HAL obo MML v MEC for Health, Free State

The judgment

The court's own file

THE SUPREME COURT OF APPEAL OF SOUTH AFRICA JUDGMENT Reportable Case No: 1021/2019 In the matter between: HAL obo MML APPELLANT and MEC FOR HEALTH, FREE STATE RESPONDENT Neutral citation: HAL obo MML v MEC for Health, Free State (Case no 1021/2019) [2021] ZASCA 149 (22 October 2021) Coram: WALLIS, MOLEMELA and MAKGOKA JJA and ROGERS and UNTERHALTER AJJA Heard: 9 March 2021 Delivered: This judgment was handed down electronically by circulation to the parties’ representatives by email, and by publication on the Supreme Court of Appeal website and release to SAFLII. The time and date for hand down is deemed to be 10h00 on 22 October 2021. Summary: Medical negligence – cerebral palsy – circumstances of brain injury – adequacy of evidence of negligence. Trial – conduct – obligation of parties to define the issues – sequence of witnesses – joint minutes of experts – agreement on facts contrasted with agreement on opinion – approach to agreements on matters of opinion 2 ________________________________________________________________ ORDER ________________________________________________________________ On appeal from: Free State Division of the High Court, Bloemfontein (Daffue J, sitting as court of first instance): The appeal is dismissed with costs of two counsel. ________________________________________________________________ JUDGMENT ________________________________________________________________ Makgoka JA (Wallis JA and Unterhalter AJA concurring) (Majority judgment):

[1]On 1 May 2005 a pregnant Ms HAL (the appellant), then 21 years old, was admitted to the Thebe Hospital (the hospital) in Harrismith, Free State, at approximately 13h00. The following morning, 2 May 2005 at approximately 05h00, she gave birth to a baby boy (MML) by way of a normal vaginal delivery. At the time neither she nor the hospital raised an alarm about his condition, but some considerable time later he showed signs of neurological regression and eventually he was diagnosed with cerebral palsy. A magnetic resonance imaging (MRI) scan of the child’s brain taken in August 2014, a little over nine years later and immediately before the commencement of this action, revealed that he had suffered a hypoxic ischemic encephalopathy (HIE), a brain injury caused by lack of oxygen and lack of blood flow in the brain. It was further confirmed that this was a partial prolonged type brain injury, which develops slowly over 30 to 45 minutes (or longer), and occurs with partial asphyxia. 1 The brain’s response 1 This must be distinguished from an acute profound injury, which is severe and of short duration, with almost total asphyxia, that is, interruption of the supply of oxygen, to the brain resulting in injury to the central deep grey matter in the brain. Its onset is sudden. It is normally caused by a catastrophic sentinel event like a mother falling, etc. 3 is to direct the flow of blood entirely to its central area, thereby depriving the outside portion of blood and oxygen and causing damage there.

[2]On 2 September 2014, the appellant instituted action against the respondent, the Free State Member of the Executive Council (MEC) for Health, in the Free State Division of the High Court, Bloemfontein (the high court). She claimed that MML had suffered the brain injury during the latter stages of the labour and birth process (ie the intrapartum period). She attributed MML’s injury to the negligence of the hospital staff, alleging they did not adequately monitor her and her unborn child, as a result of which they failed to detect foetal distress. This, she alleged, led to MML’s brain injury. The respondent denied liability. After a lengthy trial, the high court dismissed the appellant’s action with the costs of two counsel, but subsequently granted the appellant leave to appeal to this Court.

[3]From its inception, the matter was hampered by the absence of neonatal and obstetric records. The experts who compiled their reports did so on the basis of the limited available records and the appellant’s factual statements. The factual matrix that was before the high court comprised the following: the reports of the experts and in some instances, their oral evidence; the evidence of the appellant and some of the midwives who were on duty at the hospital on 1 and 2 May 2005 or at the clinic she attended after MML’s birth; the limited hospital records in the form of the Maternity Register, the Delivery Register, the Ward and Discharge Summary forms and the Road to Health Chart; the extra-judicial statements made by the appellant to some of the experts during their consultations with her as recorded in their reports; and the appellant’s statements in the form of affidavits in an application for condonation of the late delivery of the statutory notice in terms of s 3(4)(a) of the Institution of Legal Proceedings Against Certain Organs of State Act 40 of 2002 (the condonation application). 4

[4]In the absence of the full hospital records, I deem it necessary to set out in some detail the contents of the available records. But before I do so, I have to address the suggestion that somehow these records were either hearsay evidence or that their contents were disputed. This is simply incorrect. These documents were discovered and placed before the court by the appellant’s counsel during his opening address as the appellant’s merits documents. Whether formally under Rule 36(9) or informally, they appear to have been accepted as having been completed by the persons shown to have compiled them and to be what they purport to be. The Discharge Summary and the Road to Health document had been in the appellant’s possession all along. They were referred to by all the expert witnesses on the basis that they were accurate.

[5]During the course of the trial no reservation was expressed in regard to the accuracy of these documents, although in certain respects the appellant's evidence was inconsistent with them. They featured in the joint minutes agreed between Mrs Bekker and Prof Nolte and between Dr Hofmeyr and Dr Schoon, without any question being raised about the accuracy of their contents. No-one suggested that they had been formulated with a view to putting a favourable gloss on the treatment the appellant received. There is no merit in the contention that their contents were disputed or were inadmissible hearsay, as opposed to being an accurate reflection of what the lost hospital records contained.

[6]Much store was placed on the fact that the principal maternity and obstetric records were missing. This does not assist the case of either party. The records were not available, and we do not know what has happened to them. Whether this was due to incompetence in the administration of the records or something nefarious we cannot say. Nor can any inference, favourable or unfavourable, be drawn from their absence. As I explain more fully in paras 77 and 78 below, the absence of hospital records in the context of this case is a neutral point. There is 5 mutual suspicion by the parties that the other had a hand in the disappearance of the records.

[7]I therefore respectfully differ from the view expressed by my sister Molemela JA in paras 116 to 123 of her judgment, where she suggests that this warrants a ‘charitable approach, which gives cognizance to the plight of the litigant’. The MEC was as handicapped by the absence of the records as the appellant. If there was nothing untoward about MML's birth or his appearance when taken to the clinic thereafter, the MEC's key witnesses, the three midwives, would not be expected to remember the appellant and MML after 13 years. In their evidence they said this was the case. If they had purported to remember and claimed that MML's birth was uneventful and that he appeared normal when observed at the clinic, they would have been accused of fabricating evidence, because after that lapse of time no-one can be expected to remember unremarkable events. They needed to consult records in order to refresh their memory of events.

[8]All that is available are the documents to which reference is about to be made, one of which – the Discharge Summary form – can only have been derived from the entries in the missing record and the remainder of which were original records. The entries in regard to the appellant were consistent with those of other patients who gave birth at the same time and were listed on the same page of the registers. There is no reason to regard them with suspicion. Given that there is no suggestion in these records that anything was amiss in regard to the appellant’s delivery and MML, that is at most an indication that the missing records did not record a problem. However, one cannot go further than that to infer anything about the treatment received by the appellant and MML. 6

[9]In regard to these missing records the appellant’s counsel submitted that an adverse inference should be drawn against the MEC for not calling Mr Rakatsinyane, the keeper of the key to the strong room where the records were stored at Thebe Hospital. Leaving aside the fact that he did not indicate what inference the court should draw from this, it is as well to reiterate that the basis for a court to draw an adverse inference from the failure to call a witness, in accordance with the decision of this Court in Elgin Fireclays v Webb, 2 is that: ‘… it is true that if a party fails to place the evidence of a witness, who is available and able to elucidate the facts, before the trial Court, this failure leads naturally to the inference that he fears that such evidence will expose facts unfavourable to him. See Wigmore (secs. 285 and 286).) But the inference is only a proper one if the evidence is available and if it would elucidate the facts. … [T]the position … was not investigated; he may not have been available as a witness, or he may have seen no more of the occurrence than was testified to by the other witnesses. Consequently, no inference unfavourable to the respondent could properly be drawn.’ Mr Rakatsinyane may have been available, but there was no indication of what he could say about the records other than that they were missing. No inference could be drawn from that. There was nothing in the evidence to suggest that he obviously had something relevant to say and was being shielded from hostile cross-examination.

[10]The same was true of the other witnesses whom it was suggested should have been called and whose absence was said to justify an undefined adverse inference. They were Ms Hlophe who made entries in the registers reflecting that both the appellant and MML were stable on discharge; Dr Matla, who is reflected in the registers as being both the admitting and discharging doctor; 3 Sister Skosana, a general nursing assistant, who was present at the birth; and 2 Elgin Fireclays Ltd v Webb 1947 (4) SA 744 (A) at 749-750; Munster Estates (Pty) Ltd v Killarney Hills (Pty) Ltd 1979 (1) SA 621 (A) at 624B-H. 3 There is an indication that she may no longer be on the register of doctors held by the Health Professions Council of South Africa and that her whereabouts may be unknown. 7 Sisters Moloi and Xaba, who dealt with MML at the clinic on occasions. They too, would not be expected to have independent recollection of the events of 1 and 2 May 2015. Other than the respective entries made by them, in regard to which there was no dispute, they would have had very little to testify about. It would therefore have served no purpose to call them as witnesses. Accordingly, nothing turns on the fact that they were not called as witnesses.

[11]Turning to the records it is convenient to start with the Discharge Summary form and Ward Register. The person who completed these records was not identified, but Sister Mokoena said that they would have been completed from the maternity records. 4 The Ward Admission and Discharge Register showed that the appellant and MML were discharged on 3 May 2005 at 14h35. It was recorded with regard to the child that he weighed 2,9 kg, and measured 51 cm, with a head circumference of 34 cm. BCG and Polio immunisations were administered and the child had been given Konakion and Chloromax shortly after birth. Method of feeding was recorded as ‘breast’.

[12]There is also a section on postnatal advice in the Discharge Summary form, where among other things, the appellant was advised to wait two years before a further pregnancy. She was to visit a clinic on 12 May 2005 for a follow-up examination. According to the Delivery Register, the child’s Apgar scores were 7/10 and 8/10 at one and five minutes after birth, respectively. 5 Counsel submitted that this was hearsay, but all of the medical witnesses relied on these scores and there is no basis for thinking that they were incorrectly inserted in the Delivery 4 The discharge summary appears to have been signed by one Mokoena, but sister Mokoena said it was not in her handwriting and she was not on shift at that time. It may have been another nurse with the same surname, who is shown on the duty roster. 5 APGAR stands for Appearance, Pulse, Grimace, Activity and Respiration. In the test, five factors are used to check a baby’s health. Each is scored on a scale of 0 to 2, with 2 being the best score. For Appearance the skin colour is checked; for Pulse, heart rate; for Grimace, reflexes; for Activity, muscle tone; and for Respiration, breathing rate and effort. The individual scores for the five factors are added up to obtain a score out of ten. The highest score to be achieved is 10 and scores of 7, 8 or 9 out of 10 are normal or good scores. Source: kidshealth.org 8 Register, much less falsified, by the nurse aid who made the entries. Prof Nolte said in her evidence in chief that if one has a Discharge Summary then there must have been preceding records from which the information in the summary was obtained.

[13]There is a section of the Discharge Summary about the mother, in which the following was recorded in respect of the appellant: there were no surgical or obstetric problems, and she looked well. Her pulse, temperature, and breasts were all normal. She had moderate vaginal bleeding and a tear of the perineum. The urinary output was good. She was given vitamin A. It was also recorded that family planning was discussed with the appellant, and the contraceptive method accepted by her was an injection. Also, the importance of breast-feeding was discussed with her and it was recorded as ‘initiated successfully’.

[14]The Road to Health Chart is a document a mother has to present to a designated clinic or hospital for postnatal check-ups. According to the appellant’s chart, for the period June to December 2005, she visited the clinic every month, except for December. In 2006, no visits were recorded for August, October and December. The child seemed to develop normally, up to the age of approximately 18 months. His putting on weight slowed in the following six months, but there may have been other reasons for this. In May 2006 Wormstop, a proprietary remedy for the treatment of worms, was administered and in June the chart records that he had suffered from diarrhoea. In November he was again dewormed. 6 It is not clear whether this was a routine precaution or directed at a specific condition. After 18 months he weighed 8.8 kg. 6 In the course of Prof Nolte’s evidence counsel for the appellant suggested that the relevant entry in the Road to Health document read ‘Dr worried’, but it was, in fact, the more prosaic ‘Dewormed’. 9

[15]No visits were recorded thereafter for the period between December 2006 and April 2007. On the next scheduled visit in May 2007 the child weighed 9,1 kg and it was noted that he was vomiting. He again seems to have been dewormed. There were no recorded visits between June 2007 and October 2007. There was a visit in September 2008, but the note is unintelligible. In November 2007 it was recorded that the child weighed 10 kg. The next entry was in July 2008, in which it was recorded that the child had been coughing and vomiting for about four days. No visits were recorded for August and October 2008. The next scheduled appointment was on 10 November 2008 when the child was three and half years old. It was recorded that the child could not walk or stand and did not crawl. He could get from prone and supine positions to a sitting position. It was further recorded that he had low muscle tone in his legs and high muscle tone in his arms, and that he ate and drank by himself. This was the first record of his having any physical problems.

[16]At various times from about the age of two MML’s condition was investigated, but it is not clear when the diagnosis of cerebral palsy was made, or when this was communicated to the appellant. The appellant told Dr Mogashoa that the problems manifested themselves after about eight months when he started to regress. There is evidence that she was referred to Manapo Clinic but her description of events suggests that as a result of administrative issues nothing came of this. In October 2011 there is a reference in the Road to Health Chart of a physiotherapist saying that the child could not walk and crawled abnormally. The appellant was given exercises for him. When she had a second child in 2012 the hospital records did not contain any reference to MML’s condition or any issues in regard to her previous pregnancy. MML has for some time been attending the Dimakatso Disabled Centre in Ntabazwe. 10

[17]A number of joint minutes were handed in at the outset of the trial. Two of the appellant’s witnesses, Prof Andronikou, a specialist radiologist, and Prof Solomons, a paediatric neurologist, were parties to such minutes but were not called as witnesses. Three expert witnesses gave evidence on behalf of the appellant. They were Prof Nolte, a nursing specialist; Dr Hofmeyr, an obstetrician and gynaecologist, and Dr Gericke, a specialist paediatrician and geneticist. In addition, after they had given their evidence, the appellant testified. The MEC called Mrs Bekker, a midwife; three nurses, Sisters Msibi and Mokoena, who were on duty at the hospital and attended to the appellant and MML, and Sister Mosia, who was the manager of the clinic that the appellant attended with MML; Dr Kamolane, a specialist radiologist and party to the joint minute with Prof Andronikou; Dr Kganane, a paediatric intensivist; Dr Mogashoa, a paediatric neurologist and Dr Schoon, a specialist obstetrician and gynaecologist. Dr Griessel, who had subsequently become unavailable, was a party to one of the joint minutes with Prof Solomons.

[18]Having considered the available hospital records, the joint minutes and the oral evidence adduced at the trial, and the opinions of the various experts, the high court found the appellant to have been a poor witness. It made credibility findings against her on key issues relating to the nature of care, monitoring and birth while in hospital. The high court found ‘internal and external contradictions as well as improbabilities’ in her version, which it held, rendered the opinions of the experts on her behalf, unreliable. The court said: ‘[The appellant’s] version to experts and evidence in court were aimed at taking advantage of missing records. Insofar as the experts relied on her version, and also the missing records, such evidence does not pass the reasonable and logical requirement test for acceptance of their opinions…’ 11

[19]As was the case before the high court, the key issue in the appeal is the likely cause and timing of MML’s brain injury. The argument for the appellant was directed at overturning the judge's factual findings in regard to the reliability of her evidence and the acceptability of the opinions of some of her expert witnesses. Counsel for the appellant assailed the high court’s findings and conclusions on various bases. It was submitted that the uncontested evidence of the appellant, and the agreements concluded between the opposing experts, established that, in all probability, MML’s brain injury occurred in the intrapartum period. It was further submitted that he showed signs of brain injury immediately after birth, which signalled an intrapartum insult. The contention that MML showed signs of a brain injury after birth was based on three propositions: he had breathing problems, did not cry and had to be stimulated by the administration of oxygen using nasal prongs and being injected on his thigh; he could not latch, suck or swallow; and he was taken away from the appellant for five hours. If there were signs of brain injury immediately after MML’s birth that might be decisive of the appeal so I consider them first below.

[20]The appellant testified that: ‘After the child was born, that child never cried. They put two pipes in the child’s nostrils. They also injected the child on the thighs.’

[21]It was common cause during the trial that the child was given an injection shortly after birth. The appellant sought to make much of this injection, testifying that ‘[t]hey just take something like a nail or something, they just put it on there or just stab the child on [the] thighs.’ In the cross-examination of Mrs Bekker, the midwife expert for the respondent, counsel for the appellant asked a hypothetical question as to what medication would be administered to a baby with breathing problems at birth. Her answer was that a medicine called Narcan would be injected into a child’s thighs. In their heads of argument, counsel for the appellant 12 submitted that on the probabilities, ‘the pricking of both thighs could only have been the administration of Narcan …[which] only proves that [the child] must have been in distress at birth.’

[22]Context is important. Earlier in her evidence-in-chief, Mrs Bekker had been referred to the Discharge Summary form, where it was noted, among other things, that MML was given Konakion and Chloromax after birth. Mrs Bekker explained that Konakion is a vitamin K injection, which is routinely administered to babies immediately after birth to prevent any possible intracranial haemorrhage. Mrs Bekker further explained that as an intramuscular injection, Konakion is administered mostly in the upper thigh. There was nothing sinister about the injection administered to MML. In the face of the entry in the Discharge Summary form that Konakion was the injection administered to the child, it is difficult to understand how counsel can tenably insist that the child was injected with Narcan. Chloromax, Mrs Bekker explained, is an antibiotic drop to cleanse the baby’s eyes with sterile water to prevent eye infections. This too, is routine, in case the baby might have contracted an infection during birth.

[23]With regard to the alleged use of nasal prongs (the technical name of the pipes the appellant said were inserted in the baby’s nostrils), there was some uncertainty on the part of sisters Mokoena and Msibi as to whether the hospital had nasal prongs as at May 2005, but it is unnecessary to determine this. Sister Msibi, one of the midwives on duty on 1-2 May 2005, explained that it is sometimes necessary to insert a suction tube into a child’s nostril to suck mucus. If this was done, and Sister Mokoena who attended at the birth was unable in the absence of clinical records to say whether it was, the appellant could possibly have confused the suction with the nasal prongs. However, Sister Msibi gave unchallenged evidence that nasal prongs, if available at all, were only available 13 in the room for sick babies and not the delivery room. The appellant remained in the delivery room throughout, so she could not have seen the use of nasal prongs.

[24]Furthermore, both Sisters Msibi and Mokoena and Dr Hofmeyr, an obstetrician called on behalf of the appellant, explained that if the child had respiratory problems which necessitated resuscitation, a face mask would be used. According to Dr Hofmeyr, while applying oxygen through the nostrils is an alternative, this would have entailed a more prolonged administration of oxygen, and not just a quick intervention to help a baby recover from the trauma of the delivery. The appellant testified that the administration of what she called 'nasal prongs' to the child was for a short duration. On the most favourable view of the appellant's evidence, there was therefore a plausible explanation for both the injection and for the alleged use of nasal prongs. The former was a routine vitamin K injection, while in regard to the latter the appellant possibly mistakenly thought that the mucus suction was an instrument for administering oxygen. The high court cannot be faulted for not accepting the appellant's evidence on these two issues.

[25]These explanations aside, there was an objective basis to test the appellant’s assertion that the child had breathing problems at birth. All the relevant experts – Dr Hofmeyr, Mrs Bekker, Dr Kganane, and Dr Schoon – confirmed that the colour of a healthy baby is pink, which signals that oxygenation is good. A baby deprived of oxygen, according to Dr Kganane, would either be bluish or purplish, and if they remained that way after stimulation, ‘[t]hat would be an unwell child.’ A pale or blue baby is an indication of either infection or problems in breathing, according to Mrs Bekker. Dr Hofmeyr testified that if there is no sign of breathing, no crying or any attempt to make respiration, then it is highly unlikely that the baby would be pink. The 14 significance of a pinkish baby as a sign of good oxygenation was confirmed by the midwives who testified on behalf of the respondent.

[26]The appellant was clear that MML was pink at birth. Based on this, it must therefore be accepted that the child showed a strong indication of good oxygenation and the ability to breathe independently at birth. This, considered in the light of all the factors mentioned earlier, dismantles the appellant’s assertion that the child had breathing problems. It also takes care of a related contention that the child did not cry after birth. Even were it to be accepted that the child did not cry, it does not mean the child could not breathe, given the healthy colour at birth, and all other indications of a normal child, recorded in the available hospital records.

[27]I am also wary of accepting the appellant’s mere say-so on these matters. This was her first child and it was not suggested that she was familiar with either the normal steps a midwife would take after birth, or the type of equipment they might have had readily to hand. Her reference to the nasal prongs is particularly troubling once it is accepted that they were not in the delivery room and were not used. How then did she come to be aware of them, much less give evidence that they were used? The appellant testified 13 years after MML’s birth. She was a young first-time mother. One has to question whether, so long after the event, she would have recalled accurately all the details to which she testified. It is true that her evidence on several points is uncontradicted, but it does not have to be accepted for only that reason. As I demonstrate later, the appellant’s evidence on certain important aspects was found wanting, and this influences how the uncontradicted aspects of her evidence have to be approached.

[28]The appellant also testified that she could not breast-feed the child after birth as he was unable to suck properly. This was presented as being unusual and 15 something other than the ordinary process of a new mother and baby establishing breastfeeding, although Dr Hofmeyr said only that it is ‘in keeping with Neonatal Encephalopathy’, but not specific to it. The appellant’s evidence was contradicted by the contents of the Discharge Summary form, which recorded that the importance of breast-feeding was discussed with the appellant, and had been initiated successfully. If there were problems as testified by the appellant, it is likely that this would have been noted in the form. There is no suggestion that the form was tampered with or that its contents were not a reflection of the true position. It must therefore be accepted as the true and official record that there were no feeding problems worth noting at birth.

