---

 

CONSTITUTIONAL COURT OF SOUTH AFRICA

 

Case CCT 202/20

 

In the matter between:

 

NVM obo VKM                                                                                                      Applicant

 

and

 

TEMBISA HOSPITAL                                                                                First Respondent

 

MEC FOR HEALTH AND SOCIAL DEVELOPMENT

GAUTENG PROVINCE                                                                       Second Respondent

 

Neutral citation:      NVM obo VKM v Tembisa Hospital and Another [2022] ZACC 11

 

Coram:                      Zondo ACJ, Madlanga J, Madondo AJ, Majiedt J, Mhlantla J, Pillay AJ, Rogers AJ, Theron J, Tlaletsi AJ and Tshiqi J

 

 

Judgments:             Majiedt J (minority): [1] to [87]

                                    Rogers AJ (majority): [88] to [114]

                                    Zondo ACJ (concurring): [115] to [117]

 

Heard on:                 17 August 2021

 

Decided on:             25 March 2022

 

Summary:                 Jurisdiction — section 167(3)(b) of Constitution — sole issue one of factual causation — constitutional considerations not affecting resolution of issue — case not raising point of law — Court’s jurisdiction not engaged

 

 

 

ORDER

 

 

 

On appeal from the Supreme Court of Appeal (hearing an appeal from the High Court of South Africa, Gauteng Local Division, Johannesburg):

1.         Leave to appeal is refused.

 

 

JUDGMENT

 

 

MAJIEDT J (Madondo AJ, Pillay AJ and Tlaletsi AJ concurring):

 

Introduction

[1]          This case concerns the birth of a child, VKM (V) who, as a result of severe oxygen deprivation for a sustained period shortly before delivery, suffered a serious brain injury that manifested in a form of cerebral palsy.  A claim for damages on his behalf by his mother, Ms NM, was successful in the High Court of South Africa, Gauteng Local Division, Johannesburg (High Court), but was reversed on appeal to the Full Court of that Division (leave to appeal having been granted by the Supreme Court of Appeal).

 

[2]          In this Court, the sole issue on the merits is the question of factual causation.  As will appear presently, negligence was conceded in the High Court and the startling attempt by the respondents to withdraw that concession in this Court is untenable.

 

Background

[3]          In the High Court, Ms NM and several medical experts on her behalf testified in her case.  The respondents adduced no evidence at the trial.  There were, however, several agreed joint minutes prepared by the parties’ various experts.  This considerably narrowed down the issue of factual causation.  On the common cause facts, Ms NM was admitted to Tembisa Hospital, the first respondent, on 3 April 2009.  V is her first child.  The hospital records indicate that on that day she was first examined at 15h30, and again at 18h00 by a doctor.  By 21h30 she was 2 cm dilated and the foetal heart rate was 144 bpm and at 23h10, the foetal heart rate was 132 bpm.  At that juncture, she was transferred to the labour ward.  There, the nursing staff compiled a partogram[1] to track the progress of Ms NM’s labour as well as of the foetal heart rate.[2]  The partogram reflects three entries for the foetal heart rate on 4 April 2009:

 

(a)          At 01h15: 120 bpm;

 

(b)          At 02h15: 115 bpm; and

 

(c)         At 03h15: 128 bpm, when the amniotic fluid was clear, a sign of foetal well-being.

 

[4]          No further foetal heart rate monitoring occurred after 03h15.  The failure to monitor was conceded in the High Court to have constituted a negligent omission on the part of the nursing staff.  That concession appears to have been based on the obstetrics experts for both sides having agreed that the lack of monitoring was sub standard.  At 04h45, Ms NM was examined by either a doctor or a midwife.  It is recorded in the hospital notes that she was fully dilated and they reflect a diagnosis of cephalic-pelvic disproportion (CPD), which means that the baby’s head was too big for the mother’s pelvis.  The notes show that, at that time, Ms NM was booked for a caesarean section.  Ultimately, the caesarean section did not take place.  V was delivered naturally in the ward at 05h10.

 

[5]          V was born with very low Apgar scores.[3]  The readings were three at one minute after birth, three at five minutes after birth, and five at 10 minutes after birth.  On the uncontested medical evidence, the low score of three at one minute after birth was indicative of severe respiratory and heart function challenges, since it ought to be between eight and nine in a healthy new-born baby.[4]  The score of five at 10 minutes after birth was indicative of intrapartum hypoxia.[5]  It is undisputed that V suffered an acute profound hypoxic ischemic injury to his brain in the latter stages of labour, known as hypoxic ischaemic encephalopathy (HIE).[6]  This was confirmed by a cranial ultrasound performed on V on 8 April 2009 and by magnetic resonance imaging (MRI) taken when V was seven years and four months of age.  In V’s case, the HIE developed into cerebral palsy.  A follow-up examination on V at age three indicated that V had cerebral palsy, epilepsy and breakthrough seizures.

 

[6]          Ms NM sued on behalf of V and also on her own behalf for damages.  The claims emanated from the aforementioned injuries sustained by V, and that sustained by Ms NM which was caused by the alleged unlawful conduct of a nurse who applied extreme pressure to Ms NM’s abdomen, forcing her to give birth to V naturally.  It was conceded that this was not a standard medical procedure.  As stated, her claims were successful, but they were overturned on appeal to the Full Court.

 

Litigation history

High Court

[7]          The trial in the High Court proceeded on the question of liability only.[7]  Pursuant to the concession on negligence, the only issue in the High Court was causation.  That Court noted that there was very little in dispute between the experts on the basis of their joint minute.  It also noted that very little was placed in issue during the cross- examination of the experts who testified for Ms NM.  The High Court recorded the facts agreed upon between the parties’ paediatric neurologists, obstetric experts and specialist nursing experts.

 

[8]          The High Court had particular regard to the unchallenged expert evidence of Dr Pistorius (an obstetrician and maternal foetal medicine specialist) and Prof Kirsten (a neonatologist) on behalf of Ms NM that:

 

(a)          V suffered an acute profound hypoxic insult;

 

(b)          In babies, it is difficult to pinpoint when the hypoxia started in the absence of a known traumatic event, or as is often referred to, the sentinel event (like a prolapsed cord or a ruptured uterus).  This is why the monitoring of the foetal heart rate during labour is important, as it can give an indication of when the hypoxia commenced;

 

(c)          During the active phase of labour, the midwife must assess the foetal heart rate and response to contractions every 30 minutes.  This will allow for changes to be identified.  The slowing of the heart rate is a sign of hypoxia.  Before the onset of the slowing of the heart rate, there will be changes in the pattern on the cardiotocograph (CTG).[8]  In other words, there will be warning signs, which may include meconium in the amniotic fluid;

 

(d)          Towards the end of the hypoxic episode, the foetus will have a very slow heart rate and delivery needs to be done quickly.  If delivery is done quickly enough, it may be possible to avoid a hypoxic ischemic episode and the consequent brain abnormalities;

 

(e)          Once warning signs are detected, midwives should put in place emergency measures to “buy time” for the foetus while preparations are made for urgent delivery.  This includes moving the mother onto her left side and administering oxygen to her.  The midwife should also call the doctor to consider whether medication should be administered to suppress the mother’s contractions, which affect the flow of oxygen to the foetus;

 

(f)        If these measures are introduced, the foetal heart rate can be improved before an emergency caesarean section is performed.

 

[9]          That Court was satisfied that, according to the experts, the only possible cause of the brain injury on the facts was the acute profound hypoxic injury that V sustained intrapartum.  It was also accepted by the parties’ respective obstetric experts, Dr Pistorius and Dr Koll, that this occurred during the one and a half-hour period “from 03h15 until 04h45 (4 April 2009) where there [was] no recording of foetal monitoring during active labour”.

 

[10]       The High Court had regard to the joint minute by both obstetric experts.  That minute plays an important role in determining factual causation, so it is prudent to quote its relevant parts in full:

 

“4.        At 03h15, there was still evidence of foetal well-being (clear amniotic fluid and normal foetal heart rate), which makes a sentinel event severe enough to cause an acute profound hypoxic event during the active phase of labour until that time unlikely;

5.         It is therefore likely that an acute profound hypoxic event occurred in the time from 03h15 until 04h45 where there is no recording of foetal monitoring during active labour;

 

            . . .

 

7.         It is doubtful whether it would be possible to perform a caesarean section quickly enough to prevent the neurological sequelae of an acute profound hypoxic event in this time interval.”  (Emphasis added.)

 

[11]       The High Court understood paragraph 7 of the joint minute to mean that, when read in the context of Dr Pistorius’s evidence as a whole, since there was no monitoring during the critical period, there would not have been time, once monitoring resumed, to take measures to avoid the brain injury to V.

 

[12]       The High Court found that it was also clear from the uncontested evidence of Prof Kirsten and Dr Pistorius that foetal monitoring every 30 minutes is critical to detect warning signs of a possible hypoxic episode.  This was not done and, consequently, the nursing staff was unable to notice the warning signs that would probably have been evident had monitoring of the foetal heart rate been done as required.  The High Court took the view that this omission resulted in Ms NM being denied the correct treatment that would have bought time for V while urgent steps were taken to speed up the delivery in order to prevent the injury to V’s brain.  Thus, with proper monitoring the warning signs would have been detected, and with the proper emergency measures V’s brain injury would probably not have occurred.

