HKSAR v. KWAN HAU-CHI, VANESSA

CACC 244/2021 [2026] HKCA 1078

On appeal from [2021] HKCFI 2978

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF APPEAL

CRIMINAL APPEAL NO 244 OF 2021

(ON APPEAL FROM HCCC NO 200 OF 2018)

_______________

BETWEEN

_______________

__________________________________

REASONS FOR JUDGMENT

__________________________________

Hon Zervos JA (giving the Reasons for Judgment of the Court):

Introduction

1.  On 23 August 2021, the appellant, a medical practitioner, was convicted after a trial before D’Almada Remedios J (the judge) and a jury of manslaughter by gross negligence in relation to Ms Josephine Lee Kar Ying (the deceased) following a liposuction procedure.[2] On 4 October 2021, the judge sentenced the appellant to 6 years’ imprisonment[3].

2.  By a notice dated 29 October 2021, the appellant applied for leave to appeal against conviction and sentence. She appealed against conviction on the basis of alleged errors at trial and on fresh evidence from two medical experts. On 11 May 2023, a Single Judge granted the appellant bail and leave to appeal against conviction on Grounds 4 to 10, and against sentence on Grounds 11 and 12. A written judgment was handed down on 23 June 2023.[4]

3.  On 10 August 2023, Mr Martin Hui SC, with Ms Sharon Chan, for the appellant, filed re-amended perfected grounds of appeal against conviction and sentence based on the grounds for which leave was granted. It was contended that: the judge wrongly refused to allow defence counsel to cross-examine on (i) the contents of the expert report from Dr Lester Ah Critchley (Dr Critchley) that had been considered by prosecution expert witnesses; and (ii) the original unmuted CCTV footage adduced at trial (Ground 4); the judge failed adequately to sum up and direct the jury on the expert evidence (Ground 5); the judge failed to adequately direct the jury on the issue of causation (Ground 6); the judge erroneously directed on the issue of foreseeability after directing on causation (Ground 7); the judge erred in instructing the jury to find the prosecution case proved on the expert opinion of Dr Phoebe Anne Mainland (Dr Mainland) (Ground 8); the judge failed to give a Kevin Brown direction and to give the jury sufficient instructions regarding multiple particulars of breach of duty (Ground 9); and the conviction was unsafe and unsatisfactory (Ground 10). The additional Grounds 11 and 12 concerned the appellant’s appeal against sentence.

4.  Mr Jonathan Man, with Ms Chan Sze-yan, for the respondent, submitted that none of the grounds of appeal, whether individually or collectively, were sufficient to vitiate the conviction. However, it was accepted that the sentence could be reduced on humanitarian grounds in light of the appellant’s terminal illness.

The fresh evidence on appeal

5.  There were two applications to adduce fresh evidence under section 83V of the Criminal Procedure Ordinance (Cap 221), both made by way of motion dated 12 March 2024. The first was a joint application by the parties to adduce three expert reports from Dr David Andrew Ross Burd (Dr Burd) and three expert reports from Dr Critchley, for the Court’s consideration in relation to Grounds 4 to 6 of the appeal against conviction.[5] The second was an application by the appellant, seeking to adduce four affidavits[6], together with exhibits, including various medical reports relating to herself and members of her family, in support of her appeal against sentence.

6.  The Court of Appeal has power under section 83V to receive evidence, consisting of the production of documents or the examination of witnesses, that was not adduced at trial. On the subject of receiving fresh evidence on appeal, the relevant principles have been set out in R v Chattoo[7] and related authorities[8], which address the equivalent Hong Kong provision. This Court has applied R v Chattoo in HKSAR v Sysoev Igorevich Iurii[9]. The relevant principles may be summarised as follows. First, the statutory discretion to receive fresh evidence on appeal applies equally to expert and non-expert evidence, even if the governing provision was not framed with expert evidence specifically in mind. Secondly, the governing criterion is whether it is necessary or expedient in the interests of justice to receive the evidence, taking into account all the factors set out in section 83V.[10] Thirdly, save in exceptional or truly rare circumstances, an appellant will not be permitted to advance on appeal an expert case which could and should have been pursued at trial, or to rerun the same expert case with a different or additional expert, since that would subvert the trial process and convert the appeal into a second “bite of the cherry” by way of a new presentation of expert opinion. Fourthly, the same applies where expert evidence on the relevant issues was obtained for the trial but deliberately not deployed because it was thought that it would assist the prosecution; in such cases, an appellant who chose not to call such expert evidence at trial should be in no better position than one who did.

7.  The principal consideration, in deciding whether it is necessary or expedient in the interests of justice to admit fresh evidence, is whether the evidence is credible and relevant, and whether it could have materially affected the outcome of the case. Whilst the Court of Appeal has to be satisfied that there is a reasonable explanation for the evidence not having been adduced at trial, it would be difficult to exclude such evidence on this basis if it appears cogent and relevant to a material issue on appeal.

8.  We considered the fresh evidence de bene esse but, for the reasons set out in this judgment, we did not formally admit the medical expert reports or the affidavits.

The outcome of the appeal

9.  At the conclusion of the hearing, we allowed the appeal against conviction, quashed the conviction and set aside the sentence. The reasons for our decision to overturn the conviction concern the nature of the medical expert evidence and related issues, in particular the failure to put to the jury for its consideration the realistic possibility of an unknown cause of death. Following the pronouncement of our judgment, the respondent made no application for a retrial. We said that we would hand down the reasons for our judgment in due course. This we now do.

The elements of gross negligence manslaughter

10.  The offence of gross negligence manslaughter is unusual, as it has, as its essential ingredient, the civil concept of negligence, but at a level that is so egregious as to warrant criminal liability. Before a defendant is convicted of manslaughter by reason of gross negligence, the jury must be satisfied so that they are sure (R v Broughton [11]; R v Kuddus [12]; R v Rose [13]) that:

(1) the defendant owed an existing duty of care to the victim;

(2) the defendant negligently breached that duty of care;

(3) at the time of the breach there was a serious and obvious risk of death. “Serious”, in this context, qualifies the nature of the risk of death as something much more than minimal or remote. Risk of injury or illness, even serious injury or illness, is not enough. An obvious risk is one that is present, clear and unambiguous. It is immediately apparent, striking and glaring, rather than something that might become apparent on further investigation;

(4) it was reasonably foreseeable at the time of the breach of the duty that the breach gave rise to a serious and obvious risk of death (the elements in (3) and (4) do not require separate consideration or articulation in many cases (R v Broughton[14]));

(5) the breach of that duty caused or made a significant (that is, more than minimal) contribution to the death of the deceased; and

(6) the circumstances of the breach were truly exceptionally bad and so reprehensible as to justify the conclusion that it amounted to gross negligence and required criminal sanction (this element of gross negligence is proved by the application of the objective standard of reasonableness (HKSAR v Mak Wan Ling (No 2) [15]).