[29]Furthermore, there are improbabilities in the appellant’s version. The appellant claimed to have been discharged in the afternoon of 2 May 2005, that is, the day MML was born, but that was inconsistent with the Ward Admission and Discharge Register, which reflected that the appellant and MML were discharged on 3 May 2005 at 14h35. There is no reason to believe that these records were inaccurate, and the further particulars for trial on behalf of the appellant recorded that she was discharged in accordance with the Discharge Register. On that basis more than a day and a half passed before she was discharged. The suggestion in para 107 of the second judgment that there was something odd about this was not supported by the expert witnesses, none of whom said that the appellant remaining in hospital until 3 May was unusual. From a practical point of view, it is highly improbable that the child would have gone for that long without being fed. According to the appellant, this was a complaint she registered with the nurses and the doctor who discharged her, but it was ignored.

[30]The appellant said that the problem of poor feeding persisted after birth but improved over time, although the Road to Health chart showed that MML 16 appeared to gain weight normally. If there had been a serious problem, she would presumably have reported it to the clinic where she attended postnatal examinations. There is no record that she did. In fact, there is no record that the appellant ever reported any breast-feeding problem to anyone after birth, until the report of Prof Solomons was produced some eleven and a half years later. What is more, the appellant apparently told Dr Griessel a different version on this aspect. In his report, Dr Griessel noted that the appellant had told him that the child did not have any problems drinking milk, was breastfed for 18 months, and grew well for the first year. This accords with what was reflected on the Road to Health Chart.

[31]The last matter about which the appellant testified allegedly occurred immediately after birth. According to her the baby was taken away from her to another room, and only returned to her after five hours, ie at 10h00. There is a difficulty with this time period. In the condonation application, the appellant stated in her founding affidavit that the child was brought back to her within a few minutes. This was not corrected in her replying affidavit in that application, unlike her evidence in respect of the method of delivery. In the founding affidavit, she had stated that the child was delivered by caesarean section. In the replying affidavit she corrected this and stated that she gave birth by normal vaginal delivery. The fact that she did not correct the length of time she was separated from the child after birth means that there are two contrary statements by the appellant, both under oath, on the same issue.

[32]In any event, there is nothing sinister in a mother and child being separated after birth. Mrs Bekker explained that this may happen for various reasons. She explained that often the child is taken to another room to be cleaned or to be physically examined. The mother might also need attention, for example to suture an episiotomy. In the present case, it is common cause that the appellant had an 17 episiotomy which had to be sutured. Mrs Bekker also testified that although ideally one does not wish to separate the mother and child, practice at hospitals, including private hospitals, varies. Sometimes the infant will be kept apart from the mother for a day to allow the latter to sleep, only being brought to the mother for feeding. A further difficulty is that even if MML was taken from the appellant for a lengthy period, we do not know the reason for that and there is no reason to assume that it was because of a problem.

[33]Before dealing with the experts’ reports and the agreed minutes, there is one aspect of the evidence that deserves mention. The experts all agreed that MML had suffered a severe injury resulting in a severe impairment of his motor abilities. In the course of her cross-examination, Dr Hofmeyr was asked whether being born with cerebral palsy affects the baby's muscle tone. Her answer was: ‘I can refer to a baby that has probably suffered hypoxic damage because cerebral palsy is a diagnosis that is made much later, as is neonatal encephalopathy, but a baby that is believed to have suffered a hypoxic incident, during birth or prior to delivery, will typically be a floppy baby, as we would, the term we use is floppy, so it would be a limp baby with very little muscle tone. That is the typical description.’ (Emphasis added.)

[34]Neither the appellant, nor the hospital records that are available, suggest that MML presented as a ‘floppy baby’. Mrs Bekker referred to a child with severe encephalopathy as being flaccid. Again, no-one attached that description to MML when he was born, but it is what would be expected if his profoundly damaged condition were occasioned by hypoxic damage during the intrapartum period. Mrs Bekker’s report said that not even mild clinical signs were recorded and that he did not display the signs of hypoxic ischaemic encephalopathy.

[35]Turning now to the opinions of the experts, the parties’ respective radiologists, Prof Andronikou and Dr Kamolane, agreed that the brain injury 18 would probably have occurred in the perinatal period, that is, any time over 37 weeks of pregnancy and up to 30 days after birth. But as to its more accurate timing, and causes thereof, the radiologists deferred to the paediatric and obstetric experts. Prof Solomons and Dr Hofmeyr were, respectively, the experts on behalf of the appellant. Their counterparts on behalf of the respondent were Dr Griessel, a neurodevelopmental paediatrician (later replaced by Dr Mogashoa, a paediatric neurologist) and Dr Schoon, an obstetrician.

[36]According to Prof Solomons’ report, he consulted with the appellant and examined the child on 17 December 2016. It is perhaps opportune to clarify something about his report. In cross-examination the appellant was confronted with certain discrepancies between her evidence and Prof Solomons’ report concerning the factual background she conveyed to him during consultation. The appellant disputed the contents of the report, and denied that she ever consulted with Prof Solomons. The appellant was adamant that the only doctors she had consulted with were the respondent’s experts, Dr Mogashoa and Dr Kganane.

[37]However, in re-examination the appellant confirmed that the first time she consulted a doctor in relation to the matter was on 17 December 2016, although she did not remember the doctor’s name. This date ties in with what appears in Prof Solomons’ report as the date on which he consulted with the appellant and examined the child. One must accept that given the lapse of time, the appellant was mistaken in her denial of consulting with Prof Solomons. It can therefore be safely accepted that Prof Solomons indeed consulted with the appellant and examined the child on 17 December 2016.

[38]Prof Solomons stated in his report that among other things, the appellant had informed him that MML had difficulty in sucking and swallowing after birth. He referred to the American College of Obstetrics and Gynaecology (ACOG) 19 definition of neonatal encephalopathy as a clinically-defined syndrome of disturbed neurological function in the earliest days of life of an infant born after 35 weeks of gestation manifested by a subnormal level of consciousness or seizures and often accompanied by difficulty with initiating and maintaining respiration and depression of tone and reflexes. However, other than the appellant’s report of MML experiencing breathing difficulties requiring nasal prongs to be used, and having sucking and swallowing difficulties when feeding, both of which have been dealt with and rejected above, none of the other identified markers were present. (Emphasis added.)

[39]Dr Kganane used the first edition of the guidelines of the American Academy of Pediatrics (AAP) and the American College of Obstetrics and Gynecology (ACOG). The guidelines set out the following factors, which all have to be present, before a diagnosis of perinatal asphyxia, severe enough to result in an acute neurological insult, could be made: (a) profound metabolic or mixed acidemia (pH <7.00) in umbilical artery blood sample, if obtained; (b) persistence of an Apgar score of 0–3 for longer than 5 minutes; (c) neonatal neurologic sequelae (e.g., seizures, coma, hypotonia); and (d) multiple organ involvement (e.g., kidney, lungs, liver, heart, intestines).

[40]None of those factors were shown to be present in MML. Prof Solomons referred to the more recent 2014 edition of the ACOG guidelines, which identified four ‘essential criteria for an acute intrapartum hypoxic event sufficient to cause cerebral palsy’. It is unnecessary to set these out because of his conclusion that none were present in MML, a conclusion shared by Dr Mogashoa. He also 20 referred to Volpe’s 7 criteria for the diagnosis of an intrapartum insult as being the likely cause of neonatal brain injury, namely: (a) Evidence of foetal distress; (b) Depression at birth necessitating resuscitation; (c) An overt neonatal neurological syndrome during the first hours and days of life. Prof Solomons considered that, given the appellant’s statement that MML received nasal oxygen after birth, he fulfilled the second criterion, ie depression at birth necessitating resuscitation. However, his report showed that once the evidential basis for that fell away, none of Volpe's criteria would be satisfied. Because the medical records were missing, there was no evidence as to the presence or absence of foetal distress.

[41]Prof Solomons’ opinion as it appeared in his report can be summarised as follows. The MRI features are those of chronic evolution of a global insult to the brain due to hypoxic ischemic injury of the partial prolonged variety, most likely occurring at term. The history he received from the appellant was of problematic respiration and of MML receiving oxygen shortly after birth. HIE is one of the possible causes of MML's neonatal encephalopathy. The appellant said to him that MML had a sucking and swallowing abnormality. Apart from this report of breathing difficulties and receiving nasal oxygen, MML did not fit any of Volpe's criteria for an antepartum asphyxia insult or the ACOG Guidelines. Radiologically the evidence supported an injury occurring in the intrapartum period. Prof Solomons expressed the opinion that in the setting of the absence of medical reports and the maternal history indicating neonatal encephalopathy, the partial prolonged hypoxic ischemic injury was timed to the intrapartum period. 7 J J Volpe Neurology of the Newborn, 4 ed, 2001 WB Saunders Company. 21 Expert obstetric and nursing opinion needed to be obtained concerning the monitoring and management of the intrapartum period.

[42]Central to Prof Solomons’ opinion that MML’s brain injury occurred in the intrapartum period, was the appellant’s statement that the child received nasal oxygen and had difficulty in sucking and swallowing. I have already pointed to the difficulty in accepting the appellant’s evidence in regard to those matters. It was not supported by the Discharge Summary form. It also suffered from the inherent improbabilities that I have pointed out. It is also important to note that Prof Solomons did not testify. As a result, his opinion and views were not subjected to the scrutiny of cross-examination. The status of his report is unclear. The pre-trial conference minute did not reflect an agreement that it could be used without Prof Solomons giving evidence. It was part of a bundle of expert witness reports handed in at the commencement of the trial without any indication of their status.

[43]It appears that the appellant's legal representatives were prepared to stand or fall by the contents of the joint minute between Prof Solomons and Dr Griessel, whose report was also placed before the court. The minute departed in certain respects from Prof Solomons’ report. For example, he said that MML’s brain changes were indicative of injury ‘at term’, which was a more definite opinion on timing than that in his report. Dr Griessel said he would defer to the radiologist’s opinion in regard to the timing of the injury. The minute recorded Prof Solomons’ view that in the light of the maternal history of sucking and swallowing abnormality the timing of the partial prolonged HIE to the intrapartum period cannot be excluded. Dr Griessel’s opinion was that the normal growth of MML during the first year of his life made severe feeding difficulties unlikely. 22

[44]Lastly, Prof Solomons recorded in the joint minute that there was no evidence for the injury in the antepartum or postpartum periods. Dr Griessel’s response was that there was no evidence for peripartum injury and that he would defer to the radiologists. On all three points therefore the joint minute reflected disagreement, not agreement. The views and reasoning of these two experts were not canvassed in evidence, because neither of them gave evidence.

[45]Prof Solomons’ other counterpart on behalf of the respondent was Dr Mogashoa, a paediatric neurologist, who also had consulted with the appellant, examined MML, and testified. Before I set out her opinion, I must mention this. Counsel for the appellant sought to exclude her opinion because of the following. Initially, Dr Griessel, a neuro developmental paediatrician, was the respondent’s expert. He became unavailable for the trial and Dr Mogashoa replaced him. At that stage, Prof Solomons and Dr Griessel had already signed a joint minute in which they set out their points of agreement and disagreement.

[46]The two had agreed on the following: that MML suffered a partial prolonged hypoxic ischemic injury; that he had suffered spastic cerebral palsy right more than left with microcephaly and profound intellectual disability; that his motor disability was severe, and that there existed a good correlation between the child’s brain abnormalities and the type of cerebral palsy. However, Prof Solomons and Dr Griessel could not agree on whether MML had a history of sucking and swallowing abnormality, or on the timing of the brain injury to the intrapartum period. While Prof Solomons was of the view that that there was no evidence of hypoxic ischemic injury in the antepartum or postpartum periods, Dr Griessel qualified his assent to this proposition by saying that in the absence of records there was also no evidence for peripartum injury. 23

[47]Subsequently, Prof Solomons and Dr Mogashoa signed a joint minute in which Dr Mogashoa disagreed with some of the issues which Prof Solomons and Dr Griessel had earlier agreed on. Counsel for the appellant argued that Dr Mogashoa was not entitled to do so, and that the respondent was bound by the joint minute between Prof Solomons and Dr Griessel. The issues on which Dr Mogashoa differed from Prof Solomons were: the timing of the injury to the intrapartum period; whether the normal circumference of the child’s head at birth made the antenatal timing of the brain injury unlikely; whether there was a history of poor sucking, and if so, whether it was indicative of an intrapartum injury.

[48]Having done this simple comparison, it seems there is little difference between Prof Solomons' and Dr Griessel’s joint minute, on the one hand, and Prof Solomons' and Dr Mogashoa’s, on the other. Significantly, in both minutes, the key issue in the matter, that is, the timing of the brain injury, was disputed. The issue about the history of poor sucking was a factual one, which I have considered already. The real dispute about Dr Mogashoa's evidence related to her view that the nature of MML’s disability was more consistent with injury in the perinatal period rather than the intrapartum period. She was also of the view that, as his lower limbs were predominantly affected, this was not typical of injury caused by hypoxia. The appellant did not assert any prejudice as a result of Dr Mogashoa’s evidence. She and her legal representatives was aware of Dr Mogashoa’s views long before the trial, and were accordingly not taken by surprise, yet Prof Solomons was not called to refute her views. In addition, Dr Mogashoa was cross-examined at length by the appellant’s counsel.

[49]It is trite that where experts agree on a matter of fact in a joint minute, the parties are bound by the agreement and may not, without more, deviate from the agreement, without proper explanation and the consideration of prejudice. The 24 situation is different here. Dr Mogashoa was not party to the joint minute between Prof Solomons and Dr Griessel. The question is whether the respondent is bound by that agreement to the extent that Dr Mogashoa is prohibited from expressing a different view. In the light of the practical approach I have adopted above, I refrain from expressing any view on this issue.

[50]In her clinical observations of MML, Dr Mogashoa found that his motor signs were normal, with power and a slightly increased tone in the upper limbs. The child could pull himself up using predominantly his upper limbs. There was reduced power and some rigidity in the lower limbs, which also showed brisk reflexes with upgoing plantars. With regard to the contention that the child exhibited sucking problems at birth, Dr Mogashoa stated that, even if this were to be accepted, it could not necessarily be attributed to an intrapartum hypoxia. A history of poor sucking and an MRI performed at nine years, she said, were not sufficient for that conclusion, as there are other causes for poor sucking in neonates.

[51]Based on these views, Dr Mogashoa was of the opinion that the child’s brain injury was not typical of one that occurred in the perinatal period, but was more commonly seen with injury that occurred to the pre-term brain. Thus, she concluded, the injury was unlikely to have been caused by an intrapartum hypoxia. Like Prof Solomons, Dr Mogashoa referred to the AAP and ACOG guidelines, and likewise, found that the child satisfied none of the criteria for an intrapartum insult as being the likely cause of the brain injury. She also considered the following objective facts as pointing further away from an intrapartum hypoxia: MML was well enough to be discharged a day after birth; he had normal Apgar scores; he did not lose more weight than is to be expected in the first week of life; and the pattern of his regression (being able to sit at four months and leopard-crawling soon thereafter, but then experiencing a regression 25 after more than a year) was not typical of an intrapartum hypoxia. Dr Mogashoa also disagreed with Prof Solomons’ conclusion that because the child’s birth head circumference was within normal limits, any brain injury in the perinatal period prior to admission was unlikely.

[52]Although Dr Mogashoa sought to argue that the injury probably happened earlier than the perinatal period, she acknowledged that the radiologists were best qualified to interpret the MRI scans, and ultimately, she unconditionally agreed with the radiologists, save for her point that a ‘term brain’ for radiology purposes was at 35 weeks rather than 37 weeks. It is not known exactly how long the appellant’s gestation was at the time of birth. But assuming it to have been 40 weeks, there would have been at least five weeks prior to her admission to hospital when a partial prolonged asphyxia may have occurred. Viewed in this light, there is some force in Dr Mogashoa’s evidence that the insult could have occurred in the perinatal period prior to the appellant's admission to hospital, given the relatively good Apgar scores, the generally good presentation at birth, the pattern of late regression, and the atypical feature of spastic diplegia (affecting the lower limbs rather than the upper limbs) rather than spastic quadriplegia or dyskinesia which one normally associates with intrapartum brain injury.

[53]When dealing with the evidence of experts in a field where medical certainty is virtually impossible, a court must determine whether and to what extent their opinions advanced are founded on logical reasoning. The court must be satisfied that such opinion has a logical basis, in other words that the expert has considered comparative risks and benefits and has reached ‘a defensible 26 conclusion.’ 8 An opinion expressed without logical foundation can be rejected. 9 Having carefully considered the totality of the evidence of the two experts, I am of the view that Dr Mogashoa’s evidence provided the most reasoned and cogent explanation of why an intrapartum brain injury was not likely in this case. Prof Solomons’ opinion was thinly reasoned, relying as it did, on the questionable statement of the appellant. In addition, Dr Mogashoa’s opinions were subjected to an arduous, if hostile, cross-examination, during which she stood her ground. She made a favourable impression on the high court. Overall, Dr Mogashoa’s opinions offered the most rational explanation for MML’s condition and closely matched the objective facts.

[54]I turn now to the timing of the brain injury. In their joint minute, Prof Solomons and Dr Mogashoa deferred to obstetricians on this issue. Dr q Hofmeyr was that expert on behalf of the appellant. Her counterpart on behalf of the respondent was Dr Schoon. Dr Hofmeyr had not consulted with the appellant. In compiling her report, she relied on the available hospital records, Prof Andronikou’s radiologist’s report, Prof Solomons’ report; and a ‘factual statement’ by the appellant dated 31 August 2017’. The latter statement was not made available to the court, its origin was never explained and the appellant denied in cross-examination ever making such a statement.

[55]In order to place the brain injury in the intrapartum period, Dr Hofmeyr sought to establish that there was foetal distress and prolonged labour. Foetal distress is a sign that an unborn baby is not well. It happens when the baby does 8 Michael & Another v Linksfield Park Clinic (Pty) Ltd and Another 2001(3) SA 1188 (SCA); [2002] All SA 384 (A); paras 36-37 where this Court adopted the decision of the House of Lords in the medical negligence case of Bolitho v City and Hackney Health Authority [1997] UKHL 46; [1998] AC 232 (HL(E). 9 Medi-Clinic Limited v Vermeulen [2014] ZASCA 150; 2015 (1) SA 241 (SCA) para 5. 27 not receive enough oxygen through the placenta. How it should be monitored, and the frequency of such monitoring, are among the many guidelines set out in the 2002 National Guidelines for Maternity Care (the maternity guidelines), which define levels of care, obstetrical record keeping; and antenatal care, among others. Frequently, foetal distress is monitored by a CTG. The partogram is a document in which the foetal and maternal information and the progress of labour are all recorded. According to the maternity guidelines, failure to use and complete a partogram during labour constitutes ‘substandard care’.

[56]With reference to this case, and in particular whether there was evidence of foetal distress, Dr Hofmeyr mentioned that in the absence of antenatal and obstetric records, she was unable to accurately evaluate or comment on foetal heart rate patterns or the adequacy of foetal monitoring. Nevertheless, she concluded: ‘[T]he [appellant’s] statement …reflects inadequate [foetal] heart rate monitoring in direct contrast to prescribed minimum standards of care. The [foetal] heart rate should have been monitored every 30 minutes during the active phase of labour but was only performed once during the (more than) 14-hour admission, through the initial admission CTG. Without regular [foetal] heart rate monitoring attending staff would not be able to detect foetal distress depicting a possible intra-partum event.’ (Emphasis added.)

[57]With regard to the child’s brain injury and its timing, Dr Hofmeyr relied on Prof Andronikou’s initial (August 2014) opinion, that the brain injury had most likely occurred at term. Dr Hofmeyr accepted that this did not necessarily limit it to the intrapartum period. She conceded that it could have occurred in the broader peri-partum period before the onset of labour, or even after birth. This is indeed more in line with the agreement between the radiologists referred to above. In this regard, Dr Hofmeyr considered that the appellant was a low-risk patient with an uncomplicated antenatal course; the appellant had a spontaneous onset of 28 labour and that the admission CTG was reassuring. She concluded that the most likely timing of the brain injury was ‘during the labour and/or delivery…’ She based her conclusion on the following: the alleged absence of further foetal and maternal monitoring throughout the labour; and ‘probable prolonged labour’.

[58]Dr Hofmeyr was supported in these conclusions by Prof Nolte. She, like Dr Hofmeyr, had not consulted with the appellant. In compiling her report, she relied on Prof Solomons’ report and the ‘consultation records’ of the appellant. She said that ‘[i]t seemed as if no observations, except the cervical dilatation, was done between 12:00 and [2 May 2005] at 04:00’. She concluded in her report that the midwives who cared for the appellant during her labour delivered sub- standard care because, according to her, they failed to: monitor foetal and maternal condition or to record observations thereof; refer the appellant to a doctor ‘when there was prolonged labour’; and keep accurate records of the case.

[59]Thus, Dr Hofmeyr and Prof Nolte both relied on inadequate foetal and maternal monitoring, as well as prolonged labour, for their respective opinions. Their opinions on these are based entirely on the appellant’s statements availed to them for their respective reports – the appellant’s undisclosed ‘factual statement’ dated 31 August 2017 used by Dr Hofmeyr, and the appellant’s undisclosed ‘consultation notes’ used by Prof Nolte. The contents of both documents, insofar as inadequate monitoring is concerned, are consistent with the appellant’s evidence on the issue. In her evidence, the appellant testified that CTG monitoring was done only once on admission, and never again. The appellant further testified that apart from cervical examinations, no other examinations were made on her. In the absence of hospital records, the reliability of the experts’ opinions, rest on the reliability and acceptability of the appellant’s uncontradicted evidence. 29

[60]Despite the fact that the appellant’s evidence was uncontradicted, it is not without difficulties. I have already discussed the problems with it in regard to the use of nasal prongs and the ‘pricking’ of the baby's thighs. There are other instances where the appellant’s evidence was inconsistent with what she had conveyed to the experts who consulted with her. For example, in respect of CTG monitoring, the appellant testified that this was done 20 minutes after admission, which must have been at approximately 13h20. This was consistent with Dr Hofmeyr’s report. But according to Dr Mogashoa’s report, the appellant conveyed to her that CTG monitoring commenced at 15h00. To Dr Kganane, the appellant mentioned that at 18h00 ‘belts were put on her to monitor the baby.’