 

[13]       On this basis, the High Court concluded that Ms NM had established causation and her representative claim succeeded.  The High Court also upheld the claim in Ms NM’s personal capacity.

 

Full Court

[14]       With leave of the Supreme Court of Appeal, the matter went on appeal to the Full Court which, in a unanimous judgment, upheld the appeal.[9]  The Full Court reasoned that on the common cause facts, more particularly the obstetric experts’ joint minute, it was unlikely that there had been an event of sufficient severity to cause an acute profound hypoxic event before 03h15, given the fact that there was foetal well being until 03h15 on 4 April 2009.  They agreed that the acute profound hypoxic event must have occurred between 03h15 and 04h45 (the critical period).

 

[15]       The Full Court formulated the central question thus:

 

“[T]he question becomes whether, had there been adequate monitoring, warning signs would have been picked up and that there was then enough time to engage proper emergency measures which would have avoided the brain injury.”

 

It observed that this very question had presented “in many of these cases, including AN”.[10]

 

[16]       The Full Court took the view that the High Court had failed to apply the “but for” test for factual causation.  It suggested that the Court had erred in adopting, instead, the approach set out by this Court in Lee.[11]

 

[17]       The Full Court rejected the High Court’s reasoning in respect of the emergency measures.  This was because, from the point that the foetal distress was discovered at 04h45, there was not enough time to carry out an emergency caesarean section.  Therefore, the emergency measures would not have prevented the harm to V.  It argued thus:

 

“The possibility of successfully carrying out a caesarean section, and whether it would have yielded positive results, after a CPD was diagnosed at 04h45, was not seriously explored by the court a quo.  Such a possibility is what the ‘but for’ test is all about.  That possibility was canvassed with Dr Pistorius, when he testified in chief.  His evidence is that from 04h45, which is the time when a diagnosis for caesarean section would have been made, a number of standard protocols would have been expected to be undertaken to prepare for the procedure.  The list is lengthy and would have included: obtaining informed consent, preparing the patient for theatre, including applying intravenous infusion, preparing the theatre to receive the patient and to carry out the procedure, and securing the attendance of an anaesthetist, a doctor and an assistant.  Where there is a suspicion or conclusion, so he testified, of a form of foetal distress, intra-uterine resuscitation would be performed.

 

Dr Pistorius was of the view that in terms of international standards, the hospital staff had 30 minutes, from the time the decision to refer the patient for caesarean section which in this case was 04h45, to perform an emergency caesarean section.  Given that the baby was delivered vaginally at 05h10, the 25-minutes window of opportunity would have been insufficient to perform the emergency procedure.  Hence the conclusion that: ‘it is doubtful whether it would be possible to perform a caesarean section quickly enough to prevent the neurological sequelae of an acute profound hypoxic event in this time interval’.”[12]

 

[18]       This finding led the Full Court to conclude:

 

“It is undisputed that the defendant was negligent.  It is also accepted that an acute profound hypoxic event took place between 03h15 and 04h45.  When a determination was made at 04h45 to perform a caesarean section, on the evidence, there was insufficient time to carry it out.  By parity of reasoning, the plaintiff in our view has failed to show that the negligent conduct, which has been isolated as lack of sufficient monitoring, had a causal effect on the neurological sequelae.”[13]

 

[19]       Ultimately, in upholding the appeal, the Full Court concluded that “the plaintiff has not shown that the negligence caused the child’s condition: the circumstances that caused the cerebral palsy occurred too late to have taken steps that would as a matter of probability have prevented the cerebral palsy”.[14]

 

This Court

Applicant’s main submissions

[20]       The applicant supports the findings and underlying reasoning of the High Court.  In respect of the test for factual causation, it is contended that the High Court in fact applied the “but for” test and did so correctly, whereas it is the Full Court that obfuscated the test in its approach.

 

[21]       The approach of the Full Court is subjected to trenchant criticism, particularly its alleged impermissible reliance on the factual findings in AN and M v MEC.[15]  The contention is that, having outlined the “similar facts” in M v MEC, the Full Court applied “parity of reasoning” and concluded that the applicant, qua plaintiff, had failed to prove that the negligent omission was causally linked to the neurological sequelae.  In this regard, it erred by not following the legal precedent of a higher court, but in effect its factual precedent.

 

[22]       The applicant submits that the respondents’ case has transmogrified in this Court.  It is contended that, instead of attempting to defend the Full Court’s judgment, the respondents advance new propositions.  The respondents now seek to make out a case of a sudden, unexpected sentinel event having occurred at approximately 04h40 and of the damage to V’s brain having ensued within 10 to 30 minutes thereafter.  The applicant submits that this new theory postulated by the respondents is not only contrary to the facts, which are mostly common cause or not seriously disputed, but also misconstrues the experts’ evidence.

 

[23]       Thus, the applicant contends that the High Court’s reasoning is sound and asks this Court to endorse that reasoning and the resultant outcome.

 

Respondents’ main submissions

[24]       The respondents submit that the failure to monitor the foetal heart rate at 03h45 and 04h15 on the morning of 4 April 2009 was not the cause of the acute profound insult.[16]  They refer to Prof Kirsten’s evidence that the cause of the insult was a sudden, unpredictable, unidentified sentinel event.  The central question, according to the respondents, is whether, even if the foetal heart rate was monitored at 03h45 and 04h15, the hospital staff could have taken measures that would have prevented or minimised V’s brain injury.

 

[25]       According to the respondents, Prof Kirsten’s evidence shows that it takes 10 to 30 minutes before delivery for brain injury to occur.  That means that in the present instance, the insult is likely to have occurred around 04h35 to 04h40.  And the obstetricians agree that the insult happened between 03h15 and 04h45.  That, postulate the respondents, is consistent with the insult occurring around 04h35 to 04h40.

 

[26]       The respondents advance the following sequence of events during the critical period on Prof Kirsten’s evidence: the insult occurred and then continued until delivery.  Brain damage set in 10 to 30 minutes after the insult occurred.  That means that the insult must have happened between 10 to 30 minutes before delivery.  It is unlikely that the insult happened before then.  If it did, V would not have survived.  As it turned out, V did survive, but only just.  V’s severe depressed state after birth, recorded in the clinical observations, confirms Prof Kirsten’s conclusion, namely that delivery happened just before the baby would have died if he were not delivered.  That confirms that the insult is likely to have occurred 10 to 30 minutes before delivery.

 

[27]       According to the respondents, the time interval referred to in paragraph 7 of the obstetricians’ joint minute, is not the period between 03h15 and 04h45.  It is in fact the time period it takes to inflict damage to the foetal brain from the time of the acute profound insult.  They contend that the significance of the agreement by the obstetricians is that, if (as suggested by Prof Kirsten) brain damage takes 10 to 30 minutes from the time of the insult, there was no time to prevent or minimise brain damage by carrying out a caesarean section.

 

[28]       The respondents seek to withdraw the concession on negligence made at the trial.  They contend that the concession was wrong since a conclusion, rather than a fact, was conceded.  That conclusion, they submit, is not consistent with the law.  According to them, negligence “in the air” does not result in delictual liability.  Insofar as it was conceded that hospital staff did not monitor the foetal heart rate at 03h45 and 04h15, they contend that negligence does not follow.  Consequently, according to them, the concession should not stand.

 

[29]       The respondents contend that, in applying the test for causation endorsed by this Court in De Klerk[17] and Mashongwa,[18] the failure to monitor the foetal heart rate at 03h45 and 04h15 was not the reason that a caesarean section (the only measure reasonably available to hospital staff) was not performed timeously, so as to prevent brain damage.  Even if the insult was detected as it happened or within a short while after it happened, and a caesarean section ordered immediately or reasonably soon after detection, that would not have prevented brain the damage.

 

[30]       In sum, the respondents submit that a proper application of the test for factual causation to the facts ought to lead to the conclusion that the hospital (first respondent) could not have prevented or minimised brain damage, even if the foetal heart rate had been monitored at 03h45 and 04h15.

 

Jurisdiction and leave to appeal

[31]       In order for this Court to grant leave to appeal, two requirements must be met.  First, the matter must fall within the jurisdiction of this Court, in that it raises a constitutional issue or an arguable point of law of general public importance.[19]  And second, the interests of justice must warrant the granting of leave to appeal.[20]

 

[32]       There are a number of possible bases of jurisdiction.  First of all, the Full Court impermissibly imported the facts in cases of the Supreme Court of Appeal and applied them in this case to make findings on factual causation.  In doing so, to the extent that the Full Court adjudicated the present case by measuring it against the factual circumstances, rather than legal principles, of precedent, requires this Court to consider its approach.  It raises the question whether a court is permitted to elevate the findings of another court – even one of higher jurisdiction – on the facts of a separate and unrelated matter into law.  That implicates the section 34 right to a fair hearing.

 

[33]       It is true that some of the questions in this matter are factual, and factual issues do not usually engage this Court’s jurisdiction.  However, in Metrorail,[21] this Court held:

 

“[W]here, however, a separate constitutional issue is raised in respect of which there are disputes of fact, those disputes of fact will constitute ‘issues connected with decisions on constitutional matters’ as contemplated by section 167(3)(b) of the Constitution.  On many occasions, therefore, this Court has had to determine on appeal the facts of a matter in order to determine the constitutional claim before it.  Were it to be otherwise, this Court's ability to fulfil its constitutional task of determining constitutional matters would be frustrated.”[22]

 

[34]       The present provision on constitutional jurisdiction in section 167(3)(b) of the Constitution no longer has the phrase “issues connected with decisions on constitutional matters”.  However, the excision of this phrase from the section cannot mean that this Court has since been divested of the power to decide factual issues connected to issues that it has jurisdiction to determine.  In that case, this Court would be hamstrung in the conduct of its function.  I therefore conclude that the approach remains the same.

