epidural. By 2040 hours, the epidural was in situ and the first dose had beengiven. A top up dose was given at 2055 hours. The CTG, which had been temporarily discontinued, was restarted at about the same time.

At 2100 hours, Mrs Garcia was seen by Mr Mahady. By this time, she was said to be contracting once every 4 minutes. Mr Mahady viewed the CTG and described it as “normal”. A vaginal examination showed that the cervix was fully effaced, the os was 2cms and the vertex 3cms above the ischial spines. Meconium stained liquor was noted. A fetal scalp electrode was applied and a Syntocinon infusion (to strengthen and increase the frequency of uterine contractions) was prescribed. The midwifery notes record that Syntocinon was commenced at 2130 hours and that “early decelerations” were evident on the CTG.

At about 2230 hours, the midwife on duty was concerned that there were “? late decelerations” on the CTG. She asked Dr Sharkar, the registrar on duty, to view the CTG trace. He noted that it showed reduced variability and late decelerations. He faxed the CTG trace to Mr Mahady. Meanwhile, he directed that the Syntocinon should not be increased. A midwifery note made at 2255 hours records that Mr Mahady was coming in to see Mrs Garcia. The Syntocinon was stopped. Mr Mahady attended and, since the labour was not progressing and he considered that the CTG was abnormal, he took the decision to deliver the baby by caesarean section. At 2325 hours, Mrs Garcia was transferred to theatre. An emergency caesarean section was performed and Amelia was born at 2345 hours.

The operation note records that the umbilical cord was around the neck once and around the body of the baby. The placenta and membranes were complete. Meconium liquor was present.

The paediatric notes record that the baby was cyanosed and floppy at birth, with no spontaneous breathing. She immediately underwent oropharyngeal suction by the midwives. Her heart rate was 80 bpm. Oxygen was given by bag for three minutes and by mask for a further three minutes and she was stimulated and warmed. By seven minutes, she was breathing independently without oxygen. Her Apgar scores were recorded as 4 at one minute, 7 at five minutes and 9 at ten minutes. The paediatrician recorded that she was to be reviewed at 20 minutes of age and observed for respiratory distress. She was to be fed early and her blood sugars tested. On review at 20 minutes, Amelia was noted as being “alert, pink” with no evidence of respiratory distress. She was transferred with her mother to the post natal ward at 0030 hours on 9 August. I shall deal with her progress thereafter at a later stage in this judgment.

The Claimant’s case on causation is that, towards the end of Mrs Garcia’s pregnancy, placental insufficiency developed. This had the effect of impairing the supplies of nutrition (in particular glucose) and oxygen reaching the fetus. The placental insufficiency had the effect of causing the Claimant to become hypoxic, producing abnormalities in the fetal heart which were evident on the CTG from 1330 hours on 7 August. It is not known when these abnormalities in the fetal heart rate first developed but it must have been some time before they were seen on the CTG. The various abnormalities seen on the traces during the latter half of 7 August and throughout 8 August were, it is said, the effects of a fluctuating level of chronic hypoxia. It is contended that, at some time during the 24 hours immediately preceding delivery (and therefore at a time when, but for the Defendant’s breach of duty, the Claimant would have been safely delivered), the Claimant suffered a right middle cerebral artery infarction (stroke) resulting in severe disability.

The Defendant agrees that the Claimant suffered a stroke. However, its case is that the stroke occurred earlier than the time contended for by the Claimant, namely at some time between the end of the CTG on 4 August and before the first abnormal CTG on 7 August. It is contended that the abnormalities of the fetal heart rate which were evident on the CTG traces of 7 and 8 August (and all the Claimant’s disabilities) were attributable to the effects of the stroke and not to chronic hypoxia. The Claimant’s condition immediately after birth and in the ensuing hours and days, together with the damage to the brain evident on scanning are, it is said, entirely consistent with this history.

It is of course for the Claimant to prove, on a balance of probabilities, that the Defendant’s admitted breach of duty caused her injury.

In order to reach conclusions about the competing theses advanced by the parties, it is necessary separately to consider the evidence relating to

The evidence relating to the CTG traces came from two consultant obstetricians and gynaecologists, Mr John Spencer (for the Claimant) and Mr Donald Peebles (for the Defendant).

Mr Spencer is a renowned expert on fetal physiology and fetal monitoring. Between 1986 and 1994, he was a Senior Clinical Lecturer at the University of London. At the same time, he worked as an Honorary Consultant Obstetrician and Gynaecologist at two leading London hospitals. During that period, he conducted important research into the effects of hypoxia on the fetal heart rate and the use of fetal monitoring in the management of labour. He published on the topic and also edited, co-edited and contributed to leading textbooks on fetal monitoring. Between 1994 and 2006, he was a Consultant Obstetrician and Gynaecologist at the Northwick Park Hospital, Harrow, and Honorary Senior Clinical Lecturer at the Imperial College School of Medicine. He is at present working as a Locum Consultant in obstetrics at the Royal Victoria Infirmary, Newcastle-upon-Tyne.

Mr Peebles has been a Senior Lecturer and Honorary Consultant in the Department of Obstetrics and Gynaecology, University College London (UCL) since 1999. The main focus of his research has been the investigation of the causes of perinatal brain injury. He was co-founder in 2001, and is now Head, of the Perinatal Brain Research Centre at UCL. He runs an antenatal clinic for medical disorders in pregnancy, which is a tertiary referral unit for women with high risk pregnancies, and is Lead Consultant for a day assessment unit. He has published widely on relevant topics, such as fetal cerebral oxygenation and antepartum hypoxia. He was an impressive witness.

Both experts have provided reports, held a joint meeting and produced a joint statement resulting therefrom and both gave oral evidence.

It is the Claimant’s case that the chronic partial hypoxia which, it is said, developed and caused the abnormalities on the CTG traces, was caused by placental insufficiency.

In order for a fetus to grow, it requires oxygen and other nutrients (including glucose). The adequacy of the utero-placental circulation is of major importance to the supply of nutrition to the fetus. Placental insufficiency occurs when the capacity of the placenta to meet fetal demand for nutrition becomes inadequate, as a consequence of which the fetus receives insufficient supplies of food and oxygen. Placental insufficiency is usually intrinsic or chronic in that it is present from the beginning of pregnancy, although it can occur acutely, e.g. as a result of a placental abruption. In the case of chronic placental insufficiency, the fetus will tolerate the gradual reduction in the supply of oxygen and nutrients for some time by adjusting its metabolic demands and oxygen requirements. This may lead to fetal growth retardation/restriction. The fetus will redirect nutrition to the brain and heart in preference to other, less important, organs. However, it will not be able indefinitely to tolerate the insufficiency and will eventually become hypoxic. It may also become acidotic. When the fetus reaches this point, characteristic changes occur in the fetal heart rate. Chronic partial hypoxia – if permitted to continue – can lead to hypoxic-ischaemia, with consequent damage to the fetus and, ultimately, to death.

Mr Spencer identified two factors which he said were risk factors for placental insufficiency in Mrs Garcia’s case. First, she had developed hypertension (which he described as ‘mild’ in his report and ‘mild to moderate’ in his oral evidence) towards the end of pregnancy. The proper functioning of the placenta depends on the maternal blood supply. Hypertension causes restriction of that blood supply and consequent reduction in the amount of oxygen delivered to the placenta. Mrs Garcia’s hypertension was not of such severity as to require treatment. No protein was detected in her blood. It was clear therefore that she was not suffering from pre-eclampsia, which would have greatly increased the risk of placental insufficiency. Mr Peebles did not agree that Mrs Garcia’s hypertension constituted a risk factor for placental insufficiency. He said that hypertension of that level was relatively common and was to be distinguished from hypertension with proteinuria. Non-proteinuric pregnancy-induced hypertension is not, he said, associated with a worse maternal or perinatal outcome.

The second feature identified by Mr Spencer was the fact that the pregnancy was approaching ‘post-term’ (i.e. term + two (42) weeks). He pointed out that, because of the intrinsic risk of placental insufficiency in a pregnancy which proceeds beyond 42 weeks, clinical guidelines recommend that induction of labour is offered during the 41^st week. Mr Peebles agreed that there is an increased perinatal mortality rate after the 42^nd week, but not prior to that time. In cross-examination, Mr Spencer agreed that the two risk factors were relatively small.

I accept that there was a modest increase in the risk of placental insufficiency on account of the fact that the pregnancy was approaching its 42^nd week. The rise in Mrs Garcia’s blood pressure was also modest and I accept Mr Peebles’ evidence that it gave rise to no significant increase in risk. Overall, the risk factors for placental insufficiency were, as Mr Spencer himself acknowledged, relatively small.

Later in this judgment, I shall consider whether any support for the presence of placental insufficiency can be derived from the information available relating to the Claimant’s size and weight after birth and to her blood sugar levels (indicating the possible presence of hypoglycaemia) in the neonatal period. Suffice it to say here that, for reasons that I shall set out, I have concluded that there is no convincing evidence from either of these sources which would, of itself, or combined with the modest increase in risk which I have found existed, establish the presence of placental insufficiency. It is necessary, therefore, to consider the evidence shown on the CTG traces and to assess whether that evidence supports the contention that there was placental insufficiency such as to cause hypoxia.

I shall now turn to the evidence relating to the CTG traces.

A CTG is used to monitor both fetal heart rate and uterine activity. During the 1970s and 1980s, with the advent of CTG monitoring, obstetricians tackled the questions of how to detect fetal hypoxia (in particular, hypoxia resulting from placental insufficiency) and whether it was possible to identify the presence of fetal hypoxia by CTG in time to intervene and avoid damage to the fetus. Mr Spencer was responsible for much of the work done at that time and, in evidence, he produced a considerable amount of literature dating from that period.

The purpose of CTG monitoring is to assess fetal wellbeing in pregnancy and labour. A CTG can show abnormalities in the fetal heart rate pattern but cannot provide a diagnosis of why those abnormalities are occurring. The CTG trace must always be viewed in the light of the other available clinical evidence. Mr Spencer referred to this point in a Chapter entitled, ‘Assessment of fetal wellbeing in late pregnancy’, which he contributed to a leading textbook (Footnote: ¹):

“Appropriate management of a pregnancy with reduced fetal movements remains controversial … but it is widespread practice to perform cardiotocography (CTG). This begs the question of whether the CTG can identify uteroplacental insufficiency. Current experience suggests that, in the absence of ultrasound data to confirm growth retardation, the CTG is not sufficiently sensitive to recognise chronic placental insufficiency. Conversely, with an established diagnosis of fetal growth retardation, the CTG becomes an essential component of management.”

A normal CTG trace should show a fetal heart rate of 120-160 bpm, with sporadic and periodic accelerations in the heart rate associated with fetal movements. The fetal heart rate will be observed to vary around the baseline by between about 5 and 25 bpm. Periods of reduced variability with few accelerations occur normally when the baby is resting between periods of movement. Therefore, it cannot be assumed automatically that reduced variability constitutes an abnormality. However, a prolonged period (conventionally accepted as about an hour) of reduced or completely absent variability is considered to be an abnormal feature. A CTG trace which shows the presence of accelerations, together with a baseline heart rate and variability which are within normal limits, is termed ‘reactive’.

Decelerations in the fetal heart rate can sometimes be seen even on a normal CTG trace. They frequently occur in response to uterine activity, i.e. the so-called Braxton Hicks tightenings (which are gentler than the contractions of labour, are in general pain-free and occur intermittently, then with increasing regularity, before labour becomes established) or uterine contractions (which are painful and occur during the course of established labour). Each tightening/contraction causes some interruption of the maternal blood flow (and therefore the oxygen supply) through the uterus to the placenta. Decelerations which occur early in the tightening/contraction are usually considered to be benign. When decelerations are ‘late’ (i.e. the maximum fall in the fetal heart rate occurs late in, or after, the tightening/contraction), they indicate that the fetus is responding adversely to the transient fall in oxygen delivery and constitute an abnormal feature, suggestive of hypoxia.