[61]Similar problems in relation to the events of that day arise in relation to what she was told by the nurses about the progress of her labour. She testified that she was told after being examined on admission that she was 3 cm dilated and the baby was ‘high’. In his report, Dr Kganane mentioned that the appellant had informed her that this was conveyed to her at 20h00. Dr Kganane testified that there were no language barriers between her and the appellant as they communicated in the appellant’s home language. The disparity between these is a further indication that with the elapse of time the appellant's memory of events and their timing was not necessarily reliable.

[62]The same problem crops up with the appellant's evidence in regard to when she was examined by the nurses during the course of her labour. She testified that she was examined at the following intervals on 1 May 2005: 13h00, 13h20, 18h00, 20h00 and midnight. In his report, Prof Solomons recorded that the appellant was examined at 18h00, 20h00 on 1 May 2005, and again at 01h00, 02h00 and 04h00 the following morning. Dr Kganane mentioned in her report ‘at midnight the pain worsened but no examination was done.’ This is a further indication that the appellant's memory was not reliable, which as the trial judge 30 pointed out was hardly surprising given the passage of time between the events in question and her having to recall them. It is not a reflection on her honesty as a witness, but on the reliability of the information on which the experts based their reports in relation to her treatment.

[63]The appellant’s evidence of inadequate monitoring, on which the opinions of Dr Hofmeyr and Prof Nolte rest, is doubtful, given the contents of the available hospital records, which seem to have been completed fairly carefully and comprehensively. As the court a quo correctly pointed out, the appellant testified that contemporaneous notes were made by the nursing staff that examined her. It is not clear on what basis Prof Nolte limits these observations to only cervical dilation. All of the available records could only have been completed with reference to other records, not only those recording cervical dilations. With reference specifically to the Discharge Summary form, Mrs Bekker emphasised that its existence suggested that the relevant records were completed, as it formed part of the Maternity Case Record. These could not have been completed without proper monitoring and proper recording of the observations. Prof Nolte described the records of the case as inadequate, but that was plainly incorrect as the records were not available. A judgment on their adequacy could only be made if they had been produced, but that was not possible because they have in some way gone astray.

[64]Apart from the difficulties pointed out in respect of the appellant’s evidence on inadequate monitoring, the appellant had shown herself to be an unreliable witness on a number of other issues. On almost every issue where there was some form of record, her oral evidence was not borne out by it or contradicted it. Her evidence also had internal contradictions and improbabilities. The following are examples: 31 (a) In cross-examination, the appellant disavowed the very basis of Dr Hofmeyr’s opinion – her own factual statement dated 31 August 2017. (b) In cross-examination, the appellant denied that she ever consulted with Prof Solomons. (c) The appellant testified that she consulted her attorney for the first time on 31 August 2017. In the condonation application she stated that she first consulted her attorney during May 2014. A letter giving notice under the Act was sent on 30 June 2014. As the summons was issued in August 2014 ten days after MML had undergone the MRI scan there must have been some contact with the attorneys at around that time, but how it occurred is unclear. (d) The appellant testified that she and the child were discharged on 2 May 2005, in contrast to the Discharge Summary form, which showed the date of discharge to be 3 May 2005. (e) The appellant was untruthful about her HIV status to her attorney and to Dr Gericke, during May 2017 and August 2017, respectively, to whom she claimed to be HIV negative, whereas she knew in 2011 when her second child was born, that she was HIV positive. (f) Regarding the times at which she was examined whilst in hospital, the appellant gave specific and precise times in her evidence-in chief. However, the unreliability of her memory was exposed when she was tested in cross- examination with regard to the birth of her second child on 16 July 2012. For example, she testified that the ambulance was summoned to her home at 03h00 and arrived at 03h30. However, the Maternity Case Record showed that the ambulance was summoned at 04h15 and at 04h51 and arrived at her home at either 05h21 or 05h23. (g) Asked in cross-examination how she was able to recall the exact times about events which had happened 13 years earlier, the appellant explained that there was a watch on the wall in the room she was in, and that she was always checking the time. However, later, when questioned how she remembered the time when 32 the child was brought back to her after being taken away shortly after birth, she testified that she noted the time on her cell phone. The improbability of her having a precise recollection of these times is manifest. (h) With regard to the antenatal period and how her pregnancy was confirmed, Prof Solomons mentioned in his report that the appellant had performed a home pregnancy test. The appellant denied this in her evidence, and testified that her pregnancy was confirmed by a urine test done at a local clinic. (i) The appellant testified that she was told by a nurse at 16h00 not to lie in bed but to walk around. She gave Dr Mogashoa and Dr Kganane two different versions of this. According to Dr Mogashoa’s report, the appellant was advised to lie on her side shortly after her admission at 13h00. According to Dr Kganane’s report, this only happened at 21h30. (j) Dr Hofmeyr testified that when the appellant screamed for help on various occasions, she was ignored. This information, must have been contained in the appellant’s factual statement. It is in direct contrast to the appellants’ own evidence, in terms of which a nurse came to her assistance and examined her each time she screamed for help.

[65]These inconsistencies, and the others I have pointed out earlier, are material, and justify the high court’s finding that the appellant was not a reliable witness. Giving two different versions in evidence about the same occurrence is not a sign of a reliable witness. Nor did the inconsistencies in regard to the dates of key events in the litigation occurring between 2014 and 2018, give one any confidence in her ability to recall accurately the precise times of the events of 1 and 2 May 2005.

[66]I must emphasise that it is not suggested that the appellant was not a credible witness. Rather, she was an unreliable witness. There are conceptual differences between credibility and reliability, which should not be conflated. 33 Credibility has to do with a witness’s veracity. Reliability, on the other hand, concerns the accuracy of the witness’ testimony. Accuracy relates to the witness’ ability to accurately observe, recall and recount events in issue. Any witness whose evidence on an issue is not credible cannot give reliable evidence on the same point. Credibility, on the other hand, is not a proxy for reliability: a credible witness may give unreliable evidence. 10 The passage of time that we have in this case of over 13 years from the relevant events impacts on the reliability of the witness's evidence not her credibility. 11 The unreliability of the appellant’s evidence is underscored by the fact that, where relevant, her purported recollection of dates, times and events is not supported by the objective evidence. Seen in this light, the dictum in President of the Republic of South Africa and Others v South African Rugby Football Union and Others, 12 which has to do with credibility, does not find application here.

[67]It is regrettable that these inaccuracies could not be put to the appellants’ experts, as all of them testified before she did. One will never know how they would have impacted their opinions and evidence. It would have been particularly awkward for Dr Hofmeyr to learn that the very statement she had premised her opinion on, had been disavowed by its supposed source – the appellant. The net effect of all the above is that the factual basis on which Dr Hofmeyr and Prof Nolte concluded that there was inadequate monitoring, was shown to be unreliable.

[68]On the balance of probability, the appellant’s latent phase must have started mid-morning on 1 May 2005, at least by the time she was collected by the ambulance around noon. Her evidence was that the mucus plug had been 10 R v H C 2009 ONCA 56; 241 C.C.C.(3d) 45 para 41. 11 R v Morrissey 1995 CanLII 3498 (ONCA); 22 OR (3d) 514 at 526. 12 President of the Republic of South Africa and Others v South African Rugby Football Union and Others [1999] ZACC 11; 1999 (10) BCLR 1059; 2000 (1) SA 1 (CC) para 62. 34 discharged before she was collected by the ambulance, and that her waters broke in the ambulance. She testified that shortly after her admission she was told by the nurse that her cervix was 3 cm dilated. To this may be added that she was admitted to the maternity ward as being ‘in labour’, and not turned away as a person whose labour had not yet started. On that basis, I accept that by the time she was admitted to hospital her latent phase had commenced. But the duration of the latent phase is not scientifically fixed, nor was there evidence that the cervix dilates at a regular rate during the latent phase.

[69]There is evidence that the risk of hypoxia during the latent phase is not high because the contractions are weak. The appellant may have had weak, and not yet regular, contractions for some hours after her admission. Dr Hofmeyr said in her report that the latent phase of labour ‘commonly lasts about 8 hours’ and requires the patient to be monitored at four hourly intervals. On the basis of her ward admission at about 13h00 latent labour lasting to 21h00 would be normal. The appellant said that when examined during this time, which occurred twice in accordance with the guidelines, the nurses told her that the baby was ‘far’, so a latent labour after this time is likely. Assuming active labour commenced at shortly before, or about, midnight, when she said that a nurse came in response to her calls, neither the latent labour nor the active labour appears unduly protracted.

[70]If the appellant had a conventional active phase of about six hours, her active phase could not have started much before 23h00 on the Sunday night. Was there evidence that the appellant started the active phase earlier than 23h00? According to Dr Hofmeyr the most important indicator of the active phase is that the mother has three strong contractions every 10 minutes coupled with a cervical dilation of 4 cm at the commencement of the active phase. Prof Nolte said that if contractions were not severe, they were not regular. The glaring lacuna in the 35 present case is that it is not known when that started. The appellant was never asked about her contractions. In short, we do not know that the appellant did not have a slightly longer than normal latent phase followed by a normal active phase. All we know is that the appellant said she called for assistance because she was in pain at about midnight.

[71]With the unreliability of the appellant’s evidence, the opinions expressed by the experts on her behalf were found by the high court to be speculative. The high court put it as follows: ‘I mentioned supra that the conclusions of [appellant’s] experts in respect of the probable timing of the injury - to wit an intrapartum injury - are not based on a proper factual foundation. Although the experts generally made a good impression on the court, opinions based on incorrect facts and/or speculation are to be ignored. If the factual foundation is proven to be baseless, an opinion falls apart like a house of cards. In casu we have a witness – the appellant - whose credibility and reliability are in tatters. I have shown the internal and external contradictions as well as improbabilities in her version. Her version to experts and evidence in court were aimed at taking advantage of missing records. Insofar as the experts relied on her version, and also the missing records, such evidence does not pass the reasonable and logical requirement test for acceptance of their opinions as mentioned in Oppelt supra. I am not convinced of the cogency of the underlying reasoning of [the appellant’s] experts.’

[72]The presumption is that a trial court’s factual findings are correct in the absence of demonstrable error. To overcome this presumption, an appellant must convince the appellate court on adequate grounds that the trial court’s factual findings were plainly wrong. If the appellate court is merely left in doubt as to the correctness of a factual finding, then it will uphold that finding. It is only in exceptional circumstances that an appellate court will interfere with the trial court’s evaluation of oral evidence, in the light of the advantages enjoyed by the trial court of seeing, hearing and appraising the witnesses. See Sanlam Bpk v 36 Biddulph 2004 (5) SA 586 (SCA) para 5; Roux v Hattingh [2012] ZASCA 132; 2012 (6) SA 428 (SCA) para 12.

[73]In the present case, the high court’s conclusion that the appellant was not a reliable witness given the inconsistencies and contradictions in her version of events, was undoubtedly correct. This had a devastating impact on the opinions of Prof Solomons, Dr Hofmeyr and Prof Nolte, upon which appellant’s case rested. In all the circumstances, the substratum of the appellant’s case – the timing of the child’s brain injury to the intrapartum, unravelled when the appellant’s factual foundation floundered.

[74]There was another string to the appellant’s bow. Dr Gericke, a specialist paediatrician and geneticist on behalf of the appellant, introduced the possibility of delayed manifestations of an intrapartum brain injury. He placed reliance upon a passage in an article by two Dutch authors, Ms LS de Vries and Ms F Groenendaal. 13 There, the authors discuss what they describe as a ‘watershed predominant pattern of injury.’ The authors say the following (references to the footnotes and figures omitted): ‘Watershed predominant pattern of injury (WS) is the other pattern of injury which is also referred to as a pattern seen following ‘prolonged partial asphyxia’. The vascular watershed zones (anterior-middle cerebral artery and posterior- middle cerebral artery) are involved, affecting white matter and in more severely affected infants also the overlying cortex …The lesions can be uni-or bilateral, posterior and/or anterior. Although loss of the cortical ribbon and therefore the grey-white matter differentiation can be seen on conventional MRI, DWI highlights the abnormalities and is especially helpful in making an early diagnosis. A repeat MRI may show cystic evolution, but more often atrophy and gliotic changes will be recognised It is also more common after hypotension, infection and hypoglycaemia all of which may be associated with a more protracted course. Neurological manifestations at birth may be mild 13 L S de Vries and F Groenendaal ‘Patterns of Neonatal Hypoxic-Ischaemic Brain Injury’ (2010) 52(6) Neuroradiology 555. 37 and do not always meet the perinatal asphyxia criteria and onset of neurological signs can be delayed. Severe motor impairment is uncommon in this group of infants, and they are often considered to have an early normal outcome, when seen at 12-18 months. When seen up till early childhood sub-optimal head growth, behavioural problems and delay in language are, however, common. Miller et al were first able to recognise cognitive deficits associated with the watershed pattern of injury at 30 months, while the problems were largely overlooked, when seen at 12 months. More recently, they also showed a correlation with verbal IQ at 4 years of age. Symptomatic parieto-occipital epilepsy may occur later in childhood, often associated with reduced intelligence quotients and visuospatial cognitive functions.’ (Emphasis added.)

[75]It was common cause in the present case that MML had suffered a severe motor impairment. Therefore, the statement that ‘severe motor impairment is uncommon in this group of infants’ automatically excluded the appellant’s child. It was put to Dr Gericke in cross-examination that because of this, the child could not be used as an example of a child within the category discussed in the article. Dr Gericke had considerable difficulty in dealing with this. He explained, rather unconvincingly, that the final call to make an assessment whether the child has a brain injury is five years old and that if a brain injury is missed at two years old, this would bring the child within the category, as ‘undetected’. The fallacy of his argument was exposed by a hypothetical question of a child whose brain injury had manifested at six months, which is the case with the appellant’s child, which would take the child out of the stated category. Dr Gericke would blame ‘the health system’ for not picking up the manifestations early in the first year.

[76]In any event, even if all the suppositions propounded in the article are accepted, the views expressed by the authors still would not assist to answer the central issue in this case, ie the timing of the brain injury. The article is premised on children in respect of whom an intrapartum brain injury is common cause. That it is not the position here. In the end, there is not much value in this article. 38

[77]Finally, it was submitted that the respondent had failed to ensure proper safe-keeping of the appellant and the child’s hospital records in breach of a statutory duty. 14 Therefore, so went the submission, this warranted this Court to regard this as an exceptional case warranting the application of the res ipsa loquitur maxim, and find negligence on the part of the hospital staff on the mere presence of the brain injury. Reliance was placed on the majority decision in Meyers v MEC, Department of Health, Eastern Cape. 15

[78]It is necessary to set out the following factual background about the missing hospital records. The facts are distilled from the condonation application, which the respondent opposed. In the answering affidavit in that application, Mr Monyane, the Legal Administration Officer in the Department of Health, Free State (the Department) made the following averments: There were, at that stage, six cases of children with cerebral palsy, born at various hospitals in the province, and in respect of whom claims for compensation had been instituted against the MEC, for alleged negligent conduct of doctors and nurses during birth. This included the appellant’s claim. In all of the claims, legal action had only been instituted many years after the births, and the original hospital files were missing. All of the claims were instituted by the appellant’s attorneys of record, MED Attorneys (formerly known as Mokoduo Incorporated) on behalf of the claimants. In four of the cases, copies of the claimants’ missing hospital records were received from the said attorneys. 14 Section 13 of the National Health Act 61 of 2003 (NHA) provides that subject to the National Archives of South Africa Act 43 of 1996, and the Promotion of Access to Information Act 2 of 2000, the person in charge of a health establishment must ensure that a health record containing such information as may be prescribed is created and maintained at that health establishment. Section 17(1) of that (NHA) enjoins a person in charge of a health establishment in possession of a user’s health records to set up control measures to prevent unauthorised access to those records and to the storage facility in which, or system by which, records are kept. 15 Meyers v MEC, Department of Health, Eastern Cape [2020] ZASCA 3; [2020] 2 All SA 377; 2020 (3) SA 337 (SCA). 39

[79]As a result, Mr Monyane caused a letter to be directed to MED Attorneys requesting them to explain their possession of the missing hospital records. There was no response to Mr Monyane’s letter. An internal investigation revealed that the attorneys had never requested the said copies, or paid for them. It was therefore ‘a complete mystery’ to the Department as to how MED Attorneys had obtained possession of the copies of the missing hospital records. Dr Schoon, an employee of the Department and an expert witness in this case, testified that it was suspected that the missing files had been removed and sold by corrupt individuals in the Department. The fraud unit of the Department was investigating this.

[80]The second judgment states that ‘it is not inconceivable that a healthcare professional who becomes aware that his or her negligent acts might be questioned, may be motivated to spoliate the patient records so as to conceal his or her negligence.’ This is true. But the same can be said of a plaintiff or legal practitioner who, aware of the weakness of an intended claim, surreptitiously removes the hospital records to conceal that which reveals the weakness of the claim. In the light of the facts of this case, the absence of the hospital records is a neutral factor. It does not establish that the hospital staff who treated the appellant were negligent. Put differently, in this case, negligence, cannot, without more, be inferred from the absence of the hospital records.

[81]The application of the res ipsa loquitur maxim is not appropriate in this case. There is no evidence of what caused the child’s brain injury and when it occurred. In Van Wyk v Lewis 16 this Court cautioned that the maxim should rarely, if ever, find application in cases based on alleged medical negligence, where it has not been established what went wrong, and where the views of 16 Van Wyk v Lewis 1924 AD 438 at 462. 40 experts are all based on speculation – giving rise to various but equally feasible possibilities – as to what might have resulted in the injury being sustained. This is such a case. 17

[82]The general rule is that he who asserts must prove. As Innes CJ explained in Van Wyk v Lewis, 18 the question of onus is of capital importance. A plaintiff who relies on negligence must establish it. If at the conclusion of the case the evidence is evenly balanced, a plaintiff cannot claim a verdict; for he or she will not have discharged the onus resting upon him or her. While true that an intrapartum injury cannot be excluded in this case, both antenatal and postnatal injuries cannot be excluded either. Nor is any one of them more probable than any other. As such, an intrapartum injury is not the most plausible inference to be drawn from the proven facts.

[83]There are many proven and objective facts that point to MML’s brain injury as not being typical of an intrapartum one. Some of the important indicators are: a seemingly healthy child at birth, being pinkish in colour; the normal Apgar scores; the available hospital records which show that the child was well enough to be discharged a day after birth and that breast-feeding was initiated successfully; and the child’s normal growth until at least 18 months. There is also academic literature referred to by the experts, such as the ACOG guidelines and Dr Volpe’s textbook, which set out criteria to determine when a brain injury can be deemed to have occurred in the intrapartum period. MML did not fulfil any of the criteria and displays atypical features in the form of spastic diplegia affecting the lower limbs rather than the upper limbs, rather than spastic 17 See also Buthelezi v Ndaba [2013] ZASCA 72; 2013 (5) SA 437 (SCA) para 16. 18 Op cit, fn 16, at 444-445. 41 quadriplegia or dyskinesia, which is normally associated with intrapartum brain injury.

[84]Under these circumstances, I agree with the high court that it would be purely speculative to conclude that the child’s brain injury was suffered in the intrapartum period. In sum, the appellant had failed to establish the negligence of the hospital staff in the respects she alleged. Even if one assumes in the appellant’s favour that negligence on some basis has been established, the appellant would still have difficulty to establish causation. The evidence in this case is that a partial prolonged episode lasts at least 30-45 minutes. There is no evidence that regular monitoring, coupled, if necessary, with a prompt Caesarean section, would have resulted in the delivery of MML before the injury was suffered. The high court was therefore correct to dismiss the appellant’s claim. In the result the appeal must fail. Costs must follow the result.

[85]Accordingly, the following order is made: The appeal is dismissed with costs of two counsel. _____________ for T Makgoka Judge of Appeal Molemela JA (Rogers AJA concurring) (dissenting)

[86]I have read the judgment of my colleague, Makgoka JA (the first judgment). Regrettably, I am unable to agree with its reasoning and conclusion. My disagreement primarily pertains to its endorsement of the high court’s findings. For reasons that will become evident, I respectfully hold the view that the final ruling of the high court was premised on an erroneous application of the rules of engagement relating to a trial, especially in relation to the evaluation of evidence. 42

[87]Although the powers of appellate courts to overturn credibility findings made by a trial court are restricted, it is trite that where the findings of a trial court are based on false premises, or where relevant facts have been ignored, or where the factual findings are clearly wrong, the appeal court is entitled to reverse them. 19 Equally well-established is that with the benefit of a full record, a court of appeal can sometimes be in a better position to draw inferences. 20

[88]Since I take issue only with the evaluation of the high court’s analysis of the evidence, there is no need for me to traverse the factual ground already covered in the first judgment. Thus, facts will be repeated only to the extent that they are necessary to articulate the reasoning that forms the basis of my dissension.

