Introduction

There is a crisis affecting maternity care in many countries. Both caesarean section rates and medical defense premiums are increasing. New options for the care of pregnant women (e.g., birthing units) may fail because midwives, general practitioners and smaller hospitals may be unable to afford adequate insurance to continue their services.¹

The crisis is fuelled by widespread beliefs that cerebral palsy is often caused by injuries sustained during labour and birth and that cerebral palsy may therefore be the result of inappropriate obstetric care. There is now considerable evidence to suggest that these beliefs are unfounded. A conference of Australian and New Zealand specialists in this area was convened to review the relevant literature and offer a consensus statement to help parents, counsellors, lawyers and health professionals understand what is known and what is not known about the origins of cerebral palsy.

Definition

Cerebral palsy is not a single entity but covers neurological impairments characterised by abnormal control of movement or posture resulting from abnormalities in brain development or an acquired non-progressive cerebral lesion.

Prevalence

Cerebral palsy is the most common physical disability in childhood, occurring in about 2-2.5 per 1000 children born. The frequency of cerebral palsy has not changed over the last 40 years, despite a fourfold drop in both perinatal and maternal mortality. In some countries there is an increase in the occurrence of cerebral palsy, attributable mostly to the increased survival of very low birth weight infants.

CAUSE

There are many antenatal factors that may lead to cerebral palsy and often there may be several contributing factors.^2,3 Examples are shown in Box 1. There are also strong associations between cerebral palsy and intrauterine growth restriction, antenatal death of co-twin/triplet and extreme prematurity.

The fetus is designed to withstand the stress of labour, which usually involves a reduction in the amount of oxygen in the blood (hypoxaemia) and the amount of blood reaching the brain (ischaemia) during passage through the birth canal. If these reductions are too great, a normal fetus is more likely to die than survive with cerebral palsy.⁴ These normal reductions of blood supply and oxygen can compound the detrimental effects of any chronic hypoxia already experienced during the antenatal period.⁴ Cerebral palsy occurring after birth is also uncommon and is caused by, for example, complications of prematurity, untreated rhesus disease (kernicterus), meningitis, accidents or near drowning.³

Timing the onset of pathological brain lesions

It should be clearly stated that, currently, fetal brain development (or maldevelopment) cannot be monitored during pregnancy. Only examination of the brain at autopsy can identify the full extent of injury in some cases. In other cases of cerebral palsy no pathological lesion is identifiable. The neuropathological lesions leading to cerebral palsy are various and include maldevelopments (cerebral dysgenesis), germinal matrix-intraventricular haemorrhage, cerebellar haemorrhages, grey matter damage, white matter damage (periventricular leukomalacia), hypoglycaemic neuronal injury, thromboembolic injury (including vasculitis secondary to infection) and kernicterus.

The immature brain has only a limited number of ways of responding to acute or chronic injury and these essentially consist of neuronal and white matter loss and glial proliferation. These changes occur over many days and weeks. They may later be modified by secondary changes such as posthaemorrhagic or postinflammatory hydrocephaly or white matter atrophy. By the time a child presents with cerebral palsy during the first years of life, the neuropathological effects of any hypoxic-ischaemic injury or other injury will have become modified by these changes and by further postnatal brain development. Even if that child were to die during its first year and the brain was made available for expert examination, it would be impossible, on this basis, to determine the exact timing of the original neurological insult.

Recent suggestions that the condition of the placenta could be used as a surrogate marker of antepartum fetal injury are based largely on anecdotal argument and have not yet been fully evaluated; such techniques are fraught with observer and sampling bias.

Imaging of the brain

Antenatal ultrasound scans may detect gross brain changes, but a normal scan does not exclude antenatal neuropathology. Few infants undergo brain imaging, either because there is no clinical suspicion that cerebral palsy may develop or because there are no imaging facilities available. A neonatal ultrasound examination may not demonstrate longstanding microscopic changes but can detect major acute lesions or secondary changes. Acute lesions may appear as periventricular flare, oedema and intracranial haemorrhage. Periventricular and caudothalamic notch cysts, porencephaly, parenchymal calcification and intraventricular adhesions may also be demonstrated in the early neonatal period and imply events occurring more than two weeks previously.