The characteristic changes in the fetal heart rate associated with chronic partial hypoxia are a loss of accelerations accompanied by a reduction in fetal movements, a raised heart rate (tachycardia), a reduction in baseline variability to below 5 bpm and the onset of recurrent late decelerations. When these changes are present, delivery by caesarean section is indicated if the fetal is viable. The accepted sequence of changes consequent on hypoxia is that the fetus first responds to deoxygenation by decelerations in the heart rate. It then secretes stress hormones (known as catecholamines) which cause the fetal heart rate to increase. The stress response then leads to a reduction in variability of the heart rate.

I shall now briefly summarise the evidence relating to the individual CTG traces. The obstetricians’ assessments of the traces were not, of course, carried out in the same conditions that would apply during the management of labour, when each trace would have to be considered prospectively, without the benefit of knowledge about what was to happen subsequently. Mr Spencer and Mr Peebles were making their assessments retrospectively, with full knowledge of the results of all the CTG traces and of the Claimant’s condition after delivery and thereafter.

Mr Spencer and Mr Peebles agreed that this trace was within normal limits. The baseline was 150 bpm, baseline variability was normal and accelerations were present. Mr Spencer suggested that a heart rate of 150 bpm was rather higher than would be expected for a fetus at this late stage of pregnancy; he said that the average rate would have been about 120 bpm. Mr Peebles disagreed. He said that a heart rate of 150 bpm fell within every definition of a normal baseline; 120 bpm defined the lower limit of normal. He pointed out that there was no means of knowing what the ‘normal’ fetal heart rate baseline was for this baby.

The obstetricians agreed that there were significant differences between this trace and the previous trace. The baseline heart rate had risen to just above 160 bpm, i.e. just above the upper limit of the normal range. The most striking feature was the reduction in the variability in the heart rate, which was less than 5 bpm throughout the trace. There were no accelerations. Mr Spencer identified one late deceleration. Mr Peebles observed that the quality of the trace was poor but was prepared to accept, in the light of the fact that later traces showed decelerations, that there was a deceleration here. The tocograph showed no clear evidence of uterine activity save for that immediately preceding the late deceleration. In his report, Mr Spencer had contended that this trace should have led to delivery of the Claimant at no later than 1600 hours. As I have already said, this allegation was not pursued at trial.

and iv. CTG Trace of 7 August 1998 running from 1957 until 2036 hours (first part) and 2037 until 21.05 hours (second part)

This trace showed irregular uterine activity. Mr Spencer and Mr Peebles agreed that there had been an increase in the baseline heart rate of about 5 bpm to between 165 and 170 bpm. Variability had decreased further. Mr Spencer identified two clear shallow late decelerations which he said were associated with rises in the tocograph signal, probably caused by the occurrence of Braxton-Hicks tightenings. He said that the trace showed a typical pattern of chronic hypoxia caused by placental insufficiency. Mr Peebles agreed that there was one shallow late deceleration; he was uncertain about the other. Mr Spencer and Mr Peebles agreed that the heart rate variability was fractionally more evident in the second part of the trace than in the first part, although it was still minimal. Mr Peebles accepted that the abnormalities evident on this trace, taken in conjunction with earlier traces, should have led to the Claimant’s delivery by 2130 hours. He said that the indication for delivery would have been concern about the fetal condition. Chronic hypoxia would have been one possible explanation for the changes on the trace but there would have been others.

CTG Trace of 8 August running from 0335 until 0410 hours

Mr Spencer said that variability had reduced further. His view was that the tocograph showed much more frequent uterine activity, with tightenings or contractions every 3-4 minutes or so. He said that, in response, the fetal heart rate was showing more frequent late shallow decelerations. On occasion, the recovery of the heart rate was slow, in that it did not get back to a baseline above 160 bpm very often. He said that it was difficult to be clear about the baseline because of the frequency of the oscillations of the heart rate.

Mr Peebles’ opinion was that this trace showed essentially the same features as the previous trace although there was a slight decrease in the baseline fetal heart rate. There was still reduced variability. He observed that the tocograph was of poor quality, making it difficult to say whether contractions were occurring or not, although they may have been. His view was that there was only one occasion when there was clear evidence of a contraction. He said that the problems with interpreting the tocograph made it difficult to say whether or not the fluctuations in the baseline heart rate indicated late decelerations.

CTG Trace of 8 August running from 0907 until about 1010 hours

It was agreed that the trace showed uterine activity about every 3-4 minutes. The heart rate baseline was between 155 and 160 bpm. This was a slight reduction from the previous evening. Mr Spencer drew attention to the fact that, at a number of points on the baseline, the heart rate oscillated. Some of the uterine tightenings were followed by what he called ‘exaggerated oscillations’ which lasted for a short time. The oscillations were smooth with no short term variability. Mr Spencer said that these oscillations seemed to have replaced the late shallow decelerations that had been evident previously. They suggested that the fetus was ‘tolerating the situation’. Mr Peebles said that the timing of the oscillations (i.e. preceding or simultaneous with the contraction) was completely different from that of the late decelerations seen previously. He suggested that they resulted from a different mechanism. As I shall explain, he regarded the absence of late decelerations on this trace and on the next as very significant.

CTG Trace of 8 August running from 1230 until 1300 hours

Mr Peebles said that the uterine activity which was clearly present did not appear to be associated with any late decelerations. The baseline fetal heart rate was about 160 bpm. Heart rate variability was still reduced. This accorded with the description given by Mr Spencer in his report.

CTG Trace of 8 August running from 1630 until about 1720 hours

Mr Peebles said that, although the quality of the tocograph was poor, there was evidence of maternal contractions. The heart rate baseline was still less than 160 bpm and there was reduced variability. There was evidence of at least two late decelerations in the heart rate, apparently in response to uterine contractions. Mr Spencer observed that the trace showed minimal variability with shallow late decelerations after some uterine tightenings.

CTG Trace of 8 August running from 1835 until about 2230 hours

Mr Peebles said that the first part of the trace was similar to the earlier traces of 8 August. The heart rate baseline was 155-160 bpm and variability was reduced. However, he said that, despite regular and clearly indicated uterine contractions occurring on average every four minutes, he could see no convincing evidence of late decelerations. Mr Spencer agreed that there were no late decelerations in the early part of the trace.

Mr Spencer and Mr Peebles also agreed that, from about 2110 hours, there was a clear change in the appearance of the CTG trace. The baseline fetal heart rate remained at around 160 bpm with virtually no variability. However, there was now very clear evidence of shallow late decelerations with each contraction. As the trace progressed, uterine contractions were occurring approximately every 2-2 ½ minutes. The change in the appearance of the trace caused Mr Mahady to take the decision to deliver the Claimant.

Mr Spencer attributed the change in the appearance of the trace to the administration of an epidural at about 2000 hours and to the start of Syntocinon at 2130 hours. He explained that the insertion of an epidural can have an effect on fetal oxygenation. This effect is not direct. The mechanism by which it can occur is to cause a reduction in maternal blood pressure which can in turn cause an unpredictable and unquantifiable reduction in the pressure of the mother’s blood flow through the uterus to the placenta, i.e. a reduction in utero-placental perfusion. It is recognised that this can cause late decelerations and, for that reason, continuous CTG monitoring is mandatory when an epidural is in place. Mr Spencer acknowledged that the risk of a reduction in perfusion occurring as a result of an epidural was small.

Mr Peebles did not agree with Mr Spencer that the changes in the CTG trace were caused by maternal hypotension resulting from the epidural. He gave three reasons for his view. First, he said that changes in uterine perfusion related to the insertion of an epidural tend to be relatively short-lived whereas, in this case, the change in the appearance of the CTG trace was sustained for the rest of labour, a period of about two hours. Second, he doubted whether the available evidence supported the occurrence of a significant change in maternal blood pressure following the insertion of the epidural. Although there was a reduction in the diastolic component of the blood pressure readings from 90 (before the test dose was given) to 70 (after the administration of the first ‘proper’ dose), it was not in his view sufficient to cause a significant change in perfusion pressure. He said that both the diastolic and the systolic components of the blood pressure readings were relevant for these purposes, and, taken together, they were within normal limits. Third, he pointed out that the alteration in fetal heart rate pattern occurred a little more than an hour after the epidural had been commenced. He said that, usually, if there was a hypotensive response to an epidural, it would happen within an hour.

In their joint statement, the obstetricians agreed that it was not possible to determine whether or not there was a significant reduction in utero-placental perfusion in this case. They also agreed that it was not possible to identify the effect, if any, of the epidural on the fetal heart rate.

Mr Spencer said that Syntocinon acted as a stimulant to the uterus and its primary effect is to strengthen, and increase the frequency of, uterine contractions. He said that the effect of the administration of Syntocinon was evident on the CTG trace. He suggested that it probably caused a reduction in utero-placental perfusion which, when added to the existing placental insufficiency, resulted in late decelerations. Mr Peebles said that a significant reduction in utero-placental perfusion as a result of Syntocinon occurs only when the uterus becomes hyper-stimulated, causing contractions to be too frequent and too strong. Both he and Mr Spencer agreed that there was no evidence of hyper-stimulation here. However, Mr Peebles accepted that the uterine contractions did become more frequent following the administration of Syntocinon, and were accompanied by correspondingly frequent late decelerations.

CTG Trace of 8 August 1998 running from 2240 until 2317 hours

The final trace showed a continuation of the same pattern as previously. The fetal heart rate remained at around 160 bpm and there were late decelerations and reduced variability.

Although there were some differences in the obstetricians’ analyses of features shown on the CTG traces, these were relatively minor. However, their interpretation of the traces was markedly different.

They agreed that the first trace of 7 August was wholly different in appearance from that of 4 August. The most striking change was the reduction in variability of the fetal heart rate. In addition, the heart rate had increased and there was one shallow late deceleration, apparently in response to uterine activity. It is clear that something (whether the effects of hypoxia or of stroke or of some other mechanism) had occurred to produce these changes. Mr Spencer suggested that there must have been a gradual progressive change in the fetal heart rate from 4 August.

That picture of reduced variability, tachycardia and occasional shallow late decelerations remained essentially unchanged until the morning of 8 August. Then, despite the fact that the CTG showed clear evidence of uterine activity every 3-4 minutes, there were no late decelerations. Instead, there were the “exaggerated oscillations” described by Mr Spencer. There were no late decelerations in the trace taken at lunchtime. Occasional late decelerations were evident in the trace of the late afternoon. No decelerations were evident during the first 2½ hours of trace starting at 1835 hours on 8 August. Both obstetricians agreed that recurrent late decelerations were evident from 2110 hours onwards as uterine activity became more frequent.

Mr Spencer’s view was that the picture on 7 and 8 August was that of a baby suffering from a degree of sustained hypoxia caused by placental insufficiency. He suggested that the onset of Braxton-Hicks tightenings had caused the placental insufficiency – which had been present previously but undetected – to become evident. There had been a transition to a situation where fetal reserve could no longer fully compensate for the slow fall in placental function, together with reduced utero-placental perfusion. The presence of tachycardia, reduced variability and late decelerations was typical of hypoxia. He said that his impression was that, as at the morning of 8 August, the fetus seemed to be ‘tolerating the situation’. However, the further reduction in utero-placental perfusion produced by the epidural and the Syntocinon had caused the CTG appearances to revert to the pattern of the previous evening. Mr Spencer said that, even then, the CTG trace suggested that the fetus was ‘managing well’. His view was that the whole picture was consistent with hypoxia caused by placental insufficiency.

Mr Peebles said that the first CTG trace of 7 August was clearly abnormal and indicated that there had been a deterioration in fetal condition since 4 August. It was not possible to make a diagnosis of the cause of that deterioration from the CTG alone. He suggested that the picture as it developed was not consistent with chronic partial hypoxia. In the presence of uterine contractions, he would have expected a hypoxic fetus to deteriorate. Not only did this not happen, but there appeared to be some recovery. By the morning of 8 August, when uterine activity had increased and almost 21 hours had elapsed since the first abnormal trace, the CTG trace showed no late decelerations and this remained the position (apart from a few decelerations in the afternoon) until 2110 hours that day. He said that was not compatible with the worsening fetal condition which would be expected if the problems were caused by chronic hypoxia. He pointed out that two important features had remained largely unchanged from the first CTG on 7 August, i.e. the heart rate and (more particularly) the reduced variability. The major change in the evening of 8 August was in the frequency of late decelerations.