 

[35]       The question is whether the constitutional issue is a separate issue which incidentally involves factual disputes, in which case it may engage jurisdiction.  Where a factual finding is what directly implicates constitutional rights, that will be sufficient to ground jurisdiction.  In Metrorail, this Court identified that there was “a welter of factual disputes on the papers”,[23] but was nevertheless satisfied that jurisdiction was engaged on the basis that the question it was seized with was whether the conduct of Metrorail was “reasonable” and:

 

“Unlike the question of whether a particular issue has been established beyond a reasonable doubt, which turns only on an evaluation of evidence and its cogency, the question of whether conduct is reasonable in the context of a legal duty, requires the application of legal principles to a set of established facts.”[24]

 

Accordingly, the factual issues were not dispositive of the matter.

 

[36]       The factual disputes in this case are not material and do not negate the jurisdiction of this Court, as they mostly relate to differences in the parties’ interpretation of the expert evidence, which this Court is privy to.  The very nature of the main issue, being the interpretation and application of the test for factual causation, means that this Court must engage with the facts and evidence in the record to some degree.  As stated in Metrorail, therefore, these “disputes of fact will constitute ‘issues connected with decisions on constitutional matters’”.[25]

 

[37]       Additionally, in Mashongwa, this Court held that where a delictual claim is underpinned by constitutional rights and the state’s duty to take reasonable measures to respect, protect, promote and fulfil these rights, this Court’s jurisdiction will be engaged:

 

“[A]lthough it may not look like the outcome turns on the meaning or vindication of any constitutional provision or right, sections 7(2) and 12(1)(c) of the Constitution are the pillars on which the superstructure of this case rests.  Mr Mashongwa’s claim owes its origin largely to the obligations imposed on PRASA, an organ of state, by these provisions.  In addition, an enquiry into wrongfulness ‘focuses on the conduct and goes to whether the policy and legal convictions of the community, constitutionally understood, regard it as acceptable’.  On these bases this Court does have jurisdiction in terms of section 167(3)(b)(i) of the Constitution.

 

This Court also derives jurisdiction from the realisation that this matter raises an arguable point of law of general public importance, which deserves the attention of this Court.  In this country, trains are generally used by the overwhelming majority of people who fall within the low income bracket.  These are the proverbially voiceless and in reality vulnerable members of our society.  Furthermore, incidents of crime on trains and related issues have in the past been sufficiently raised before our courts to warrant a pronouncement by this Court.  The safety and security of the poor people who rely on our train network to go to work or move from one place to another does raise an arguable point of law of general public importance.”[26]

 

[38]       Importantly, in that case this Court accepted that its jurisdiction was engaged in terms of sections 167(3)(b)(i) and 167(3)(b)(ii). It held that, although at first blush the matter did not appear to turn on constitutional issues, it in fact did.  The constitutional issues in that matter were central to its determination, while the factual issues were ancillary.  That was held to be sufficient to engage the jurisdiction of this Court.

 

[39]        Regarding this Court’s extended jurisdiction, it emphasised the general importance of public safety as it relates to issues of negligence on the part of organs of state.  This Court went as far as to say that although the point of law was not novel, having been dealt with in Van Duivenboden^[27] and Van Eeden,^[28] it was still necessary to address it because it raises an arguable point of law of general public importance.  And, it continued, the public “needs a pronouncement by this Court on whether PRASA can be held delictually liable for its failure to provide safety and security measures”.  These findings are apposite, as here the factual issues are framed by the right of access to healthcare services.  Cerebral palsy cases which raise the question of medical negligence are, unfortunately, seen quite frequently in our courts.  Similarly, these matters are of great public importance, and the question of the delictual liability of medical professionals in this context also requires a pronouncement by this Court.

 

[40]       As in Mashongwa, on the face of it, the issues in the present matter appear to entail purely factual disputes.  However, the delictual claim here is also underpinned by constitutional rights, such as the right of access to healthcare services[29] and the state’s duty under section 7(2) of the Constitution to “respect, protect, promote and fulfil the rights in the Bill of Rights”.  On that basis, analogous to Mashongwa, this Court’s jurisdiction is engaged.

 

[41]       While Mashongwa was decided on wrongfulness and jurisdiction was partly established on the basis of wrongfulness,[30] this Court explicitly held further that a delictual matter that appears factual in nature may still engage this Court’s jurisdiction where constitutional rights “are the pillars on which the superstructure of th[e] case rests”, and particularly where state obligations are involved.[31]  It follows, therefore, that this Court’s jurisdiction is usually engaged where there is a delictual claim against the state – as long as it is undergirded by constitutional rights.  It is notable that in Mashongwa, the point that a consideration of wrongfulness establishes jurisdiction is prefaced by the words “in addition”.[32]  The plain meaning is that the point on wrongfulness is an additional, separate basis for jurisdiction, and that what was described directly beforehand is the first basis for jurisdiction.  What appears before the point on wrongfulness is the point that the Court’s jurisdiction is engaged where a delictual claim against the state is underpinned by constitutional principles.

 

[42]       In Alexkor,[33] this Court held that it had jurisdiction to determine anterior issues relating to land rights, dispossession and restitution – including the nature and the content of the land rights that the Richtersveld Community held in the subject land prior to annexation and whether such rights survived annexation – notwithstanding that such issues were factual in nature.[34]  And it subsequently amended portions of the Supreme Court of Appeal’s order.

 

[43]       The cases adumbrated demonstrate a more nuanced view of this Court’s jurisdiction with respect to factual issues.  First, where a matter does not only involve constitutional issues, but is framed by constitutional issues, this may be enough to establish jurisdiction and address related factual issues.  Secondly, while it is generally accepted that this Court’s jurisdiction will not be engaged on factual disputes,^[35] where there are disputes of fact that are ancillary to constitutional issues, this Court may deal with them, as they are “issues connected with decisions on constitutional matters”.  The Court has to grapple with such factual issues if they are inextricably connected to constitutional issues, so that it would be impossible to resolve these constitutional issues without also resolving the factual issues.  Finally, the limits to the jurisdiction of this Court are not inflexible, and this is important given the necessity for this Court to meet the evolving needs of context in the interests of justice.

 

[44]       In this matter, the legal question goes beyond the ordinary application of factual causation, since what must be considered is the flexibility of the test for factual causation, and how it may be used to accommodate a set of facts that is inherently subject to uncertainty.  Cases of medical negligence involve such facts, given the great uncertainty that exists in any – or at least many – medical treatment cases.

 

[45]       This legal question – concerning the flexibility of the test for factual causation – is closely connected to the right of access to healthcare services in this case.  It is a centrally important constitutional entitlement, and it bears noting that there is a dearth of jurisprudence on the content of the right to access to healthcare services,[36] besides instances where the state had refused to act in particular cases.[37]  There have not been cases regarding the standard of care required under section 27(1).  However, Ngwenya convincingly argues that inasmuch as section 27 does not define the quantity or quality of health care services to be accessed, it is open to courts to develop principles to ensure that the state diligently adheres to the spirit and intent of the Constitution.[38]

 

[46]       Furthermore, in this case, I would add, specifically the protection of the best interests of the child is of paramount importance. This points to the important role of the courts in giving meaning to section 27, and fortifies this particular ground of jurisdiction.  Questions of accountability and responsiveness in a healthcare system that is able to meet constitutional standards, must surely raise constitutional issues.  As a result, questions of medical negligence in state-operated hospitals that implicate the rights of women and children to healthcare and the rights of new-born babies to have their best interests protected, certainly engage this Court’s jurisdiction.

 

[47]       Lastly, there is a legal point of general public importance that requires this Court’s attention.  As stated, judging from reported cases, many cerebral palsy cases appear to be brought before the courts.[39]  This case, one of many, concerns the test for factual causation in relation to negligent omissions by public healthcare workers, when they care for mothers in labour.  Having admitted negligence, the question arises whether, as a matter of public policy, the state as provider of healthcare services ought to be allowed to escape liability for its omissions.  The Full Court plainly misapplied the test for factual causation expounded in Lee.  Although this Court has emphatically in Mashongwa and De Klerk clarified the misconceptions that inexplicably followed its judgment in Lee, some confusion still abounds.  The question of whether and how the flexible test for factual causation should be applied still does not yield answers that are clear and consistent.  The differing approaches in M v MEC and in AN attest to that.  It is in the interests of justice that this confusion be laid to rest by this Court, and this forms an additional ground of jurisdiction.

 

[48]       I have read the judgment by my Brother Rogers AJ and I disagree with his reliance on Booysen,[40] as this matter is distinguishable.  In Booysen the issue was far narrower.  This matter goes beyond mere factual findings or an incorrect application of the law.  The factual questions are framed by the constitutional right of access to courts and the right of access to healthcare services.  Additionally, this matter is not a mere application of an established legal test for factual causation, but is a consideration of how the test should be applied within the broader scheme of rights violations in healthcare.