11.  The offence of gross negligence manslaughter in the healthcare context arises where a healthcare professional or provider breaches an existing duty of care by a grossly negligent act or omission in the course of medical treatment or care, which causes or significantly contributes to the death of a patient in circumstances involving a serious and obvious risk of death, and where that breach is truly exceptionally bad and so reprehensible as to warrant a criminal sanction.

12.  There has been a growing number of cases in which medical malpractice has attracted criminal liability, and this case illustrates the need for greater protection of patient safety and for improved standards of medical care.

The prosecution case

13.  The prosecution alleged that the appellant failed properly to manage the deceased, who was put under deep sedation for a liposuction procedure, and that this failure resulted in the deceased’s death.

14.  The charge in the indictment alleged that the appellant unlawfully killed the deceased by gross negligence on 26 June 2014, in that:[16]

“(i) Being a registered doctor responsible for conducting the liposuction procedure to the said LEE Kar-ying, Josephine, owed her a duty of care;

(ii) In breach of that duty of care, failed to take reasonable care for the safety of the said LEE Kar-ying, Josephine, by:-

(a) Failing to ensure the presence of a properly qualified person to administer and monitor sedation when the said liposuction procedure was performed to the said LEE Kar-ying, Josephine;

(b) Failing to ensure sufficient oxygen supply to the said LEE Kar-ying, Josephine, during sedation;

(c) Failing to follow the paragraphs 3.3.1 to 3.3.3, 3.3.7, 3.4, 4, 5.1 to 5.3, 5.4.1.4 to 5.4.1.8, 7.3, 7.5, 10 and 11.2 of the ‘Guidelines on Procedural Sedation’ of the Hong Kong Academy of Medicine endorsed by the Hong Kong Academy of Medicine Council on 22 December 2009 when she was responsible for conducting the liposuction procedure for the said LEE Kar-ying, Josephine;

(d) Failing to provide the said LEE Kar-ying, Josephine, with proper and sufficient monitoring after the liposuction procedure; and

(e) Failing to provide adequate timely resuscitation to the said LEE Kar-ying, Josephine;

(iii) Her aforesaid breach of duty amounted to gross negligence on her part; and

(iv) Her aforesaid negligence was a substantial cause of the death of the said LEE Kar-ying, Josephine.”

15.  It is important to recognise at the outset that the allegation against the appellant was complicated because the forensic pathologist could not ascertain the cause of death. As explained by the judge to the jury, an autopsy, including a toxicology report, did not reveal a specific cause or mechanism of death.[17] The prosecution relied on the evidence of medical experts in the fields of plastic surgery and anaesthesiology, who were critical of the conduct of the liposuction procedure, each expressing an opinion on the cause of death with varying degrees of qualification.

16.  The main issues at trial were whether the prosecution had proved that the appellant breached her duty of care as alleged, and whether any such breach amounted to gross negligence that substantially caused, or made a significant contribution to, the deceased’s death. Other related issues concerned whether the deceased was under deep (unconscious) sedation or conscious sedation after the appellant had administered a combination of sedative drugs to facilitate the liposuction procedure she conducted[18]. There were a number of alleged failings in the clinical setting of the operating room and in the conduct of the liposuction procedure, which in various respects were said to have breached professional standards. The question was whether, individually or collectively (although this was unclear), these alleged failings caused, or made a significant contribution to, the death of the deceased. However, there was no allegation against the appellant regarding her surgical skills, the adequacy of the facilities of the operating room, the sterility of the surgical instruments, or an overdose of drugs.

17.  In this labyrinth of medical evidence and expert opinions, a key question was whether the deceased died of an unknown cause. This realistic possibility was not raised before the jury for its consideration.

The background and facts

18.  This was a tragic case, made more complicated at trial because of the uncertainty and divergence of expert opinions as to the deceased’s cause of death.

19.  The appellant and the deceased were friends[19]. At the time of the offence, the appellant, a medical practitioner of seven years’ standing, worked part-time at a beauty centre and frequently performed liposuction procedures in an operating room that had been specifically set up for surgeries at the centre[20].

20.  The deceased was a large woman for her height, and her weight was recorded as 116 kilogrammes[21]. The appellant performed the liposuction procedure on the deceased to remove fat from her abdomen on 30 April 2014 without any apparent incident[22]. The procedure was repeated on 26 June 2014, this time to remove fat from the deceased’s back[23]. The deceased was placed in a prone position on the operating table[24] and connected to a Mindray machine to monitor her vital signs and oxygen saturation during the procedure[25]. At the conclusion of the procedure, the deceased did not regain consciousness and died.

21.  Inside the operating room were two CCTV cameras - Cameras 4 and 8, which visually recorded most but not all of the room. They did not capture the whole body of the deceased, only the back of her legs. Camera 4 had both audio and visual recording functions[26], whereas Camera 8 only had a visual recording function[27]. During the trial, the footage from the two CCTV cameras was combined to synchronise the visual and audio recording of one camera with the visual recording of the other[28]. This became a contentious issue between the parties due to the late disclosure of the combined CCTV footage and the quality of the synchronisation of the two recordings.

22.  A police officer (PW2) put the footage of Camera 4 and Camera 8 on a synchronised screen[29]. When played, the footage recorded by the two cameras ran at different speeds, and the time difference or time lapse ranged from a few seconds to 10 seconds for a 5-minute segment[30]. PW2 cut up the CCTV footage recorded by Cameras 4 and 8 into 5-minute segments, which resulted in a series of 5-minute clips[31]. She put each of the clips together on the same screen[32]. The start time for each clip was the same[33]. The combined clips were Exhibit P55, which was played before the jury[34]. Part of the CCTV footage recorded by Camera 4 was muted upon agreement of the parties[35]. There was also a transcript of the CCTV footage[36].

23.  In performing the liposuction procedure, the appellant was assisted by four female employees of the centre. Two of them were employed as receptionists (Ms Cheng Hoi Ki (Kay, PW4) and Ms So Ka Yan (Yen, PW6)) and the other two as beauticians (Ms Lau Tsz Kwan (Rossa, PW5) and Ms Wong Kai Pui (Addplus, PW8)). At issue in the trial was whether the assistants were suitable and had the proper medical training[37].