[89]One of the trite principles of our law is that every case must be decided on its own merits. As regards the fact that the claim on behalf of MML was instituted nine years after his birth, a crucial consideration is that it was always open to the appellant to lodge MML’s claim until such time as he attained the age of majority. 21 Her averment regarding how she became aware that she could lodge a claim on behalf of the MML is reminiscent of the following observation made by the Constitutional Court in Mohlomi v Minister of Defence: 22 ‘That disparity must be viewed against the background depicted by the state of affairs prevailing in South Africa, a land where poverty and illiteracy abound and differences of culture and language are pronounced, where such conditions isolate the people whom they handicap from the mainstream of the law, where most persons who have been injured are either unaware of or poorly informed about their legal rights and what they should do in order to 19 R v Dhlumayo 1948 (2) SA 677 (A) 705-706; Santam Bpk v Biddulph [2004] 2 All SA 23 (SCA); 2004 (5) SA 586 (SCA) para 5; RB v Smith [2019] ZASCA 48; 2020 (4) SA 51 (SCA). 20 Minister of Safety and Security and Others v Craig and Others [2009] ZASCA 97; [2010] 1 All SA 126 (SCA); 2011 (1) SACR 469 (SCA) para 58. 21 It must be borne in mind that all private and public health-care facilities are enjoined to retain patient records relating to minors until they reach the age of 21 years. See the guidelines of the Health Professional Council of South Africa 2009 para 9. 22 Mohlomi v Minister of Defence 1996 (12) BCLR 1559; 1997 (1) SA 124 (CC). 43 enforce those, and where access to the professional advice and assistance that they need so sorely is often difficult for financial or geographical reasons.’ 23 I echo the same sentiments. It is simply not hard to fathom why an indigent, ill- informed mother who has given birth to a child with severe disabilities would delay litigation. Unjustified credibility findings

[90]The salient findings of the high court are set out in paragraphs 108 – 116 of its judgment. It was contended on behalf of the appellant that the high court’s credibility findings are not borne out by the record. Having gone through the record, I can only agree with that contention. The reasons for this view are set out below. The proper test for evaluating a witness’ testimony is not whether a witness is truthful or indeed reliable in all that he or she says, but whether on a balance of probabilities, the essential features of the story which he or she tells are true. 24 Courts engaging in the analysis of evidence adduced in a trial must be careful not to fall into the trap of evaluating it in a piecemeal fashion; rather, the mosaic of the evidence that was adduced, must be considered as a whole. 25

[91]It is important to bear in mind that the credibility of witnesses and the probability of what they say should not be regarded as separate enquiries to be considered piecemeal, as they are part of a single investigation into the acceptability or otherwise of the appellant’s version. 26 In that investigation, the importance of any discrepancies or contradictions is assessed. The story presented by a litigant ‘is tested against facts that cannot be disputed and against the inherent probabilities, so that, at the end of the day, one can say with conviction that one 23 Ibid para 14. 24 Santam Bpk v Biddulph [2004] 2 All SA 23 (SCA); 2004 (5) SA 586 (SCA) paras 10 and 13. 25 S v Shilakwe [2011] ZASCA 104;2012 (1) SACR 16 (SCA) para 11. 26 Mabona and Another v Minister of Law and Order and Others [1988] 3 All SA 408 (SE); 1988 (2) SA 654 (SE). 44 version is more probable and should be accepted, and that therefore, the other version is false and may be rejected with safety’. 27

[92]In S v Mkohle, 28 this Court held that not all contradictions affect a witness’ credibility. The court cautioned that in each case, the trier of fact has to make an evaluation, taking into account such matters as the nature of the contradictions, their number and importance and their bearing on other parts of the witness’ evidence. In my opinion, the credibility findings made against the appellant were not justified. There were no material contradictions in her evidence. I am of the view that the cardinal rules of cross-examination were not observed. These were reaffirmed by the Constitutional Court in President of the Republic of South Africa and Others v South African Rugby Football Union and Others, 29 as follows: ‘. . . As a general rule it is essential, when it is intended to suggest that a witness is not speaking the truth on a particular point, to direct the witness’s attention to the fact by questions put in cross-examination showing that an imputation is intended to be made and to afford a witness an opportunity, while still in the witness-box, of giving any explanation open to the witness and of defending his or her character. If a point in dispute is left unchallenged in cross- examination, the party calling the witness is entitled to assume that the unchallenged witness’s testimony is accepted as correct. . . . The precise nature of the imputation should be made clear to the witness so that it can be met and destroyed, particularly where the imputation relies upon inferences to be drawn from other evidence in the proceedings. It should be clear not only that the evidence is to be challenged but also how it is to be challenged. This is so because the witness must be given an opportunity to deny the challenge, to call corroborative evidence, to qualify the evidence given by the witness or others and to explain contradictions on which reliance is to be placed.’ 30 27 Ibid p 662. 28 S v Mkohle 1990 (1) SACR 95 (A) at 98E-.F. 29 Op cit fn 12, para 62. 30 Ibid paras 61 and 63. 45

[93]It seems to me that the high court did not realise that the appellant’s cross- examination did not conform to the above-stated tenets of fair play. A crucial aspect that comes out clearly from the appellant’s cross-examination is that the version of the respondent was never put to the appellant for her comment. It is also noteworthy that none of the questions posed to the appellant (during cross- examination) directed her attention to the fact that an imputation was intended to be made suggesting that her evidence was untruthful. For instance, no questions were put to the appellant to suggest that at the time of her admission, she had not yet shown the signs of labour, yet this is what was contended for, on behalf of the respondent. This contention found favour with the high court, which also found that there was no proof that the appellant’s labour was prolonged. The basis of that conclusion is unclear to me. I will demonstrate that both findings were not borne out by the record.

[94]In my view, there is nothing inherently improbable about the appellant’s version. The essential features of the story told by the appellant are, on the whole, probable. It seems to me that the aspects of the appellant’s evidence, which were corroborated by the respondent’s witnesses, did not receive sufficient consideration. In particular, I could not find any justification for the high court’s conclusion that the appellant’s credibility was ‘in tatters’. Neither do I agree with a view that characterises her version as falling within the category of unchallenged evidence that is ‘so improbable as not to discharge the onus’ resting on her. 31 The high court’s purported reliance on McDonald v Young 32 and the judgments relied upon in that judgment, was misplaced.

[95]In evaluating the appellant’s version, the high court remarked that the complainant was a single witness in relation to the factual matrix. It ought to have 31 See paras 16 and 110 of the judgment of the high court and passages from the judgments cited therein. 32 McDonald v Young [2011] ZASCA 31; 2012 (3) SA 1 (SCA) para 6. 46 heeded this Court’s warning, that ‘the evidence of a single witness to a fact, there being nothing to throw discredit thereon, cannot be disregarded’. 33 Moreover, no matter how serious the allegations might be, the onus of proving facts in a civil case is discharged on a preponderance of probabilities and not on any higher standard. 34

[96]In my opinion, the theory of the appellant’s case as presented to all the experts remained the same despite minor discrepancies in respect of some detail, and the facts on which the opinions of the different experts were based, were thus established. 35 With respect, the high court’s conclusion that the appellant’s version as presented by her experts collapsed like a pack of cards is not supported by evidence. This is because it is quite clear that the difference of opinion between the appellant’s experts and those called by the respondent, was mainly because the latter accepted the correctness of the contents of the hospital records relating to the Apgar score and MML’s alleged stability upon discharge even though the veracity of such documents was in dispute. I will revert later to this aspect.

[97]It has been found that there was a discrepancy between the evidence of the appellant’s expert, Prof Nolte, and the appellant regarding whether the appellant’s pregnancy was confirmed by a home pregnancy test or by a urine test done at the local clinic. This discrepancy is self-evidently inconsequential. In so far as doubt was expressed as to whether the appellant had attended an antenatal clinic, it bears noting that at no stage was it put to the appellant that she had not attended an antenatal clinic during her pregnancy. To the contrary, a note in the Maternity Register reflected that she had been ‘booked’, which was understood to mean that 33 Da Mata v Otto N O 1972 (3) SA 858 (AD) at 869C, approving Wigmore Wigmore on Evidence 3 ed, vol VII at 260. 34 Ley v Ley’s Executors and Others 1951 (3) SA 186 (A) at 192-3. 35 See PriceWaterhouseCoopers Inc v National Potato Co-operative Ltd [2015] ZASCA 2; [2015] 2 All SA 403 (SCA). 47 she had attended the antenatal clinic. Besides, the appellant’s attendance at the antenatal clinic is one of the facts agreed upon in the joint minute of the obstetric experts submitted by Dr Schoon and Dr Hofmeyr and therefore cannot be one of the features forming a basis for an attack on the appellant’s credibility.

[98]The 2007 Guidelines for Maternity Care in South Africa 36 (maternal guidelines) clearly stipulate that a patient’s antenatal record must be handed over to the hospital upon a patient’s admission to the labour ward. This practice was confirmed by Sister Msibi, who also testified that the same antenatal folder becomes part of the patient’s hospital record upon admission. On acceptance of this undisputed practice, it stands to reason that the antenatal clinic records would have formed part of the appellant’s hospital records that could subsequently not be located.

[99]The high court found that the appellant was not in labour at the time of her admission to the hospital. I disagree. The definition of labour in the maternal guidelines is helpful in this regard. According to those guidelines, labour is diagnosed if there are ‘persistent painful uterine contractions’ accompanied by at least one of the following: cervical effacement and dilatation, rupture of the membranes and a ‘show’. Considering that the appellant had attended antenatal classes, where pregnant women are educated about the onset of labour and the timing and intensity of contractions, there is nothing improbable about the appellant’s evidence that she felt abdominal pains at approximately 01h00, went back to sleep when the pain subsided, and summoned the ambulance only after noticing a pinkish mucous plug (‘show’). 36 The appellant’s obstetric expert, Dr Hofmeyr, alluded to these guidelines, and their applicability was never disputed. The 2007 Guidelines for Maternity Care in South Africa were admitted into evidence as Exhibit F. 48

[100]It is worth noting that in explaining the entry of ‘in labour’, made in the Ward Admission and Discharge Register, Sister Msibi testified that before recording that a patient was in labour, she would first ask a patient for the history pertaining to the onset of contractions and also physically examine the patient. It can thus be safely inferred that she wrote ‘in labour’ after making an assessment and satisfying herself that the appellant was indeed in labour.

[101]Whereas the appellant’s unchallenged evidence revealed that by the time she arrived at the hospital at approximately 13h00 she had manifested all three symptoms of labour, the high court found that her labour could have commenced at 20h00. To the extent that the high court accepted that labour started at 20h00, it misdirected itself, as that conclusion is in direct contrast to the appellant’s uncontested evidence and is not supported by the objective evidence in the form of the maternal guidelines. The high court’s suggestion that the appellant could have mistaken Braxton Hicks pain for contractions and that it was not shown that the appellant was in labour at the time of her admission are without foundation. 37 I am also unpersuaded by any suggestion that Dr Hofmeyr, a qualified obstetrician and gynaecologist, and Prof Nolte, a specialist midwife, could make a mistake in relation to distinguishing between abdominal pains and uterine contractions. The first judgment’s preparedness to accept that the appellant was indeed in labour upon admission attests to the unsustainability of the high court’s credibility finding insofar as that aspect is concerned.

[102]The high court criticised Prof Nolte’s evidence that the appellant’s labour lasted for approximately 28 hours, and suggested that it was one of the glaring contradictions between her evidence and that of the appellant as regards the onset of labour. This criticism fails to take into account that the first signs and 37 Further and in any event, the contractions that appellant experienced after her admission at the hospital ought to have been investigated, as the maternal care guidelines stipulate that abdominal pain suffered by a pregnant woman admitted to hospital must be investigated to exclude foetal distress. 49 symptoms of labour are also taken into account when the total duration of labour is considered. This is borne out by the medical record with the caption ‘Labour- Initial Assessment’ relating to the birth of the appellant’s second child, which records the appellant’s time of admission as ‘06h10’ but records the onset of labour as ‘01h30’ on 16 July 2012 (long before her admission to the hospital). Furthermore, in a document with the caption ‘Summary of Labour’, the total duration of labour is calculated from 01h30 and not from the time of admission, which shows the relevance of the time at which the signs and symptoms of labour were observed. It is therefore obvious that Prof Nolte’s calculation of the duration of labour commenced from the time the appellant experienced contractions at home and not from her time of admission. There is therefore no contradiction on this aspect. The number of hours that elapsed since the manifestation of labour up to MML’s delivery (28 hours) indeed supports the proposition that the appellant’s delivery fell within the category of prolonged labour within the contemplation of the maternal guidelines.

[103]An issue was raised about a statement made by Dr Hofmeyr in her report, in terms of which she stated that the appellant screamed for help on various occasions during her labour but was ignored or not examined. This statement must be considered in proper context. In her evidence, the appellant remained steadfast that, although there were numerous vaginal examinations, the foetal heart rate was monitored only once (by CTG) and was not monitored in any other way. Furthermore, the appellant testified that on one of the occasions after screaming for help, a nurse did come, but only told her where the toilet was situated. To state that she was ignored or not examined is therefore not incorrect.

[104]The discrepancies about the exact number of times the appellant was physically examined during her labour do not detract from her evidence that the foetal heart was monitored only once between her admission and MML’s delivery. In my view, there was no material discrepancy between Dr Hofmeyr’s 50 evidence relating to what was presented to her as the appellant’s version and the appellant’s testimony in court.

[105]The finding that the appellant contradicted herself in relation to how she ascertained the time at certain stages of her labour is, in my respectful view, unfounded. Crucially, it was never put to the appellant that she was being untruthful when she said that there was a wall-clock in the delivery room. Her truthfulness in this regard was corroborated by one of the nurses who testified on behalf of the respondent; she confirmed that there is indeed a clock that is hanging on the wall of the delivery room. Against that background, there was nothing odd about the appellant noting the time from the wall-clock while she was in the delivery room. The appellant’s allusion to checking the time on her cellular phone does not relate to the time when she was in the delivery room but rather to the time when she was in the post-natal ward while waiting for MML to be brought to her.

[106]The appellant did contradict herself as regards the date of her discharge from the hospital. The mistake regarding the date of discharge does not negate or detract from her evidence that MML did not cry; that he had to be resuscitated; and that by the time the appellant and MML were discharged, MML had not started suckling. That contradiction is therefore not material and should not serve to tarnish the appellant’s reliability as a witness. It is trite that not every error made by a witness affects their credibility. 38 Of importance is that cross- examination of the appellant by the respondent’s counsel did not shake her evidence in any way. Furthermore, she was not confronted about the fact that the version she presented to her experts was perceived to differ materially from her 38 S v Oosthuizen 1982 (3) SA 571 (T) at 576G. 51 evidence-in-chief, or that any of the respondent’s witnesses would differ with her version.

[107]In criticising the appellant’s evidence that she and her baby were discharged on 2 May 2006 when the hospital records showed the discharge date to be 3 May 2006, the high court said the following: ‘I do not understand why plaintiff insisted that she and the minor were discharged on the same day that she had given birth notwithstanding the documentary evidence. It is highly unlikely that she would have been discharged on her version, if it is accepted for the moment that the minor had to be resuscitated and taken away from her for five hours, yet, a few hours later they were discharged on her version.’ I accept that on this aspect the appellant was mistaken. However, the high court failed to follow through its criticism of the appellant to its consistent conclusion. Since the birth took place at 05h00 on the Monday morning, and the appellant was only discharged at 14h30 on the Tuesday afternoon, she was kept in hospital overnight and for a total period of nearly 33 hours from the time of birth. As the high court seemed to appreciate, this delayed discharge is consistent with the fact that the baby had to be resuscitated and was taken away from the mother for five hours. I say discharge was delayed, because Dr Mogashoa testified that if all is well, the mother and child are usually discharged 6 to 24 hours from birth, yet in this case the mother and child were not only kept in hospital for a further night but were not immediately discharged the next morning.

[108]In my view, there is nothing extraordinary about a first-time mother remembering the details of the birth of her child. On the probabilities, her labour and the delivery are likely to be imprinted in her mind. 39 For the same reason, I would not deem it odd for a woman who has attended an antenatal clinic to have 39 IK obo KK v MEC for Health, Gauteng Province [2018] ZAGPJHC 580 para 230. 52 taken note of the newborn baby not crying after birth or being anxious about that fact. Still at the level of probability, I note that it was not suggested to the appellant that she could not have known that a newborn baby’s cry is generally regarded as one of the indicators of good health. Notably, none of the experts deemed it odd that the appellant was able to observe that MML did not cry after his birth. For these reasons, I am unable to agree with the proposition that the appellant, on account of her youthfulness, could not have known about the importance of a baby’s cry as at the time of her admission to the hospital.

[109]Although the fact that MML did not suckle and was always sleepy after joining the appellant in the postnatal ward is an aspect that the experts considered to be in keeping with neonatal encephalopathy, it was downplayed in the evaluation of evidence. In my view, there is nothing implausible about the appellant’s evidence that MML had not cried after birth and that there was an attempt to resuscitate him by using nasal prongs. The manner in which he was resuscitated by using nasal tubes, and the fact that he was injected in the thighs, are aspects that seem to have been rejected on the basis that such a method of resuscitation did not fall within the realm of standard practice. This method of resuscitation could not simply be dismissed as a figment of the appellant’s imagination because Sister Msibi confirmed that nasal prongs were indeed being used on sick babies who required oxygen at Thebe hospital.

[110]Sister Mokoena, too, stated that oxygen is administered to babies either by way of nasal prongs or an oxygen mask. However, she could not recall whether the hospital was using nasal prongs at the time of the appellant’s accouchement. Sister Mokoena was asked about the procedure that would be followed if a baby did not cry despite having been suctioned after birth. The exchange between Sister Mokoena and the respondent’s counsel on that aspect was as follows: 53 ‘Counsel: Now Sister if that still does not work you don’t get the baby to cry even after that what do you do? Sister Mokoena: If now we’ve performed everything that we did to the child and the child is not crying then we will have to inform the doctor. What we usually do if the child is not crying then because there is this umbilical cord that is attached to the child and the mother that umbilical cord will be cut and the child will be placed aside on the warm area and there at the warm area she will also be given oxygen while we are waiting for the doctor. Counsel: Now explain very carefully this warm area you are talking about where is the warm area is it in the maternity delivery room or is it somewhere else? Sister Mokoena: At this present moment there is that warm area in that maternity ward but in the past I don’t remember whether we were having those warm areas in the ward but I remember there were times when the child was not feeling well after birth then the child will be taken to a separate place where there are incubators and the child will be put into those incubators. Counsel: Do you take this child to the incubators outside of the delivery room before or after the doctor has seen the child? Sister Mokoena: Like I already indicated we had that warm area in that maternity delivery room then the child will be placed in that room in that warm area but I don’t remember at that time whether we were using that but if that was the case the child will be placed there on that warm area and the oxygen will also be on the child while we were waiting for the doctor. Counsel: Now let’s go to the provision of this oxygen that you are talking about, how do you administer and how do you give this child oxygen? Sister Mokoena: When coming to the oxygen issue this is in two ways and at this moment we are using two things we call those nasal prongs and they have two holes and you put those things into the nostrils of the child, at this present moment we are having those prongs because I don’t know I don’t remember at that time in the past time that we were using those things but what I remember is that when we were giving the child the oxygen we were using an oxygen mask at the time. 54 Counsel: Now sister the [appellant] informed the court that when her baby was delivered the baby was not crying and the baby was immediately given oxygen with the nostril is it the nostril tube that you just spoke about? Sister Mokoena: Nasal [prongs].’

[111]Dr Hofmeyr’s evidence that ‘if there is no sign of breathing, no crying or any attempt to make respiration, then it is highly unlikely that the baby would be pink’ was taken out of context and fails to take into account that the appellant never suggested that there was ‘no attempt to make respiration’. Dr Hofmeyr did not discount resuscitation by nasal prongs as something that cannot be initiated in the delivery room under any circumstances. She considered applying oxygen through the nostrils as a possible alternative, but added that ‘it speaks to me of a more prolonged administration of oxygen, not just a quick intervention to help baby recover from the trauma of the delivery.’ Notably, Dr Kganane stated that an injection known as naloxone, was ‘a reversal injection just to wake a baby up . . . [which] might happen if you worry that a baby is not responsive’. In this regard, the evidence of Sister Mokoena about the steps actually followed whenever resuscitation of a new-born was necessary at Thebe hospital cannot be disregarded.

[112]MML’s initial Apgar score of 7 is at the lowest level of ‘normal’, and after five minutes the score was only 8. One can see from the hospital’s records that most of the babies delivered at around the same time had initial Apgar scores of 9 or 10. Sister Mokoena testified that she did not know why they did not score ‘score the child well’ because ‘usually it will have to be 9 out of 10 or 10 out of 10’. Dr Hofmeyr said that in the absence of maternal records one did not know why the baby initially scored only 7, but that based on the limited information available to her, the suppressed score was probably attributable to the fact that the baby did not cry and that there was thus a low respiration count. She also said 55 that Apgar scores are usually matters of impression recorded retrospectively rather than contemporaneously.

[113]In the course of decrying the appellant’s credibility as a witness, the high court inter alia found that the appellant’s version to the experts and her evidence in court were aimed at taking advantage of the missing records. This finding is refuted by the fact that both the Discharge Summary Form (Discharge Summary) and the Road to Health Chart were made available by the appellant despite some of the information set out therein, like the high Apgar score, being considered to militate against the presence of an intrapartum brain injury.

[114]It is undisputed that the Road to Health Chart was not completed in detail. One of the nurses who attended to MML at the clinic, Sister Mosia, admitted that besides recording the weight, none of the assessments of the tests referred to in the Road to Health Chart were recorded in that document. The high court found that even though parts of the information were ‘clearly missing’, the health chart provided objective evidence of MML's growth and confirmed that he had reached his developmental milestones until the age of at least fourteen months. 40 It seems to me that this is one of the reasons why the high court regarded the appellant’s evidence pertaining to MML’s inability to suckle and constant sleepiness, as unreliable.

[115]I am of the view that in the face of some information not having been fully recorded in the Road to Health Chart, there is no reason to doubt the appellant’s assertion that she had previously raised a concern about MML’s inability to sit when he was seven months old. Furthermore, the apparent surprise about MML’s normal weight gain in his first year does not take into account the appellant’s 40 Paragraph 112 of the high court’s judgment. 56 evidence that when MML could not suckle, she expressed breast milk and cup- fed him. In the face of a concession that some of the assessments were not recorded, there was no basis for considering the Road to Health Chart to be objective evidence that served to prove that MML had not manifested the symptoms of cerebral palsy in his infancy; nor was there any basis for finding that MML’s growth and development were ‘not indicative of a child that suffered from an intrapartum injury’. Missing hospital records and the failure to call the author of the disputed hospital records as a witness

[116]It is trite that documents may be used in evidence for a number of purposes other than establishing the truthfulness of the contents thereof. 41 The high court considered the Discharge Summary as objective evidence despite the fact that the veracity of some of the contents thereof was never admitted, and that certain entries in that form remained in dispute throughout the proceedings. It is common cause that three nurses attended to MML’s delivery. However, only one of these nurses, Sister Mokoena, was called to testify. Although she stated that she could not dispute that she was one of the midwives who had delivered MML, she stated that the handwriting on the Discharge Summary was not hers. She neither completed nor signed the Discharge Form.