The period from 26 to 34 weeks' gestation is critical for neurodevelopment. The patterns of brain injury, and therefore imaging features, are similar in fetuses and neonates, and late investigations cannot separate brain injury occurring in utero from perinatal or postnatal events. Magnetic resonance imaging of limited numbers of cerebral palsy patients^5,6 suggests that the adverse neurodevelopmental event occurs prenatally in up to 50% of cases.

Can obstetric care prevent cerebral palsy?

If better obstetric care could prevent cerebral palsy then lower rates of cerebral palsy would be associated with good obstetric care rather than with bad obstetric care. There is very little good evidence for this supposition. Several decades ago, new technologies such as electronic fetal monitoring were introduced into obstetrics, without adequate assessment by randomised trials. It was assumed that early detection of fetal distress would allow early delivery (e.g., by caesarean section), thus avoiding damage to the fetal brain and cerebral palsy. Although it is probable that some perinatal deaths have been prevented by such techniques, there is no evidence that they have reduced the prevalence of cerebral palsy, despite increased use of fetal monitoring and rising caesarean section rates.⁷ In view of the likely antenatal origins of most cases of cerebral palsy, this is not surprising. However, the mistaken belief that birth injury is a major cause of cerebral palsy may have stemmed from:

• The high rate of non-specific fetal heart rate aberrations detected by electronic fetal monitoring;⁸
• The lack of tests which are specific for damaging levels of intrapartum hypoxia, and;
• The more intense medical observation during labour and delivery than antenatally. Thus, any signs of possible fetal distress resulting from antenatal neurological injuries may not be detected until the commencement of labour or until the normal stresses of labour uncover signs of a compromised fetus.

These factors may lead to the assumption that the fetus was healthy before labour and that it was only in labour or at birth that problems occurred. However, events causing or predisposing to cerebral palsy may occur from conception onwards, as shown in the Figure. Any one event or any combination of events could be the cause of the cerebral palsy.

Fetal distress

The term fetal distress is imprecise and non-specific.⁹ No antenatal or intrapartum monitoring technique can clearly tell whether a fetus is unhealthy or whether the condition is acute or chronic. Tests such as electronic fetal heart rate monitoring poorly predict fetal outcome. Up to 79% of fetal heart rate traces during labour show some type of variation that has been described as abnormal,⁸ but the vast majority of such infants are born without signs of perinatal asphyxia or cerebral palsy.¹⁰ Clinical signs and test results predict only the possibility of distress and, as distress has no agreed definition in this context, it is preferable to avoid the term fetal distress and instead describe the observed signs or the variation in the test results.

A normal fetal heart rate is generally predictive that there is no acute hypoxaemia, but variations from the normal pattern are not good predictors of hypoxaemia. Even prolonged late decelerations with reduced variability have a less than 50% chance of being associated with major fetal acid-base changes, which themselves are only poorly correlated with cerebral palsy.¹¹ Similarly, meconium-stained liquor is a common finding in labour, but only a minority of babies from these labours are born with a low umbilical arterial blood pH. Finally, Illingworth has suggested that there is an undue readiness to ascribe brain damage to umbilical cord problems.¹² Earn found a loop around the neck of the fetus in 23% of 5676 consecutive births, without significant effect on fetal outcome, except for one neonatal death where the cord was wound eight times around the baby's neck.¹³

Birth asphyxia

Birth asphyxia is not a well-defined term.¹⁴ It implies some sort of dysfunction resulting from a lack of oxygen supply to the baby's tissues during the birth process. The term should not be used clinically because of the difficulty in ascribing clinical signs and symptoms in the neonate to an event during birth. Low pH and/or low Apgar scores at birth are supportive evidence of asphyxia but should not be used alone to make the diagnosis.^15,16 Until more information is available, perinatal asphyxia is the preferred term to describe a neonate in whom there is:

• An event or condition during the perinatal period that is likely to severely reduce oxygen delivery and lead to acidosis (e.g., major antenatal haemorrhage or cord prolapse); and
• A failure of function of at least two organs (usually the brain and kidneys) consistent with the effects of asphyxia.¹⁷

However, even these criteria are not specific for recent hypoxia and can be the result of any of the antenatal causes of cerebral palsy. They are minimum requirements to suggest the possibility of asphyxia but do not prove its intrapartum origin.

Does cerebral palsy ever originate in labour?