24.  The deceased was sedated by an infusion of propofol and a combination of other sedative drugs[38]. Propofol was administered to the deceased at various times throughout the procedure.

25.  Although there was an audio and partial visual recording of the procedure, there were no written notes monitoring or recording the deceased’s vital signs, nor was there a record of the drugs administered to her[39]. Dr Mainland, an expert in anaesthesiology, from her viewing of the CCTV footage, was of the opinion that seven syringes of propofol were administered to the deceased[40]. She estimated that a total of 14 ampoules, amounting to 2,800 mg of propofol, had been administered to the deceased[41]. This matter only emerged during the trial from the synchronised CCTV recording[42].

26.  The timing of propofol infusion was as follows: Syringe 1 was started at 11:34:46 hours[43]; Syringe 2 at 11:54:33; Syringe 3 at 12:21:57; Syringe 4 at 12:50:32; Syringe 5 at 13:16:33; Syringe 6 at 13:44:18; and Syringe 7 at 14:09:17[44]. Dr Mainland opined that the deceased was unable to communicate or respond from about the beginning of the operation except for a short period between the 5^th and 6^th Syringes when she came out of her sedation. From the audio recording, at 13:44:40, the appellant instructed “add 2 cc bolus”[45]; at 13:44:52, the deceased said “Don’t press my hand”[46]; and at about 13:48:16, she said “Painful, painful, painful”[47]. By 14:31:50, no anaesthetic had been infused[48]. The appellant was still operating after the infusion had finished[49].

27.  At about 14:40, the appellant sutured the deceased on completing the liposuction procedure and left the operating room at about 14:44[50]. According to the prosecution, the deceased was still sedated and unconscious[51]. Before leaving the operating room, the appellant told the assistants to hand a key to the deceased as she had something to do, but to tell the deceased to call her[52]. The deceased was detached from the Mindray machine and left prone on the operating table[53].

28.  At about 15:19, one of the assistants noticed that the deceased was not snoring and was unresponsive to physical contact or verbal communication[54], and also looked pale[55]. The appellant was contacted, and she returned to the operating room at about 15:34[56]. She instructed that an emergency 999 call be made at about 15:35[57]. She administered oxygen to the deceased and tried to revive her[58]. She also telephoned another medical practitioner to seek assistance[59]. The police officer, who responded to the 999 call, saw the appellant doing compressions on the deceased’s back as he entered the operating room at about 15:45[60]. At about 15:59, the deceased’s blood pressure could not be measured, and at about 16:00, there was a call for “adrenaline, adrenaline”[61]. At that time the ambulance officers arrived[62]. They removed the deceased from the operating table and placed her in a supine position[63]. The deceased had no pulse, but they still tried to resuscitate her[64]. She was transported to the hospital at about 16:12 and certified dead at about 17:06[65].

The defence case

29.  The appellant did not testify, nor call any witnesses. It was the defence case that the deceased was under conscious sedation[66], neither deep sedation nor general anaesthesia[67], and that the Guidelines on Procedural Sedation of the Hong Kong Academy of Medicine (the Guidelines) were neither mandatory nor applicable to the appellant[68]. The defence position was that the assistants were not completely untrained[69]. It was submitted that the appellant made an assessment of the deceased after the liposuction procedure[70], and that there was a resuscitation attempt on the deceased[71]. The CCTV footage only showed the legs of the deceased on the operating table and there was a period when the appellant was out of view, at which time it was contended that the appellant was assessing the deceased following the liposuction procedure[72].

30.  In summary, the defence case was that nothing the appellant did or failed to do substantially caused the deceased’s death[73].

The finding that the cause of death was unascertained

31.  Dr Chiao Wing Fu (Dr Chiao), the forensic pathologist, conducted an autopsy on 28 June 2014, two days after the deceased’s death. He concluded that an examination of the deceased’s body revealed no injuries or any form of trauma that could have accounted for the cause of death; there was no infection, perforation or massive haemorrhage over the operated parts. An examination of the vital organs showed no disease or potentially fatal pathology to which the cause of death could have been attributed. The toxicology analysis revealed the presence of an analgesic, local anaesthetic and sedatives, but the blood levels of the drugs could not have accounted for the cause of death. He also discounted the possibility that the cause of death was associated with surgical intervention. He was of the opinion that the cause of death was “unascertained” because the autopsy did not reveal an abnormality or a fatal pathology that could have satisfactorily accounted for the cause. However, he suggested that expertise from a qualified plastic surgeon and anaesthetist should be sought in order to examine the appropriateness of the clinical settings and judgments and to exclude the possibility of any medical mishap[74].

32.  In his report, Dr Chiao set out the following relevant findings and opinions:[75]

“2. Examination of the body of the deceased revealed no injuries or any forms of trauma that could have been accounted for the cause of death in this case. There was no infection, perforation or massive haemorrhage over the operated parts.

3. … Considering the nature and extent of the pathological changes in the heart, lungs and liver they were not considered to be severe enough to cause a sudden death. There was no other identifiable disease or potentially fatal pathologies to which the cause of death could have been attributed.

4. … as the vitreous glucose level was not very high and no ketone bodies was found in the blood, it was unlikely that the deceased had been died of certain kind of fatal complication of her glucose metabolic disorder.

…

6. … Considering the blood levels of the drugs, they could not have been accounted for the cause of death. No significant finding for the blood and urine by a general screening procedure for other common drugs and poisons.

…

8. Basing on the available information and the autopsy findings, the sudden death of the deceased was unlikely to have been caused by injury. There was also no evidence to indicate that her death was caused directly by the surgical intervention. Although the deceased was found to have been suffering from several medical problems with acute pathologies (mild pulmonary oedema, mild healing bronchopneumonia, mild myocarditis and possibly mild pulmonary aspiration) and chronic medical problems (enlarged tonsils, marginally heart enlargement, severe fatty change of liver and probably diabetes mellitus), these pathologies were not considered to be severe enough to cause a sudden death in general situations. However, it was unascertained whether the summative effects of all these pathologies could have had any significant contribution to the cause of death when a person was sedated as in the present case.

9. On the other hand, as the deceased was under the medical care of the in-charge doctor during the whole process of sedation and surgical intervention, the appropriateness of the clinical setting, clinical judgements made a medical treatments given by the in-charge doctor were critical and should be subjected to a detailed review by a qualified plastic surgeon and a qualified anaesthetist in order to exclude medical mishap.