[117]Although it is trite that a document that is introduced as evidence in court proceedings must be identified by a witness who is either the writer or signatory thereof, the other two midwives who were on duty at the time of MML’s delivery, including a chief professional nurse, were not called as witnesses. 42 No explanation was given for not calling them as witnesses. Surprisingly, not only 41 ABSA Bank Ltd v Ons Beleggings BK 2000 (4) SA 27 (SCA) ; [2000] 3 All SA 199 (A) para 6. 42 See Howard and Decker Witkoppen Agencies and Fourways Estates (Pty) Ltd v De Sousa 1971 (3) SA 937 (T) at 940E; Maize Board v Hart 2005 (5) SA 480 (O) at 484E-J; CRC Engineering (Pty) Ltd v J C Dunbar & Sons (Pty) Ltd 1977 (1) SA 710 (W); [1977] 1 All SA 146 (W) at 147. 57 did the high court find that the Discharge Summary constituted objective evidence, it also found that ‘the objective evidence supports the respondent’s contention of proper record-keeping’. It was never the appellant’s case that the nurses never made any notes at any stage. The fact that some entries were made in some of the documents cannot be equated with ‘proper’ record-keeping. Nor does that mitigate the loss of patient records. It is necessary to briefly consider the precepts that impel healthcare facilities to preserve patient records.

[118]Section 27 of the Constitution guarantees for everyone the right to have access to health care services, including reproductive health care. The foreword to the maternal guidelines recognises that maternal health care is one of the priority reproductive health issues that have been identified as requiring urgent attention in South Africa. It bears emphasising that hospitals have a constitutional obligation to dispense reasonable care. It therefore comes as no surprise that hospitals have a statutory duty to create and maintain proper records. Medical records frequently serve as medico-legal documents.

[119]The international standard for medical records has been adopted as a national standard in South Africa. 43 The National Health Act 61 of 2003 (the National Health Act) obliges the person in charge of a health facility to set up control measures to prevent unauthorised access to the storage facility in which those records are kept. In terms of that Act healthcare professionals have a legal duty to make contemporaneous notes of the medical care given to a patient. Section 17(2) of the National Health Act imposes stiff penalties for the failure to keep medical records and for their disappearance, falsification or alteration. All these prescripts speak to a recognition of the prejudice that a plaintiff in a medical 43 International Organisation for Standardisation (ISO/IEC) 15489.2001; South African National Standards (SANS) 15489:2004. 58 negligence claim stands to suffer if the healthcare professional or facility fails to create or safeguard patient records.

[120]The disappearance of hospital records has been lamented in a plethora of judgments in this country; in some instances, the records disappeared permanently and, on a few occasions disappeared but mysteriously re-surfaced; and, in other instances, the hospital records were altered. Various articles have been written about missing hospital records in this country, and it is quite evident that records have gone missing even in circumstances where no litigation was involved. 44 The problem has become endemic. Plainly, it is not inconceivable that a healthcare professional who becomes aware that his or her negligent acts might be questioned, may be motivated to spoliate the patient records so as to conceal his or her negligence. To the extent that there was an insinuation that the appellant’s legal representatives were behind the disappearance of the appellant’s hospital records, counsel for the appellant categorically raised this issue and placed the following on record: ‘If it is the [respondent’s] case that our attorney was involved [in the unlawful removal of the appellant’s file] or anybody else for that matter and they want to advance that case, it has not been pleaded. Now the mere suggestion that our attorney might have been involved is a very serious one and then it must be pleaded in detail . . . We can continue to deal with the issues. If that is not going to happen, they must refrain from making any suggestions to any witnesses, accusing my attorney of the unlawful removal of these files.’ The amendment that was subsequently made did not implicate the appellant’s attorneys. Reference was made to the hospital’s filing system, and the key-holder of the storage room was identified. This person was not called as a witness. Instead, the respondent decided to deal with the matter by adducing the evidence of Ms Tshabalala. 44 See L Wegner and A Rhoda ‘Missing Medical Records: An Obstacle to Archival Survey-Research in a rural community in South Africa’ (2013) 69 (2) SA Journal of Physiotherapy at 15-19. 59

[121]The high court remarked that there was no proof that a register was kept in respect of the files stored at the hospital’s filing or archive room. It also found that there was no proof that the appellant’s file in respect of the birth of MML was ever taken to the filing or archive room. These findings are borne out by the record and are therefore correct. The high court further said: ‘[The respondent] applied for amendment of its plea as indicated supra. The alleged [filing] system at the hospital and the identity of the key holder of the storage facilities were pleaded. This person was not called without any explanation about his unavailability. Ms Tshabalala not only contradicted the pleaded version, but her testimony was also difficult to [comprehend]’.

[122]Since it is clear that the methods for the preservation of the records were inadequate, the blame for the missing records cannot be laid at the door of the appellant. Although the high court’s judgment criticised the respondent’s filing system, and rightly so, 45 it stated that it would be wrong to blame the hospital staff for ‘improper care and record-keeping’.

[123]In the absence of medical records, claimants who lodge a medical negligence claim might stand to be prejudiced, as they might not be in a position to substantiate their assertions, thus hampering them in discharging the burden of demonstrating the negligence of the defendant healthcare professionals. Given that the creation and safeguarding of the records fulfils a right guaranteed by the Constitution, I venture, as an aside, to opine that where healthcare records were not properly safeguarded, it is only just and equitable that the healthcare facility in question must bear the evidentiary burden of showing that the care and service given to the patient were consistent with good medical practice. If this approach 45 S v Mkohle op cit, fn 27, para 50. 60 is not followed, the prejudice to the appellant (and any plaintiff in a matter in which the respondent has not produced medical records) would be immeasurable.

[124]In the present matter, a significant consideration regarding the missing records is that most of the expert witnesses who compiled medicolegal reports stated that the record-keeping was suboptimal and indicated that they were hamstrung by the lack of availability of the hospital records. The prejudice the appellant stood to suffer was perhaps downplayed by the obfuscation of matters insofar as reference was made to unrelated cases where the hospitals in the Free State had mysteriously lost patient records. It is quite interesting that the respondent was able to deliver its plea in this matter despite the absence of medical records.

[125]The respondent in this matter is a state respondent and all the hospital staff members are employees of a hospital falling within the respondent’s jurisdiction. There was no suggestion that the other staff members who attended to the appellant were not available to testify. It would seem that there were simply no efforts expended in securing their attendance. This is a disconcerting state of affairs. In MEC for Health, Eastern Cape and Another v Kirland Investments (Pty) Ltd 46 the Constitutional Court made the following insightful observation: ‘. . . [T]here is a higher duty on the state to respect the law, to fulfil procedural requirements and to tread respectfully when dealing with rights. Government is not an indigent or bewildered litigant, adrift on a sea of litigious uncertainty, to whom the courts must extend a procedure- circumventing lifeline. It is the Constitution’s primary agent. It must do right, and it must do it properly.’

[126]In circumstances like the present, where hospital records are not properly safeguarded and are lost to the detriment of a litigant, a charitable approach that gives cognisance to the plight of the litigant is required. Where some of the 46 MEC for Health, Eastern Cape and Another v Kirland Investments (Pty) Ltd [2014] ZACC 6; 2014 (5) BCLR 547 (CC); 2014 (3) SA 481 (CC) para 82. 61 hospital staff members who treated a litigant are, despite their availability, not called as witnesses to explain some of the entries they made in the limited records that are available, a court should not hesitate to draw an adverse inference from the failure to call such a witness.

[127]It was important to call the midwife who completed the Discharge Summary so that she could explain her basis for stating that MML’s breastfeeding had been successfully initiated. This was an important aspect of evidence, as the appellant’s version was that one of the symptoms exhibited by MML, which was consistent with HIE, was MML’s inability to suckle. It is well-established that a court is entitled to draw a negative inference from a party’s failure to call a relevant witness. Given that no reason was proffered for not calling the author of the Discharge Summary as a witness, there was justification for inferring that the reason for not calling her as a witness was that her evidence might not have been in the respondent’s favour.

[128]The high court stated that it was not prepared to draw any adverse inferences ‘in view of the time lapse and the obvious and probable lack of independent memory of such witnesses’. Considering that there is a statutory duty to safeguard a minor patient’s records until the minor has reached majority, the time lapse between the date of the treatment and the institution of the claim should not be a reason to assume that there will be a lack of independent memory of witnesses who attended to the patient. I agree with the contention that the high court ought to have drawn an adverse inference from the respondent’s failure to call the other midwives as witnesses. Regrettably, its remissness in failing to proffer a valid reason for not presenting the oral evidence of all the relevant witnesses, in order to supplement the inadequate record, was condoned. In the end, the respondent benefitted from the hospital’s failure to secure the appellant’s medical records. 62

[129]It must be borne in mind that the appellant vehemently denied that breastfeeding had been ‘successfully initiated’ as claimed in the Discharge Summary. It was thus crucial for the respondent to call the nurse who completed and signed the Discharge Summary and recorded that breastfeeding had been successfully initiated. Since this witness was not called, the recordal that breastfeeding had been successfully initiated remained inadmissible hearsay evidence. Despite this, the high court found that ‘the information contained in the Discharge Summary form is factually correct’.

[130]Much was made about an entry in the Delivery Register, describing the condition of the appellant and MML as ‘stable’. Both Sister Mokoena and Sister Msibi confirmed that the hospital staff member who made that entry, and signed next to it (Ms Hlophe), was an assistant nurse who only did administrative work in files and was not authorised to examine patients. She could accordingly not have made the entry on the basis of a medical examination that she had personally performed. The primary source of Ms Hlophe’s information was not identified, nor was Ms Hlophe called as a witness. Under the circumstances, the description of the appellant and MML’s status at discharge as ‘stable’ clearly constitutes inadmissible hearsay evidence. However, this piece of hearsay evidence was used to discredit the appellant’s evidence that up to the date of discharge, the baby had not been able to suckle from her breast. Further, and in any event, even if it is accepted, without so deciding, that MML was described as ‘stable’ at the time of discharge, two days after his birth, that does not, without more, diminish the value of the appellant’s evidence that after his birth, MML did not cry, had to be resuscitated and was constantly sleepy. 47 47 It is noteworthy that in IK obo KK v MEC for Health, Gauteng Province [2018] ZAGPJHC 580, an infant who was later confirmed to have suffered intrapartum HIE was discharged from hospital within two days of his birth, the hospital staff apparently not having noticed immediate manifestation of adverse sequelae. 63

[131]The high court placed a high premium on the fact that the appellant testified that at birth, MML’s colour was pink. It is self-evident from how an Apgar score is assessed that skin colouration of the new-born patient is but one of several features that serve as indicators of the health status of a new-born infant. The appellant’s evidence that at birth her minor child did not cry despite being resuscitated in the delivery room is a relevant factor in the Apgar scoring but seems to have been downplayed. I pause to mention that it is worth noting that in the 1996 article alluded to by Dr Kganane, it was also observed that 75% of children with cerebral palsy had normal Apgar scores at birth. For its part, the 2014 ACOG Report acknowledges that an Apgar score assigned during a resuscitation is not equivalent to a score assigned to a spontaneously breathing infant.

[132]What remains clear is that three crucial aspects remained unproven throughout the trial: evidence of a reliable assessment of the Apgar score, evidence of Sister Hlophe as the person who recorded that the appellant and MML were stable upon discharge, and evidence of the witness who recorded that breastfeeding was successfully initiated. In the absence of that proof, the foundation on which the respondent’s case rested, crumbled. The joint minutes of the experts

[133]It is trite that admissions made by the parties significantly narrow down the issues in dispute. In relation to the status of joint minutes filed by experts, this Court in Bee v Road Accident Fund, 48 held that where experts in the same field reach agreement, a litigant cannot be expected to adduce evidence on the agreed matters. It cautioned that, unless a trial court that was for any reason dissatisfied with the experts’ agreement, had alerted the parties to the need to adduce evidence 48 Bee v Road Accident Fund [2018] ZASCA 52; 2018 (4) SA 366 (SCA). 64 on the agreed material, it would be bound to accept the matters as agreed by the experts. 49

[134]In this matter, several joint minutes were handed in as exhibits. There can be no debate that the agreed-upon aspects recorded in the joint minutes that were furnished, were binding on the high court. In evaluating the evidence, the high court should have taken cognisance of the various joint minutes furnished by the parties. Notably, two radiologists, namely Dr Kamolane and Prof Andronikou, had signed a joint minute in terms of which they agreed that MML sustained a hypoxic ischemic injury of the partial prolonged variety, which occurred when the appellant’s pregnancy was equal or greater than 37 weeks and that there were no MRI features that suggested intracranial congenital infection, congenital anomalies, metabolic disorders, inflammatory disorders or haemorrhage. The high court did not quibble with any part of the joint minute and did not ask for any further evidence to be adduced on that aspect. Dr Kamolane also testified that ‘an MRI pattern of a child up to the age of 12 months, with the same insult, taken at the age of nine will still show the same pattern’. That opinion was not contested. Notwithstanding this, the high court found that the probabilities did not suggest that MML had suffered an HIE. 50 This finding in my view, constitutes a misdirection. Negligence

[135]In the present matter, some of the hospital records that could have substantiated the appellant’s version and proven negligence more conclusively were unavailable. This is through no fault of the appellant. I align myself with the approach suggested in Monteoli v Woolworths (Pty) Ltd, 51 where the court 49 Ibid para 73. 50 Paragraph 113 of the judgment of the high court. 51 Monteoli v Woolworths (Pty) Ltd 2000 (4) SA 735 (W) paras 25 and 29. 65 observed that notwithstanding that a plaintiff bears the onus of proving negligence on a balance of probabilities, a plaintiff is sometimes not in a position to produce evidence on a particular aspect. It suggested that in those instances, less evidence would suffice to establish a prima facie case, especially where the matter is peculiarly in the knowledge of the defendant. In my view, that approach is consistent with the following finding of this Court in Sea Harvest Corporation (Pty) Ltd and Another v Duncan Dock Cold Storage (Pty) Ltd and Another (Sea Harvest): 52 ‘It is probably so that there can be no universally applicable formula which will prove to be appropriate in every case. As Lord Oliver observed in Caparo Industries plc v Dickman and Others [1990] 2 AC 605 (HL) at 633 F-G [1990] 1 All ER 568 at 585 in fine-586a), “the attempt to state some general principle which will determine liability in an infinite variety of circumstances serves not to clarify the law but merely to bedevil its development in a way which corresponds with practicality and common sense”. I agree. A rigid adherence to what is in reality no more than a formula for determining negligence must inevitably open the way to injustice in unusual cases. Whether one adopts a formula which is said to reflect the abstract theory of negligence or some other formula there must always be, I think, a measure of flexibility to accommodate the “grey area” case. . . . Inevitably, the answer will only emerge from a close consideration of the facts of each case and ultimately will have to be determined by judicial judgment.’

[136]As stated before, the appellant’s factual evidence regarding her labour, which culminated in MML’s birth, is probable. Dr Hofmeyr’s evidence regarding the provisions of the maternal guidelines about the monitoring of a woman in labour was not challenged. She opined as follows: 52 Sea Harvest Corporation (Pty) Ltd and Another v Duncan Dock Cold Storage (Pty) Ltd and Another [2000] 1 All SA 128 (A); 2000 (1) SA 827 para 22. 66 ‘In the absence of antenatal and obstetric records, I cannot accurately evaluate or comment on fetal 53 heart rate patterns or the adequacy of fetal monitoring. However, the [appellant’s] statement regarding the events surrounding labour reflects inadequate fetal heart rate monitoring in direct contrast to prescribed minimum standards of care. The fetal heart rate should have been monitored every 30 minutes during the active phase of labour but was only performed once during the (more than) 14 hour admission, through the initial admission CTG. Without regular fetal heart rate monitoring attending staff would not be able to detect fetal distress depicting a possible intra-partum event.’ (Own emphasis).

[137]No credible and persuasive evidence was put forward to challenge Dr Hofmeyr’s evidence that the appellant’s labour was prolonged, thus requiring interventions aimed at expediting MML’s delivery. Furthermore, Prof Nolte’s evidence that a prolonged labour leads to foetal compromise was not challenged. In underscoring the importance of monitoring a woman in labour and the foetal, Dr Hofmeyr stated that the general principle is that ‘if the baby is in trouble, the baby needs to be delivered’. Having considered the absence of the further monitoring of the foetal heart rate after the CTG which was performed immediately after the appellant’s admission, inadequate maternal monitoring and probable prolonged labour, Dr Hofmeyr concluded that ‘the most likely timing of an hypoxic event capable of causing neonatal encephalopathy in [MML] (as was formally diagnosed by the relevant experts), was during the labour and/or delivery on 1-2 May 2005 at Thebe Hospital’. That conclusion is justified by the facts.

[138]The high court found that the appellant’s labour was not protracted. I disagree with the proposition that on the assumption that active labour commenced at approximately midnight, neither the latent nor active phase of labour appeared to be ‘unduly protracted’. With respect, this hypothesis does not 53 There are different ways of spelling the medical term ‘foetus’ (fetus) and the corresponding adjectives. In this judgment I have use 'fetus' where that usage appears in a quotation and the English spelling 'foetus' and 'foetal' elsewhere. 67 take sufficient account of the maternal guidelines and the entirety of Dr Hofmeyr’s evidence, in particular the correlation between uterine contractions, the supply of blood to the foetal and the foetal heart rate. The maternal guidelines unequivocally provide that ‘the latent phase is prolonged when it exceeds eight hours’. The guidelines then go on to set out how poor progress in labour is to be managed; that management includes investigating the cause of abdominal pains, among others.

[139]It is clear from the maternal guidelines that foetal distress can be reliably detected by monitoring the foetal heart rate. Once there are indications of foetal distress, the attending midwives must expedite the baby’s delivery to eliminate the eventuation of harm. The maternal guidelines stipulate that where foetal distress is detected when delivery is imminent (where the cervix is fully dilated), the baby must be delivered immediately; where delivery is not imminent, the patient must be prepared for immediate caesarean section.

[140]My understanding of Dr Hofmeyr’s unchallenged evidence on this aspect is that the duration of labour is not a stand-alone factor; the condition of the woman during that time, which can only be ascertained through adequate monitoring, is a crucial consideration. The healthcare professionals attending to a woman’s labour are duty bound to monitor the labour and to take steps to ensure that the foetal is not in distress. In stressing the importance of monitoring the foetal heart rate, Dr Hofmeyr explained that as the uterine contractions increase in strength (during labour), the blood vessels in the placenta become constricted and the blood supply to the foetal via the umbilical cord contains increasing levels of carbon dioxide and less oxygen. A series of late decelerations of the heartbeat, which can only be detected when the foetal heart is monitored, are a cause for concern, as they may suggest that the foetal is in distress. Absent timeous 68 intervention, the increasing levels of reduced oxygen supply to the foetal (hypoxia) will result in brain damage.

[141]Dr Hofmeyr’s evidence about the different phases of labour was not disputed. Similarly, her evidence about the interventions that should be made when there is no satisfactory progress in labour were not disputed. Moreover, her evidence about the progression of labour after the first phase, ie that the dilation is expected to be at the rate of 1cm per hour, was not disputed. The evidence of Dr Hofmeyr leaves no doubt that the appellant’s labour was indeed prolonged.

[142]A further consideration is that, against the backdrop of the undisputed evidence regarding the correlation between uterine contractions and foetal distress, the appellant’s evidence of more severe contractions that caused her to scream, and rendered her unable to walk five hours before MML’s delivery suggested the presence of foetal distress. Even under those alarming circumstances, the foetal heart rate was not monitored. That there are no records regarding the severity of her uterine contractions at that stage is neither here nor there. It cannot be right that the absence of the complete patient records is simply lamented but the claimant is faulted for not substantiating the allegations of ischemia with CTG tracings showing patterns suggestive of foetal distress during labour. What is clear from the appellant’s evidence is that (i) the contractions she was experiencing at that stage were much more severe than the ones she had experienced earlier, and (ii) that the vaginal examination that was done by the nurses at that particular stage led one of the nurses to conclude that her baby’s delivery was imminent. To my mind, these are circumstances that called for the invocation of the interventions aimed at excluding foetal distress, as laid down in the maternal guidelines. However, these interventions were not invoked. 69

[143]The available hospital records do not suggest that the labour ward was particularly busy at the time of the appellant’s admission up to the time of MML’s delivery. Furthermore, insufficiency of resources to follow the steps set out in the maternal guidelines was not pleaded in this matter. On the probabilities, nothing precluded the invocation of any of the interventions, set out in the maternal guidelines, to expedite MML’s delivery once the appellant’s labour became prolonged. In my view, it can be accepted on the basis of the evidence adduced, that the notional reasonable healthcare professional in the position of the nurses who attended to the appellant would have monitored the appellant in accordance with the maternal guidelines and would have intervened once her labour became prolonged.

[144]The evidence canvassed in the foregoing paragraphs permits a finding that the inadequate monitoring during the appellant’s prolonged labour led to MML suffering a partial prolonged type of brain injury, culminating in cerebral palsy. In my view, the general manner of the occurrence of harm to the foetal as a result of inadequate monitoring was reasonably foreseeable. Despite this, the hospital staff failed to act with the required degree of care, skill and diligence that was warranted in circumstances where similarly qualified healthcare workers would have taken steps to prevent harm by expediting the delivery of MML.