 

[49]       To conclude on jurisdiction – it is clear that the factual disputes in this matter are both underpinned by the right of access to healthcare, and are also ancillary to this constitutional issue.  The question of delictual liability for medical negligence is squarely located amongst the tools of citizens seeking to uphold their section 27 rights.  More specifically, the test that is used for factual causation in medical negligence cases will likely determine whether it is at all possible for delictual liability to follow, given the uncertainty which is all but endemic in this context.  Therefore, the question of how to apply the test in medical negligence cases is an arguable point of law with constitutional implications.  There are factual issues to consider, but these are ancillary to the broader constitutional issues, so that this Court is empowered to consider them.  If this Court were to draw a hard line in this regard – which it has not in the past – this would severely limit its powers to determine important constitutional and broader legal questions, which would otherwise engage its jurisdiction.  It would also likely limit the adaptive potential of this Court, as it must constantly turn its attention to evolving legal questions which map social developments, and were likely not considered by lawmakers.  That is why this Court is clearly empowered to determine its own jurisdiction, and why there is not – and can never be – a closed list of what is a constitutional issue.

 

[50]       For the reasons expounded, this Court has jurisdiction in this matter.  The approach of the majority is of some concern – it is hard to discern how a case of this nature and on these type of facts would ever engage this Court’s jurisdiction.  Law would not serve the cause of justice on that approach.

 

Merits

The law: general requirements of delict

[51]       The requirements for a successful claim in delict are well-established.  A plaintiff must prove positive conduct or an omission, causation, wrongfulness, fault and harm.  As stated, the only issue in the High Court was causation, more particularly factual causation.  The hospital staff’s failure to conduct adequate monitoring of the foetal heart rate between 03h15 and 04h45 during the day of V’s delivery was conceded in the High Court to have been a negligent, wrongful omission.

 

[52]       The startling attempt by the respondents in this Court to withdraw that concession is based on fallacious grounds.  The respondents’ contentions are that “a conclusion (rather than a fact) was conceded.  That conclusion is inconsistent with the law.”  They submit that “negligence ‘in the air’ does not result in liability in delict”.  They contend that the hospital staff’s failure to monitor the foetal heart rate at 03h45 and 04h15, contrary to the Maternal Guidelines, does not constitute negligence.  Consequently, they say that the concession should not stand.

 

[53]       The concession was clearly a conclusion on the facts, drawn after having regard to the respondents’ own expert reports and the largely unchallenged evidence of the applicant’s experts, more particularly Prof Kirsten and Dr Pistorius.  In its judgment, the High Court recorded the concession in these terms:

 

“The defendants initially defended the action on two material bases: it denied that there was negligence on the part of the relevant hospital staff, and it pleaded that there was no causal connection between any negligence established at trial, and V’s cerebral palsy.  However, in the defendants’ heads of argument they conceded the following:

 

‘The obstetric experts agree that the inadequate monitoring was sub standard, more especially because the midwives did not record any monitoring (of the foetal heart) after 03:15, leading to the conclusion that there was no monitoring.  Their negligent conduct accordingly consists of an omission.’”[41]

 

[54]       The High Court recorded further:

 

“In the circumstances, the issue of negligence is no longer in dispute between the parties.  The defendants have accepted that the failure to monitor the foetus from 3h15 constituted negligence.  As the defendants point out in their heads of argument, the question is whether this negligent conduct caused V’s cerebral palsy.  In other words, the sole issue is that of causation.”[42]

 

The concession cannot be more unequivocal than this.  That remained unchanged in the Full Court, where it was recorded that “[t]he court a quo found, correctly in our view, that negligence had been conceded before it and that the only issue to consider was causality”.[43]  Self-evidently, negligence is but one of the requirements for delict.  On its own, it does not establish liability.  It is trite that a plaintiff must prove further that the proved negligence, be it positive conduct or an omission, is causally linked to the harm.  If that is what the respondents meant by “negligence in the air”, there can of course be no quarrel with it.  But to seek to withdraw an informed, unequivocal concession now on the basis postulated is an altogether different matter.  It is deeply flawed and must be rejected.

 

[55]       What was conceded was not, as the respondents seek to persuade us, “negligence in the air”, but negligence through a pertinent and plain omission – a clear failure to perform the monitoring that reasonable nursing staff would have performed in a maternity ward in respect of a mother in active labour.  The concession cannot just be jettisoned willy nilly, without any adequate, satisfactory explanation for it.  That brings me to factual causation.

 

Factual causation

[56]       It is trite that the enquiry into factual causation asks the question whether the wrongful conduct or omission was a factual cause of the loss.  After citing Siman,[44] this Court in Lee described that enquiry as follows:

 

“The enquiry as to factual causation generally results in the application of the so-called ‘but for’ test, which is designed to determine whether a postulated cause can be identified as a causa sine qua non of the loss in question.  This test is applied by asking whether but for the wrongful act or omission of the defendant the event giving rise to the loss sustained by the plaintiff would have occurred.”[45]

 

And in the Appellate Division, in Bentley, Corbett CJ enunciated that enquiry thus:

 

“The enquiry as to factual causation is generally conducted by applying the so-called ‘but for’ test, which is designed to determine whether a postulated cause can be identified as a causa sine qua non of the loss in question.  In order to apply this test one must make a hypothetical enquiry as to what probably would have happened but for the wrongful conduct of the defendant.  This enquiry may involve the mental elimination of the wrongful conduct and the substitution of a hypothetical course of lawful conduct and the posing of the question as to whether upon such a hypothesis the plaintiff’s loss would have ensued or not.  If it would in any event have ensued, then the wrongful conduct was not a cause of the loss; aliter, if it would not have ensued.”[46]

 

[57]       In applying this test, no mathematical or scientific exactitude is required.  As this Court said in Lee:

 

“Application of the ‘but for’ test is not based on mathematics, pure science or philosophy.  It is a matter of common sense, based on the practical way in which the ordinary person’s mind works against the background of everyday-life experiences.”[47]

 

[58]       In Lee, this Court emphasised that the test is not inflexible and had to make provision for situations where “the use of the substitution of notional, hypothetical lawful conduct for unlawful conduct in the application of the “but for” test for factual causation” may lead to an injustice.[48]  This Court held that in some circumstances factual causation would be established where the plaintiff has proved that, but for the negligent conduct, the risk of harm would have been reduced.[49]

 

[59]       In Mashongwa, this Court explained that Lee never sought to replace the pre existing common law “but for” approach to factual causation, but rather to recognise the flexibility in the “but for” test.[50]  It held that where the traditional “but for” test was adequate to establish causation, it may be unnecessary to resort to the Lee test.[51]  That has been confirmed in De Klerk.[52]

 

[60]       Where Mashongwa states that the Lee test can be used where the ordinary “but for” test is not suitable, this should not be read to suggest that a litigant can try both tests in turn.  The appropriateness of the Lee test must be determined independently of whether the test succeeds.  In this context, the ultimate consideration is the interests of justice.  By this, I do not mean that the interests of justice test should not be applied if the traditional “but for” test fails, and there are no other factors to consider.  Rather, there are categories of harm which are better suited to the Lee test, and these categories may be determined in an abstract fashion, and based on the interests of justice.  An example would be systemic state failures, as seen in the Lee case itself,[53] where the harm was caused by dispersed, overlapping and polycentric institutional omissions.  The nature of these factors was such that one cause of harm could not be identified, but the interests of justice nevertheless demanded that the state be held liable for the harm suffered by Mr Lee.  Another example may be environmental harms, which exist within ecological systems, those complex webs of interacting natural and social components.  For example and purely hypothetically, there may be no single cause of the abysmal air quality in large parts of the Highveld,[54] but coal-fired power stations certainly contribute to the problem, and circumstances may exist in which they should be held accountable for this.  In this way, the flexibility identified in Lee allows the “but for” test to be expanded to meet the reality of certain categories of social harm which may not have been envisioned when the test was first developed.

 

[61]       If used consciously, it may be that the Lee test can add to the transformative potential of delict – and might even allow us to reconsider the nature of harm.  Matsuda has written about the many factors that commonly contribute to social harms, but are not recognised in either our social or legal systems.[55]  For example, mass shootings in the United States of America are generally attributed to the individual shooter, while enabling legal systems around gun use, media portrayals of violence and school culture are not considered.  Whether and how the law should intervene in broader social systems is a question that must be answered contextually, but for the moment it might be said that a flexible causation test may allow for systems of harm to be seen more completely.

 

[62]       I refer to Lee and the support it found in both Mashongwa and De Klerk, to emphasise the flexibility of the test for causation.[56]  As this case demonstrates, there is an apparent continuing confusion about the impact of Lee, an aspect which not only engages this Court’s jurisdiction as I have said, but also requires discussion on how that flexible test ought to be applied.  Thus, although in the present matter the High Court, despite the Full Court’s misconceived criticism, correctly applied the traditional “but for” test, the application of the more flexible Lee test would lead to the same result.  The question is whether factual causation is established where probable cause is shown, or whether it is enough to show that there is an increase in risk.  As stated, Lee suggests that it is enough to prove contribution to risk to establish factual causation.[57]  As far as systemic failures are concerned, and these failures contribute to or increase risk, factual causation can be established.  In this context, probable causation and risk reduction are closely linked – an increase in risk is the probable cause of the harm, but there are too many factors in the system to identify one cause of the harm.  That would provide a further basis, beyond the “but for” test, correctly applied by the High Court, to find factual causation here.  In this way, this case demonstrates that medical negligence is one of the categories of harm that may be better covered by the more flexible test for factual causation, as described above.