10. In summary, the cause of death in this case has to be labelled as ‘Unascertained’ because the autopsy did not reveal an abnormality or a fatal pathology that could have been satisfactorily accounted for the cause of death. However, as the death occurred shortly after receiving liposuction operation before recovering her consciousness from the sedation in the beauty centre, expertise from a qualified plastic surgeon and a qualified anaesthetist must be sought in order to exclude the possibility of medical mishap.” (Emphasis added)

The appeal against conviction

33.  The mainstay of the prosecution case was the expert medical evidence, and in addressing this appeal, it is necessary to examine the nature and scope of the expert evidence and the related issues regarding the liposuction procedure, including how, in the process of presenting the expert medical evidence, the real possibility of the deceased’s cause of death being unknown was not canvassed with the jury when assessing the elements of the offence of gross negligence manslaughter.

34.  As Dr Chiao was unable to ascertain the deceased’s cause of death, this was problematic for the prosecution in establishing a case against the appellant. The prosecution engaged various medical experts in exploring the cause of death, which proved to be difficult, as revealed by the length of time the various medical experts in the fields of plastic surgery and anaesthesiology took to develop and formulate their opinions in light of additional information and considerations. This difficulty was exacerbated by the way the prosecution at trial presented the medical evidence and the medical opinions relied upon to show that the appellant’s breach or breaches of duty caused or significantly contributed to the deceased’s death. The element of causation was an important issue. In this context, the prosecution had to prove that the alleged breach or breaches of duty gave rise to a reasonably foreseeable serious and obvious risk of death, and that such alleged breach or breaches caused or made a significant contribution to the death of the deceased.

35.  There are seven grounds of appeal against conviction, Grounds 4 to 10. We first address Grounds 5 and 8 together, as they concern the expert medical evidence, particularly Dr Mainland’s evidence (A. The expert medical evidence and related issues). We next examine Ground 4 regarding the judge’s rulings on the references to Dr Critchley’s expert report and the unmuted CCTV footage (B. Dr Critchley’s report and the unmuted CCTV footage). We turn to consider Grounds 6 and 7 together, which are concerned with the elements of foreseeability and causation (C. The elements of foreseeability and causation). We then examine Ground 9 regarding the judge’s failure to give a Kevin Brown direction (D. The failure to give a Kevin Brown direction). Finally, we analyse the issue and implications of an unknown cause of death, which had not been raised as a ground of appeal (E. The issue of an unknown cause of death).

A. The expert medical evidence and related issues[76]

36.  Following Dr Chiao’s report, the prosecution obtained from Dr Critchley, a professor and specialist anaesthesiologist, two draft reports, dated 22 and 28 May 2015, and a final report, dated 3 July 2015.[77] That final report was not adduced in evidence, but was considered by Dr Mainland. It was an issue at trial whether the defence could cross-examine Dr Mainland regarding her consideration of Dr Critchley’s report and his findings.

37.  We have as a matter of necessity, examined the reports of Dr Critchley and Dr Burd, notwithstanding that they did not give evidence at trial, in order to understand the background and circumstances of the issues raised in relation to how the medical expert evidence evolved and was presented to the jury, and to address related matters concerning the nature and basis of the opinions on the cause of death.

38.  Dr Critchley questioned the procedures that were followed by the appellant in performing the liposuction procedure and the lack of documentation. In his first draft, he said that it was his “guess” that the cause of death was a combination of factors that in isolation were innocent but together led to significant cardiorespiratory depression followed by cardiorespiratory arrest that was not promptly treated. [78] In his second draft, he questioned the procedures that were followed and commented that the level of care was below standard. He said that the facilities in the clinic performing liposuction under heavy sedation and the level of staffing were reasonable. He was critical of leaving the unconscious deceased in the care of an inexperienced assistant and the lack of documentation of the procedure. He said that the deceased appeared to have cardiac arrest following the procedure because she was over-sedated and due to other contributing factors.[79] In his final and signed report, he said that the cause of death was cardiorespiratory depression caused by excessive intravenous sedation, possibly complicated by a minor degree of gastric aspiration. He said that the facilities in the clinic for performing liposuction under heavy sedation were reasonable. However, he was critical of carrying out the procedure without a qualified medical assistant and leaving an unconscious patient in the care of untrained personnel.[80]

39.  The prosecution also obtained from Dr Chan Yu Wai (Dr Chan), a plastic surgeon, a report dated 16 July 2015 (which was replicated as a statement to the police dated 20 March 2018[81]). He criticised the absence of a formal protocol and a detailed operative record[82]. He noted that from the post-mortem report and the toxicology examination, the apparent cause of death was “not clear”[83]. He said that without certain crucial information, “one cannot judge whether the surgical procedure has been properly performed”[84]. However, he was very critical of the casual approach to the surgical procedure[85]. He explained that before surgery, a medical examination and an adequate explanation of the procedure and its risks should be provided[86]. He also explained that it would be necessary, if the patient did not have communication with the surgeon, for a trained person to assist in monitoring the patient[87]. He was of the view that because of the significant amount of aspirate, the surgery could not have been done under conscious sedation[88]. He said that if a patient was lying prone for the surgery, he would have immediately turned the patient over after the surgery[89]. He expressed the view that the deceased was liable to suffer cardiovascular and respiratory complications that were potentially lethal[90].

40.  On 27 July 2015, the reports of Dr Chan and Dr Critchley were submitted to Dr Chiao for his comments, particularly in respect of their views about the cause of death. On 3 August 2015, Dr Chiao submitted a memorandum commenting on the two expert opinions regarding the deceased’s cause of death. He qualified Dr Chan’s opinion on the cause of death in various respects, namely that the deceased was liable to suffer from cardiovascular and respiratory complications under sedation in a prone position which was potentially lethal, stating that there was no direct clinical information or post-mortem findings to further confirm “this speculation”. He agreed with Dr Chan’s opinion that comments from an anaesthetist would be helpful regarding cardiovascular and respiratory protection during surgery under sedation. In his memorandum, Dr Chiao also commented on Dr Critchley’s report, but this was redacted because the prosecution did not call him to give expert evidence at the trial.