Most infants who develop cerebral palsy are born from uncomplicated pregnancies and are delivered without signs of fetal heart rate abnormalities. They do not have low Apgar scores or acidosis at birth or any abnormal neurological signs in the neonatal period.^19,20 Two major studies of children with cerebral palsy have been reported in which their labour, delivery and neonatal records were in agreement in finding that, for 90%-94%, their disability could not be related to intrapartum hypoxia.^20-22 This does not mean that in 6%-10% of labours hypoxia beginning in labour is the cause of cerebral palsy. In cases in which severe intrapartum hypoxia was documented, it may not have been preventable and earlier delivery may not always have been possible. Pre-existing neurological deficit can contribute to intrapartum hypoxia, or be associated with chronic hypoxia.²³ It is our opinion that the lesions causing cerebral palsy are rarely initiated in labour and are rarely preventable.^24,25 What little evidence exists suggests that less than 2% of cerebral palsy could be attributed to suboptimal intrapartum care.²⁶ It is the opinion of this conference that this figure could be lower.

Causation and fault

To determine causation, Australian courts have had regard to what is known as the but for test, where a plaintiff seeks to establish that, without (or but for) a defendant's breach of duty, he or she would have remained uninjured. It is recognised that applying this test becomes difficult when there are a number of events which may have influenced an outcome.³¹ There must also be an application of common sense to the facts of each particular case.³² Causation will be a very difficult matter to determine in cerebral palsy cases because most are caused by an antenatal condition which often may not have declared itself before or immediately after birth. An absence of major risk factors, such as intrauterine growth restriction, extreme prematurity, twin gestation or known antenatal viral infections, does not exclude congenital or acquired neurological deficit during fetal development.

The Chief Justice of the Australian High Court, Justice Mason, has talked about important conceptual differences between causation in law and causation in science.

. . . [In] science, the concept of causation has been developed in the context of explaining phenomena by reference to the relationship between conditions and occurrences. In law, on the other hand, problems of causation arise in the context of ascertaining or apportioning legal responsibility for a given occurrence . . . a person may be responsible for damage when his or her wrongful conduct is one of a number of conditions sufficient to produce that damage. [However,] . . . it is for the plaintiff to establish that his or her injuries are "caused or were materially contributed to" by the defendant's wrongful conduct. . . . Generally speaking, that causal connection is established if it appears that the plaintiff would have not sustained his or her injuries had the defendant not been negligent.³³

Given our understanding of the rarity of preventable intrapartum causes of cerebral palsy and the difficulty of detecting antenatal causes before labour, it should not be necessary for a defendant to prove the likelihood of neurological abnormality of the fetus before labour or birth. The inability of a defendant to provide such details in retrospect should not allow the assumption that the origins of the cerebral palsy began in labour. It is for a plaintiff to prove the causative link between a putative breach and injury, not for a defendant to prove, in hindsight, the precise antenatal cause and timing of this condition.

The standard of care required of health professionals is " . . . that to be expected by an ordinarily careful and competent practitioner of the class to which the practitioner belongs".^34,35 Whether a medical professional has acted in accordance with a standard of reasonable care " . . . is a question for the Court [to decide] and the duty of deciding cannot be delegated to any profession or group in the community".³⁶ Standards of care will be determined by the courts, not in isolation, but with regard to the evidence of expert witnesses who should give opinion based on scientific evidence.

Conclusions

There is no evidence that current obstetric practices can reduce the risk of cerebral palsy. The origins of many cases of cerebral palsy are likely to be antenatal. While obstetric interventions in the presence of signs of possible hypoxia may prevent fetal death, there is no evidence that they will limit the prevalence or severity of cerebral palsy. The antenatal signs of hypoxia and the methods to monitor hypoxia in labour are still imprecise. This can lead to over-diagnosis of severe hypoxia and, even when correctly diagnosed, early delivery by caesarean section may not change the risk of cerebral palsy. All expert witnesses and the public should recognise that the belief that caesarean section will prevent many cases of cerebral palsy is incorrect. There is a great need for further research into the antenatal origins and the prevention of cerebral palsy. This will include more cerebral palsy registries, improved methods of assessing the development and well being of the fetus throughout pregnancy and labour, and better methods of assessing the neurological integrity of the baby before and after birth.