 

Was factual causation proved?

[63]       In applying the law to the proved facts, it is convenient to commence with the approach adopted by the Full Court.  As stated above, its approach elicited trenchant criticism from the applicant’s counsel.  And rightly so.  The nub of its impugned reasoning follows after the Full Court’s evaluation of the issues, the judgment of the High Court and the evidence.  It is necessary to record the reasoning in some detail.

 

[64]       The Full Court’s reasoning has already been alluded to in respect of what it saw as the central question relating to factual causation.  And the rest of its reasoning and conclusion has also been outlined.  A disturbing feature is that the Full Court appears to have placed significant reliance on the two judgments of the Supreme Court of Appeal referred to earlier, the majority judgment in M v MEC and the unanimous judgment in AN.  That reliance did not concern the legal principles that emanate from those dicta, but on what the Full Court regarded as “comparable facts”.

 

[65]       Self-evidently, every case must be decided on its own facts – that principle is so well established that no authority need be cited for it.  The perils of ignoring that trite principle emerge starkly here.  Firstly, as the applicant’s counsel correctly submitted, in AN, the sentinel event was a compressed umbilical cord.  The Supreme Court of Appeal said that:

 

“[A] sudden, sustained, total interruption to the blood supply caused by cord compression occurred in this matter.  For the sake of convenience, I shall refer to this as the sentinel event.  This caused the damage.”[58]

 

And that Court importantly pointed out:

 

“Unlike the other sentinel events listed above, a cord compression cannot be detected after the fact.  This is because the cord does not remain compressed when the pressure on it is alleviated.  It leaves no indication that it was compressed.  Any pressure on the cord ceases once a baby has been born.”[59]

 

[66]       The “other sentinel events”, alluded to by the Supreme Court of Appeal in the passage cited, include placental abruption, uterine rupture, umbilical cord prolapse, shoulder dystocia or maternal collapse.  That Court explained that, on the evidence before it, each of these sentinel events are capable of subsequent verification, since they leave some trace, what was termed “a footprint”.[60]  Thus, the common cause sentinel event in that case was determinable – the sudden, complete and ongoing interruption of blood supply to the foetus’ brain, caused by the occlusion by the compressed umbilical cord.  That common cause fact was agreed upon by the experts in that case, notwithstanding the absence of evidence afterwards of that cord compression, or a “footprint”.  And there the time of the sentinel event could not be determined.  Not so here.  The evidence cannot establish a known traumatic event and the experts were unable to suggest one.  What they could unequivocally agree on, however, was that “an acute profound hypoxic event occurred”, and that this must have been during the critical period.

 

[67]       There is a second – even more telling – distinction between the facts of this case and AN.  In the latter, there was no clear, acceptable evidence that there would have been prior warning signs of the total interruption to the blood supply.  The Supreme Court of Appeal was unable to make that finding on the evidence and on authoritative, peer-reviewed literature referred to by both parties.  In fact, that literature pointed in the opposite direction.  In the present instance, the experts are agreed that a slowing foetal heart rate is a sign of the onset of hypoxia.  The unchallenged evidence of Prof Kirsten and Dr Pistorius confirmed that fact.

 

[68]       A related distinction is whether adequate monitoring would have detected the warnings.  Here the uncontested evidence is that there would be warning signs before the hypoxia, and if emergency measures were put in place to “buy time” for the foetus and the delivery was done quickly enough, it might be possible to avoid a hypoxic ischemic episode and the consequent brain damage.  In AN, that Court asked, whether there were in all probability no prior warning signs (unlike here)—

 

“the issue [was] whether, when the sentinel event occurred, there would have been sufficient time to avoid the damage by expediting the delivery.  The obvious first factor in this enquiry is that counsel for the appellant candidly admitted that it could not be proved when the sentinel event occurred.  Without being able to do so, it could not be said at what time monitoring would have alerted the staff to this event.”[61]

 

As a consequence, that Court found that it was not proved that there would have been sufficient time in which to deliver the baby so as to avoid damage, and causation could not be determined.[62]

 

[69]       There is also one crucial difference between the facts of this case and those in M v MEC.  It is this.  In the latter instance, the majority found on the evidence that the baby suffered an HIE “immediately before delivery”.[63]  In the matter before us, the HIE occurred during the critical period, thus well before V’s delivery at 05h10.

 

[70]       In sum, the differences between the factual scenarios in this case and in AN, and those in M v MEC, are not only striking, but they are also material when it comes to factual causation.  This Court reminded us in Lee that “there is no magic formula by which one can generally establish a causal nexus.  The existence of the nexus will be dependent on the facts of a particular case.”[64]  But the Full Court erred even further when it sought to overturn the High Court on this aspect insofar as the application of the correct test is concerned.  The Full Court held that the High Court had erroneously adopted the reasoning of the minority in M v MEC in applying the test enunciated in Lee, instead of the conventional “but for” test.  The Full Court adumbrated the majority and minority judgments and “by parity of reasoning” on the facts in M v MEC, which it thought “are to an extent the same as those in casu”, adopted the majority reasoning expounded in that case.[65]

 

[71]       There are two difficulties with this approach.  First, the High Court unequivocally adopted the “but for” test in its reasoning.  At the outset it correctly identified the only remaining issue as factual causation, after the respondents’ concession on negligent omission (inadequate monitoring) and wrongfulness.  In respect of that remaining issue, the High Court correctly eliminated in its mind the common cause negligent omission, substituted in its place lawful conduct and asked the question whether the harm to V would still have ensued.  Based on the undisputed expert evidence, particularly that of Prof Kirsten and Dr Pistorius, the High Court answered that question in the negative.  It held that, if Ms NM had been properly monitored, the warning signs of foetal distress would have been detected and appropriate action would have been taken.  This included emergency measures to “buy time” for the foetus and for a caesarean section to be done.  Had proper monitoring occurred, there would, on the probabilities, have been enough time to take the emergency measures and the harm to V would probably have been averted.

 

[72]       The reasoning of the High Court is quintessentially a “but for” approach: the mental elimination of the failure to conduct adequate monitoring and its replacement with the hypothetical correct conduct (proper monitoring) to ask whether the harm would on the probabilities still have ensued.  It found on the facts that the answer was “no”.  Had there been adequate monitoring, the foetus’ slowing heart rate would have been detected, symptomatic of the onset of hypoxia, and emergency measures could have been taken to “buy time” for the foetus while preparations were made for a caesarean section to be performed.  There is no suggestion of flexibility in the test, no consideration of a probable cause.  The Full Court was wrong in associating the approach of the High Court in this matter with the one set out by this Court in Lee.  The High Court’s reasoning was classical “but for” stuff as it has conventionally evolved over time.

 

[73]       The second difficulty is that the doctrine of stare decisis in our law, that courts must follow the precedent of higher courts, is based on findings of law, not findings of fact.  In Walters,[66] this Court cited the following passage in Hahlo and Khan to explain the doctrine:

 

“In the legal system the calls of justice are paramount.  The maintenance of the certainty of the law and of equality before it, the satisfaction of legitimate expectations, entail a general duty of Judges to follow the legal rulings in previous judicial decisions.  The individual litigant would feel [themselves] unjustly treated if a past ruling applicable to [their] case were not followed where the material facts were the same.  This authority given to past judgments is called the doctrine of precedent.”[67]  (Emphasis added.)

 

[74]       While this Court in Walters does not explicitly make the point that stare decisis applies to legal findings and not factual findings, there is never any question of factual findings being followed by lower courts.  Instead, this Court discussed “legal interpretations”, stating for example:

 

“High Courts are obliged to follow legal interpretations of the Supreme Court of Appeal, whether they relate to constitutional issues or to other issues, and remain so obliged unless and until the Supreme Court of Appeal itself decides otherwise or this Court does so in respect of a constitutional issue.”[68]

 

[75]       To bolster the point, this Court in Walters also referred to Shabalala,[69] where the Court made reference to stare decisis in the context of the interpretation of constitutional provisions (referring to the interim Constitution):

 

“I appreciate that section 4(1) of the Constitution provides that ‘[t]his Constitution shall be the supreme law of the Republic . . .’ and that section 4(2) provides that ‘[t]his Constitution shall bind all . . . judicial organs of state at all levels of government’; but those provisions do not in my view mean that the established principles of stare decisis no longer apply.  Such an approach would justify a single Judge departing from a decision of a Full Bench in the same Division because he considered the interpretation given to the Constitution by the Full Bench to be in conflict with the Constitution, with resultant lack of uniformity and certainty until the Constitutional Court, whose decisions in terms of section 98(4) bind, inter alia, ‘all judicial organs of state’, had pronounced upon the question.”[70]

 

[76]       In Turnbull-Jackson, this Court also explicitly held that the principle of stare decisis[71] applies to the ratio decidendi[72] of a judgment, explicating that “[t]he doctrine of precedent decrees that only the ratio decidendi of a judgment, and not obiter dicta,[73] have binding effect”.[74]  It is true that this finding on the precedential value of ratio decidendi was contrasted with obiter dicta, rather than factual findings, but the same principle applies.