41.  On 15 May 2015, the prosecution also instructed Dr Mainland to provide an expert report. She produced a first draft report dated 11 January 2016[91]. She was asked to address matters raised by advising counsel of the Department of Justice on 12 February 2016, resulting in a second draft report dated 15 February 2016[92]. There were further communications with Dr Mainland about her report, and she was provided with additional documents for her consideration in August and October 2016[93]. She was also given a copy of a recent judgment of R v Rudling[94], and asked to address whether the appellant, being in breach of an existing duty of care to her patient, which had been reasonably foreseeable, had given rise to a serious and obvious risk of death, and had caused the death of the deceased.[95] She produced a third draft report dated 17 January 2017[96], in which she stated that the cause of death could not be established with certainty[97]. However, she was critical of the surgical procedure and the way the appellant conducted it. In her opinion, it was likely that the deceased would not have died if the sedative medications had not been administered[98].

42.  In an email dated 13 July 2017, Dr Mainland was provided with CCTV footage of the operating room, which included the audio recording that had been “lately found”.[99] Sometime later, in July 2021, there was a series of emails from Dr Mainland to the prosecution, in which she referred to a report and supplementary report from Dr Burd, who had been instructed by the defence.[100] Dr Burd was highly critical of the opinions of Dr Chan and Dr Mainland, essentially arguing that the appellant had met the appropriate requirements for the liposuction procedure and agreeing with Dr Critchley’s opinion that the level of patient monitoring for sedation procedures provided was reasonable. He spent some time addressing Dr Mainland’s opinion on the issue of supplemental oxygen when administering propofol, with which he strongly disagreed.

43.  In an email dated 23 July 2021, Dr Mainland referred to Dr Burd’s supplementary report and said that she should amend her report[101]. On the same day, she produced her final report.[102] There were email communications between prosecuting counsel and Dr Mainland, on 25 July and 1 August 2021, in which Dr Mainland stated that she observed seven syringes of propofol on viewing the CCTV footage and that there was no delay between syringes as she had initially thought.[103]

44.  In the Executive Summary of her final report, following an introductory paragraph, Dr Mainland stated:[104]

“2. Ms LEE was deeply sedated. The patient monitor alarmed throughout the procedure. There was no trained person monitoring her. Dr KWAN left after the procedure without assessing Ms LEE. She returned after a phone call reporting that Ms LEE’s condition had deteriorated with a low oxygen level. Dr KWAN called for help but did not provide resuscitation according to Basic Life Support.

3. Although the cause of death cannot be established with certainty it is my opinion that with her body positioned face down, deep sedation and lack of supplementary oxygen would result for death due to low oxygen levels in the blood and vital organs because of airway obstruction and reduced breathing effort.

4. Other contributing factors include the lack of monitoring by a trained person so delayed recognition of a problem and inadequate timely resuscitation. Resuscitation equipment was available in the facility but not all were utilized. The facility could not transfer an unconscious person onto their back and Dr KWAN went ahead with the surgery with this capability unavailable.

5. In my opinion if Dr KWAN had followed Hong Kong Academy of Medicine Guidelines on Procedural Sedation (2009) Ms LEE would not have died.” (Emphasis added)

45.  As Dr Mainland could not see the head of the deceased on the CCTV footage, she assumed in paragraph 3 that the deceased had low oxygen levels because of airway obstruction and reduced breathing effort.

46.  In her Summary of Conclusions, Dr Mainland stated:[105]

“6.1 The cause of death of Ms LEE cannot be established with certainty. However exclusion of causes such as myocardial infarction (heart attack), significant aspiration (inhalation of stomach contents into the lungs), stroke, massive pulmonary embolism or other catastrophic medical events, supports cause of death to be the result of over-sedation, respiratory depression, hypoxia and cardiac arrest.

6.2 It is my opinion that it is likely Ms LEE would not have died if the sedative medications had not been administered.

6.3 The actions of Dr KWAN contributed to a significant extent to the death of Ms LEE. As the supervising doctor, she was responsible for drugs administered or not, used for sedation, including oxygen, the monitoring of Ms LEE during and after the procedure and management of complications resulting from the procedure, including complications of medications administered.

6.4 Dr KWAN did not ensure supplementary oxygen was administered during sedation, she arranged administration of propofol by a bolus doses and infusion to Ms LEE without ensuring a trained person was responsible for monitoring the patient and her resuscitation efforts once an emergency had been identified were inadequate. In my opinion, these actions and inactions of Dr KWAN contributed directly to the death of Ms LEE.

6.5 It is my opinion that a reasonable doctor would have arranged for an anaesthesiologist or other trained person to perform sedation for and monitor a patient such as Ms LEE, undergoing liposuction in the prone position, if such a deep level of sedation was required. Alternatively a reasonable doctor would have not prescribed propofol for such a patient and would have administered an amount of sedation that ensured that Ms LEE maintained verbal communication with the doctor.

6.6 The conduct of Dr KWAN was reckless unsatisfactory[106] with regard to her duty of care as the responsible medical practitioner and in breach of her duty it was reasonably foreseeable had given rise to serious and obvious risk of death and caused death of Ms LEE.

6.7 Lack of embracement by individual medical practitioners of their responsibility to provide safe patient care, not limited to competence in delivering medical care including medications, resuscitation and use of emergency equipment, is more likely to contribute to patient demise even when the facilities including equipment are adequate.” (Emphasis added)

47.  In paragraph 6.1, Dr Mainland’s reasoning as to the cause of death was based on a process of elimination that had been employed by medical experts in cases involving shaken baby syndrome where the baby displayed non-accidental injuries which could only be the result of deliberate shaking. However, in this case, the forensic pathologist found no signs of any trauma or injury, rendering this process of elimination logically flawed. She did, however, identify that the deceased’s prone position, deep sedation and lack of supplementary oxygen may have resulted in the deceased’s death, although she accepted that this conclusion was not definitive. In paragraph 6.6, she made legal findings regarding the elements of the offence of gross negligence manslaughter apparently by reference to Rudling, which she had been asked to address.

48.  The prosecution requested Dr Chan and Dr Mainland to consider the defence medical expert reports. Dr Chan was asked to provide a supplementary report and to comment on Dr Burd’s expert opinion. He produced a report dated 17 July 2021 in which he commented on Dr Burd’s report, sometimes agreeing with Dr Burd’s opinions.[107] Dr Mainland was asked to address whether she agreed with Dr Critchley’s opinion regarding over sedation as the cause of death and to also comment on Dr Burd’s expert reports. She noted this in the introductory remarks of her final report dated 23 July 2021. Even though the reports of Dr Critchley and Dr Burd were not produced, there was a proper basis on which to cross-examine both Dr Chan and Dr Mainland in respect of their various references to the opinions of Dr Critchley and Dr Burd.