Mr Peebles said that he found the suggestion that the fetus had been suffering from chronic hypoxia caused by placental insufficiency for a period of at least 34 hours before delivery (i.e. from 1330 hours on 7 August) implausible. He described the sequence of events that occurs in a growth-restricted fetus that becomes chronically hypoxic. He explained how the fetus makes adaptive physiological compensations for the chronic hypoxia in order to preserve function of the two components involved in the fetal heart rate tracing, i.e. the brain and the heart. In that way, the chronically hypoxic fetus maintains oxygen delivery to the brain and heart until it is in a pre-terminal state. Changes in the fetal heart rate occur late in the progression of growth restriction. Mr Peebles said that, if the abnormal heart rate features evident on the CTG traces had been related to chronic hypoxia and fetus growth restriction resulting from placental insufficiency, he would have expected the fetus to be severely compromised. In that situation, any hypoxic stress (such as uterine contractions) would be expected to lead to a rapid deterioration in the fetal condition. He said that this was recognised in some of the earlier publications about abnormal fetal heart rate tracing in growth-restricted fetuses. They showed that those fetuses which had the combination of abnormal features shown on these CTG traces as a result of hypoxia either died in utero or had to be delivered by elective caesarean section before they were subjected to the hypoxic stress of labour. Mr Peebles said that he did not think that a CTG whose main features did not change significantly over a period of more than 24 hours, against a background of clear evidence of increasing uterine activity, was compatible with this being a chronically hypoxic growth-restricted infant.

I note that this view derives some support from Mr Spencer’s report, in which he said:

“It is possible that delivery on 07.08.98 would have resulted in avoidance of fetal damage. It is recognised that delivery by caesarean section at an early stage of chronic fetal hypoxia secondary to placental insufficiency has a good outcome. Thus delivery after the CTG on 07.08.98 (i.e. at 1600 hours, more than 31 hours before the Claimant was actually delivered) would be expected to result in avoidance of fetal cerebral damage associated with chronic hypoxia.” [my emphasis]

Mr Spencer’s explanation for the fact that the Claimant’s condition did not deteriorate during the day of 8 August, as would be expected, was that the fetal reserve, although markedly reduced, must have been sufficient to protect the fetal brain from direct hypoxic damage. He agreed that the improvement in the appearance of the traces on 8 August was difficult to understand. He said that the improvement indicated that the fetus was getting used to and tolerating the situation. It was continuing to draw on its reserves and, since there was no increase in hypoxia, it was able to do that.

Mr Peebles suggested that by far the more likely explanation was that the CTG changes were caused by the fact that the fetus had already suffered a major stroke. He believed that a major stroke such as that which the Claimant is now known to have suffered could have caused sufficiently abnormal cerebral function as to produce abnormalities on CTG. That would, he said, account for the relatively unvarying nature of the CTG abnormalities over a period of 34 hours.

Mr Peebles explained that there was evidence – although not from fetal literature – to suggest that the location of the damage to the brain caused by the stroke was important in this context. The results of studies of stroke in adults (where the stroke can be precisely timed) have suggested that stroke is associated with a higher incidence of increased heart rate and cardiac arrhythmia. Similar results have been obtained in studies using rats. He referred to one paper (Footnote: ²) which suggests that adults who have had a right-sided stroke affecting the cortex of the brain (which receives its blood supply from the middle cerebral artery) have reduced heart rate variability compared with those who suffer strokes on the left side or in another part of the right side of the brain. He suggested that this was relevant since the Claimant’s stroke had occurred on the right side, involved the territory supplied by the middle cerebral artery and caused widespread cortical damage confirmed by neuroradiological evidence. Mr Peebles explained that the control mechanisms determining heart rate variability are extremely complex and are not yet completely determined. However, there is clear evidence that connections from the higher part of the brain can exert an influence on the variability of the heart rate.

There is a considerable body of evidence dealing with the CTG appearances indicative of the presence of fetal hypoxia. However, the standard textbooks and authorities do not refer to any characteristic heart rate appearances that are known to be the result of cerebral damage. Nevertheless, Mr Peebles referred to some literature which, he argued, supported his contention that the abnormal CTG appearances were caused by the stroke which the Claimant undoubtedly suffered.

Mr Peebles relied on a paper (Footnote: ³) by Sreenan et al. They carried out a study of 46 children born between 1983 and 1997 who had been diagnosed by CT scans as having single artery distribution cerebral infarctions. Of the 46 children, 27 had had abnormal fetal heart rate patterns on perinatal monitoring. The relevant passage states:

“Adverse perinatal events were present in almost three quarters of the infants. In 24 cases there was meconium-stained amniotic fluid. The abnormal fetal heart rate patterns in 27 included marked tachycardia or bradycardia in 2, moderate tachycardia or bradycardia with reduced variability in 3, absent variability in 3, variable decelerations in 7, and late decelerations in 12. Over half of the infants required operative delivery because of fetal distress or failure to progress……”

Mr Peebles acknowledged that the study did not show that stroke had caused the CTG changes. It merely demonstrated an association. However, the study did show that in excess of 50% of the children studied had abnormal CTGs. He pointed out that many large case-controlled studies of children with cerebral palsy had shown signs of CTG abnormalities in less than 20% of cases, so that 50% was quite a large proportion.

Mr Spencer said that he inferred from the Sreenan paper that the fetuses affected by stroke were more vulnerable to chronic hypoxia and that what was shown was not cause and effect. He also pointed out that the Sreenan paper did not make clear whether the strokes had occurred perinatally or after delivery. He inferred the latter. Mr Peebles did not agree with that interpretation.

Mr Peebles also referred to a retrospective study (Footnote: ⁴) by Shifrin et al of 44 children with cerebral palsy who had CTG trace patterns during labour which suggested previous injury. The study did not identify the precise nature of the conditions from which the children were suffering. It described a ‘fixed’ non-reactive heart pattern or ‘pattern of autonomic imbalance’ which occurred in ‘an already injured but not necessarily currently asphyxiated fetus’. This pattern contrasted with the normal appearances on CTG monitoring carried out only a short time earlier.

A similar pattern was found by Phelan et al (Footnote: ⁵) who identified it in 45% of 300 infants suffering from a perinatal brain injury. They concluded that a persistent non-reactive heart rate pattern resulted from an existing brain injury and was not a sign of ongoing brain injury. Mr Peebles suggested that the patterns observed in these studies were similar to that shown on the Claimant’s CTG traces.

In the obstetricians’ joint statement, Mr Spencer agreed that stroke can cause fetal tachycardia and reduced variability. In evidence, he said that he was not aware of any changes in heart rate pattern that were specifically referable to stroke. However, he said that he would not be surprised if a direct link between stroke and certain heart rate changes were to be shown in the future. What size of stroke would be capable of causing such changes had yet to be determined. He suspected that it would have to involve a fairly major loss of brain tissue. He did not accept that stroke could have produced the CTG abnormalities evident on the Claimant’s traces.

Mr Spencer accepted that there had for 20 years been papers in the literature containing anecdotal case reports of babies with cerebral damage who had exhibited a characteristic ‘fixed’ fetal heart rate pattern. The fixed pattern is caused by a gross disturbance of the fetal heart rate, i.e. by intrinsic problems within the fetus, as opposed to by the fetus responding to external factors such as hypoxia. He acknowledged that there was no generally accepted definition of a ‘fixed’ pattern. He described it as a flat, steady heart rate with no variations. He said that it could be distinguished from the pattern characteristic of hypoxia. Hypoxia produces minimal or no variability along a baseline. However, the baseline itself is not fixed; it can undulate and there can be accelerations. He said that he would not describe the first CTG trace of 7 August as showing the ‘fixed’ heart rate pattern described in the literature. Moreover, he said that he would not expect either the baseline heart rate or the variability to fluctuate if there had been an existing cerebral injury causing the CTG abnormalities.

Mr Peebles acknowledged that there were some fluctuations in fetal heart rate and in variability over the traces of 7 and 8 August although he regarded these as slight. He said that it was difficult to be dogmatic about the precise nature of the changes that might be produced by a stroke since these are not well understood. He suggested that it was possible that the stroke was evolving during the relevant period and that this may have accounted for the slight changes in variability and baseline heart rate which were evident.

On behalf of the Claimant, it was contended that the late decelerations seen on some of the CTG traces on 7 and 8 August (although absent from others) were incompatible with the suggestion that the CTG abnormalities were caused by a pre-existing stroke. Mr Peebles did not accept that contention. He suggested that the late decelerations evident on 7 August represented a response to the Braxton-Hicks tightenings. He said that the fetal response to a tightening was to lower the heart rate during the time it took for a wave of deoxygenated blood to reach the heart. He suggested that a stroke affecting the cortex of the brain might have affected the mechanisms responsible for controlling heart rate and heart rate variability. This could in turn have had the effect of producing an exaggerated response to a stimulus that might otherwise have caused little or no response. There was, he said, some suggestion from animal studies that such an exaggerated response could occur in certain circumstances. If the fetus had already been damaged by stroke, it might still respond to contractions, but in an exaggerated fashion, thus producing a late deceleration. The reappearance of late decelerations after 2110 hours on 8 August might just have reflected the more intense nature of the stimulus, i.e. the more frequent uterine contractions.

I accept that - viewed prospectively - it would have been entirely reasonable and proper for a treating obstetrician to form the view that the CTG traces showed evidence of chronic hypoxia and, for that reason, to take action to deliver the baby. Indeed, the Defendant has accepted that, by 2130 hours on 7 August, it amounted to a breach of duty not to take such action. The treating obstetrician might well have suspected that there was hypoxia caused by placental insufficiency in the light of the fact that the pregnancy had gone beyond term and there was therefore a slightly increased risk of the development of placental insufficiency. He or she might, as Dr Peebles suggested, have been uncertain of the diagnosis, but should nevertheless have decided that the appearances were such as to warrant delivery.

We now have considerably more evidence against which to view the CTG traces. There are in my view two factors that militate against the abnormalities in the fetal heart rate having been caused by chronic hypoxia. First, there appeared to be a recovery on 8 August in that, despite the increase in uterine activity, late decelerations were not evident on the CTG traces (save for a few in the afternoon) until after the marked increase in the frequency of contractions at 2110 hours. Increased uterine activity would have been expected to produce an increase in hypoxia and thus an increase in late decelerations. I have difficulty in accepting Mr Spencer’s explanation that the apparent ‘improvement’ resulted from the fetus drawing on its reserves. I could understand this happening in the absence of an increase in uterine activity, but not when uterine activity was becoming more frequent. I accept Mr Peebles’ evidence that, if the Claimant had been hypoxic, he would have expected her condition to have deteriorated rapidly as a result of additional hypoxic stress in the form of more frequent uterine contractions.

Second, if Mr Spencer is right, the fetus continued in a state of hypoxia sufficient to cause changes in the fetal heart rate for a period of 34 hours from 1330 hours on 7 August until delivery (in addition to an unknown period before that) without apparently suffering any marked deterioration evident from the fetal heart rate, at least until 2110 hours on 8 August, when the deterioration was plainly associated with the increased frequency of contractions. For the reasons advanced by Mr Peebles, I accept that this is a highly unlikely scenario.

If Mr Spencer is correct in saying that the fetal heart rate on 4 August was abnormally high for such a late stage of pregnancy (thus implying that, even then, the fetus was responding to the effects of hypoxia), the problems become even greater. An increase in heart rate unaccompanied by decelerations would not be consistent with the accepted sequence of events signalling the onset of the effects of hypoxia that I have described previously. Furthermore, the fact that the fetal heart rate was already affected by hypoxia on 4 August would render it even more unlikely that the fetus could have withstood ongoing hypoxia without any deterioration in the fetal heart rate until 2110 hours on 8 August.

The alternative explanation is that the CTG changes were caused by the stroke which the Claimant undoubtedly suffered at some point. The effects of stroke on fetal heart rate (if any) are much less well-researched and well-defined than are the effects of hypoxia on fetal heart rate. What literature there is tends to relate to cerebral damage generally, rather than specifically to stroke. It is only recently, with the advent of MR scanning, that the true incidence of perinatal stroke has been recognised. It is clear that, as yet, no CTG appearances have been identified that are characteristic of stroke. This may be because the effects of stroke in an individual case would depend very much on the precise area and extent of the brain affected by the stroke.