[145]As regards the yardstick by which the conduct of healthcare professionals is gauged, it is trite that the law expects of them to act in accordance with a notional standard set by a reasonable healthcare professional with their experience and qualification in their circumstances. Thus, the question is whether healthcare professionals in the position of the hospital staff would have foreseen the reasonable possibility of their conduct causing harm and, if so, whether they 70 would have taken steps to guard against that harm. 54 The team of nurses who attended to the appellant’s delivery were midwives, and one of them was a chief professional nurse. A midwife in their position would have foreseen the reasonable possibility of the lack of adequate foetal monitoring leading to foetal distress not being detected, which in turn could lead to brain damage resulting in the sequelae suffered by MML, and would have taken steps to prevent the harm. I am fortified in this view by the following remarks made by this Court in Sea Harvest: 55 '. . . [I]t should not be overlooked that in the ultimate analysis the true criterion for determining negligence is whether in the particular circumstances the conduct complained of falls short of the standard of the reasonable person. . . . [I]t has been recognized that while the precise or exact manner in which the harm occurs need not be foreseeable, the general manner of its occurrence must indeed be reasonably foreseeable.’ (Own emphasis.)

[146]With the benefit of the maternal guidelines and the expert evidence of Dr Hofmeyr and Prof Nolte regarding how inadequate monitoring can lead to foetal distress not being detected, I see no reason why that evidence, coupled with the appellant’s unchallenged evidence on the inadequate medical care she received during her labour (which fell far short of the monitoring stipulated in the maternal guidelines), should not suffice in proving the respondent’s negligence. I am satisfied that on the probabilities and circumstances of this case, the conclusion that is more plausible, from several conceivable ones, is that MML suffered hypoxic ischemia which was not timeously detected during labour as a result of the failure of the hospital staff to adequately monitor the appellant’s 54 This is in accordance with the two-step enquiry laid down in the seminal judgment of Holmes JA in Kruger v Coetzee 1966 (2) SA 428 (A) at 430 E-G. 55 Sea Harvest Corporation (Pty) Ltd and Another v Duncan Dock Cold Storage (Pty) Ltd and Another paras 21- 22. 71 labour. What remains is to show the nexus between the negligent conduct of the hospital staff and MML’s brain damage and its sequelae. Causation

[147]It is trite that where the defendant has negligently breached a legal duty and the plaintiff has suffered harm, it must still be proved that the said negligence caused the harm suffered. It is well-established that causation has two elements, namely: (i) the factual issue, the answer to which can be determined by applying the ‘but for’ test; and (ii) legal causation, which answers the question whether the wrongful act is linked sufficiently closely to the harm suffered; if the harm is too remote, then there is no liability. 56

[148]The high court held that ‘it will be highly speculative to pin-point the precise timing of the injury’. In my respectful view, this finding is misconceived at two levels: (i) the applicable test for causation and (ii) the mechanism of MML’s brain injury, which speaks to the link between the negligent conduct and the harm suffered. It is to those interlinked aspects that I now turn. Various judgments of this Court have cautioned against a rigid application of the ‘but-for’ test. 57 In Minister of Finance and Others v Gore, 58 this Court cautioned that the application of the ‘but for’ test does not require the precision of mathematics, pure science or philosophy; instead, it requires the invocation of common sense, where things are viewed against the backdrop of everyday life experiences.

[149]Similarly, this Court in Minister of Safety and Security v Van Duivenboden 59 observed that a determination of a causal link was not an exercise 56 International Shipping Company (Pty) Ltd v Bentley 1990 (1) SA 680 (A) at 700E-I. 57 See Mashongwa v Passenger Rail Agency of South Africa [2015] ZACC 36; 2016 (3) SA 528 (CC); 2016 (2) BCLR 204 (CC) and the judgments quoted therein. 58 Minister of Finance and Others v Gore N O 2007 (1) SA 111 (SCA) ; [2007] 1 All SA 309 (SCA) para 33. 59 Minister of Safety and Security v Van Duivenboden 2002 (6) SA 431 (SCA); [2002] 3 All SA 741 (SCA). 72 in metaphysics; rather, it ought to be based on the evidence adduced and what can happen in the ordinary course of human affairs. Therefore, the appellant did not need to prove the causal link with certainty, but only needed to establish that the wrongful conduct of the hospital staff was the probable cause of the loss. 60 The finding that establishing causation necessitated the pin-pointing of the precise timing of the injury goes against the grain of many judgments of this Court and the Constitutional Court and misconceived the nature of the enquiry. This much is also attested by the fact that the Task Force made this telling conclusion in the 2014 ACOG Report: 'The multidimensional aspect of the assessment process is key to recognising that no single strategy to identify hypoxic-ischemic encephalopathy is infallible and will achieve 100% certainty of the cause of neonatal encephalopathy in all cases.’

[150]In this matter, the nature and mechanism of MML’s brain injury was described as the partial profound type, which is a type of brain injury that generally occurs in instances where the insult on the foetal had lasted longer than in the acute profound type of injury. On that score, this case is distinguishable from the line of judgments of this court which dealt with causation in the context of an acute profound brain injury. That distinction is crucial. 61

[151]As mentioned before, the nature of the injury is a crucial determination in this matter. It bears emphasising that the paediatricians, Dr Gericke and Dr Kganane, were agreed that the pattern of the injury sustained by the infant was consistent with a partial prolonged type of hypoxic ischemic injury. It was common cause that a partial prolonged type of brain injury generally occurred in instances where the episode of the insult on the foetal had lasted at least 30 – 45 minutes. Expert evidence revealed that severe uterine contractions have an impact 60 Sea Harvest op cit, fn 54, para 6. 61 Compare AM obo KM v Member of Executive Council for Health, Eastern Cape [2018] ZASCA 141 para 65. 73 on the foetal heart rate and are discernible in CTG tracings. I have already alluded to the severity of appellant’s contractions, which worsened at approximately midnight. MML’s foetal heart rate was not monitored again after the CTG monitoring that was done immediately after her admission. The more plausible inference is that subsequent to the admission CTG monitoring, MML suffered foetal distress. On probabilities, his non-reassuring heart rate was not detected because the nurses had stopped monitoring the foetal heart rate. The inescapable inference is that the non-detection of foetal distress ultimately caused MML to sustain a brain injury of the partial prolonged type.

[152]It is important to note that the conclusion made in the 1996 article 62 mentioned by Dr Kganane, stating that an infant who has suffered hypoxia proximate to delivery should demonstrate all the four criteria mentioned in the first judgment, had been revised and was no longer the applicable benchmark at the time of the hearing of the case in the high court. It was also recognised that using the Sarnat classification system to stratify the severity of neonatal encephalopathy was insufficient as a stand-alone test. Prof Solomons’ conclusion that MML did not meet the essential criteria must be seen in context; it was clear that the absence of the relevant criteria was linked to the absence of hospital records, hence his qualification that MML’s case was ‘complicated by the absence of any antenatal, obstetric and resuscitation records’ .

[153]The 2014 ACOG Report states as follows: ‘Thus, for the current edition, the Task Force on Neonatal Encephalopathy determined that a broader perspective may be more fruitful. This conclusion reflects the sober recognition that knowledge gaps still preclude a definitive test or set of markers that accurately identifies, with high sensitivity and specificity, an infant in whom neonatal encephalopathy is attributable to an acute intrapartum event. 62 American Academy of Paediatrics ‘Use and Abuse of the Apgar score’ Paediatrics Vol 98 No 1 July 1996. 74 . . . [T]o determine the likelihood that an acute hypoxic-ischemia event occurring within close temporal proximity to delivery contributed to neonatal encephalopathy, it is recommended that a comprehensive multidimensional assessment be performed of neonatal status and all potential contributing factors, including maternal medical history, obstetric antecedents, intrapartum factors (including foetal heart rate monitoring results and issues relating to the delivery itself) and placental pathology.’

[154]In this matter, there were no antenatal, intrapartum or postnatal records, save for the Discharge Summary and the Road to Health Chart, which were made available by the appellant, as well as the Maternity Register and Delivery Register made available by the respondent. Almost all the experts lamented the absence of the other hospital records. The prejudice suffered by the appellant on account of the missing hospital records is self-evident. 63

[155]The ACOG Report lists a variety of ‘markers’ regarded to be consistent with intrapartum brain injury. It goes on to mention that ‘when more of the elements from each of the item categories are met, it becomes increasingly more likely that peripartum or intrapartum hypoxia–ischemia played a role in the pathogenesis of neonatal encephalopathy’. 64 Due to the absence of some hospital records, some of the intrapartum factors, which constitute important variables in the equation, are ‘unknown’. In particular, information pertaining to (i) the foetal heart rate patterns as observed on CTG tracings, (ii) the issues pertaining to the delivery, and (iii) placental pathology 65 was not available. It can be discerned 63 The prejudice suffered by the appellant on account of the missing records is also evident from the finding in (i) paragraph 39 of the first judgment, where it is stated that none of the factors mentioned in the 1996 article are present despite there being no evidence suggesting that a blood sample was obtained for purposes of determining the presence of metabolic or mixed acidemia; as well as the finding in paragraph 40, where it is stated that ‘[b]ecause the medical records were missing, there was no evidence as to the presence or absence of foetal distress.’ 64 2014 ACOG Report at 208. 65 The 2007 Guidelines for Maternity Care in South Africa provide that in addition to recording the heart rate patterns, the CTG tracings must be retained. The same guidelines also stipulate that one of the things that have to 75 from the 2014 ACOG Report that the scientific community considers that if the cord arterial gas pH levels are above 7.2, it is unlikely that intrapartum hypoxia played a role in causing neonatal encephalopathy. 66 Information pertaining to this fourth ‘marker’ was also not available, apparently because no cord blood sample was analysed. Dr Hofmeyr’s undisputed remark on this aspect is critical: she stated in her report that the recordal of blood gas analysis of cord blood at birth ‘was not prescribed care and not standardly available in all healthcare facilities.’ The point here is that the existence of gaps in the necessary information as a result of ‘unknown’ factors is through no fault of the appellant.

[156]The high court expressed its dissatisfaction with the cogency of the underlying reasoning of the appellant’s experts and also stated that their evidence ‘does not pass the reasonable and logical requirement test for acceptance of their opinions’. It went on to find that the experts’ testimony was based on incorrect facts. With respect, I could not find any justification for that finding. It is clear that heavy reliance was placed on the evidence of Dr Mogashoa. The appellant’s counsel contended that Dr Mogashoa’s evidence was not satisfactory; most criticism related to her evidence under cross-examination. I have to assess whether this contention holds water. It bears emphasising that the function of an expert is to assist the court to reach a conclusion on a matter in respect of which the court itself does not have the necessary knowledge to decide. It is not the mere opinion of the witness which is decisive but his or her ability to satisfy the court that because of his or her skill, training or experience, the reasons for the opinion he or she has expressed are acceptable. 67 This Court in Coopers (South Africa) be done once the baby has been delivered (third stage of labour) is to ‘examine the placenta for completeness and any abnormalities.’ 66 2014 ACOG Report at 208. 67 Glenister v President of the Republic of South Africa and Others [2013] ZACC 20; 2013 (11) BCLR 1246 (CC) para 7. 76 (Pty) Ltd v Deutsche Gesellschaft für Schädlingsbekämpfung MBH (Coopers) 68 held that a proper evaluation of the opinion of an expert can only be undertaken if the process of reasoning which led to the conclusion, including the premises from which the reasoning proceeds, are disclosed by the expert.

[157]I accept Dr Mogashoa’s qualifications and respect her training as a paediatric neurologist. However, having gone through the record, I had difficulties with certain parts of her evidence which, in my view, lend credence to counsel’s criticism. With respect, I found it quite astounding that having stated in her medico-legal report that formulating an opinion on this matter was ‘impossible’, Dr Mogashoa, in the same report, went on to venture an opinion in terms of which she categorically found that ‘[MML’s] impairments were not caused by intrapartum hypoxia’.

[158]In my view, Dr Mogashoa’s conclusion that the injury was not an intrapartum event based on a ‘lack of an overt neonatal syndrome, history of regression and current clinical features with predominance of lower limb involvement’ clearly disregarded the appellant’s evidence of MML having been resuscitated and the poor reflexes he presented after his birth. Insofar as she limited herself to only a few criteria, her approach seems out of sync with the scientific community, as the 2014 ACOG Report recommended a multidimensional assessment of all contributing factors. 69

[159]As regards one of the ‘markers’ considered to be consistent with HIE, Dr Mogashoa inexplicably made an assumption in favour of the respondent. This is evident from the fact that after noting (in her medico-legal report) that there were 68 Coopers (South Africa) (Pty) Ltd v Deutsche Gesellschaft für Schädlingsbekämpfung MBH 1976 (3) SA 352 (A) at 371A-H. Also see Bee v Road Accident Fund op cit, fn 48, para 73; Michael and Another v Linksfield Park Clinic (Pty) Ltd and Another op cit, fn 8, paras 36-7. 69 2014 ACOG Report at 4. 77 no neonatal records that evidenced the presence of multisystem organ failure consistent with HIE, she recorded that ‘one can assume that blood workup was not done because the baby was well at birth'.

[160]The joint minute of the radiologists confirmed that MML had HIE-related cerebral palsy which was probably sustained during the perinatal period (up to one month after birth). This means that by the time Sister Mosia examined him, MML had already sustained his brain injury. It is noteworthy that Sister Mosia, a qualified health professional, did not observe any of the clinical features associated with cerebral palsy when she assessed MML at the age of three months, on 8 August 2005. This loudly attests to the ACOG Task Force’s acknowledgment that ‘the clinical features of neonatal encephalopathy can be difficult to recognise reliably and consistently in newborns’. 70 The fact that Sister Mosia did not, during her assessments, observe these features in MML does not mean that he never exhibited the signs and symptoms. The preoccupation with obvious signs and symptoms of neonatal encephalopathy was clearly misplaced.

[161]A number of the respondents’ experts appeared to consider the high APGAR score allocated to MML as a basis for excluding intrapartum hypoxia. Dr Hofmeyr stated that the assessment of an Apgar can be subjective. Support for this statement can also be found in the evidence of Dr Bekker, who was called as the respondent’s witness. Dr Hofmeyr pointed out that Apgar scores are assessed by the attending midwives, such that the score reflected in the Discharge Summary and the Road to Heath Chart was probably based on the midwives’ score. Her evidence that APGAR scores are not diagnostic in nature was uncontroverted. Moreover, that evidence is borne out by the 2014 ACOG Report. 70 2014 ACOG Report at 4. It is noteworthy that in IK obo KK v MEC for Health. [2018] ZAGPJHC 580, an infant who was later confirmed to have suffered intrapartum HIE was discharged from hospital within two days of his birth, the hospital staff apparently not having noticed any manifestation of an overt neonatal syndrome. 78

[162]It is of significance is that in the 2014 ACOG Report, it was acknowledged that; ‘an APGAR score assigned during a resuscitation is not equivalent to a score assigned to a spontaneously breathing infant. There is no accepted standard for reporting an APGAR score in infants undergoing resuscitation after birth because many of the elements contributing to the score are altered by resuscitation . . . In order to correctly describe such infants and provide accurate documentation and data collection, an expanded APGAR score report is encouraged’. In my view, the fact that the accuracy of the assessment of MML’s APGAR score was never proven is an important aspect that must be borne in mind when considering the opinion of various experts in relation to the high APGAR score allocated to MML.

[163]There was nothing to gainsay the appellant’s evidence that she had not bumped her tummy against any hard object and that she had not been involved in any accidents during her pregnancy. The appellant’s obstetric expert, Dr Hofmeyr, concluded from the available records that there were no documented concerns suggesting that MML was born prematurely. This conclusion was not disputed. The conclusion that the appellant’s admission CTG reading was reassuring was deduced from the fact that after approximately 15 minutes, the CTG monitoring of the appellant was terminated. It can be accepted that had there been any problems that had been observed in the appellant’s antenatal clinic records at the time of the appellant’s admission or in the CTG tracings pertaining to the monitoring that was done upon her admission, a senior midwife or medical practitioner would have been summoned so as to follow the interventions stipulated in the maternal guidelines. Dr Hofmeyr’s conclusion that the appellant was a low-risk patient with an uncomplicated antenatal course, that she had a spontaneous onset of labour and that the admission CTG was reassuring was not attacked under cross-examination. At no stage did Dr Hofmeyr, in her oral 79 evidence, change the views expressed above. This unrefuted evidence is vital in relation to the timing of MML’s brain injury.

[164]Moreover, Dr Schoon and Dr Hofmeyr, in their joint minute, accepted that the appellant booked and attended an antenatal clinic during her pregnancy. During the appellant’s cross-examination, she was cross-examined on whether she had displayed any symptoms inconsistent with a normal pregnancy, like bleeding, abnormal vaginal discharge etc. She answered in the negative. Her evidence that she had not experienced any problems with her pregnancy and was not on any chronic medication was not challenged. The ineluctable inference is that the appellant had had an uneventful pregnancy. At no stage was it put to her that she had any pre-existing conditions that posed a risk to the foetus. In her report, Dr Mogashoa observed that MML was not dysmorphic and opined that a genetic abnormality was unlikely. In their joint minute, Dr Gericke and Dr Kganane, too, were agreed that there were no neurocutaneous lesions or dysmorphic features suggestive of genetic chromosomal abnormalities. Based on all those factors, any probability of a medical predisposition to antepartum injuries was therefore eliminated.

[165]At the end of the day, the appellant’s factual evidence remains uncontroverted. There is nothing improbable about the appellant’s evidence that MML did not cry at birth and had to be resuscitated immediately after delivery. There is also no evidence to disclaim the appellant’s evidence about MML’s poor reflexes, which manifested themselves in his inability to suckle and being constantly sleepy. The appellant’s evidence that at the time when MML was unable to suckle, she fed him by expressing milk into a cup and that his feeding was gradually supplemented is the most logical explanation for MML’s apparent normal weight gain in the first 12 months of his life. She did not assert that MML was only breastfed throughout his infancy. To make much about MML’s apparent 80 normal weight gain in that period ignores the alternative method of feeding employed and presupposes that MML was only being breastfed for that entire period, when there was no evidence to that effect.

[166]In their joint minute, Dr Gericke and Dr Kamolane were in agreement that ‘neurological manifestations of partial profound HIE at birth may be mild and do not always meet the perinatal asphyxia criteria and neurological signs can be delayed’. This conclusion was based on an article that Dr Gericke had relied on, authored by Ms LS de Vries and Ms F Groenendaal. 71 Moreover, Dr Kamolane expressly reaffirmed his concurrence on this aspect during his testimony in court. In any event, when reading that text in the context of the entire article, it is evident that the article does not suggest that severe motor impairment is absent in the group alluded to, but merely acknowledges that it is uncommon. I am therefore unable to agree with the first judgment’s conclusion that the statement that ‘severe impairment is uncommon in this group of infants’ automatically excludes MML.

[167]I have already alluded to the fact that a number of ‘markers’ consistent with an intrapartum brain injury are missing due to the unavailability of hospital records. They remain ‘unknown’ factors. In this case, the appellant is unable to show the existence of more ‘markers’ because the patient records that were supposed to be safeguarded by the hospital were not available. The fact of the matter is that there are circumstances that point to the injury having occurred intrapartum. This Court must do the best it can based on all the material presently before it, mindful of the fact that the unavailability of some of the information is through no fault of the appellant. 71 L S de Vries and F Groenendaal ‘Patterns of Neonatal Hypoxic-Ischaemic Brain Injury’ (2010) Neuroradiology. 81

[168]In my view, the interests of justice permit an approach consistent with the following dicta of Holmes JA, expressed in his dissenting judgment in Ocean Accident and Guarantee Corp. Ltd v Koch: 72 'The fact that, scientifically speaking, the aetiology of the disease is uncertain, does not hamper the court in deciding, on the facts and on the expert evidence adduced in a given case, whether a likely cause was proved in such a case. Judicial decisions reflect the particular facts and testimony of each case, and are not intended and cannot be regarded as scientific treatises. Accordingly, the possibility of future scientific disproof of the opinion of one or the other of the expert medical witnesses is, judicially, a matter of no moment - the Court must do the best it can on the material presently before it in each case.’ (Own emphasis.) Conclusion

[169]To sum up, each case must be decided on its specific facts as gleaned from the evidence. In this matter, there is evidence of negligence, on the one hand and there is evidence of harm, on the other hand. I am satisfied that these elements of a delictual claim have been proven on a balance of probabilities. The experts who submitted medico-legal reports and/or testified deferred to obstetric experts regarding the timing of the injury. There was unanimity among all the experts that foetal distress is one of the most common causes of intrapartum HIE.

[170]As stated before, the maternal guidelines provide that a woman in labour must be regularly monitored so as to ensure the safe delivery of the baby. Dr Hofmeyr testified that the intrapartum hypoxic ischaemia was probably as a result of foetal distress that was not detected during the appellant’s prolonged labour because of inadequate monitoring of the foetal heart rate, which constitutes negligence. Alongside this evidence is the joint minute of the radiologists, confirming that the partial prolonged type of the injury suggested that the insult 72 Ocean Accident and Guarantee Corporation Ltd v Koch 1963 (4) SA 147 (A) at 159E-F; Hulse-Reutter and Others v Godde 2001 (4) SA 1336 (A); [2002] 2 All SA 211 (A) para 14. 82 on the brain was as a result of hypoxic ischaemia. Dr Gericke and Dr Kganane were agreed that the type of the brain injury sustained by MML was consistent with a partial prolonged type of hypoxic ischemic injury.

[171]The credible and unchallenged expert evidence of Dr Hofmeyr and Dr Gericke militates against the probability of the occurrence of an antepartum injury. Their evidence fortifies the view that MML’s brain injury occurred intrapartum as a result of the failure of the hospital staff to (i) adequately monitor the foetal heart rate during the appellant’s prolonged labour, (ii) to detect foetal distress and (iii) to timeously intervene to prevent the brain injury. This is the more probable version.

[172]A significant aspect in relation to the timing of MML’s brain injury is that, under cross-examination, one of the respondent’s experts, Dr Kganane, conceded that if MML was indeed not suckling after birth that would presuppose that the severe brain injury must have happened before birth, as opposed to after birth. Given this concession by the respondent’s witness, as well as the appellant’s evidence that MML had not cried after birth, had to be resuscitated and was constantly sleepy, as well as her unchallenged evidence that MML sustained no injuries after her discharge from the hospital, there is no basis for entertaining the possibility of MML having sustained a postpartum injury after the appellant’s discharge. Since there is no evidence of an injury having occurred before or after MML’s birth, inferentially the highest risk period during which the injury occurred, was during labour.