49.  One matter that concerns us is how the medical experts prepared their final reports. For instance, Dr Chan produced a report dated 16 July 2015, which was later considered by Dr Chiao. However, neither Dr Chan nor Dr Mainland explained how their opinions had evolved from their earlier drafts. Where relevant, expert reports should record material changes of opinion and the reasons for them. There had been an exchange of opinions and an ongoing debate between the parties’ experts which was not properly reflected in the final reports, although some of it emerged in cross-examination of the expert concerned.

50.  The prosecution challenged the admissibility of Dr Burd’s report, which had been highly critical of the opinions of Dr Chan and Dr Mainland, particularly Dr Mainland’s opinion on the issue of supplemental oxygen when administering propofol. This was the main basis on which the judge excluded most of his report from being admitted into evidence. She ruled that his opinions were largely concerned with the subject of anaesthesiology and were outside his field of expertise. The ruling limited Dr Burd’s report substantially[108], and he was not called by the defence. Consequently, the debate between the experts in respect of several pertinent matters was not played out before the jury. However, it has to be said that Dr Burd’s comments were, at times, expressed in strong terms and in areas apparently outside his field of expertise.

51.  Initially, the judge took the view that the defence could cross-examine the prosecution expert witnesses on Dr Critchley’s report, as they had made references to his opinions, but she later ruled that the defence could not do so. She directed that the parties could not make any reference to Dr Critchley and further instructed the jury to delete references to Dr Critchley contained in any of the expert reports adduced at trial.[109]

The presentation of the medical evidence at trial

52.  The prosecution at trial called the expert witnesses in the following order. The first was Dr Chiao (PW10 at trial); the second was Dr Chan (PW11 at trial); and the third was Dr Mainland (PW12 at trial).[110] As explained by the judge, Dr Chiao’s evidence was in respect of the autopsy he conducted on the deceased and his finding that the cause of death of the deceased was “unascertained”. Dr Chan’s evidence concerned the surgical aspects of the liposuction procedure. Dr Mainland’s evidence was in regard to matters such as whether the deceased was in deep sedation during the procedure and all other matters relating to the sedation. The evidence of Dr Chan and Dr Mainland in the areas on which they testified was relevant to the issue of whether the appellant had breached her duty of care owed to the deceased and whether the breach or breaches were the substantial cause of her death.[111]

53.  Dr Chan, in addition to giving evidence about the requirements and standards governing surgical procedures, also expressed an opinion about the cause of death which was submitted to Dr Chiao for his opinion. Dr Chiao corrected Dr Chan’s opinion as to the cause of death, dismissing his suggestion that the cause of death was attributable to cardiovascular and respiratory complications, as there was no direct clinical information or post-mortem findings to further confirm “this speculation”.

54.  The prosecution case principally centred on the expert opinion of Dr Mainland.[112] She gave expert evidence about the requirements and standards that would be expected when sedating a patient, and about breaches of those requirements and standards that took place during the liposuction procedure on the deceased. She also gave an opinion as to the cause of death, which was expanded and adjusted to some extent from her previous opinions. Her opinion was qualified and assumed various factual matters that were in dispute.

55.  Significantly, Dr Chiao was not asked to comment on Dr Mainland’s opinion as to the cause of death, which she qualified by stating that it could not be established “with certainty”. The case was further complicated by the fact that Dr Chiao gave his evidence before that of Dr Mainland, so her opinion as to the cause of death was not considered by him, whose expertise lay in determining causes of death.

Dr Mainland’s expert evidence

56.  In her evidence, Dr Mainland qualified some of the conclusions in her report. She explained various levels of sedation as defined under the Guidelines[113] and the breaches that were particularised in paragraph (ii) of the charge[114]. She emphasised that conscious sedation and deep sedation were not a step or cut-off, but a continuum.[115] The deceased could go between conscious and deep sedation, especially when there was an interruption of drug infusion, when the level of sedation would be lightened[116]. She explained that, according to the Guidelines, while a doctor could look after and provide for sedation of a patient, there must be rational verbal intercommunication during sedation and monitoring of those verbal commands must be continuously possible during the procedure. If continuous verbal communication was lost, the doctor should devote attention to monitoring and treating the patient until recovery or until such time as another appropriately trained medical practitioner became available[117]. If deep sedation was intended, an anaesthesiologist or an appropriately trained medical practitioner must be present to monitor the patient throughout the procedure[118]. She was of the view that when propofol was administered, there should have been a trained person monitoring the deceased’s vital signs, with supplemental oxygen administered as necessary[119].

57.  When Dr Chiao testified, he was asked if he could tell if a lack of oxygen caused the death, and he explained that it would take hours to show a hypoxic change[120]. Dr Mainland said on this issue that she was not suggesting that the deceased was hypoxic to the catastrophic level for three hours[121]. There was a difference between a lack of oxygen and low oxygen saturation[122]. In her opinion, the deceased’s saturation level was below what it should have been and there was a compromise of oxygen to the deceased’s body[123]. She opined that extra oxygen should be supplied to the patient throughout the procedure, even though the patient could breathe on her own, and cited the Guidelines[124].

58.  Dr Mainland agreed there was no legislation requiring the Guidelines to be followed. However, she testified that “a general principle is if there are guidelines, they’re there for a good reason and if someone chooses not to follow the guidelines they’re – they should be able to justify why the guidelines were not followed.”[125]

59.  Dr Mainland was unclear in her opinion as to the actual cause of death, conceding her lack of certainty, and instead referred to the “additive effect”, focusing mainly on hypoxia as the cause of death. This was reflected in the following passage of her evidence in court:[126]

“A. I think, in my opinion, looking at the CCTV, there were opportunities where there could have been intervention that might have prevented the death. I think, perhaps, I’m incorrect in saying that it caused -- caused the death, but I think the conduct contributed to the risk of death.

Q. And, I think, am I correct to say all the things that you have mentioned in your report and in your evidence, in fact, acted collectively and not any -- and not singularly insofar as towards the death of Ms Lee is concerned, is that correct?

A. Yes, it was the additive effect or the combination of several failures that together contributed to the situation where death occurred. I can’t say if one thing was done that would have stopped it, her dying, or it was more that there was a collection of actions and inactions that created the situation where death was a high possibility, probability.