 

[77]       It is trite that a decision on a legal principle made by a superior court must be followed by all courts of lower or equal status until it is overruled or modified by a court of higher authority.  The doctrine is self-evidently intended to provide certainty on principles of law.  But findings of fact stand on a different footing.  Those findings are unique to the facts adduced before a court in a particular case.  They cannot be ferried wholesale to another case, regardless of how closely the cases may resemble each other.  On first principles, the obvious bears repetition: each case must be decided on its particular facts.  And the position is no different in respect of medical expert evidence.  Medical experts must base their opinions on the facts in the particular case.[75]

 

[78]       The Full Court’s erroneous approach implicates the right of access to courts entrenched in section 34 of the Constitution.  The reason is simple: the applicant did not have the opportunity, through counsel, to test the opinions of the experts indirectly relied upon by the Full Court.  Those experts testified in M v MEC and in AN, not in this matter.  And yet, the Full Court placed reliance on it through “parity of reasoning”, purely by reason of the fact that the facts in those cases were “to an extent similar” to those of the present matter.  Recently, this Court in Van der Walt made these remarks where medical literature was relied on for an accused’s conviction in circumstances where he did not have an opportunity to engage with that literature during the trial:

 

“The relevant question is whether the applicant had the opportunity to challenge the textbook evidence.  The applicant was plainly denied that opportunity.  Likewise, not knowing that such evidence would be relied upon, he was denied the opportunity – if so minded – to adduce controverting evidence.  The right to challenge evidence requires that the accused must know what evidence is properly before the court.  In the applicant’s case, the medical literature relied upon was never adduced at all.  This goes to the heart of a fair trial.”[76]

 

[79]       For the reasons that follow, I regard the High Court’s approach to be sound and its findings on factual causation to be unassailable.  The common cause and initially conceded facts, the unchallenged expert evidence, particularly that of Prof Kirsten and Dr Pistorius, together with the joint minute of Drs Pistorius and Koll, establish the following:

 

(a)          The foetus was evidently in a healthy condition at 03h15, judging by the clear amniotic fluid and the apparently normal foetal heart rate.

 

(b)          During the critical period, between 03h15 and 04h45, there was no foetal heart rate monitoring at all by the hospital staff.  Monitoring should have occurred at half hourly intervals, that is, at 03h45 and again at 04h15, as Ms NM was in active labour by that time.

 

(c)          V suffered an acute profound hypoxic injury intrapartum during the critical period.

 

(d)          There were no emergency measures adopted by the nursing staff – none were recorded in the hospital records and none were mentioned at the trial.

 

(e)          Absent any foetal heart rate monitoring during the critical period, no warning signs of a possible hypoxic event were capable of being detected.  This was very properly conceded at the trial to constitute a negligent omission and the attempt to withdraw that concession in this Court falls to be rejected.

 

(f)        Had the monitoring been done, on the probabilities the hospital staff would have picked up the warning signs (that probably would have been present) to indicate foetal distress caused by hypoxia.  In the face of these warning signs, the staff would on the probabilities have taken urgent steps to “buy time” for V and to make arrangements for an urgent caesarean section so as to prevent the injury to V’s brain.  It is probable that with the proper emergency measures, V’s brain injury would not have occurred.

 

[80]       The argument by the respondents in this Court, that the sentinel event must have occurred at between 04h40 and 04h45, is not only completely novel, but is also unsustainable on the evidence.  At its core, that argument is based on a misconception of the sequence of events as outlined in the evidence.  The argument misconstrues Prof Kirsten’s evidence.  And it attaches an erroneous interpretation to paragraph 7 of the joint minute.

 

[81]       The exact time of the sentinel event is unclear and not capable of more precise ascertainment, exactly because of the inadequate monitoring.  The respondents’ argument appears to confuse the various stages: there is first the sentinel event, evident from the foetal heart rate having dropped markedly and taking too long to recover; then follows the damage to the brain; and thirdly, foetal bradycardia ensues, with the foetal heart rate constantly below 110 bpm.  Properly understood, Prof Kirsten’s evidence bears this out.  He first expounds the sentinel event (citing examples), then the damage to the brain and finally the foetal bradycardia.  The evidence is:

 

“V had [an] acute profound hypoxial insult and in babies . . . it is very difficult to pinpoint exactly when the hypoxia actually started.  It is easy if there is cord prolapse.  You know when the cord prolapse[s] or if [there is a] ruptured uterus, but during labour it is not that easy and that is why foetal heart rate monitoring and foetal frequent observations are so important . . . .  And then the duration of the actual final bradycardia[77] varies and it depends on if there is total loss of oxygen provided to the foetus.  If there is an abruption[78] . . . a total complete abruption that lady will have signs of [bradycardia] within 10 minutes . . . but in other situations where the cause for the acute profound hypoxia is not clear cut . . . [t]he foetal bradycardia can last for 30 minutes, but that is the foetal bradycardia.  The slow heart rate.  If there is hypoxia the heart muscle requires oxygen.  So the final sign will be that the foetus will have a very slow heart rate, but before the onset of that slow foetal heart rate the foetus from the CTG will have changes in the pattern of the CTG.”

 

[82]       This evidence is to the effect that foetal bradycardia takes 10 to 30 minutes.  Thus, counsel for the respondents is wrong in the assertion that brain damage occurs 10 to 30 minutes after the sentinel event.  This is the mistaken basis on which an estimation is then made that there was only 10 to 30 minutes available to “buy time” for the foetus and that, based on (a) the time of delivery at 05h10; and (b) the distressed state of V after birth (as Prof Kirsten opines, “he was delivered very close to foetal death”), this is consistent with the insult occurring around 04h35-04h40.  And, ultimately then, counsel’s contention is that “[i]f (as according to Prof Kirsten) brain damage takes 10 to 30 minutes from the time of the insult, there was no time to prevent or minimise brain damage by carrying out a caesarean section procedure”.  Thus, so the argument for the respondents goes, even with proper monitoring, the harm would probably not have been averted.

 

[83]       The entire construct of this reasoning, which culminates in the respondents’ counsel’s ultimate contention, is thus fatally flawed, proceeding as it does from a misconception.  Read in context and in full, Prof Kirsten’s report together with his oral evidence, firmly and finally dispels that misconception.

 

[84]       The reliance on paragraph 7 of the joint minute of Drs Pistorius and Koll is likewise based on a false premise.  The two experts stated: “[i]t is doubtful whether it would be possible to perform a caesarean section quickly enough to prevent the neurological sequelae of an acute profound hypoxic event in this time interval.”  The reference to “this time interval” is to the period from 04h45.  That is the clear implication of that entry when the joint minute (particularly point 5) is read as a whole and Dr Pistorius’ addendum report is considered.  He notes:

 

“There was clearly insufficient monitoring during the latent and active phase of labour.  No ‘sentinel event’ was recorded, but a sentinel event would easily have escaped notice, given the insufficient monitoring.  The available evidence indicates that there was suboptimal care during labour, resulting in foetal asphyxia and subsequent hypoxic ischemic encephalopathy, which would have been avoided by appropriate monitoring and action.”

 

[85]       Moreover, Dr Pistorius’ evidence on this aspect was left unchallenged and the reading of this entry on the basis now postulated by respondents’ counsel was never broached with him in cross-examination.  Both the High Court, when this point was raised before it in the application for leave to appeal, and the Full Court, saw the entry in the joint minute in the correct light and not on the misconceived basis now advanced in this Court on behalf of the respondents.  This is the scenario in which the “buying of time” for the distressed foetus starts running from the time when the slow recovery of the foetal heart rate is picked up by adequate monitoring – it does not relate to 04h45 when CPD was diagnosed and the caesarian section was booked.  When Dr Pistorius’s evidence is viewed in its entirety, the passage from the addendum report meant that given that there was no monitoring, there would not have been time once monitoring resumed, to take measures to avoid the medical consequences to V.  In this regard, therefore, counsel for the respondents’ contentions are devoid of merit.

 

[86]       A plaintiff is not required to show a causal connection between the conduct or omission and the eventual harm with certainty.  All that is required is “to establish that the wrongful conduct was probably a cause of the loss, which calls for a sensible retrospective analysis of what would probably have occurred, based upon the evidence and what can be expected to occur in the ordinary course of human affairs rather than an exercise in metaphysics”.[79]  And cerebral palsy cases where the brain damage is caused by HIE, like the present one, fall into a particularly challenging field of medicine, “where medical certainty is virtually impossible”.[80]  It is a form of harm that calls for a more flexible understanding of factual causation.

 

[87]       In this matter, the applicant adduced sufficient evidence to prove factual causation, in the context of a harm which is replete with uncertainties.  Absent any countervailing evidence from the respondents, the unchallenged evidence of the applicant’s medical experts, particularly that of Prof Kirsten and Dr Pistorius, together with the admitted facts and the joint minute of the obstetricians, proved the applicant’s claim for damages.  The negligent failure by the hospital staff to conduct adequate monitoring of the foetal heart rate during the critical period, denied them the opportunity to detect the warning signs of the onset of hypoxia.  That, in turn, resulted in the failure to take emergency measures to afford V more time until a caesarean section could be arranged.  On the probabilities, the brain injury would not have occurred had all of this been done, or the risk of this brain injury would have been significantly reduced.  In the premises, had I commanded the majority, I would have upheld the appeal with costs.