[173]The persuasive evidence that the partial prolonged type of injury occurred intrapartum, viewed against the backdrop of the radiologists’ joint minute confirming that the injury arose as a result of hypoxic ischemia and the rest of the joint minutes, collectively constitute sufficient scientific evidence of a probable 83 connection between the negligence of the hospital staff and the harm suffered by MML.

[174]On acceptance of the appellant’s evidence and expert evidence, supported by scientific data, the mosaic of evidence points to the presence of the following ‘markers’, recognised by the ACOG Task Force as being consistent with an intrapartum event: (i) the neuroimaging patterns seen on MRI are consistent with hypoxic ischemia (the respondent’s expert, Dr Kamolane, testified that the MRI patterns after the first year of life remain the same); (ii) there is no evidence of other factors that could be regarded as contributing factors (on the unchallenged evidence of the appellant and the available medical records as interpreted by Dr Hofmeyr, 73 there was nothing to suggest the presence of any predisposing intrapartum events); (iii) on the available medical records, there was no evidence of any sentinel event occurring immediately before or during labour and delivery; (iv) the developmental outcome is spastic quadriplegia. 74

[175]Pursuant to physically examining MML, a paediatric neurologist and neurodevelopmental paediatrician, Prof Solomons, concluded that he had spastic quadriparetic cerebral palsy. It is clear from the record that, during the trial, the respondent’s counsel agreed to the admission of the joint minute of Prof Solomons and Dr Griessel. While there are areas where the two experts were not in agreement, it is of significance that they did pertinently agree that ‘there exists a good correlation between [MML’s] brain abnormalities [which showed features of chronic evolution of a hypoxic ischaemic injury] and the type of cerebral palsy MML was suffering from’. They also agreed that MML’s motor disability is severe and classified it as Gross Motor Function Classification System V. Notably, the joint minute of Prof Solomons and Dr Mogashoa, acknowledged that 73 See para 140 – 1 of this judgment. 74 See the 2014 ACOG Report at 208 – 211. 84 ‘the MRI findings of white matter abnormalities do correlate with the severe spasticity’. To my mind, their joint minute should therefore be given the status of any joint minute, such that their points of consensus should be regarded as proven, in line with the principle laid down in Bee v Road Accident Fund. 75 These constitute a sound basis for finding that MML suffered an intrapartum HIE. The ACOG Report speaks for itself. To state, as a fact, that MML did not fulfil ‘any of the criteria’ is clearly incorrect.

[176]Had the appellant’s labour been properly monitored, interventions aimed at expediting delivery in accordance with the maternal guidelines would have been taken once the first stage of labour became prolonged. Clearly, there was sufficient opportunity to intervene by performing a caesarean section timeously. As stated before, the type of brain injury suffered by MML (partial profound) generally occurs in instances where the insult on the foetal had lasted longer than in the acute profound type of injury. The mechanism of the injury is different from the acute profound type of injury, which typically happens moments before the delivery of the baby. It is for those reasons that I am unable to agree with the first judgment’s conclusion that there is no evidence that suggests that regular monitoring coupled with a caesarean section would have resulted in the delivery of MML before the injury was suffered. In my view, all the evidence canvassed in this part of the judgment suffices to prove all the elements of delictual liability on a balance of probabilities.

[177]To the extent that further substantiation in the form of additional ‘markers’ may be considered lacking, it has to be borne in mind that this is attributable to the hospital’s non-compliance with applicable prescripts. Thus, sight should not be lost of the fact that the respondent did very little to shift the evidentiary burden 75 Bee v Road Accident Fund op cit, fn 48, para 73. Also see Coopers (South Africa) (Pty) Ltd v Deutsche Gesellschaft für Schädlingsbekämpfung MBH op cit, fn 68, 371A-H. 85 of showing that the care and service given to the appellant were consistent with good medical practice. 76 Regard being had to those aspects, this matter neatly falls within the ambit of the test laid down in Lee v Minister for Correctional Services 77 on the basis that the conspectus of the evidence has shown on a balance of probabilities that the harm suffered by MML is closely connected to the omissions of the hospital staff in relation to the monitoring of the appellant’s labour.

[178]For all the reasons set out above, my conclusion is that a balanced evaluation of all the evidence reveals that the more probable inference is that MML’s intrapartum brain injury could have been avoided by expediting delivery if the hospital staff had properly monitored the appellant’s labour. 78 In the result, causative negligence has been proven on a balance of probabilities, thus rendering the respondent vicariously liable for damages. I would therefore uphold the appeal with costs. _______________ for M B Molemela Judge of Appeal Wallis JA (Makgoka JA and Unterhalter AJA concurring)

[179]I have had the privilege of reading the judgments of my colleagues Makgoka JA and Molemela JA and am in full agreement with that of Makgoka JA. I write separately to express my disquiet over two matters. They are the circumstances in which these proceedings were brought and the approach 76 Compare Athey v Leonati [1996] 3 SCR 458, 1996 CanLII 183 (SCC), where the Supreme Court held that where the court is satisfied that a claimant would otherwise be unjustly deprived of a remedy by reason of the inability to establish direct causation, the claimant should not be held to a strict standard; rather, the claimant could succeed by establishing that the healthcare practitioner’s breach of the standard of care materially contributed to the occurrence of the injury. 77 Lee v Minister for Correctional Services [2012] ZACC 30; 2013 (2) SA 144 (CC); Also see Mashongwa v PRASA [2015] ZACC 36 para 60. 78 Life Healthcare Group (Pty) Ltd v Suliman [2018] ZASCA 118; 2019 (2) SA 185 (SCA) para 16. 86 to the conduct of this trial, typical of others in the medical negligence cases that are now burgeoning in our courts. Institution of the litigation

[180]MML was born on 2 May 2005. The action was instituted on 2 September 2014, over nine years later. There are two different descriptions in the record of how his mother (HL), who my sister describes in her judgment as indigent and ill-informed, came to commence proceedings.

[181]The first description was in her affidavit in an application for condonation of her failure to give notice in terms of s 3(2) of the Institution of Legal Proceedings Against Certain Organs of State Act 40 of 2002. HL testified that she had been unaware of the possibility of making a claim arising out of MML's cerebral palsy until early 2014, when she met a woman, apparently at Thebe Hospital, who indicated that she had instituted an action for damages against the government due to negligence on the part of a hospital when she gave birth. Her affidavit then reads: She suggested that I contact her attorneys and after receiving details of my current attorneys from her, I consulted Kagiso Mokoduo of Mokoduo Incorporated currently known as MED Attorneys during May 2014.' The oddity of a chance encounter with an unidentified stranger 79 leading an indigent, unemployed person living in or near Harrismith, to consult an attorney practising in Johannesburg, was not explained. According to her the attorney advised that there needed to be further investigation of a potential claim, but that they should in the meantime give notice of her intention to institute an action. 79 HL said in her replying affidavit that she did not have the details of this woman who was a 'complete stranger' who helpfully provided her with the details of her attorney. 87

[182]Notice was given by registered post on 30 June 2014. As was to be expected, the letter was bereft of any detail about the claim. It read: 'We are advised that the medical and nursing staff at the hospital failed to render and provide the necessary medical, surgical and nursing care, advice, treatment and supervision with such skill and diligence as is reasonably required and expected of doctors, nurses and other medical and administrative staff acting within the course and scope of their employment with the Department of Health of the Free State Provincial Government … As a result of this the minor sustained various birth injuries, as a result of which the minor now suffers with inter alia Cerebral Palsy and mental retardation …'

[183]Notwithstanding the failure to make any attempt to comply with s 3(2)(b) of the statute by providing 'the facts giving rise to the debt', a claim for R20 million was made. The letter's speculative nature was demonstrated by what followed, which was twenty-five paragraphs of demands for information from the MEC. The letter manifestly did not serve its statutory purpose of informing the MEC of sufficient particulars to enable the matter to be investigated and consideration to be given to whether to resist the claim. 80 It was largely a fishing expedition.

[184]The Constitutional Court has explained the purpose of provisions such as these in Mohlomi. 81 It is that: 'Inordinate delays in litigating damage the interests of justice. They protract the disputes over the rights and obligations sought to be enforced, prolonging the uncertainty of all concerned about their affairs. Nor in the end is it always possible to adjudicate satisfactorily on cases that have gone stale. By then witnesses may no longer be available to testify. The memories of ones whose testimony can still be obtained may have faded and become unreliable. Documentary evidence may have disappeared. Such rules prevent procrastination and those harmful consequences of it.' 80 Avex Air (Pty) Ltd v Borough of Vryheid 1973 (1) SA 617 (A) at 621H-I; 81 Mohlomi v Minister of Defence [1996] ZASCA 20;1997 (1) SA 124 (CC) para 11. 88 I appreciate that there are good reasons why HL had not approached attorneys at an earlier stage, but that does not mean that the problems of litigating in relation to a delayed claim such as the present one disappear. They remained and were manifest when the matter came to trial thirteen years after MML was born.

[185]At the trial an entirely different and more plausible explanation emerged when HL was cross-examined about the circumstances in which she consulted her attorneys. She said that she first saw her attorney on 31 August 2017 and described what happened as follows: 'And how did it happen that you decide to visit an attorney, to see an attorney, in 2017? _ _ _ What happened exactly My Lord is that there was a certain lady who arrived at Dimakatso Disabled Centre. At the what, sorry … _ _ _ Dimakatso Disabled Centre and she told us about their attorneys and there were children there at that centre and five children were selected and we were taken for an interview. Sorry, who selected the children _ _ _ That lady selected those five children and my child was also amongst those five. We went for an interview but four of those five did not qualify … Yes, only my child qualified. That is why he was picked.'

[186]HL explained that the woman who came to the school was called Charmaine. When asked who made the decision of who qualified and who did not qualify, her answer was: 'We were told at school that a certain lady arrived and she picked up [Quaere: out] five children and we as the parents to those children, we also accompanied our children and [on] our arrival there, there were these two ladies… that Charmaine lady and the other white lady. And what happened is that we got into a room, but we did not got in [at] the same time, we got into that room one by one for that interview.' 89

[187]No doubt concerned by this turn of events, which was wholly inconsistent with the explanation given in the condonation application and involved an egregious example of touting for work among vulnerable children, HL's counsel questioned the relevance of this line of cross-examination. The objection was correctly rejected and HL was asked on what basis the selection had taken place. She was hard put to say, her evidence being: 'After I explained everything to that lady, she said to me that child might have got injured during birth, because I have already told her everything about the birth of the child, the whole process. That is how she said that she qualified to take this matter to the lawyers based on the information I furnished to her.' HL said that the lady later called and arranged for her to see a doctor, but not at that stage – which she placed in July or August 2015 – an attorney.

[188]If the version to which HL attested in court is correct, other than in respect of the dates, it seems to me to be a matter for investigation by the Legal Practice Council. While many rules have been relaxed to aid access to justice, such as the rules governing advertising and those permitting the charging of contingency fees, active touting is not permitted. Going to a school for disabled children to select promising cases for litigious purposes strikes me as something that should not be condoned. Rule 18(10) of the Code of Conduct for attorneys promulgated in terms of the Legal Practice Act 28 of 2014, provides that attorneys may not: 'buy instructions in matters from a third party and may not, directly or indirectly, pay or reward a third party, or give any other consideration for the referral of clients other than an allowance on fees to an attorney for referral of work.' Rule 18.22 provides succinctly that attorneys may not: 'Tout for work' and provides that: 90 'An attorney will be regarded as being guilty of touting for professional work if he or she either personally or through the agency of another, procures or seeks to procure, or solicits or for, professional work in an improper or unprofessional manner or by unfair or unethical means …' On the face of the appellant's evidence it appears that these rules may have been breached in this case. Conduct of the trial The pleadings

[189]The action commenced by the issue of summons on 2 September 2014. According to the medical reports in the record the only investigation of MML's condition at that stage was an MRI scan in respect of which a report by Prof Andronikou dated 20 August 2014 was available. The report's conclusion was that the MRI scan showed: 'Features are those of chronic evolution of a global insult to the brain due to hypoxic ischaemic injury, of the partial prolonged variety, most likely occurring at term.' As was to be expected the report said nothing about the cause of the injury or what, if anything could have been done by the hospital staff to prevent the injury or ameliorate its consequences.

[190]The necessary consequence of this was that the particulars of claim were based entirely on the imagination of the attorney who drafted them, rather than any endeavour to comply with the Uniform Rules governing pleadings. Rule 18(4) requires a pleading to contain a clear and concise statement of the material facts upon which the pleader relies for the claim, with sufficient particularity to enable the opposite party to reply thereto. The latter is obliged by rule 18(5) not to plead evasively, but to meet the point of substance. If that point does not emerge from the particulars of claim they cannot do this. 91

[191]In breach of Rule 18(4) the particulars of claim made no attempt to identify the facts giving rise to the claim. It consisted entirely of vague generalities summarised in the following paragraphs: 'During the course of 1 and 2 May 2005 the Plaintiff endured prolonged periods of labour in circumstances where more and/or alternative and dedicated medical attention, treatment and/or advice was required to ensure the safe birth of a healthy child and in particular a timely Caesarean Section. As a result of the prolonged labour, a lack of attention and medical care as may be reasonably required in the circumstances and in particular a failure to timeously perform Caesarean Section to deliver the Minor, the Minor suffered a hypoxic-ischemic insult due to peri-natal asphyxia and/or hypoxia, causing the Minor to sustain severe brain damage, as a result of which the Minor is permanently suffering from cerebral palsy and mental retardation ("the Complications"). The Complications occurred as a result of the negligence of the Defendant, alternatively as a result of the negligence of the Defendant's employees and/or representatives and/or agents, alternatively, as a result of the combined and cumulative negligence of the Defendant and the Defendant's aforesaid employees, representatives and/or agents.'

[192]The pleaded particulars of negligence were if anything even vaguer. It was alleged that 'the Defendant and/or the Defendant's aforesaid employees, representatives and/or agents' were negligent in one or more or all of the following respects: '10.1 they failed to employ and/or ensure medical attention by suitably qualified and/or proficient and/or experienced medical practitioners and/or nursing staff who would be available, able and/or capable to examine, treat and/or provide whatever reasonably required assistance and/or advice to the Plaintiff as may be reasonably required and/or appropriate regarding her labour and delivery, and in particular in respect of performing a Caesarean Section if and when required, either at the Hospital or at all; 92 10.2 they failed to ensure that such medical practitioners and/or nursing staff were in attendance at all material and relevant times; 10.3 they failed to employ and/or ensure medical attention by suitably qualified and/or proficient and/or experienced medical practitioners v nursing staff who were able to assess, monitor and manage the Plaintiff's labour and delivery; 10.4 they failed to ensure that the Hospital was suitably, adequately, appropriately and/or properly equipped to provide such medical attention as was reasonably required by the Plaintiff at all relevant times hereto, and in particular to allow the timeous and proper performance of a Caesarean section when it was required; 10.5 they failed to take any and/or any reasonably required steps to ensure the proper, timeous and professional assessment of the Plaintiff, her monitoring and management of labour and/or assistance during the Plaintiff's labour and her process of birth; 10.6 they failed to implement such steps as could and would reasonably be required to prevent the occurrence of the Complications; 10.7 they failed to avoid the Complications when by the exercise of reasonable care, skill and diligence they could and should have done so.'

[193]Not content with this exercise in obfuscation, the pleader added a further paragraph with eighteen sub-paragraphs of equally general allegations about the hospital staff, culminating with the allegation that they failed to prevent MML from suffering a hypoxic-ischemic incident, causing him to suffer severe brain damage, as a result of which he suffers from cerebral palsy and mental retardation, when by the exercise of reasonable skill, care and diligence they could have done so. One can allow a measure of generality in pleading allegations of negligence, but simply to allege everything the pleader can conjure up as potential negligence is unacceptable. There needs to be clarity as to the case being made and the nature of the impugned conduct on the part of the defendant, or those for whose conduct the defendant is said to be liable, who must at the least be identifiable. 93

[194]There is much to be said for the proposition that these particulars of claim could have been set aside as an irregular proceeding under Rule 30, but instead the defendant pleaded a bald and general denial of all these allegations. An attempt was then made by way of a detailed request for further particulars for trial to ascertain what the case was that the defendant had to meet. By way of example, para 4 of the request read: '4.1 What were the risk factors presented by the mother indicating the need for a caesarean section? 4.2 What were the risk factors presented by the unborn child indicating the need for a caesarean section? 4.3 Exactly in what time, during 1 st and 2 nd May 2005, was each indicator present? 4.4 Who was the surgeon who should have performed the caesarean section? 4.5 At what hospital was the operation to be performed? 4.6 The exact date and approximate time and duration of Plaintiff's stages of labour. 4.7 The exact date and approximate time and duration of Plaintiff's second stage of labour. 4.8 The exact date and approximate time and duration of Plaintiff's third stage of labour.' The utterly unhelpful response to the first of these questions was: 'In so far as the particulars sought are not contained in the Notices filed and to be filed by the Plaintiff in terms of Rule 36(9)(b), they remain matters for evidence.' That answer was repeated in response to a question whether MML suffered any birth injury during the delivery process and to the detailed questions in paras 10 to 27 of the request. The end result was that after this the defendant was no wiser as to the factual basis for the claim. Neither the defendant, nor the court, should be required to analyse the expert notices and accompanying reports in order to ascertain what the case is about. 94

[195]Two other questions and answers deserve mention. In response to a question about the qualifications that the relevant medical practitioners and nursing staff should have possessed – relevant given the allegations that they were not suitably qualified – the answer was: 'The Defendant is the party better suited to answer this question.' In response to questions about which doctors and nurses attended to HL and the date and time of her admission to a general ward, the maternity ward and the labour ward, the answer was: 'The Defendant is directed to the Thebe Hospital records held under number 840821 in the possession of the Thebe Hospital.' At the time this answer was given discovery had been made of the records that were available and HL's attorney knew that the remaining records were missing. The impression this leaves is that HL's attorneys were unwilling, or unable, to clarify their case.

[196]The pre-trial conferences did nothing to address the problem that the issues in dispute were wholly undefined. Both the pleadings process and the pre-trial procedures failed to serve their purpose of clarifying the issues in dispute between the parties. HL's legal representatives seem to have laboured under the misapprehension that everything was resolved by way of a series of joint minutes by medical experts consulted by the parties. I will revert to those later in this judgment, but for the present it suffices to say that minutes of experts are no substitute for a proper definition of the issues in the pleadings, preferably narrowed by the proper conduct of pre-trial conferences under rule 37. As matters stood, when the trial commenced all that could be said about the nature of the claim was that MML had cerebral palsy caused by a hypoxic-ischemic incident, which may or may not have occurred during labour. It was alleged that some unidentified member or members of the medical or nursing staff should have 95 prevented this by unspecified means, possibly including delivery by Caesarean section.

[197]It has on several occasions been said by this court that litigation is not a game. This case resembles nothing so much as a game commenced by hopefully kicking the ball of a summons into play, without any factual basis for a claim beyond knowing that MML had cerebral palsy caused by a hypoxic-ischemic incident and, if I may mix my metaphors, hoping Micawber-like that something would turn up in the course of pre-trial preparation. The end result was that, when counsel opened the case before the judge, he was unable to identify the issues to be decided crisply and coherently. Beyond saying that the insult suffered by MML had probably occurred intra-partum, he said nothing about the specific acts of negligence that were to be relied on, who was responsible for those acts, and what should have been done if there had been no negligence.

[198]This diffuse, unfocussed approach to the conduct of complex litigation is to be deprecated. If the issues are not properly and clearly defined the conduct of the trial cannot be controlled in a properly efficient manner. On appeal, by which stage the issues should have been clear and the alleged negligence defined in terms of the acts or omissions of specific individuals, HL's counsel contented themselves with saying that the appellant's case was that: '… as a result of her prolonged labour and the lack of attention and medical care she received, in particular the failure by the personnel to properly monitor her and the foetal in order to either expedite delivery or perform an emergency caesarean section delivery upon detecting foetal distress, [MML] suffered a hypoxic ischemic insult in the intrapartum period.' The heads of argument proceeded by saying that the trial focussed on an almost complete absence of hospital records and submitting that the res ipsa loquitur 96 principle should be applied. 82 The impression is that even at this stage the precise basis of the claim is uncertain.

[199]The remedy is straightforward. In any case where the pleadings and pre- trial procedures have not resulted in a clear statement of the issues, the trial judge should require the parties to deliver a statement of the issues in accordance with Rule 37A(9)(a), that is, a statement of what is not in dispute and a statement of what is in dispute, setting out the parties' respective contentions on those issues. If the matter is subject to judicial case management under that rule such a detailed statement is a requirement. If it is not, it is within the judge's powers, under Rule 38(8)(c) and their inherent power to regulate the proceedings, to require that such a statement be provided. Sequence of witnesses

[200]This is where the next problem arose. The first three witnesses for the appellant were experts – a professor of nursing, an obstetrician and a specialist paediatrician and medical geneticist. Only the last of these had consulted with the plaintiff and her son, and the usual documentary material that is sometimes a sufficient basis for the experts to consider and express their opinions was not available. The only available hospital records reflected that the appellant experienced a normal delivery with nothing untoward occurring. Her baby was delivered at 5.00 am on 2 May 2005 and she was discharged the following afternoon. Nothing untoward was noted in the Road to Health Chart of her clinic visits until some 18 months later. 82 Counsel cited Meyers v MEC, Department of Health, Eastern Cape, op cit, fn 15, paras 71, 80 and 82 in support of this submission, but the case is of no assistance. It dealt with a surgical procedure that had clearly gone wrong in a way that should not have occurred. In those circumstances the failure to provide an explanation for how the injury was suffered was held by the majority to support an inference of negligence. The law remains that, save in extreme and unusual cases, the principle has no application in medical negligence cases. Buthelezi v Ndaba, op cit, fn 17, para 16. 97

[201]The judge needed to determine when the insult that caused MML's cerebral palsy occurred; the underlying cause; whether foetal distress should have been detected during the appellant's labour and, if so, what should have been done about it. The experts were unable to answer these questions on the basis of an MRI scan; the scanty medical records, which reflected that his birth was normal and not attended by any complications; the examinations of MML in the presence of the appellant undertaken by Prof Solomons, Dr Gericke and Dr Kganane and the interview of the appellant by Dr Mogashoa. All of the experts based their reports and, where they gave oral evidence, that evidence, on information obtained from the appellant either directly, or by reference to Prof Solomons report and notes or a statement apparently provided by the appellant's attorneys. This underpinned all the opinions being expressed.