The Defendant’s thesis involves putting together various pieces of research evidence and drawing inferences from them. I accept that the thesis is plausible. I accept also the Claimant’s contention that the pattern of small fluctuations in the baseline heart rate and variability is not wholly consistent with the ‘fixed’ heart rate pattern described in the literature as resulting from an existing cerebral injury. It may be that Mr Peebles’ explanation that the fluctuations resulted from the fact that the stroke was evolving is correct. This would account for those features of the CTG traces that are otherwise difficult to understand. However, the explanation is necessarily somewhat speculative.

It is clear that the CTG trace appearances are in some respects inconsistent with the usual pattern caused by chronic partial hypoxia. That being so, I would be unable, without more, to conclude that hypoxia was the cause of these abnormal appearances. On the other hand, the evidence in support of the abnormalities having been caused by stroke is not sufficiently strong for me to be confident that this was the case. It is necessary, therefore, to look at the other available clinical evidence in order to ascertain whether or not it supports the Claimant’s contention that the abnormalities on CTG were caused by chronic hypoxia.

I move on now to consider the evidence relating to the Claimant’s condition at birth and in the neonatal period and the disability and damage from which she has suffered thereafter, and to consider what light that evidence sheds on the causation of her injuries. The evidence relating to these matters was given by two experts in neonatology and, in relation to the radiological findings, by two neuroradiologists.

Both the experts in neonatology are practitioners of great stature. Dr Michael Smith, the Claimant’s expert, has been a consultant paediatrician for 23 years. From 1983, he was Consultant Paediatrician (Child Development and Neurology) and Consultant Neonatologist at major hospitals in Sheffield. Since July 2000, he has been Consultant Paediatrician at the Regional Neonatal Intensive Care Unit, Sheffield. He has enormous clinical experience and publishes widely. His current research interests include MR imaging in the newborn and he has a long standing interest in cerebral palsy disorders.

Dr Janet Rennie has been a Consultant in Neonatal Medicine since December 1988. From 1988 until 1995, she was Director of Neonatal Services at the Cambridge University Teaching Hospitals Trust. From 1995 until 2004, she was Consultant and Senior Lecturer in Neonatal Medicine at the Kings College Hospital NHS Trust. Since October 2004, she has been Consultant and Senior Lecturer in Neonatal Medicine at the University College London Hospitals. She is the author of textbooks on paediatrics and neonatology, including the standard work on neonatology.

The neonatology experts have produced reports, held a joint meeting and produced a joint statement recording their responses to questions put to them by the parties’ lawyers and both gave oral evidence.

The subject matter of this case concerns a very difficult area of medicine, in which leading practitioners can hold genuinely differing views. My clear impression is that, when giving evidence, both experts were striving to assist the Court to the best of their ability. The fact that, in the event, I have preferred the evidence of one of the experts to that of the other is no reflection upon their respective expertise or conduct. Both were highly impressive witnesses.

The neonatologists agreed that the Claimant had suffered a large right-sided stroke. The Defendant’s case was that stroke was the only pathology which had been present and that it had been wholly responsible for the Claimant’s injury. The Claimant’s case was that there had been two co-existing pathologies, namely stroke and hypoxic-ischaemia, the latter resulting from a period of chronic partial hypoxia prior to delivery. Dr Smith’s evidence was that the hypoxic-ischaemia (as well as the stroke) had caused permanent damage evident on MR scanning. This was in distinction to the evidence of Mr Spencer, the Claimant’s obstetric expert, whose thesis was that the Claimant had suffered chronic partial hypoxia which had resulted in no lasting damage.

It is necessary first to say a little about the mechanism of stroke, its incidence in babies and its association, if any, with hypoxia and/or hypoxic-ischaemia.

The Claimant’s stroke is likely to have been caused by an occlusion (obstruction) of the right middle cerebral artery. The occlusion would have been caused either by a blood clot which formed in situ (a thrombus) or by a clot (embolus) which travelled through the fetal circulation from elsewhere, probably from the placenta. The occlusion would prevent blood from flowing beyond it and would lead to critical under-perfusion of the territory of the brain served by the artery. This would in turn give rise to focal infarction, causing damage sited in the right hemisphere of the brain, within the distribution of the right middle cerebral artery.

Although stroke is generally considered to be a disease of old age, it is in fact quite common in babies. The true incidence of neonatal stroke has only become evident since sophisticated MR scanning became available. Thus, while the incidence was estimated in the mid-1990s to be one in 10,000 births, it is now thought to be about one in 4,000. It is likely that some strokes still go entirely unrecognised since strokes have sometimes been found by chance on scanning children with no symptoms. It is not clear why stroke is so common in babies; it may be due to the fact that neonatal blood tends towards clotting easily, perhaps as a protective mechanism against blood loss. In addition, the ageing placenta is thought to be a source of emboli that can pass spontaneously into the fetal circulation. Neonatal strokes include a wide spectrum of lesions which are difficult to classify.

In opening the case for the Claimant, Mr Redfern QC suggested that the Claimant’s case would be that the reduction in oxygenation and under-perfusion of the fetal brain resulting from chronic partial hypoxia caused ‘sludging’ of the blood and formation of a thrombus which in turn caused the right middle cerebral artery infarction. In the event, Dr Smith’s evidence did not go so far as to suggest that hypoxia had caused the stroke. Dr Smith said that he believed there was an association between stroke and perinatal hypoxia. He acknowledged that the process by which the two conditions were linked was not known. Dr Smith said that he had himself seen children with evidence of a middle cerebral artery infarction which had occurred ‘in the context of’ generalised hypoxic-ischaemia. He believed that the coincidence of the two conditions was more than one would expect by chance. His view was that hypoxia is a risk factor for stroke.

The Claimant’s neuroradiologist, Dr Wellesley St Clair Forbes, suggested that, on a balance of probabilities, the Claimant’s stroke had been caused by chronic hypoxia or that, at the very least, hypoxia was a contributory factor. However, in doing so, it seems to me that he moved well outside his area of expertise. In my view, this issue falls squarely within the province of the neonatologists.

In support of his view, Dr Smith relied on an extract from a Chapter in the current (2004) edition of a textbook by Volpe, ‘Hypoxic-Ischemic Encephalopathy: Neuropathology and Pathogenesis’. The Chapter discusses the aetiology of cerebral infarction in neonates. A Table (Table 8-15) sets out the aetiologies in 249 neonatal cases of cerebral infarction which had been identified by brain imaging and derived from the author’s personal experience and from the reports of others. Table 8-15 shows that, in 35% of cases, the aetiology was identified as ‘perinatal asphyxia’. The author comments as follows:

“late intrauterine, intrapartum, or neonatal generalized hypoxic-ischemic events are important causes of focal cerebral ischemic lesions … Indeed, in the composite series of 249 cases of focal cerebral lesions identified in the newborn by brain imaging shown in Table 8-15, approximately 35% appeared to be related to perinatal asphyxia, presumed to be intrapartum because of fetal heart rate abnormalities, depressed Apgar scores, and neonatal neurological features. The reasons for the focal cerebral lesions, in a vascular distribution, in the setting of an apparent generalized insult, are unknown. However, it is noteworthy that the vessels of anterior circulation, unlike those of the vertebrobasilar circulation, have a dense sympathetic innervation. Thus asphyxia, a potent sympathetic stimulator, may be particularly likely to induce vasoconstriction in the anterior circulation and thereby result in the preponderance of focal cerebral ischemic lesions in the distribution of the middle cerebral artery. A speculation would be that this innervation may exhibit asymmetries during development and that this developmental feature may explain some of the unilateral lesions seen with this apparently generalised insult.”

Dr Rennie suggested that the extract from Volpe’s book did not represent current thinking. She said that, in the past, a turbulent perinatal period, followed by encephalopathic damage, had often been ascribed to birth asphyxia. She said that it had not been recognised that a baby who suffered seizures did not necessarily have hypoxic-ischaemia. In the past, therefore, many babies with non-specific signs such as seizures would have been diagnosed incorrectly as suffering from hypoxic-ischaemia. Dr Rennie observed that the neonatal cases referred to in Volpe’s Table 8-15 included many dating from the 1980s, before MR scanning was available objectively to confirm the presence of hypoxic-ischaemic damage. In the absence of objective evidence of hypoxic-ischaemia in an individual case, the assertion that a stroke had been caused by birth asphyxia in that case was unreliable.

Volpe refers to the phenomenon whereby an apparently generalised impairment of perfusion of the brain produces a focal cerebral lesion, observing:

“The reason for the strikingly unilateral predominance in the distribution often of only a single vessel, in the presence of an apparently generalized disturbance of perfusion, is puzzling. The phenomenon has been observed in the fetal and neonatal monkey. Possible explanations include variations in the development of the cerebral vessels, or their responses to regulatory effectors, associated vascular spasm because of concomitant subarachnoid blood or infection, and variations in the development of metabolic capacities of brain regions with consequent variations in vulnerability to ischemia.”

Dr Smith said that he would add to that list of possible causes the augmentative effect on coagulation of hypoxic-ischaemia, together with individual susceptibility to that effect. He agreed with Volpe that it was puzzling that a generalised disturbance should be associated with damage to a single vessel but was confident that this did on occasion occur.

Dr Rennie did not accept that a global insult caused by hypoxic-ischaemia could be associated with asymmetric damage affecting only one major artery, without affecting the border zones between the areas of distribution of the major cerebral arteries. She said that she did not find the possible explanations advanced by Volpe convincing. She did not accept that hypoxic-ischaemia could cause vasospasm sufficient to close a major artery, while causing no damage elsewhere. She suggested that it was because of the implausibility of this occurring that many people considered that a prolonged period of partial hypoxia was not likely to cause a stroke. She said that the literature was moving away from a link between the two conditions because of the failure to identify any satisfactory mechanism for the link.

Dr Rennie relied on a review (Footnote: ⁶) by Nelson and Lynch published in March 2004. In dealing with the reported link between birth asphyxia and stroke, the authors observed:

“Signs interpreted to indicate birth asphyxia might be either a result or a cause of prenatal stroke. Early reports of infants with perinatal stroke included a history of birth asphyxia in most of these infants, but this has not been confirmed in more recent studies. In a study of full-term infants with cerebral infarction, there was no significant difference between cases and controls in pregnancy complications, intrauterine heart rate monitoring, and mode of delivery, umbilical artery pH, or Apgar score at 5 min.”

Dr. Rennie was asked about an observation in her own book “Neonatal Cerebral Ultrasound”, written in 1995 and published in 1997. The book is now out of print and Dr Rennie is in the process of preparing a revised version for publication. In the book she stated:

“Neonatal strokes are rare, with an estimated incidence of 1 in 10,000 deliveries (Uvebrant, 1988). Arterial occlusion can result from an embolus, hyperviscosity, sepsis, stretching and damage to the artery from birth trauma or secondary to the oedema associated with birth asphyxia.”

It was suggested by Mr Redfern that, in that passage, Dr Rennie had acknowledged that birth asphyxia could cause cerebral infarction. She said that, since she wrote the book, thinking had moved on; an example of this was the greater recognition of cerebral infarction which is now thought to occur in 1 in 4,000 (rather than 10,000) births. At the time she wrote the passage, she had believed that external pressure caused by brain swelling could compress and exert pressure on an artery, causing a thrombus to form and the artery to be occluded. She said that she no long held that view, which did not accord with current thinking, and the passage will not appear in the new version of her book. In any event, the oedema to which she had referred would have occurred after, not before, delivery, so the passage referred to strokes occurring in the neonatal period, not perinatally. It was not, therefore, relevant to the Claimant’s case. I accept her evidence on that point.

I also accept that the literature currently supports an association between stroke and hypoxia. I further accept that, as Dr Smith fairly conceded, there is no clear evidence of an established causative link between hypoxia and stroke and no generally accepted mechanism for the association between the two conditions. It may be that, as Nelson and Lynch suggest, future studies informed by MR imaging will show that the link is less strong than was previously believed – possibly that it does not exist at all. For the purposes of this case, however, I find that an association exists, although that association falls short of establishing, in an individual case, that a baby’s stroke was caused by hypoxia.

I shall now move on to consider the evidence relating to the Claimant’s condition immediately after birth and thereafter.