 

ROGERS AJ (Madlanga J, Mhlantla J, Theron J and Tshiqi J concurring):

 

[88]       I have had the pleasure of reading the judgment by my colleague Majiedt J (first judgment).  I disagree that this case engages this Court’s jurisdiction.  In order for a case to be a “constitutional matter” within the meaning of section 167(3)(b)(i), the resolution of a constitutional issue must be reasonably necessary in order to determine the case’s outcome.  Similarly, a case only “raises an arguable point of law” within the meaning of section 167(3)(b)(ii) if the answer to that question is reasonably necessary to determine the case’s outcome.  A peripheral constitutional issue or arguable point of law is not a justification for embarking on a factual reappraisal of a case where the reappraisal is not rendered reasonably necessary by the answer to the constitutional issue or arguable point of law.

 

[89]       This Court has consistently held that it does not have jurisdiction to decide purely factual matters, and this is so even where a lower court has gone badly wrong on the facts.  The Court can analyse evidence and make factual findings where the determination of such facts is reasonably necessary in order to answer or to give practical effect to the Court’s decision on the constitutional matter or arguable point of law, but not otherwise.  As this Court said in Mbatha in relation to its constitutional jurisdiction, it will only engage in contested factual issues if they are “connected with a well grounded constitutional issue”.[81]

 

Sections 27 and 7(2) of the Constitution, accountability and responsiveness

[90]       The first judgment holds that this case is a constitutional matter because it implicates the health care rights guaranteed by section 27 and the State’s related duty under section 7(2) to respect, protect, promote and fulfil the rights in the Bill of Rights.  The first judgment posits that, “[q]uestions of accountability and responsiveness in a healthcare system that is able to meet constitutional standards, must surely raise constitutional issues”; and that questions of medical negligence in public hospitals, implicating the “rights of women and children” and “the rights of new born babies to have their best interests protected, certainly engage this Court’s jurisdiction”.[82]  These are far-reaching propositions with which I disagree.

 

[91]       It is not in dispute that the applicant had the right to have access to health care services.  She had access to them at Tembisa Hospital.  It is not in dispute that Tembisa Hospital had a private-law duty to provide her with a reasonably competent level of care, the breach of which would be wrongful.  It was also common cause that the service provided to the applicant fell below a reasonably competent standard, in other words that the Tembisa Hospital staff were negligent.  The negligence related to the absence of FHR monitoring between 03h15 and 04h45.  There is no dispute that the required standard was half-hourly FHR monitoring.  The issue is whether the wrongful and negligent conduct caused the injury suffered by the applicant’s baby, and that is a purely factual question.  Sections 7(2) and 27 of the Constitution, and considerations of accountability and responsiveness, shed no light on its answer.

 

[92]       To a greater or lesser extent, the rights guaranteed in the Bill of Rights cover the whole field of human existence.  Almost any case could be framed as touching on one or other fundamental right.  This is not enough to make the case a constitutional matter.  This is shown by Boesak.[83]  A sentence of imprisonment, following upon a conviction that was not justified by the evidence, might be said to implicate the convicted person’s right not to be deprived of freedom without just cause (section 12(1)(a)) and his right to a fair trial (section 35(3)), yet a contention that the conviction was not justified on the evidence is not a constitutional matter but a factual one.[84]  If section 27 were implicated in the present case, it would apply to every medical negligence case, even though the issues raised by the case were purely factual.  The same would be true, analogously, of every defamation case, on the basis that it implicates the right to freedom of expression guaranteed by section 16.

 

[93]       The first judgment calls in aid this Court’s decision in Mashongwa.[85]  Paragraph 13 of the latter judgment should not be parsed as if it were a statute; it must be understood in the context of what was in issue, namely the legal question whether a transport utility ought to be held delictually liable for damages that flow from a breach of its public-law duty to provide safety and security measures for its rail commuters.[86]  In other words, the key disputed issue was delictual wrongfulness.  It was in that context that the “pillars on which the superstructure” of the claimant’s case rested were identified as sections 7(2) and 12(1)(c) of the Constitution.  Additionally, the enquiry into wrongfulness, in accordance with this Court’s case law, focused on “whether the policy and legal convictions of the community, constitutionally understood” regarded the impugned conduct as acceptable.  It was these considerations in combination which gave this Court jurisdiction.

 

[94]       In Mashongwa, sections 7(2) and 12(1)(c) were thus, along with the policy and legal convictions of the community, constitutionally understood, matters which featured centrally in the Court’s assessment of whether PRASA’s public-law duty should be matched by a private-law duty, as is apparent from the Court’s reasoning.[87]  It was in that sense that those two sections of the Constitution were the “pillars on which the superstructure” of the claimant’s case rested.

 

[95]       The present case is quite different.  Here it has from the outset been common cause that Tembisa Hospital owed the applicant a private-law duty to provide reasonable care.  And in the respects relevant to this case, the standard of care (that is, the negligence aspect) is also common cause.  Neither the High Court nor the Full Court was called upon to examine sections 7(2) and 27 of the Constitution in order to decide whether they justify imposing a private-law duty on Tembisa Hospital or to decide where the standard of reasonable care should be set.  That remains the position in this Court.  The only question is whether the admitted breach of the required standard factually caused the baby’s injury, a question which sections 7(2) and 27 do not help us to answer.

 

[96]       The point can be illustrated with reference to the type of case with which Mashongwa dealt.  Following the decision in that case, it is settled law that PRASA owes a private-law duty to safeguard passengers by deploying security guards on trains and by ensuring that trains do not travel between stations with coach doors open.  If a claimant were now to sue PRASA on the basis that he fell from a moving train because the coach door was open, and if the only contentious issue was whether, factually, the coach doors were open or whether the commuter sustained his injuries by falling from an open door, the case would not be a “constitutional matter”.  The “pillars” of the claimant’s case would not be sections 7(2) and 12(1)(c), given that the existence of the private law duty was uncontentious.  One could say, in a very general sense, that the constitutional considerations which gave rise to the private-law duty still “frame” the case, but that would not bring the case within this Court’s constitutional jurisdiction, any more than a person’s right to bodily integrity frames, in any meaningful jurisdictional way, his or her right to claim damages for injuries suffered in a road accident or shopping mall fall.

 

[97]       If it were otherwise, every delictual case would be a constitutional matter, because wrongfulness is an element of every delictual claim, and wrongfulness depends on the policy and legal convictions of the community, constitutionally understood.  Mashongwa is not authority for such a sweeping proposition.  The claim in Mashongwa was a constitutional matter, because the very existence of the legal duty, in other words wrongfulness, had to be decided.  Once the constitutional matter was determined in the claimant’s favour, the outcome in that particular case required the law to be applied to the facts, but that factual investigation standing on its own was not a constitutional matter.  The same can be said of the earlier decision of this Court in Metrorail.[88]

 

[98]       This point can be made with reference to Metrorail, which the first judgment cites.  This Court said that the question whether a particular issue has been established beyond reasonable doubt (or, I may add, on a balance of probabilities) is a factual one, turning on an “evaluation of evidence and its cogency”.[89]  This was contrasted with the question whether conduct was “reasonable in the context of a legal duty”, the latter involving the application of legal principles to established facts.  The latter question is a legal (and normative) one.  As I have said, here the legal and normative aspects are uncontentious.  The question is whether a particular issue (factual causation) was established on a balance of probabilities, a question turning on “an evaluation of evidence and its cogency”.

 

[99]       This Court’s decision in Booysen[90] illustrates the point.  That was a claim against the Minister of Police, who was alleged to be vicariously liable for the delictual conduct of a police officer.  The applicant (the claimant) did not ask the Court to develop the test for vicarious liability laid down in its earlier decisions.  This Court held that the application of that test to the facts of the case was not a constitutional matter.  The differing judgments in the High Court and Supreme Court of Appeal were simply the result of the differing weight which the Judges in question accorded to the normative considerations underpinning vicarious liability.[91]

 

[100]    My approach also finds support, I believe, in the majority judgments in Mbatha.  In that case, the applicant was seeking to reverse the defeats he had suffered in the Labour Court and Labour Appeal Court.  Although he was relying on legislation which gave effect to the rights of employees, and thus ultimately section 23 of the Constitution, the contest between the parties in this Court was purely factual.  This was not affected by obiter observations which the Labour Appeal Court had made about section 197 of the Labour Relations Act.[92]  In finding that this Court lacked jurisdiction, Cameron J for the majority said:

 

“No constitutional point can be located in the fact that Mr Mbatha claims he is an ‘employee’ of the university under legislation that protects employment.  His dispute with the university raises no issue of interpretation or disputed application of the statutory definitions, or any contested claim about the court’s jurisdiction over employees and employment disputes.  It is a simple factual dispute about who his employer was.  If it were otherwise, every dispute about an employee’s true employer could reach this Court.  That cannot be.