Q. Your next paragraph then, ‘It was foreseeable that the airway of Ms Lee and her breathing would be compromised with heavy sedation, resulting in unconsciousness, as well as being in the prone position, resulting in hypoxia (inadequate oxygenation) and brain damage or death and cardiac arrest. Dr Kwan acknowledged this before commencing and during the procedure, as demonstrated by her statements about the risks to the patient and the attendants. When the procedure was finished Dr Kwan did not assess Ms Lee’s condition.’ I think we have gone through where you told us earlier that Dr Kwan told the attendants to get -- the assistants to get the airway out. We have gone through that. ‘Dr Kwan did not assess Ms Lee’s condition after the procedure was finished.’ What should have been done?

A. Having already acknowledged that she might have had problems, after the operation was finished would be an appropriate time to assess her airway breathing, blood pressure, particularly if leaving the -- leaving the operating room. And from the CCTV there was no effort to do that and no -- well, we know at some stage the monitor was turned off. There was no instruction to keep monitoring. I don’t know what the usual practice was of those who were experienced in the room, but it seemed that once the procedure was done within a short period of time we can’t – we don’t know exactly when the monitor was turned off, so there was no -- she was not connected to the parameters of measuring her oxygen saturation or pulse rate.” (Emphasis added)

60.  In this exchange, Dr Mainland accepted that the deceased’s death was not attributable to a singular failure, but rather to the cumulative effect of the breaches she identified.

61.  Dr Mainland’s remarks in her Summary of Conclusions were highlighted by the judge in her summing-up.[127] Her opinion was equivocal, being based on excluding other causes of death to conclude that this “supports cause of death to be the result of over-sedation, respiratory depression, hypoxia and cardiac arrest”.[128] It is relevant that similar remarks made by Dr Chan were excluded as a cause of death by Dr Chiao, who stated that the blood levels of the drugs could not have accounted for the cause of death, and that there was no direct clinical information or post-mortem findings to confirm cardiovascular and respiratory complications.

62.  Dr Chiao testified that there was no sign of hypoxia. However, he explained that if a person died from hypoxia, a hypoxic change in the brain or other organs would only show after the deceased had survived for several or more hours following the hypoxic event. Hence, if death occurred within a few hours of the hypoxic event, a histological examination may not reveal specific signs of damage.

The prosecution closing address

63.  A crucial feature in this appeal is how the prosecution presented and closed its case to the jury. Prosecuting counsel addressed the jury by taking them through the charge in the indictment. She commenced by listing the sedative medications and the dosages that were administered to the deceased as described in Dr Mainland’s report.[129] However, she stated throughout her address that there was no allegation of “overdose”.[130] She also stated that there was no allegation against the appellant in respect of sterilisation or the set-up of the operating room.[131] She then systematically summarised the evidence of the prosecution witnesses, the last three being Dr Chiao, Dr Chan, and finally Dr Mainland, whose expert evidence was the main foundation of the prosecution case. It was unclear whether the prosecution was alleging that one, some or all of the breaches substantially caused the deceased’s death. She told the jury that if they were to come to the conclusion that “there were in fact all these breaches”[132] this was not sufficient to convict, as they had to further address the elements of gross negligence, causation and the grossness of the negligence.[133] She later emphasised to the jury that “the charge is where everything is anchored, so you have to always go back to the charge.”[134] However, she did not address each of the five breaches of duty in respect of the remaining elements, particularly how they individually or cumulatively caused or significantly contributed to the deceased’s death.

64.  Both Dr Chan and Dr Mainland identified serious failings in how the procedure was carried out. There was, in particular, a lack of any formal procedure and of a detailed operative record in respect of a patient undergoing surgery under sedation. Notwithstanding these deficiencies, the key issue at trial was the cause of the deceased’s death, or more specifically, which breach or breaches of duty caused or significantly contributed to the deceased’s death, or whether there was a real possibility that the cause of death was unknown and could not be attributed to a breach or breaches of duty. As we have already noted, the forensic pathologist, Dr Chiao, was unable to determine the cause of death and concluded that it was “unascertained”. There were no signs internally or externally of any trauma or injury, and no abnormality or fatal pathology that could have satisfactorily accounted for the death. In our view, there were justifiable criticisms of the way the liposuction procedure was conducted, but there still remained the fundamental issue of how the deceased died. The medical experts were unable to answer these questions with certainty, relying on an overall impression in postulating possible causes.

65.  The prosecution identified a number of matters where medical requirements or standards were not, or not properly, adhered to during the liposuction procedure. However, the key focus of the prosecution case was the appellant’s breach of duty in failing to ensure sufficient oxygen supply to the deceased during sedation. When prosecuting counsel addressed this matter in her closing address, she noted that there was no evidence of any additional oxygen being supplied to the deceased, pointing out that the alleged breach was that there was no “sufficient” oxygen supply to the deceased during the operation. She stated that if a patient was breathing normally on her own, she may not need “additional or supplementary oxygen”, but asked the jury to consider the size of the deceased and that she was in a prone position during a three-hour operation.[135] She also addressed the medical guidelines for sedation and the risk of cardiovascular, respiratory or airway compromise during procedural sedation.[136] Further on in her closing address, she referred to Dr Chiao’s evidence, stressing that there was no allegation of an overdose. She noted that Dr Chiao’s evidence was that there were no signs of hypoxia, adding that in Dr Mainland’s report hypoxia was not excluded. After referring to Dr Chiao’s evidence as to how to diagnose signs of hypoxia, she told the jury, “so he said that even if a person died of hypoxia, you may not be able to find signs. It depends on how he died, even if it were from hypoxia. So that was the evidence.”[137]

66.  The prosecution, in its closing address to the jury, directed its case on the appellant’s breach of duty in failing to ensure a sufficient supply of oxygen to the deceased during sedation (particular (ii)(b)). This was a critical issue because, in Dr Mainland’s opinion, death resulted from “the additive effect or the combination of several failures”.

The judge’s summing-up

67.  The judge gave a comprehensive and structured overview of the elements of the offence and the evidence in her directions to the jury, drawing their attention to particular matters. The cause of death was the central issue in this case. At the outset of the summing-up, prior to summarising the prosecution case, she indicated that all four assistants had no medical qualifications nor were they qualified in sedation or qualified to monitor a sedated patient[138]. As for the Guidelines, the judge directed the jury “on the basis if you do not conclude they are applicable then you need not consider the guidelines and you will find that particular (ii)(c) has no relevance and is not proved”[139]. Even when directing the jury on the other particulars, the judge repeatedly referred to evidence presented for particular (ii)(c). The directions for particular (ii)(a), (ii)(d) and (ii)(e) referred respectively to paragraphs 3.4[140], 11.2[141], and 5.4.1.7 and 5.4.1.8 of the Guidelines[142].