Dr Smith said that, in his view, the clinical history and the radiological findings were consistent with the Claimant having suffered hypoxic-ischaemic damage in addition to damage resulting from the stroke. Dr Rennie’s opinion was that all the evidence pointed to one ‘unifying diagnosis’ of stroke.

The Claimant did not breathe spontaneously at birth and required oxygen by bag and mask before independent breathing was established. She also needed her airways sucked out. The Claimant’s Apgar scores were 4 at one minute, 7 at five minutes and 9 at seven minutes. Dr Smith argued that, having regard to what was known of her condition at birth, a score of 2 or 3 at one minute would have been more appropriate. Dr Tumi, the senior paediatrician who attended the delivery, gave evidence. He was not asked about the Apgar scores which were presumably awarded by him or under his supervision; at the very least they were adopted by him.

Dr Smith described the Claimant as having been born in a significant degree of cardio-respiratory depression. He acknowledged that the Claimant seemed to have made a good recovery after the initial resuscitation. He said that there was no record of any detailed examination having taken place, as a result of which the information about her condition was limited and it was not possible to say whether the subsequent Apgar scores were accurate. It was clear, however, that she was not thought to be encephalopathic (see below) at that stage. If she had appeared to be encephalopathic, she would have been admitted to the NICU for observation. Instead, she was allowed to accompany her mother to the post natal ward.

Dr Rennie said that she found no reason within the medical records to believe that the Apgar scores were unreliable. The scores at five and ten minutes were satisfactory and, within a short time of delivery, the Claimant was said to be ‘completely well’ and was considered fit to be transferred to the ordinary post natal ward. Dr Rennie said that there was in her view no evidence of significant birth asphyxia. She regarded the Claimant’s condition at birth as inconsistent with that of a baby who had suffered a prolonged period of chronic partial hypoxia sufficient to cause damage evident on MR scanning. Dr Rennie said that the Claimant’s condition at birth was entirely consistent with that of a baby born by emergency caesarean section through meconium-stained liquor. Mr Peebles agreed. He said that, if the fetus had been chronically hypoxic and this had got worse at the end of labour, he would have expected the Claimant to be born in a worse condition. Mr Spencer described the Claimant as ‘well’ at birth. He said that she was not hypoxic. He observed that it was because of her good condition on delivery that he had concluded that her chronic partial hypoxia had been well tolerated and cannot have caused permanent damage.

Acidosis (evidenced by a low pH value on testing of blood gases at delivery) is a sign that a baby has suffered perinatal hypoxia. The Claimant’s blood gases were not measured at delivery. Consequently, it is impossible to say whether she was acidotic. Dr Rennie said that, bearing in mind what was known about her condition, she was not likely to have been acidotic. Dr Smith said that it should not necessarily be assumed that she was not. He referred to a study (Footnote: ⁷) of 60 children who had suffered perinatal fetal hypoxia. The eight children who had suffered major motor and cognitive deficits had a mean Apgar score of 7.4 and a mean pH of 7 (i.e. they were acidotic). He said that the fact that the Claimant had an Apgar score of 7 at five minutes and appeared well did not necessarily mean that she was not acidotic and had not suffered from hypoxic-ischaemic injury.

Although the Claimant required some resuscitation, she recovered well and was, within a short time, considered fit enough to be transferred to the post natal ward. I accept the evidence of Dr Tumi and of the midwives on duty on the ward that, had she not been considered to be in good condition, she would not have been sent to, or accepted onto, the ward but would have been sent to the NICU for observation. It does not seem to me that there is sufficient information available for me to infer that the first Apgar score was an over-estimate. In any event, I find that there is no reason to believe that the later scores (which are probably more significant) were not accurate. They suggest that the Claimant was in a reasonably good condition within five minutes of delivery. Nor can it be inferred from the available evidence that the Claimant was acidotic.

I accept the evidence of Dr Rennie and Mr Peebles that the Claimant was in a condition which was wholly consistent with the circumstances of her birth. I also accept that her condition, as evidenced by the contemporary medical notes, was inconsistent with her having suffered a lengthy period of chronic partial hypoxia which became more acute in the two hours preceding delivery, particularly if the hypoxia resulted (as it is contended on her behalf that it did) in damage evident on MR scanning.

The neonatologists agreed that, during the hours following her birth, the Claimant developed an encephalopathy. ‘Encephalopathy’ is a general term used to describe a clinical syndrome arising from an acute or chronic disturbance of normal brain function. It has many causes, including focal infarction and more generalised disturbances of cerebral function such as hypoxic-ischaemia. The neonatologists agreed that the encephalopathic signs exhibited by the Claimant could have been wholly accounted for by the effects of her stroke. In their joint statement, they said this:

“MS (Dr Smith) agrees with JR (Dr Rennie) that Amelia’s clinical neonatal course was consistent with a baby who had recently sustained a large middle cerebral artery infarction with or without a more generalised cerebral pathology and cannot be used to diagnose HIE (hypoxic-ischaemic encephalopathy) in isolation.”

Dr Rennie argued that there were features of the Claimant’s encephalopathy which were inconsistent with it having been caused by hypoxic-ischaemia.

It is necessary to set out briefly the course of the Claimant’s progress immediately after her birth. A nursing note made at 0100 hours on 9 August (i.e. 1 ¼ hours after birth) records that she was put to the breast but was ‘not interested’. Dr Rennie said that this, together with the fact that she subsequently had to be fed by syringe, was not normal. At 0150 hours, she was described as ‘clammy’ and ‘jittery’. Dr Smith said that it was difficult to say what caused these signs. She had a blood sugar level of 1.1 mmols (measured on the ward). It is not unusual for a baby to have low blood sugar levels at the age of 1-2 hours after birth, following the cut off of supplies from the placenta. However, the reading was low and could, he said, indicate the presence of hypoglycaemia, which could have caused the abnormal signs. In her report, Dr Rennie had suggested that the Claimant’s signs at this time were probably due to hypoglycaemia. In oral evidence, she doubted whether that was the case. At the time, the paediatric staff suggested that the Claimant should be given formula milk, in an attempt to raise her blood sugar levels. At 0210 hours, she was reported to have taken 50 mls of formula milk well by syringe. By 0255 hours, her blood sugar levels (measured on the ward) had risen to 1.8 mmols. A sample taken at the same time and sent to the laboratory for analysis showed a level of 1.5 mmols.

At about 0313 hours (i.e. at 3½ hours of age), the Claimant suffered a ‘dusky episode’ while feeding. She required resuscitation and oxygen. Dr Smith said that he did not know what caused this episode. It could have been a seizure although the typical features of a seizure were not described in the medical notes. Alternatively, it could have been a mechanical problem caused by feeding. Dr Rennie’s view was that this was probably a seizure. Afterwards, the Claimant appeared ‘quite well’ and ‘active’. Dr Tumi, who saw the Claimant after this episode, was concerned that she might have an infection (her mother had a pyrexia at the time) although, in the event, this proved not to be the case. He arranged for the Claimant to be given antibiotics. She was taken to the NICU for a venflon to be sited and bloods taken. At 0400 hours, a blood sugar reading taken on the ward was 2.5 mmols. A sample was sent to the laboratory and showed a level of 2.1 mmols. She was then allowed to return to the post natal ward. It seems, therefore, that there can have been no suspicion that the Claimant was in the throes of an encephalopathy at that time. If there had been, no doubt Dr Tumi would have wanted to keep her under observation in the NICU.

At 0430 hours, it was reported that the Claimant was still not interested in breast feeding. By 0450 hours on 9 August, she was said to be looking ‘very sweaty’ with ‘loss of colour’. Dr Smith said it was not possible to say what caused this. It could have been related to hypoglycaemia (a reading taken on the ward at the time showed a level of 1.6 mmols and a sample sent to the laboratory at the same time gave a reading of 2 mmols.). It could have been caused by the effects of the stroke. The nursing staff responded by feeding the Claimant formula milk by syringe, obviously recognising that her problems might be caused by hypoglycaemia.

At 0500 hours (i.e. at about 5¼ hours of age), the Claimant was said to have gone ‘v. dusky’ and ‘floppy’ while feeding. Her eyes were noted to be rolling and she was ‘quite twitchy’. She required resuscitation and oxygen. A paediatric team was summoned and she was transferred to the NICU. There, the Claimant was started immediately on dextrose (to increase her blood sugar levels) and oxygen was administered. Antibiotics were continued. The neonatologists agreed that this was definitely a seizure.

Following admission to the NICU, at 0600 hours on 9 August, the Claimant was noted to be very ‘jittery’, with her eyes rolling and her limbs stiff. The neonatologists agreed that she had had another seizure then. However, there is no record of any abnormality when she was being handled and it seems that she was behaving normally between seizures. At 0800 hours, her blood sugar levels were measured at 3.8 mmols A further probable seizure was noted at 1000 hrs. At that time, she was noted to be hypertonic, with possible seizure activity.

At 1100 hours, the Claimant was seen by Dr Erhardt, Consultant Paediatrician. She was tending to irritability and was lying in extension with her head extended. Her movements were not abnormal when she was not being handled. At 1200 hours, she had occasional eye twitching and staring. The nursing notes record some hypertonus at 1300 hours.

It is clear that, as from about 1000 hours on 9 August (or, at the very latest, 1100 hours), the Claimant was exhibiting signs of abnormal tone. She was also having occasional staring and eye twitching movements. Dr Smith said that these signs indicated the presence of a developing encephalopathy, in addition to the seizures.

After that, there appeared to be an increase in the frequency of seizure activity, together with irritability, hypertonus and extension. By about 1540 hours on 9 August, the frequency of seizures was such that phenobarbitone was prescribed. At 2030 hours on 9 August, the Claimant was reported to have had no further seizures since the administration of phenobarbitone although she was still exhibiting a high pitched cry and hypertonus when handled. She was reported to have had three short seizures during the night of 9/10 August. No obvious seizure activity was reported thereafter save for a brief episode on 13 August. The phenobarbitone was reduced on 11 August and stopped altogether on 13 August. By the morning of 11 August, she was showing interest in breast feeding. Later that day, she was reported to be handling well, with normal tone. On 13 August, she had one ‘dusky episode’ which was self correcting and of short duration. She was otherwise well. She was considered well enough to return to the ward with her mother on 14 August and was discharged from hospital on 17 August.

Dr Smith said that the fact that the Claimant exhibited signs such as irritability, abnormality of tone and extension at times when she was not having seizures was not typical of the encephalopathy associated with a stroke. He said that the postural and tonal signs were more typical of a bilateral hypoxic-ischaemic encephalopathy. It was not clear to me whether he was suggesting that there were two distinct encephalopathic processes at work, one (causing the seizures) resulting from the stroke and the other resulting from hypoxic-ischaemia or whether he was saying that the effects of the stroke and the hypoxic-ischaemia had merged to produce a collection of encephalopathic signs. It probably does not matter. In either event, his thesis was that there were two pathologies at work. However, he agreed with Dr Rennie that the encephalopathic signs which were present could be attributable solely to the stroke and had indeed encountered such signs associated with stroke in the course of his own clinical practice.

Dr Rennie’s view was that the signs exhibited by the Claimant were wholly consistent with a large volume stroke. She referred to a review article (Footnote: ⁸) by Mercuri & Cowan, in which the authors stated:

“In our experience in the majority of infants with cerebral infarction, the first hours after birth can be quite unremarkable and these infants are generally thought well enough to go to the postnatal wards with their mothers. On questioning retrospectively, mild sleepiness, poor feeding and irritability are frequently described in the first 24 hours. Early seizures in the first 24–48 hours are often the first clinical signs. A few infants will require early respiratory support and these are the ones who have signs of perinatal asphyxia and encephalopathy, such as more marked hypotonia, irritability and reduced visual alertness.”

She said that, in her view, the description contained in that passage precisely corresponded with the Claimant's condition in the neonatal period.

The question then arose as to whether the Claimant’s signs were consistent also with an encephalopathy resulting from hypoxic-ischaemia. Dr Rennie did not accept that they were, primarily because of the absence of convincing evidence of multi-organ damage.