 

Mr Mbatha cannot gain constitutional access on the basis that his case involves the interpretation or application of section 197 of the Labour Relations Act.  The Labour Appeal Court mentioned this provision in its judgment, and expressed the opinion that this was a ‘classic case’ of its application.  But its reference was incidental, and immaterial to the basis of its decision.  Though expressing its view, the Court did so only after saying clearly that this is ‘not an issue before us’ and that it was therefore ‘not necessary to deal with the issue whether there was a transfer of a business’.  A court’s expression of view on a matter immaterial to its reasoning cannot confer jurisdiction on an appellate court.  The university abandoned reliance on section 197 in argument before us.  In any event, even if section 197 were in play, the sole issue, again, is its application to the facts.  The possible peripheral relevance of the provision, where it has not been given any weight by the lower courts, cannot strengthen Mr Mbatha’s claim to jurisdiction.”[93]

 

[101]    Madlanga J, in whose judgment Cameron J concurred, put the matter thus:

 

“The contest between the parties is on the facts, nothing more.  And that is what the decisions of the Labour Court and Labour Appeal Court turned on.  There is no contest between the parties about the interpretation of the provisions of the Basic Conditions of Employment Act.  Whether or not an employer/employee relationship existed between Unizul and the applicant is a question of fact.  An interpretation of the provisions is not in issue.  The application of the provisions is axiomatic, if the applicant was an employee of Unizul.  Thus the antecedent factual question is, was he?

 

It is exactly contests of this nature that Boesak has decreed do not raise a constitutional issue.  Writing for a unanimous Court in Boesak, Langa DP says that there is no constitutional issue if all there is, is a challenge to a decision on the sole basis that it is wrong on the facts.”[94]

 

[102]    Ultimately, the character of a proposed appeal to this Court turns on the findings of the lower court which the would-be appellant seeks to challenge.  If the challenge is purely to factual findings, this Court does not have jurisdiction.[95]  When all is said and done, the applicant’s case here is that the Full Court’s factual finding on causation was wrong and the trial court’s factual finding right.  That this is the crux of the matter is, I respectfully suggest, borne out by the first judgment.  Ultimately, the conclusion which the first judgment reaches in favour of the applicant depends solely on the first judgment’s evaluation of the evidence.  Unlike Alexkor, which the first judgment instances, this factual evaluation is not a precursor to reaching and deciding a constitutional issue.  The factual evaluation is the “only show in town”, because it is the sole determinant of whether the appeal fails or succeeds.

 

The Full Court’s invocation of M v MEC

[103]    As an independent ground of constitutional jurisdiction, the first judgment holds that the Full Court’s invocation of a passage from M v MEC[96] violated section 34 of the Constitution by importing evidence from another case into this case.  For two reasons, I cannot agree.  The first is that the first judgment places on the Full Court’s invocation of this passage a weight it cannot bear.  The first judgment considers that the Full Court imported into the present case expert evidence given in M v MEC, in circumstances where the applicant and her experts were not confronted with the evidence.  Since it is elementary law that a case must be decided on its own evidence, we should not readily suppose that the three judges in the Full Court had any such intention.

 

[104]    In my view, the Full Court’s purpose was more modest.  It is apparent from the passage immediately preceding its quotation from M v MEC that the Full Court was treating 04h45 as the critical time for decision making, and its focus was whether a decision at that time to perform a caesarean section would have resulted in delivery sooner than the vaginal delivery which occurred 25 minutes later at 05h10.  It was in this respect that the Full Court regarded the facts in M v MEC to be “to an extent the same” as those in this case.  In this limited respect, the facts in M v MEC were to an extent similar, and the passage was quoted merely by analogy.  The Full Court was not saying anything about whether warning signs would have been detected before the unidentified sentinel event or whether the harm could have been avoided if the time for decision-making had been earlier than 04h45.

 

[105]    This leads to my second reason for disagreeing with the first judgment’s reliance on section 34.  The Full Court, in treating 04h45 as the critical time for decision-making, went awry on the facts.  It should have asked whether, with proper FHR monitoring, a critical time for decision-making would have occurred at an earlier time, namely after 03h15 but before 04h45.  However, the Full Court’s factual error in treating 04h45 as the critical time for decision-making is simply that – a factual error, not a constitutional matter.  Having adopted this as its factual premise, the Full Court said that a caesarean section could not have been performed earlier than the vaginal delivery which occurred 25 minutes after 04h45.  The invocation of M v MEC (where an “acceptable” period of 60 minutes rather than 30 minutes for a caesarean section was adopted) is neither here nor there, because the applicant accepts that a decision to perform a caesarean section at 04h45 would not have resulted in delivery sooner than the vaginal delivery which occurred at 05h10.  It is the applicant’s case that this is the precise meaning of item 7 of the obstetricians’ joint minute.

 

[106]    Whether the invocation of M v MEC was a violation of section 34 is thus peripheral.  An affirmative answer would change nothing, because the point which the Full Court was making with reference to the analogous situation in M v MEC is in any event common cause on the evidence in the present case.  In order to reverse the decision of the Full Court, it would first be necessary to interfere with its finding that the critical time for decision making was 04h45.  The invocation of M v MEC is irrelevant to that question.  It is only by a factual reappraisal, independent of any supposed constitutional issue, that one can get out of the starting blocks to reverse the decision of the Full Court.[97]

 

[107]    I must nevertheless conclude on this issue by observing that it is usually neither helpful nor prudent for a court to quote the factual findings in another case, even if only by way of analogy.  In this instance the quotation added nothing to the cogency of the Full Court’s judgment, and it introduced confusion rather than shedding light.

 

Lee and the test for causation

[108]    The third ground of jurisdiction asserted in the first judgment is the need to clarify Lee.  In my view, there is no such need.  Neither the trial judge nor the Full Court applied any innovation which Lee may have heralded.  Both courts saw themselves as applying the conventional but-for test for factual causation.  In this Court, the first judgment likewise reappraises the facts and comes down in favour of the applicant by applying the conventional test.  This is in line with the submissions made at the hearing.

 

[109]    The first judgment states that the Full Court “plainly misapplied the test for factual causation expounded in Lee”.  I cannot discern signs of this in the Full Court’s judgment.  The Full Court’s only reference to Lee was to say that in M v MEC the minority, “[d]rawing on the reasoning in Lee”, made the statements from M v MEC which the Full Court then quoted.  That is accurate and unobjectionable.  As I have already said, the Full Court went awry on the facts when it found the critical decision-making time to be 04h45.  After quoting from both the minority and majority judgments in M v MEC, the Full Court said[98] that the trial court had “not seriously explored” the possibility of performing a caesarean section after there was a CPD diagnosis at 04h45 or the question whether a caesarean section would have yielded positive results.  Such a possibility, according to the Full Court, “is what the ‘but for’ test is all about”.  This led to the findings[99] which I have already discussed at some length.

 

[110]    On the Full Court’s factual premise, namely that the critical decision-making time was 04h45, it applied the conventional factual causation test in an unobjectionable way and reached an answer with which nobody could quibble.  The Full Court did not, for example, say that a claimant had to prove her case with certainty.  Questions of “flexibility” did not arise, because on the Full Court’s factual premise, it was clear that a decision at 04h45 to perform a caesarean section, and the undertaking of that operation within a reasonable period of time, would not have led to the baby’s delivery sooner than 05h10.  It followed that such a course of action could not have averted the injury which the baby suffered.  As the first judgment states, the inquiry into factual causation is “trite”, and the Full Court applied the trite test, albeit with reference to a factually wrong premise.

 

[111]    There is, in my opinion, nothing in the judgment of the trial court or the Full Court calling for clarification, and the first judgment’s observations on Lee seem to me to be obiter dicta (non-binding observations made in passing).  They could be omitted without affecting the first judgment’s reasoning and ultimate conclusion.  This shows that the supposed need to clarify Lee does not make the present case a constitutional matter or raise an arguable point of law which this Court ought to hear.  I prefer to express no opinion on the first judgment’s observations about the types of situations in which application of the Lee test might be more appropriate than the conventional test, because we heard no argument on the matter and it has no bearing on this case.

 

[112]    The first judgment posits that the application of the test for factual causation in medical negligence cases is an arguable point of law with constitutional implications.  I do not consider this to be a question of law.  The applicant did not argue for any special test for factual causation in medical negligence.  Difficult or borderline cases of factual causation can arise in any kind of case for delictual or contractual damages.  Conversely, factual causation is quite often straightforward in medical negligence cases.  The application of an established test to particular facts is not a question of law.  The first judgment’s analysis of factual causation bears out the proposition that each case will depend on its own facts.  It is precisely on this basis that the first judgment criticises the Full Court for having regard to the factual findings in M v MEC.

 

Conclusion

[113]    In view of the conclusion I have reached on jurisdiction, it is unnecessary to discuss the merits.  In my view, in this type of case an order of costs would not be appropriate.

 

Order

[114]    In the result, leave to appeal is refused.

 

ZONDO ACJ:

 

[115]    I have had the benefit of reading the judgment prepared by my Colleague, Majiedt J (first judgment) and the judgment prepared by Rogers AJ (second judgment) in this matter.

 

[116]    Subject to what I say below about the cases of Mbatha[100] and Booysen[101] to which the second judgment refers, for the reasons given in the second judgment I agree with the second judgment that this Court does not have jurisdiction to entertain this matter.

 

[117]    The second judgment refers to the cases of Mbatha and Booysen.  In regard to those judgments I confine myself to what I have to say to the fact that I wrote the minority judgments in those matters in which I expressed my views but I accept that I am bound by the majority decisions in both cases.

For the Applicant:

For the Respondents:

S Budlender SC and E Webber, instructed by Edeling Van Niekerk Incorporated

T Bruinders SC and A Mofokeng, instructed by the State Attorney, Johannesburg

---

---