[202]The evidence of the appellant as to what had occurred was essential for the conduct of her case. Nine years had elapsed since MML's birth before the action was instituted and the trial took place four years after that. In the absence of medical records, it could not be expected that the nursing staff would have any independent recollection of the plaintiff, or MML's delivery, especially if the delivery was normal as reflected in the available records. Only the appellant could give any direct evidence. The reliability of her evidence was accordingly fundamental. If it was reliable then the opinions of the experts based upon it would be acceptable. If it was not, that evidence could not be accepted.

[203]The report of the first witness, Prof Nolte, dated 20 September 2017, was based on information provided by third parties in the form of a report by Prof Solomons and what she described as 'consultation records' of the plaintiff. The second witness, Dr Hofmeyr, whose report was dated 6 September 2017, relied on the Road to Health Chart, the Maternity Register, appellant's obstetric 98 Discharge Summary, a report on the MRI scan on MML by Prof Andronikou, the report of Prof Solomons and a 'factual statement' dated 31 August 2017 by the appellant. Neither the statement, nor the consultation records, which may have referred to the same document, were made available to the court.

[204]The third witness, Dr Gericke, whose report was dated 21 September 2017, had at least seen HL and examined MML on 31 August 2017. In addition, he relied on 'the personal injury claims consultation notes', which may have been the same document as was provided to Prof Nolte and Dr Hofmeyr, but was not disclosed, the Road to Health Chart, the report by Prof Andronikou and the report by Prof Solomons.

[205]Prof Solomons did not give evidence because the parties agreed, during the course of the trial and before the appellant testified, that the joint minute between him and Dr Griessel could stand as a record of what they agreed, subject to qualifications expressed by Dr Griessel. That agreement did not render anything other than Prof Solomons' opinions admissible. While his report was before the court, insofar as it contained factual matter on which the evidence of these three witnesses was based, it was inadmissible hearsay. The following facts set out in his report were accordingly not facts on which the other experts were entitled to rely without proof, namely that: (a) On 1 May 2005 at 01h00 HL presented with mild lower abdominal pain; (b) At 07h00 this had 'increased in severity' indicating that this was a continuation of the pain at 01h00; (c) HL's membranes ruptured at 11h00 that morning and she went to Thebe Hospital at 12h00 arriving at 12h30; (d) At the hospital she was assessed by nursing staff and was told that the baby was 'still far' and the cervix was 3 cm dilated; 99 (e) On one occasion a CTG was placed on her whilst she was in the labour ward, but she did not recall any abnormalities; (f) HL was assessed by the nursing staff at 18h00 and 20h00 on 1 May and at 01h00; 02h00 and 04h00 on 2 May 2005; (g) On each occasion she was told that the baby was 'far' and that she should not push, but lie on her left side; (h) The abdominal pains were severe from the previous evening; (i) The baby's head crowned at 04h45 and the mother shouted for assistance and three nurses came to help; (j) MML was born at 05h00 on 2 May 2005; (k) At delivery MML did not cry and two tubes were placed in his nostrils indicating nasal prong oxygen; (l) The nursing staff did not inform HL of MML's condition, which suggested that there was something to inform her about.

[206]Almost all of those facts depended upon HL. The only ones derived from the hospital records were that she had been admitted to the hospital on 1 May 2005 at about 13h00 'in labour' and that MML was born at 05h00 on 2 May 2005. Yet the basis for Prof Nolte's report was that labour commenced at 01h00 on 1 May 2015. She said that there was prolonged labour of about 28 hours duration. Her description of HL's labour was taken directly from Prof Solomons' report, although she omitted the assessments by the nursing staff at 01h00 and 02h00 that he reported. Did she in that regard rely on the statement by HL in preference to the report of Prof Solomons? We do not know. What we do know is, for example, that she was apparently unaware that HL would testify that she felt a minor pain on her bladder at 01h00 on 1 May, but it passed and she went 100 back to sleep until 07h00. Prof Nolte did not mention that and she could not be cross-examined on it. Nor did she or any of the experts mention that between 20h00 and midnight HL would say that she slept again and awoke because of a pain. The implications of this could not be explored because that evidence had not been given. Nor could any of them take into account her answer under cross- examination that she was not told on her initial examination at the hospital that she was 3cm dilated.

[207]The materiality of the appellant's evidence was apparent from the reports of these two experts. Dr Hofmeyr based her opinion on an absence of foetal monitoring during labour; inadequate maternal monitoring and probable prolonged labour. Her report and the joint minute compiled by her and Dr Schoon noted the absence of records and said that it hindered their ability as witnesses to fairly assess the circumstances surrounding the claim of obstetric negligence. Prof Nolte said that HL was in prolonged labour for 28 hours; that the active stage of labour was only 20 minutes; and that the nursing care was sub-standard. These opinions were all based on hearsay material said to emanate from the appellant.

[208]In those circumstances HL had to be the first witness in order to set the stage for the experts. When the trial commenced and appellant's counsel indicated that Prof Nolte would be the first witness, respondent's counsel objected that, because HL would not have given evidence, he did not know on which data the expert would base her evidence. He correctly pointed out that the facts on which the expert evidence was based needed to be admitted or proved. 83 That is clear. This court has said that 84 before any weight can be given to an expert’s opinion, 83 Coopers (South Africa) (Pty) Ltd v Deutsche Gesellschaft für Schädlingsbekämpfung MBH, op cit, fn 68, 1976 (3) SA 352 (A) at 371 A-H. 84 PriceWaterhouseCoopers Inc and Others v National Potato Co-operative Ltd and Another, op cit, fn 35, para 99. 101 the facts upon which the opinion is based must be found to exist and an opinion based on facts not in evidence has no value for the court.

[209]The judge responded to the objection by saying: 'Well your viewpoint might be that the evidence that is relied upon would be hearsay evidence and that you object thereto and you have got an assurance that the witness be called.' Counsel replied that this was an expert witness who had to report on a factual basis, or on some data, if there was to be any merit to their evidence. He pointed out that in her expert summary Prof Nolte referred to Prof Solomons' report and a 'kind of statement' of HL that someone else had drafted. He expressed concern over the value of evidence given on that basis.

[210]The judge then said that he could not tell the plaintiff how to call the witnesses and that experts rely on factual foundations provided to them so that if the factual foundation falls away the expert evidence is worthless. He then said, with counsel's acquiescence, that the hearsay evidence would be admitted provisionally and could be struck out later. Counsel for HL then intervened to say that the experts on both sides had used 'basically the same information' and, if a dispute arose, he would then call her. In other words, there was no certainty that HL would give evidence.

[211]This was not an appropriate way in which to conduct the trial. In my view the judge erred in his response to the objection. It was an objection to the expert witnesses being permitted to give evidence on the basis of factual hearsay. Given the fundamental importance of HL's evidence in this case, the objection should have been upheld. Until the factual basis for the experts' evidence had been established their opinions were inadmissible. Judging by his comment that he could not tell the HL's counsel in what order he should call his witnesses, the 102 judge regarded this as a matter of counsel's discretion in regard to the presentation of a case. I think that was wrong, as the objection raised issues of the admissibility of the experts' evidence. Rejecting it placed counsel for the respondent in an impossible position, where he was unable to challenge HL's experts on the basis that the facts on which they relied were not supported by her evidence. Nor could he test her evidence against the evidence by the experts of what they had been told, either by her or by Prof Solomons, or by the attorneys in the mysterious 'consultation record' 85 or the factual statement by HL dated 31 August 2017. 86 I leave aside for present purposes the question whether the obligation of the experts to set out the materials on which their opinions were based, meant that any privilege that might otherwise have attached to these documents was waived. 87

[212]In AM v MEC for Health, 88 another medical negligence case, I had occasion to describe the functions of an expert witness in the following terms: 'The functions of an expert witness are threefold. First, where they have themselves observed relevant facts that evidence will be evidence of fact and admissible as such. Second, they provide the court with abstract or general knowledge concerning their discipline that is necessary to enable the court to understand the issues arising in the litigation. This includes evidence of the current state of knowledge and generally accepted practice in the field in question. Although such evidence can only be given by an expert qualified in the relevant field, it remains, at the end of the day, essentially evidence of fact on which the court will have to make factual findings. It is necessary to enable the court to assess the validity of opinions that they express. Third, they give evidence concerning their own inferences and opinions on the issues in the case and the grounds for drawing those inferences and expressing those conclusions.' (Footnotes omitted.) 85 Prof Nolte. 86 Dr Hofmeyr. 87 It appears that extensive consideration has been given to this in the United States of America under Federal Rule 26(a)(2)(B). See Roger S Heydock and David F Herr Discovery Practice §5.03; Jerome G Snider, Howard A Ellins and Michael S Flynn Corporate Privileges and Confidential Information §3.08(2) at 3-49 to 3-52. 88 AM v MEC for Health, Western Cape [2020] ZASCA 89; 2021 (3) SA 337 (SCA) para 17. See also The Member of the Executive Council for Health, Eastern Cape v DL obo AL [2021] ZASCA 68 paras 10 and 11. 103

[213]In dealing with the necessity for the expert's opinion to be based on admitted or proved facts, the judgment continued: 89 The opinions of expert witnesses involve the drawing of inferences from facts. The inferences must be reasonably capable of being drawn from those facts. If they are tenuous, or far-fetched, they cannot form the foundation for the court to make any finding of fact. Furthermore, in any process of reasoning the drawing of inferences from the facts must be based on admitted or proven facts and not matters of speculation.' (Footnotes omitted.)

[214]There may be cases where it is permissible, or even necessary in order to set the scene for the court to appreciate the issues, for experts to give evidence at the outset of the proceedings when the factual evidence on which they base their opinions may still need to be led. That will ordinarily be so where the factual dispute is narrow and clear-cut and the expert can properly express an opinion on all relevant factual scenarios, without relying on disputed facts. This was not such a case and nor are most similar cases. 90

[215]It is not apparent whether HL's counsel adopted this approach as a tactical device, or with a view to meeting the convenience of expert witnesses, but in my view it was impermissible. Where the facts are central to the opinions of the experts, courts should require that those facts be led in evidence before the experts express their opinions. Primarily that is for the benefit of the court, which is thereby placed in a position where the expert's opinion can be assessed, and, if need be, queried or elucidated, in the light of the factual material before it. It is also conducive to fairness in cross-examination of the experts on behalf of the defendants. Where the case comes on appeal it facilitates a reading of the record. Lastly, if this principle is borne in mind and objections are upheld to leading the expert evidence without a proper factual foundation being laid, that should avoid 89 Ibid para 21. 90 AB obo KM v Member of Executive Council for Health, Eastern Cape [2018] ZASCA 141 paras 47-50. 104 situations, such as that in Madikane, 91 where the case was conducted entirely on the basis of expert evidence without any factual foundation at all for the opinions being expressed. Expert minutes

[216]It has become a practice in medical negligence cases for parties to arrange for the expert witnesses to meet and to file agreed minutes of their opinions. In some divisions of the high court this may be a requirement. It is a useful practice that may facilitate the running of the litigation by narrowing the issue and enabling the court and the parties to focus on the central issues in the case. That is reflected in the decision of this court in Bee. 92 That was a case involving the computation of damages for loss of past and future earnings. Forensic accountants were employed by the parties and they signed a joint minute setting out the facts on which they were agreed and the areas where they were unable to agree. At the trial the Road Accident Fund's forensic accountant sought to depart from the factual agreement by relying on a report not available at the time the joint minute was signed and using that to recalculate the agreed figures on which the joint minute had been based. Contrary to the agreement, he also sought to contend that there was no gratuitous element to the remuneration Mr Bee had been receiving since the accident.

[217]In Bee the majority judgment, authored by my brother Rogers AJA, rightly held that this could not be countenanced. The trial had been prepared and conducted under this head of damages on the limited issues identified in the joint minute. Those included an agreement as to the basis for calculating the loss of earnings of the business in which the plaintiff was involved and an agreement 91 Road Accident Fund v Madikane [2019] ZASCA 103. Molemela JA and I were both parties to that decision. 92 Bee v Road Accident Fund, op cit, fn 48, para 66. The judgment broadly endorsed the approach in Thomas v BD Sarens (Pty) Ltd [2012] ZAGPJHC 161 (Thomas). 105 that a proportion of his earnings after his injuries was gratuitous and paid only because it was a family business involving him and his brother. To permit a departure from that course would have required an adjournment and probably the filing of a supplementary expert's opinion. The decision was expressly based upon the need for fairness in the conduct of legal proceedings and the avoidance of trial by ambush.

[218]The effect of Bee in relation to the agreed minutes of experts in this case involved two misconceptions. The first related to the need to call the experts to give oral evidence in support of their opinions and, where experts were called, their entitlement to expand upon and explain the basis for their opinions. The second related to the weight to be attached to the opinions themselves.

[219]There appeared to be a perception, reflected in both the record and the heads of argument that such agreements are contractual in nature. The agreements were described as having been 'struck' and not having been 'repudiated'. That is the language of contract, and the give and take of negotiation, to arrive at a compromise. It is wholly inappropriate to describe the endeavours of independent experts to explain for the benefit of a court the matters on which they hold the same view and those on which they differ. That is why it was suggested in AM v MEC for Health 93 that the experts should be required to draft these minutes themselves and that the lawyers should play no part in that process.

[220]A clear distinction in principle needs to be drawn between factual evidence given by an expert witness and the opinions expressed by that witness. As to the former, there is no difficulty in applying Bee to the facts on which the experts agree, any more than there is a difficulty where the parties themselves reach 93 Op cit, fn 88, para 26. 106 agreement on factual issues. The opinions of the experts stand on a completely different footing. Unlike agreements on questions of fact, the court is not bound by such opinions. It is still required to assess whether they are based on facts and are underpinned by proper reasoning. Bee 94 endorsed a remark by Sutherland J in Thomas 95 that the occasions on which that occurs are likely to be rare, but that will only be in cases where the opinion is clear and there is nothing in the evidence to controvert it. Before a court accepts an opinion, it must pay close attention to the qualifications attaching to it. Furthermore, agreement by two experts on an opinion cannot preclude another expert with appropriate qualifications from expressing a different view, either in a report or in oral evidence. That is especially so when the third expert's views are based on their own speciality, which differs from that of the other two. The only constraint on that is that it should not result in unfairness to the party that has relied on the agreed opinion.

[221]This point can be illustrated by reference to the agreed minute signed by the specialist radiologists, Prof Andronikou and Dr Kamolane. The minute read: 'The radiologists agree that the MRI demonstrates features of chronic evolution of hypoxic ischaemic injury, of the partial prolonged variety, occurring in a brain of term maturity (≥ 37 weeks), probably occurring in the perinatal time-period and should therefore be correlated with clinical parameters by paediatric and obstetric experts for the establishment of a more exact time-period and causes that led to this result.'

[222]In accordance with her report, Dr Kganane said in evidence that MML had spastic diplegic cerebral palsy that was not typical of HIE. That had been recorded in her joint minute with Dr Gericke. When she explained this in her evidence in chief, she said that if there had been a profound insult to MML's brain during or around the time of his birth, his clinical features would have been different. This 94 Bee op cit, fn 48, para 64. 95 Thomas op cit, fn 92, para 13. 107 attracted an objection that this evidence was inconsistent with the agreement between the radiologists. Counsel's approach was that the radiologists' agreement confined the insult suffered by MML to the perinatal period and it was not open to Dr Kganane to question the severity of the insult at that time. But the agreed minute related to their opinion in regard to the period when the insult occurred, not to a question of fact. It was always clear that there was a dispute about when MML suffered the insult, and evidence that MML's condition was not typical of HIE was relevant to whether the judge should accept the joint opinion. In addition, Dr Kganane was expressing a view based on her own speciality and, as has been pointed out elsewhere, experts in the medical field do not operate in hermetically sealed compartments. The court is entitled to the full picture.

[223]Similarly, when Dr Mogashoa came to give her evidence, in accordance with her report and joint minute with Prof Solomons, both of which were in existence before the trial commenced, her evidence was objected to on the grounds that it was inconsistent with the joint minute between Prof Solomons and Dr Griessel. 96 The latter referred to MML's injury being 'at term' and Dr Mogashoa sought to explain that a 'term brain' was from 35 weeks, while a 'term pregnancy' was from 37 weeks.

[224]In pursuing this objection, counsel referred to item 6 of the Solomons/Griessel joint minute, which read: 'In the setting of absent medical records and maternal history of sucking and swallowing abnormality, timing of the partial prolonged hypoxic ischemic injury to the interpartum period cannot be excluded. – Agree. If disagree state reasons for same. Additional comment: D Griessel: The normal growth first year of life makes severe feeding difficulty unlikely.' 96 The objection is dealt with in paras 47-49 of my brother's judgment. 108

[225]Counsel relied on the first part as having been agreed, but this disregarded the qualification added by Dr Griessel, which challenged the factual basis for the opinion. As this court recently pointed out, the necessary corollary to an agreement in a joint minute limiting the issues on which evidence is necessary is that, where there is no agreement, the minute can be disregarded and, if a party wishes to pursue the disputed point, evidence will be necessary. 97

[226]In Huntley v Simmons 98 Waller LJ said in relation to expert minutes that: 'The evidence of experts is important evidence but it is nevertheless only evidence which the judge must assess with all other evidence. Ultimately issues of fact and assessment are for the judge. Of course if there is no evidence to contradict the evidence of experts it will need very good reason for the judge not to accept it and he must not take on the role of expert so as to, in effect, give evidence himself. So far as Joint Statements are concerned parties can agree the evidence but (as happened in this case) it can be agreed that the joint statements can be put in evidence without the need to call the two experts simply because they do not disagree; but either party is entitled to make clear that the opinion expressed in the joint statement is simply evidence that must be assessed as part of all the evidence.'

[227]Reference to the record indicates that this was the basis upon which the joint minute of Prof Solomons and Dr Griessel was placed before the trial court. Counsel for HL said that 'the defendant accepts and abides by those agreements, as qualified by Dr Griessel'. Counsel for the MEC confirmed that: 'we agree that the joint minute can stand, as per between Dr Solomons and Dr Griessel. Important is that Dr Griessel did make some qualifications that must be included of course.'

[228]It is clear that the joint minute was simply evidence of the opinions of the two signatories to it, subject to the qualifications raised by Dr Griessel, and was to be taken into account along with all other evidence bearing upon the issue of 97 MEC for Health and Social Development, Gauteng v MM on behalf of OM [2021] ZASCA 128 para 16. 98 Huntley (aka Hopkins, by his litigation friend) v Simmons [2010] EWCA Civ 54 para 7. 109 when the insult causing MML's condition occurred. That evidence included the expert testimony of Dr Kganane and Dr Mogashoa as well as the factual evidence derived from the testimony of the factual witnesses and the medical records that were available.

[229]In summary, the position in regard to agreements between experts, is as follows. In accordance with Bee, if they agree on issues of fact and the appropriate approach to technical analysis, the litigants are bound by those agreements, unless they have been withdrawn in circumstances where no prejudice results, or any prejudice can be cured by an adjournment or other means. If the experts have reached agreement on a common opinion on a matter within their joint expertise, that is merely part of the total body of evidence. The court must still determine whether to accept the joint opinion. The existence of that agreement between the experts will not ordinarily preclude evidence that qualifies or contradicts their opinion, unless the case has been conducted on the basis of the agreement and the admission of that evidence will prejudice the other party in a manner that cannot be cured. If the parties choose to place an agreed minute before the court reflecting both shared opinions and areas of disagreement and do not call the parties to the minute to deal with the areas of disagreement, the minute will do no more that reflect that there is disagreement on the point. While it is for the parties to determine which witnesses they call, if they fail to call the authors of a joint minute they cannot object when other witnesses express views that qualify or dissent from the views in the minute.

[230]The existence of joint minutes may not be used to prevent witnesses from explaining the reasons for the conclusions expressed in the minute. For example it would have been most helpful for one or both of Prof Andronikou and Dr Kamolane to have explained how they arrived at the view that the injury occurred in the peri-natal period. That is the sort of question that a court would 110 ask in order to understand the degree of certainty about this opinion. They could also have been asked to comment on Dr Mogashoa's view that the nature of MML's disability was more consistent with injury occurring to the preterm brain and inconsistent with hypoxia. The passage from AM cited in para 212 identifies the second purpose of expert evidence as being 'to provide the court with abstract or general knowledge concerning their discipline that is necessary to enable the court to understand the issues arising in the litigation'. The existence of a joint minute of experts cannot be used to prevent that function from being fulfilled, whether by the experts who were party to the minute or by another expert. The decision in Bee does not relate to the admissibility of expert opinions, but to the fairness of the trial. Expert opinion evidence should only be excluded when it impacts adversely on the latter.

[231]My final point is that the joint minute does not render the whole of the expert's report admissible in evidence. Unless the expert gives evidence, or it is agreed that the report will be admissible, it remains inadmissible. The deficiencies in a joint minute cannot be resolved by reference to the report of the expert. As the trial judge remarked in Huntley 99 a joint minute is a useful document, but by its nature it is never more than a summary. Conclusion

[232]Had the issues in this case been properly narrowed prior to trial, the evidence been led in its correct sequence and the function and standing of joint expert minutes been properly appreciated, I venture to suggest that the proceedings would have been curtailed and would have been completed with far greater expedition than a ten day trial, with a six month adjournment after the first week. 99 Ibid, para 12. 111

[233]As noted at the outset of this judgment I concur in the judgment of Makgoka JA. _______________ M D J Wallis Judge of Appeal 112 Appearances For Appellant: G J Strydom SC (with him A Viljoen) Instructed by: M.E.D Attorneys, Johannesburg McIntyre & Van der Post, Bloemfontein. For Respondent: J Y Claasen SC (with him L Manye) Instructed by: State Attorney, Bloemfontein.