Dr Rennie said that such multi-organ damage would be expected to result from a hypoxic-ischaemia severe enough to result in damage evident on MR imaging. She said that, if a hypoxic insult (particularly a chronic partial hypoxic insult) had been severe enough to damage the brain, it would also have been sufficient to damage other organs. A fetus which is experiencing a lack of adequate blood supply suffers a whole body problem and, in general, will attempt to protect the brain by reducing the supply of blood to other organs. The most vulnerable organs are the kidneys; she would have expected there to be signs of damage to them, or signs of damage to other organs, such as abnormal liver function or an abnormal platelet count. There was no evidence of such damage. She said that she would certainly expect evidence of multi-organ damage in a baby who (on the Claimant’s case) had been suffering from hypoxic-ischaemia for 34 hours or more and had suffered hypoxic-ischaemic brain damage as a result. She accepted that there was literature showing that multi-organ damage occurred in only 70% of cases of hypoxia. However, she said that the incidence was higher than 70% in cases of chronic partial (as opposed to acute) hypoxia and much higher in cases (as here) where the hypoxia was said to have persisted for more than 24 hours.

For the Claimant, Mr Redfern placed reliance on the references within the medical notes made at the time to ‘hypoxic-ischaemic encephalopathy’ as one of the Claimant’s problems and/or diagnoses. It does not seem to me that those references can be in any way determinative of the correct diagnosis. Those treating the Claimant recognised correctly that she had an encephalopathy. They had no reason at that time to suspect the presence of a stroke. They were aware that she had been born by emergency caesarean section which had been carried out because of fetal distress in the first stage of labour. It is not perhaps surprising that they assumed that her encephalopathy was hypoxic-ischaemic in nature. Dr Rennie explained to me the difficulties that she experiences in ensuring that staff on her own Unit use the correct terminology when describing an encephalopathy. She said that there was a tendency to assume that every neonate who had a seizure was suffering from a hypoxic-ischaemic encephalopathy. I accept what she says. In essence, it accorded with Dr Smith’s evidence. He observed that the medical notes in this case reveal very little day-to-day evidence of detailed neurological examination. Such examination would form the basis of an accurate diagnosis of hypoxic-ischaemic encephalopathy. He said that hypoxic-ischaemic encephalopathy is a well known disorder but can be mistaken for an encephalopathy caused by stroke. I did not get the impression that he was seeking to persuade me to place any reliance on the mention of hypoxic-ischaemic encephalopathy in the medical notes. I do not do so.

I find that the Claimant suffered her first frank seizure (the so-called ‘dusky episode’) just after 0313 hours, i.e. at about 3 ½ hours of age. Dr Smith suggested that it might have been caused by a mechanical problem associated with feeding. However, the unchallenged evidence of the midwife who was feeding the Claimant at the time was that she did not think that her breathing problems were caused by her feed; if they had been, she would have made a note indicating the fact. In any event, viewed in the context of later events, it seems highly likely that this was indeed a seizure. More generalised encephalopathic signs had become apparent by 1000 hours – 1100 hours at the latest. I shall deal with the conclusions to be drawn from the Claimant’s condition in the neonatal period later in this judgment.

I have already mentioned that I would consider whether any support for the presence of placental insufficiency could be derived from the available evidence relating to the Claimant’s size and weight after birth and to her blood sugar levels (including the possible presence of hypoglycaemia) in the neonatal period. I shall first consider the evidence relating to the possible presence of hypoglycaemia.

Hypoglycaemia, if present and of a significant extent, would provide some support for the suggestion that there was placental insufficiency. It would be consistent with the fetus having used up its reserves of glycogen during the period when it was exposed to chronic partial hypoxia. Mr Spencer referred to this in his report, observing that the Claimant had suffered ‘severe’ hypoglycaemia in the neonatal period, caused by prolonged chronic hypoxia. In oral evidence, however, neither he nor any other witness suggested that the hypoglycaemia was severe.

The Claimant’s lowest blood sugar level reading was 1.1 mmols, although the reading may not have been accurate. (Measurements of blood sugar taken on the ward are not as accurate as those tested in the laboratory. A sample of blood was sent to the laboratory for testing at the same time that the reading was taken but the results are not available.) Otherwise, the lowest reading was 1.5 mmols. Dr Smith said that it was only these two readings that showed evidence of hypoglycaemia. Thereafter, the Claimant’s blood sugar levels rose after feeding and, later, the administration of dextrose. Dr Smith said that the hypoglycaemia could have been caused by hypoxic-ischaemia. It was clear, however, that he did not attach any great significance to it since he pointed out that the hypoglycaemia, if present, was relatively short-lived and was not very severe.

As I have already mentioned, the Claimant’s blood sugar levels were low in the first few hours after delivery and there was obvious concern on the part of those treating her that some of the signs which she was exhibiting were attributable to hypoglycaemia. In fact, it is now clear that the majority of them were not, since they persisted well beyond the time that her blood sugar levels had risen to normal levels. It is however possible that the first signs of clamminess and jitteriness at 0150 hours were caused by hypoglycaemia. Dr Smith said that hypoglycaemia might have produced the signs described even in a normal baby. Dr Rennie agreed, although she said that many babies with a blood sugar level of even 1.1 mmols are not jittery. Having regard to the subsequent history, I regard it as very unlikely that the later episode at 0450 hours was related to hypoglycaemia. It seems to me much more likely that it was related to seizure activity.

The degree of hypoglycaemia suffered by the Claimant was of a minor order and lasted only a short time. It does not seem to me to have been of such an order as to lend any real support to the contention that there was placental insufficiency such as to cause hypoxia.

It was argued on behalf of the Claimant that her weight and size on and after delivery were less than would have been expected, given the information about her likely size which was available from ultrasound scans carried out during Mrs Garcia’s pregnancy. It was said that this tended to confirm the fact that there had been placental insufficiency. Based on measurements taken at 32 and 36 weeks’ gestation, the Claimant was predicted to have a head and abdominal circumference at or near the 90^th centile. She was thought to be a ‘big baby’. At delivery, her weight was 3.32 kg (i.e. on about the 30^th centile). It was suggested that the difference between her predicted size and her birth weight constituted evidence of growth restriction/retardation resulting from placental insufficiency. Mr Mahady’s evidence was that it was ‘notoriously difficult’ to assess fetal weight from antenatal growth charts. He said that the purpose of the charts was merely to alert those attending the mother to the possibility of a large baby. Mr Spencer did not contend otherwise. In his report, Mr Spencer said that it was ‘possible’ that fetal growth after 36 weeks was curtailed by restricted placental function of gradual onset. He did not refer further to this possibility in his oral evidence. It did not seem to me that he considered that it was a matter of any real significance.

According to a chart contained in the medical records, the Claimant’s head circumference fell just on the 50^th centile at delivery. At 13 weeks after delivery, her body weight was still on a slightly lower centile than her head circumference. It was suggested that this asymmetry was significant and may be attributable to the effects of growth restriction/retardation. Dr Smith said that this type of difference was seen quite commonly in clinical practice. He was uncertain about the margins of error in taking the relevant measurements. He agreed with Mr Redfern that, if the measurements were reliable, they would suggest a reduction from the Claimant’s expected weight which could be due to placental insufficiency. However, he pointed out that the margins of error would be very small. It did not seem to me that Dr Smith adopted this point with any enthusiasm. Mr Spencer did not mention it in his report or his oral evidence. Dr Rennie said that the degree of disparity was tiny and not clinically significant. Even if there was a degree of growth retardation (which she did not accept), she said that it would not be inconsistent with the stroke having occurred spontaneously as a result of an embolus becoming dislodged from the ageing placenta.

If Mr Spencer or Dr Smith had thought that this was a significant point, I have no doubt they would have made it more strongly. My impression was that they were not persuaded by it. Neither am I. The evidence on the point is very limited and the accuracy of the measurements taken is unknown. Had there been any significant asymmetry at delivery, I should have expected this to have been recorded in the medical notes. No such note was made. It does not seem to me that the evidence is sufficient to establish the existence of any fetal growth retardation/restriction consequent upon placental insufficiency.

Both neonatologists agreed that the Claimant’s end stage disabilities did not assist in determining whether she had suffered a stroke alone or a combination of the stroke and hypoxic-ischaemic injury. Her principal problems are severe left hemiparesis, epilepsy and learning difficulties. They could have been caused by the right-sided stroke alone. Dr Smith’s evidence was that it was possible that there was some contribution from the injury to the left side of the brain which was evident on MR scanning and which he attributed to hypoxic-ischaemic damage. Dr Rennie said that the Claimant’s disabilities were entirely explicable by the right-sided stroke so there was no need to look for another cause.

The evidence relating to the Claimant’s condition immediately after birth and in the neonatal period seems to me to fall far short of establishing that she was suffering from the effects of hypoxic-ischaemia, whether damaging or not. The Claimant’s condition at birth was not consistent with her having suffered a long period of chronic hypoxia which was, during the last few hours, “cranked up” (to use Mr Redfern’s expression) as a result of the increased frequency of uterine contractions. Moreover, her condition in the neonatal period was, I find, more consistent with the aftermath of a stroke alone than with the effects of hypoxic-ischaemia. I accept that her encephalopathic signs were not altogether typical of stroke. However it was agreed by all that they were not inconsistent with stroke and they corresponded strikingly to the description contained in the review by Mercuri and Cowan. Even more significantly, however, she had no damage to organs other than the brain. I accept that such damage does not occur in every case of hypoxic-ischaemia, even when the hypoxic-ischaemia is caused by chronic hypoxia. However, I accept Dr Rennie’s evidence that, given the circumstances of this case, it would have been remarkable if no such damage had been sustained.

On the basis of the evidence relating to the Claimant’s condition immediately after birth and in the neonatal period, I regard it as more probable than not that it was stroke alone that caused her clinical signs, rather than a combination of the stroke and hypoxic-ischaemia. However, the clinical signs must be viewed in conjunction with the radiological evidence.

In the neonatologists’ joint statement, Dr Smith stated that he was of the view that the ultrasound scan and the MRI were the strongest evidence for hypoxic-ischaemia. In evidence, he described the radiological evidence as a ‘key feature’ and said that, if the left hemisphere had been normal on scanning, he would have accepted that there was no evidence of hypoxic-ischaemic damage. The evidence relating to the scanning is, therefore, crucial to the determination of the case. The evidence came mainly from Dr Forbes (for the Claimant) and Dr Brian Kendall (for the Defendant).

Dr Kendall is currently Consultant (or Honorary Consultant) Neuoradiologist at several major London hospitals, including Great Ormond Street. He is the author of over 200 papers in the field of neuroradiology. He has enormous experience and stature in the field of paediatric neuroradiology. Dr Forbes has, since 1975, been Senior Consultant Neuroradiologist at Hope Hospital, Manchester and part time Lecturer at the University of Manchester. He specialises in paediatric neuroradiology and has published on all aspects of adult and paediatric neuroradiology.

The neuroradiologists provided their own reports, together with a joint report following discussions and gave oral evidence about the imaging they had viewed.

The Claimant underwent five relevant scans, namely:

The neuroradiologists agreed that the scans showed widespread damage in the right cerebral hemisphere with dilatation of the body of the right lateral ventricle, in the distribution of the right middle cerebral artery. Dilatation of a ventricle would be caused by atrophy of the white matter of the brain surrounding the ventricle. The neuroradiologists agreed that the relevant damage had been caused by occlusion of the right middle cerebral artery. They agreed that it was not possible to say from the scans whether the damage was caused by a thrombus occurring in situ or by an embolus travelling from elsewhere.

The dispute between the neuroradiologists related to the appearances in the left cerebral hemisphere.

This was the first scan, taken at about 34 hours after delivery. The available images are of poor quality. The neuroradiologists agreed that they showed swelling of the right cerebral hemisphere. This was evidenced by increased echogenicity (i.e. brightness) above and lateral to the right lateral ventricle, together with compression of the right lateral ventricle. Echogenicity is a sign of swelling in the brain tissue, caused by the presence of fluid. A stroke can produce brain swelling, usually localised on the side of the brain affected by the stroke. Brain swelling can also cause compression or flattening of the ventricles.

In his report, written in February 2003, Dr Forbes stated his opinion that there was radiological damage evident in the left (as well as the right) cerebral hemisphere. He said that there was apparent increased echogenicity in the left hemisphere which indicated generalised brain swelling. This was very important because, if bilateral brain swelling were present, it would be consistent with the Claimant having suffered hypoxic-ischaemic damage affecting the whole brain. In a letter written in April 2003, Dr Forbes confirmed his opinion that the ultrasound ‘showed increased echogenicity bilaterally indicating generalised brain swelling’. Dr Kendall’s view was that the ultrasound scan showed no evidence of hypoxic-ischaemic damage.

The joint medical report of 11 May 2006 was drafted by Dr Forbes following a discussion between himself and Dr Kendall and was signed by both doctors. The report contained the doctors’ responses to a number of questions posed by both parties. Those questions included the following:

“3h In addition to the neuro-imaging demonstrating increased echogenicity compressing the right lateral ventricle consistent with an area of infarct, does it also demonstrate increased echogenicity bilaterally indicative of brain swelling?”

To that question, the neuroradiologists answered:

“Both Neuroradiologists agree that the cranial ultrasound showed swelling of the right hemisphere. Dr W St C Forbes considers that the first cranial ultrasound showed increased echogenicity of the left without significant brain swelling. See response to question 4.” [my emphasis]

(In Question 4, Dr Forbes had referred to the ‘bilateral periventricular FLAIR (in fact, ‘flares’)’, which had been mentioned by the radiologist who reported on the scan (see below) and which Dr Forbes said were evidence of hypoxic-ischaemia.)

Question 12 asked, in connection with the ultrasound scans of 10 and 14 August 1998:

“Is there evidence on either set of scans of generalised brain swelling? If so, please give the precise description of the relevant findings.”

To that question, the neuroradiologists answered:

“Both Neuroradiologists agree that there is swelling of the right cerebral hemisphere and no evidence of generalised brain swelling.” [my emphasis]

It seemed clear from Dr Forbes’ answers to questions 3h and 12 that, at the meeting with Dr Kendall, he had changed his original opinion that the ultrasound scan of 10 August showed generalised brain swelling. Shortly before he gave evidence, however, the Court and the Defendant were given notice of Dr Forbes’ wish to amend his answer to Question 12. An amended answer was submitted in these terms:

“Both Neuroradiologists agree that there is swelling of the right cerebral hemisphere and no evidence of generalised brain swelling. Dr Forbes states this was a mistaken answer. There is evidence of generalised brain swelling on 1^st ultrasound see P193 (page 193 refers to his letter of April 2003 in which he had mentioned the generalised brain swelling) of bundle.”

In evidence Dr Forbes explained how the amendment had come to be made. He had been present at Court two days previously. During a discussion with Leading Counsel, he had realised that his answer to Question 12 was inconsistent with his answer to Question 4. He said that he had therefore sought to change it. No attempt was made then or subsequently to amend the answer to Question 3h which appears inconsistent with the amended answer to Question 12.

Dr Forbes said that the appearances on the ultrasound scan of 10 August were asymmetrical. He explained that there was in his view increased echogenicity in the left (as well as the right) hemisphere. This was due to non-specific oedema (swelling), affecting the hemisphere generally and the white matter in particular.

Dr Kendall was adamant that the scan showed no increased echogenicity or swelling on the left side. He explained that, because the ventricles in a young baby vary greatly in volume and are often very small, it is difficult to diagnose brain swelling by reference to the size of the ventricles. Certainly, the left lateral ventricle was not obliterated; there was an obvious difference between the ventricles on the two sides.

Dr Kendall went on to say that it is also difficult to judge whether a structure taken in isolation is echogenic. In order to gauge echogenicity, it is, he said, necessary to compare the brightness of the relevant area with the brightness of a ‘marker’. In this case, the choroid plexus was an appropriate marker. The white matter was much less echogenic than the choroid plexus. Consequently, it was not possible to say that there was any increase in echogenicity in the white matter on the left side.

Dr Kendall was asked about the report on the ultrasound scan which had been written at the time by a radiologist who had probably carried out the scan himself. The report recorded:

“1.

The brain is very ECHOGENIC ‘BRIGHT’

2.

The ventricles are flattened and slit-like

3.

No haemorrhage seen

4.

Periventricular flares bilaterally may represent brain congestion and oedema

Impression: Post-asphyxial (??)

Brain swelling”.

Dr Kendall accepted that a consultant radiologist who carried out an ultrasound scan himself would have had the benefit of viewing the images over a period of time, rather than just seeing a series of snapshots in the images available to the neuroradiologists. However, he pointed out that the objective of taking and keeping the images was to demonstrate what the operator had seen. He would expect the operator to have selected images that illustrated the most important features that he had observed. He pointed out that, if the operator had indeed obtained a better view than that afforded by the available images, he would have expected him to have seen the right-sided infarct. Apparently he had not. Moreover, the operator had not mentioned the asymmetry in the appearances of the right and left hemispheres, which the neuroradiologists agreed were very evident. Furthermore, the ventricles were described in the report as “flattened and slit-like”, whereas this was not the appearance on the available images. He suggested that the operator may well have been aware that it was suspected that the Claimant was suffering from a hypoxic-ischaemic encephalopathy and, as a consequence, may have mistakenly assumed that the damage was bilateral when in fact it was unilateral.

Dr Smith had not seen the original scans but relied on the radiologist’s report. He agreed with Dr Kendall that the volume of the ventricles in a term baby were difficult to assess as they were quite small. Dr Rennie, who is the author of a textbook on the topic of cranial imaging and is therefore well qualified to speak on the subject, had viewed the originals. Her opinion was that the ultrasound scan of 10 August 1998 was asymmetrical, with the ventricular cavity visible on the left. The ventricle was not obliterated as, she said, would be expected if cerebral oedema was present on that side. She was unable to see any abnormal echo reflectivity in the substance of the brain on the left side. She said that the scan did not show generalised global change. She did not accept that there had been any bilateral brain swelling.

The neuroradiologists agreed that these showed no left-sided abnormalities. Dr Forbes said that the oedema in the left hemisphere had resolved but the abnormality in the right hemisphere was evolving. The left lateral ventricle could be seen more clearly. There were similar appearances on the third ultrasound. I note that the same radiologist who reported on the first ultrasound scan also reported on the third ultrasound. He observed that the left lateral ventricle was ‘slightly dilated’ on the third ultrasound whereas the neuroradiologists agreed that it is in fact normal.

These showed clear evidence of right-sided damage. The neuroradiologists agreed that both scans also showed a minor degree of dilatation of the left ventricle with no generalised brain swelling. Both observed in their reports that the MR scan (alone) also revealed areas of abnormal signal intensity in the periventricular white matter on the left.

Dr Smith said that the damage evident to the left side of the brain fell outside the distribution of the right middle cerebral artery so that it was ‘anatomically unfeasible’ that it could have been caused by the infarction. He said that he did not understand what was meant by the suggestion (made in the joint statement of the neuroradiologists: see below) that the damage to the left side of the brain might be ‘secondary to’ the right-sided damage. Furthermore, the occurrence of bilateral damage caused solely by a middle cerebral infarction did not accord with his clinical experience. He said that it had been his experience that children who had middle cerebral obstructions did not have damage to, or enlargement of, the ventricles on the opposite side to where the obstruction occurred. He referred to various photographs of MR scans showing what he said was the typical unilateral damage caused by cerebral artery infarction. He said that a major cerebral artery infarction on one side would produce atrophy only on that side unless there was some other process going on.

Dr Smith said that the left-sided damage required some explanation. His view was that there had been a hypoxic-ischaemic injury which had affected both hemispheres of the brain. The damage to the right hemisphere had been obscured by the effects of the stroke. In the left hemisphere, there was damage to the white matter of the brain, which had atrophied, causing the left ventricle to dilate.

Dr Smith acknowledged that it would be more usual for a hypoxic-ischaemic injury resulting from chronic partial hypoxia to cause damage to the border zones or watershed areas of the brain, rather than to the white matter. However, he said that that was not the only pattern of injury that could occur due to hypoxic-ischaemia. He said that damage to the white matter can also occur without showing a classical border zone appearance. This was accepted by Dr Rennie and by both neuroradiologists.

In evidence, Dr Forbes said that the dilatation of the left lateral ventricle indicated that something had affected the left central hemisphere, causing volume loss. He said that the periventricular changes represented scarring. He acknowledged that they were non-specific findings and that he could not say for sure what had caused them. In his report, he had suggested that both the dilatation of the ventricle and the periventricular signal changes were probably a ‘secondary effect’ related to a generalised asphyxial insult which had caused the middle cerebral artery infarct.

In his report and in the neuroradiologists’ joint report, Dr Kendall expressed the view that the dilatation of the left lateral ventricle was likely to be a secondary effect resulting from the right middle cerebral artery infarction. He said that it sometimes happened that a large infarction on one side of the brain was accompanied by changes (associated with atrophy) on the opposite side of the brain. The explanation for this was thought to be the loss of some of the fibres connecting the two hemispheres.

Dr Rennie said that, in the course of her clinical practice, she occasionally saw contralateral features in children with one-sided strokes. She said that her understanding, based on discussions with neuroradiological colleagues, was that the damage to one side of the brain caused by the stroke meant that there was a reduction in need for the associative fibres or ‘pathways’ whose function was to exchange information between the two sides of the brain. In effect, therefore, her explanation was the same as that of Dr Kendall. Dr Rennie pointed out that she and Dr Smith had agreed in their joint statement that the literature describes damage in the contralateral hemisphere (in two separate papers) in, respectively, 12% and 34% of cases of neonatal stroke.

In the neuroradiologists’ joint report, Dr Forbes said that, in view of the increased echogenicity he believed was present on the ultrasound of 10 August 1998, he “could not exclude the possibility” that the dilatation of the left lateral ventricle evident on the CT and MR scan was the result of asphyxial brain damage, rather than of the infarction. However, he agreed with Dr Kendall (see his answer to Question 9 below) that the more likely explanation was that the dilatation of the left lateral ventricle was secondary to the right-sided brain damage. It was clear from his oral evidence that he accepted the feasibility of the mechanism for damage described by Dr Kendall and Dr Rennie.

In oral evidence, Dr Forbes sought to depart somewhat from his answer to Question 9 of the joint report. Question 9 asked:

“What is the significance of the minor degree of dilatation of the left ventricle shown on the CT and MRI scans? Does it suggest generalised brain swelling, or is it more likely to be secondary to the right-sided brain damage and therefore irrelevant?”

The answer given was:

“Both Neuroradiologists agree that the dilatation of the left lateral ventricle is more likely to be secondary to the right-sided brain damage”.

Dr Forbes said that, because the question referred only to the CT and MR scans, he had not taken into account the appearances of the ultrasound scans when answering it. I find this surprising. The question asked about the ‘significance’ of the appearances of the CT and MR scans. I would have expected that, when considering the ‘significance’ of the appearances, the doctors would have placed the scans in the context of the other evidence available to them, including the earlier ultrasound scans.

Moreover, I note that, when answering Question 3i (which also asked about the ‘significance’ of changes seen on the left side of the brain on the MR scan), Dr Forbes specifically referred to ‘the increased echogenicity on the initial cranial ultrasound’, observing that, in the light of that feature, he could not ‘exclude the left-sided abnormality as being the sequela of asphyxial brain damage on the left’. In giving that answer, he plainly took account of the ultrasound scan despite the fact that it had not been referred to in the question. Furthermore, the fact that he spoke in his answer to Question 3i in terms of being unable to ‘exclude’ the explanation that the left-sided damage had been caused by asphyxial brain damage does not suggest to me that he was putting that explanation forward with any great vigour at that point in his meeting with Dr Kendall. That being the case, it seems to me likely that he went on to accept (as it appears from his answer to Question 9 he did) that it was more likely than not that the dilatation was secondary to the damage caused by the stroke.

The neuroradiologists disagreed about the cause of the abnormal signal intensity in the periventricular area on the left side. In his report, Dr Kendall suggested that this was also probably secondary to the right-sided brain damage. The joint report did not specifically refer to the cause of the abnormal signal. As I have already said, Dr Forbes attributed it to the effect of a generalised asphyxial insult.

In evidence, Dr Kendall said that the abnormal signal intensity was difficult to explain. It was visible on only one (the axial T1 FLAIR) sequence of images from the MR scanning. He said that, although that sequence is very valuable because it enables the viewer to see abnormalities very near to the ventricles, it is also prone to artefact. A well known artefact is the brightness which sometimes occurs around the ventricles. He said that he believed that the signal was a result of artefact around a large ventricle. It was not visible on any other (e.g. the T2) sequence of images and there was nothing in the other sequences to suggest that there was any hypoxic-ischaemic damage in either cerebral hemisphere. He said that, even if the abnormal signal did indicate that there was damage in the periventricular area, it was not hypoxic-ischaemic in nature. When asked why he had not mentioned the presence of artefact in his report, Dr Kendall replied that he had not thought that there would be any controversy over the left-sided appearances.

In their joint report, the neuroradiologists had agreed that there was no radiological evidence of chronic partial or profound asphyxial brain damage. In oral evidence, Dr Forbes explained that, in agreeing that there was no evidence of chronic or partial hypoxia, he had meant that there was no evidence typical of chronic partial hypoxia. Damage caused by chronic partial hypoxia typically affects the border zones or watershed areas of the brain. As I have already said, however, the neuroradiologists and the neonatologists agreed that there are cases when an asphyxial insult can cause damage to the white matter of the brain, without damage extending to the border zones. Dr Forbes suggested that, where damage was confined to the white matter, this was probably an indicator of the fact that the hypoxic-ischaemia which had caused the damage had been of a relatively mild degree and had not been severe enough to affect the border zones. Dr Kendall said that the abnormal signal intensity adjacent to the left lateral ventricle in the Claimant’s MR scans was just a line around the ventricle. He did not see how this could have been caused by hypoxic-ischaemia.

I have no hesitation in accepting the evidence of Dr Kendall in relation to the scanning, in preference to that of Dr Forbes. Dr Kendall gave his evidence with considerable authority and clarity. Much reliance was placed by the Claimant on the increased echogenicity on the left side of the brain and on the bilateral swelling which, according to Dr Forbes, could be seen on the ultrasound scan of 10 August. If present, these were clearly most important findings. So far as the swelling was concerned, Dr Forbes’ evidence was very unsatisfactory. In his report and subsequent letter, he asserted that swelling was present. However, at a joint meeting, it seems that he was persuaded by Dr Kendall’s arguments and changed his view. I do not accept that his answer to Question 12 (which he drafted himself) was ‘mistaken’. I find that it represented his views at the time. I think that, when he realised the implications of his change of mind, he persuaded himself that his answer in the joint report must have arisen as a result of a mistake. I do not suggest that Dr Forbes was attempting deliberately to mislead the Court but I think he could not bring himself to believe that he had indeed made the concession that he had.

I recognise that the radiologist who reported on the ultrasound scan referred to left-sided damage which he suggested was “Post-asphyxial(??)”. I note, however, that he failed to mention abnormalities, including asymmetry, which were clearly present. I regard it as highly likely, as Dr Kendall suggested, that he was aware of the differential diagnosis of hypoxic-ischaemic encephalopathy and mistakenly assumed that the damage was bilateral.

I therefore accept the evidence of Dr Kendall and Dr Rennie in relation to the appearances of this scan.

So far as the CT and MR Scans are concerned, Dr Forbes again had an apparent change of mind. I am confident, having regard to his responses to Questions 3i and 9, that, at the joint meeting, he accepted that the most likely cause of the dilatation of the left lateral ventricle was a secondary effect of the right-sided brain damage. Again, I think that, once the implications of his answer became clear to him, he persuaded himself that his response had been more restricted that it in fact was. In any event, it is clear from his answer to Question 3i that his view about the cause of the left-sided changes to the brain was heavily influenced by his opinions about the appearances of the ultrasound scan of 10 August. I have already rejected his evidence relating to that scan.

Dr Smith rightly demanded an explanation for the left-sided appearances on the MR scan. In my view, Dr Kendall has provided entirely plausible explanations for both the apparent abnormalities observed. So far as the abnormal signal intensity in the periventricular area on the left side is concerned, there is also the fact that this would be atypical of damage caused by chronic partial hypoxia. I accept that, in certain cases, damage to the white matter in the absence of typical damage to the border zones can occur. However this would be unusual and it seems to me that the chances of it coinciding with stroke damage are small. In addition, there is Dr Kendall’s evidence, which I accept, that, even if the abnormal intensity did indicate damage, the damage was of a very specific type and could not have been caused by hypoxic-ischaemia.

I therefore find that there is no radiological evidence of damage to the left side of the brain consequent upon hypoxia or attributable to hypoxic-ischaemia. I further find that the overwhelming likelihood is that all the damage evident on scanning was caused by the right-sided stroke.

I have found that:

In the circumstances, I am unable to find on a balance of probabilities that the Claimant suffered chronic partial hypoxia (whether damaging or not) in utero and/or any hypoxic-ischaemic damage. There is of course clear evidence, agreed by all the experts, that the Claimant suffered a major perinatal stroke. Since the Claimant has not proved to the required standard of proof that she suffered hypoxia and/or hypoxic-ischaemic damage, it must be assumed that all her various signs and her damage were caused by the stroke. Indeed, having heard the evidence, I am satisfied that it is more probable than not that that was the case.

I shall now proceed to consider the evidence relating to the timing of the stroke.

In order to prove causation, the Claimant must establish on a balance of probabilities that the stroke occurred after 2130 hours on 7 August, that being the time at which she would, but for the Defendant’s breach of duty, have been delivered.

In their joint statement, the neonatologists agreed that the presence of neonatal seizures was strong evidence against a long-established lesion. They agreed also that, on balance, the brain injury had occurred no more than 72 hours before delivery.

Dr Smith expressed the view that the damage (by which he meant the damage resulting both from stroke and from hypoxic-ischaemia) had occurred within the last 12–24 hours before delivery. He accepted that a longer timeframe was possible. Dr Rennie’s opinion was that, on a balance of probabilities, the stroke had occurred between the end of the CTG trace on 4 August (which it was agreed was within normal limits) and the start of the abnormal CTG trace at 1330 hours on 7 August.

The timing of the occurrence of a perinatal stroke is not an exercise generally carried out in a clinical context. Most perinatal strokes are discovered some time after delivery on ultrasound/or MR scanning. Some are discovered much later. Some may go entirely undetected. As a consequence, there is only a limited amount of data relevant to the issue of timing of perinatal stroke.

Dr Smith based his opinion about the timing of the Claimant’s stroke on the timing of the onset of encephalopathic features, in particular seizures. He argued that, if the brain injury had occurred longer than 12–24 hours before delivery, there would have been established encephalopathic features from the outset. As it was, the onset of seizures was some hours (on my finding about 3½ hours) later. Postural and tonal abnormalities were not observed until the Claimant was about 10–11 hours old.

When asked to explain the basis of his timing, Dr Smith referred to the probable association between stroke and hypoxic-ischaemic encephalopathy. His view was that the changes in the brain resulting from stroke and from hypoxic-ischaemia had occurred at approximately the same time, only a few hours before delivery. In cross-examination, Dr Smith accepted that the range of times for the onset of signs resulting from stroke is wider than that for hypoxic-ischaemic damage. He said that, nevertheless, he would expect most babies to show signs – including brain swelling – within 24 hours of the occurrence of a stroke.

Dr Smith explained that, in his view, the waxing and waning of the Claimant’s signs provided a clue to the evolution of her condition. Her condition after resuscitation was satisfactory. Even after she suffered the first “dusky episode” (which I have found was a seizure), she was handling well and displaying no other encephalopathic signs. She then developed hypertonus, irritability and a worsening of her seizures. Her seizures peaked and then subsided, after which the other signs began to settle also. Dr Smith said that the fact that her signs had grown, then subsided, caused him to believe that the time of injury was relatively near to the time of delivery, rather than more remote from it. He believed that the injury was likely to have occurred during the last 24 hours before birth.

Dr Rennie said that the time between the occurrence of a stroke and the onset of seizures was longer than the onset of signs in cases of hypoxic-ischaemia and was also more variable. She did not accept that the time of onset of the Claimant’s first frank seizure could be used to establish the time of her stroke. She referred to the fact that, from the outset, the Claimant would not suckle, which was abnormal. She said that she may have had other abnormal signs which would have been evident on detailed neurological examination. I note that Dr Smith agreed that, if someone with an experienced eye had examined the Claimant earlier, s/he may have seen some abnormal neurological signs.

Dr Rennie said that it was quite possible that the seizure observed at 0313 hours was not in fact the Claimant’s first seizure. She could have had seizures in utero prior to that time. Dr Smith agreed that it was possible for seizures to occur in utero.

Dr Rennie referred to the available literature. In the review by Mercuri and Cowan (Footnote: ⁹), the authors observed that early seizures in the first 24–48 hours were often the first clinical sign of cerebral infarction. In their paper (Footnote: ¹⁰), Sreenan et al observed that seizures in the first 72 hours of life were the most frequent presentation in cases of stroke.

The first ultrasound scan was performed 34 hours after delivery. At that time the damage referable to the stroke appeared well-established. Dr Rennie said that there was some literature which suggested that scans within the first 24 hours can be normal. She said that the appearances on the ultrasound scan suggested to her that the injury (i.e. the stroke) had occurred more than a few hours before birth.

In support of her view that the stroke occurred more than 24 hours before delivery, Dr Rennie relied on the early appearance of seizures, the early changes evident on ultrasound scan, and the relatively rapid resolution of the encephalopathy. In addition, there was the change in the appearance of the CTG between 4 and 7 August. All these factors suggested to her that the stroke had occurred more than 24 hours before delivery and before the abnormal scan of 7 August.

In their joint report, the neuroradiologists agreed that the damage visible on scanning had occurred at or around the time of birth. The radiological evidence did not allow precise timing. Both neuroradiologists elaborated upon the issue of timing in their oral evidence.

Dr Kendall said that the only facts available to the radiologist were that abnormal signs attributable to the stroke were evident on ultrasound scan at 34 hours after delivery and that abnormal signs (i.e. seizures) had started about four hours after birth. Thus, the stroke must have occurred before four hours after birth. It could have occurred within days before the scan. Dr Kendall pointed out that a thrombosis or embolism starts with a clot, then an occlusion, then ischaemic damage, then clinical signs; radiological evidence may not appear for days after that. He said that the neuroimaging would be consistent with the stroke having occurred some days prior to the first ultrasound scan.

Dr Forbes’ view, which he said was based on the ultrasound appearances, was that the damage had occurred at some time within about 24 hours before delivery. However, it became evident that – like Dr Smith – Dr Forbes’ views were predicated on the known timing in cases of hypoxic-ischaemia. He said that the timing of stroke was not within his field. He did, however, observe that the length of time between the occurrence of a perinatal stroke and the appearance and resolution of oedema on an ultrasound scan would depend on the size of the stroke.

Were it not for one factor, I would conclude that it was impossible to say when the stroke occurred. It is clear from the literature I have seen that the onset of signs referable to a stroke can occur at any time up to 72 hours after delivery. Where damage is caused by hypoxic-ischaemia, it is possible to time the onset of injury with a fair degree of accuracy. Once one removes the hypoxic-ischaemic element from the equation, it becomes very difficult to determine the timing. However, in this case, the abnormal CTG trace of 7 August is in my view determinative. There must have been some mechanism which caused changes in the CTG traces between 4 and 7 August. Since I have found that the Claimant has not established the presence of hypoxia, that mechanism can only have been the stroke. Otherwise, there would have had to have been some other coincidental mechanism. I regard that as entirely implausible.

I find, on a balance of probabilities, that the stroke occurred some time prior to 1330 hours on 7 August. It follows, therefore, that it did not occur after 2130 hours on 7 August and that the Claimant has failed to establish the necessary causation.

I am well aware that my judgment will come as a dreadful disappointment to the Claimant’s family, who care for her so devotedly. It must seem puzzling to them that a delay in delivering the Claimant of as much as 26 ¼ hours can be found not to have caused her damage. I just hope that they will understand the reasoning behind my decision. I hope also that they will be able to take some comfort in the future from the knowledge that they have done all that they could to secure the Claimant’s interests by bringing this action and pursuing it to its conclusion. It is not through any fault on their part that it has failed.