68.  The prosecution’s reliance on Dr Mainland’s evidence on the issue of the cause of death was emphasised by the judge in her summing-up as follows:[143]

“It is for you to decide, members of the jury, whether you can rely on Dr Mainland’s opinion. If you cannot rely on Dr Mainland’s opinion, you may find the prosecution has not proved its case. If you find that you can rely on Dr Mainland’s opinion, you may find the prosecution has proved its case.”

69.  The judge went on to examine in detail Dr Mainland’s expert evidence on the failings or breaches of duty and the cause of death.[144]

70.  The judge reminded the jury of Dr Mainland’s opinion that “exclusion of causes such as … or other catastrophic medical events supports the cause of death to be the result of over sedation, respiratory depression, hypoxia and cardiac arrest”[145]. She supplemented this with the forensic pathologist’s evidence, indicating that the jury “may find that Dr Chiao was not excluding death by hypoxia”[146], even though the cause of death “has to be labelled as unascertained”[147].

71.  On the question of causation regarding the alleged breaches of duty, the judge directed the jury as follows:[148]

“Members of the jury, as to the failures or breaches, as particularised in para (ii) of the indictment, if you are not sure that the defendant was in breach in any of the particulars that we have mentioned in the indictment, that is the end of the matter. You must acquit or find the defendant not guilty.

Remember, members of the jury however, you need not find that the defendant was in breach of all of the particulars. It can be one, some or all. If you are sure she was in breach in one, some or all, then you will have to go on to consider the third ingredient which is what is particularised however, here in the charge, in the indictment as particular 4, ‘Her aforesaid negligence was a substantial cause of the death of the said Lee Kay-yin’.”

72.  Although the breaches of duty were particularised separately, it would appear the prosecution presented its case on the basis that it was the cumulative effect of all the breaches of duty that substantially caused death. However, the prosecution also emphasised that the lack of supply of oxygen was the main breach of duty, with the others not being capable of substantially causing death. We will discuss this in further detail later in the judgment as it did present a confusing and contrary position regarding the prosecution case. It explains why the judge instructed the jury that any one of the breaches would suffice to find the appellant guilty as long as they also were satisfied that it was a substantial cause of the death of the deceased.

Finding guilt on Dr Mainland’s opinion

73.  Mr Hui argued that the judge erred in instructing the jury that they could find the prosecution’s case proved if they relied on Dr Mainland’s opinion, thereby risking the expert’s views supplanting the jury’s independent evaluation of all the evidence. He submitted that the judge’s direction, “If you find that you can rely on Dr Mainland’s opinion, you may find the prosecution has proved its case”[149], was inadequate and erroneous, as it essentially invited conviction based solely on the expert’s opinion, notwithstanding that the defence had suggested a direction reminding the jury that the final issue was for them to decide[150]. He further submitted that the risk was compounded by the peculiar features of Dr Mainland’s evidence, including: (i) that her opinions were formed over an extended period from 2015 to 2021[151], based on instructions and materials, including the decision in Rudling[152], which had been provided by the prosecution to her to address; (ii) her expertise was as an anaesthesiologist rather than that of a pathologist; and (iii) her consideration of the expert opinions of Dr Critchley and Dr Burd, in relation to which the judge erroneously directed the jury to ignore and improperly restricted the defence’s cross-examination on the matter[153].

74.  The respondent submitted that the judge gave proper directions to the jury on expert evidence. Regarding Dr Mainland’s opinion, the respondent disagreed with the appellant’s arguments. First, the judge’s directions were consistent with Specimen Direction 110,[154] reminding the jury that they were not obliged to accept expert opinion,[155] and that expert evidence formed only part of the case to be considered alongside all other evidence. Secondly, the prosecution’s reliance on Dr Mainland’s evidence did not require the strong warning suggested by the appellant, as Dr Mainland provided detailed and well-explained opinions.[156] Moreover, the defence did not request a strong warning before or after the summing-up, nor in closing submissions. Thirdly, the judge fairly addressed the defence’s concerns about Dr Mainland’s objectivity and reliability[157], reminding the jury to assess whether Dr Mainland was independent or biased, reflecting the defence’s arguments[158]. Fourthly, the appellant’s claim that the judge undermined the general direction on expert evidence was not correct. The judge’s comment about the jury potentially finding the prosecution had proved its case was limited to the issue of deep sedation[159], not guilt, [160] and she reiterated that the level of sedation was entirely for the jury to decide[161]. A similar argument was rejected in HKSAR v Chow Heung Wing Stephen[162], where the Court of Appeal confirmed that the specimen direction ensures the jury does not allow expert opinions to override their own judgment. Fifthly, when it was assumed that the defence’s witness Dr Burd might give evidence, the defence raised concerns about the timing of Dr Mainland’s response to Dr Burd’s report and her suitability as an expert. After discussion, certain parts of Dr Mainland’s report were deleted[163]. Ultimately, the appellant chose not to call Dr Burd. The judge then properly instructed the jury to disregard the name of an expert who was not called[164], as with Dr Critchley’s reports.

75.  We note at this stage that the issue with the directions to the jury lay in the state and presentation of the expert evidence, and in the nature and degree of the uncertainty surrounding the cause of death.

76.  The cause of death focussed on by Dr Mainland was oxygen deprivation, and in respect of this matter, Mr Hui contended that the judge erred by:

(i) instructing the jury[165] to disregard Dr Chiao’s comment that there was no clinical information or post-mortem findings of the deceased’s predisposition to fatal cardiac or respiratory complication[166], which could have undermined Dr Mainland’s opinion that the cause of death was the result of over sedation, respiratory depression, hypoxia, and cardiac arrest;[167]

(ii) treating the Mindray monitor alarms as proof of low oxygen, without reminding the jury that there was no direct evidence of unsatisfactory oxygen saturation level and that the monitor’s settings were unknown[168];

(iii) failing to present possible innocent explanations for the alarms, such as a high baseline setting (“even 98, 87 yelled”[169]) and for the appellant’s instruction to silence the alarm, which could indicate she considered the oxygen level satisfactory; and

(iv) failing to remind the jury that Dr Mainland described her view that oxygen was required was her “speculation”[170].

77.  Mr Hui also argued that the judge failed to instruct the jury not to hold the appellant to a specialist’s standard, and this was compounded by repeatedly referring to specialist practices, such as Dr Mainland’s evidence that anaesthetists provide oxygen even when saturation exceeds 95 per cent.[171] The respondent submitted that the judge fairly reminded the jury of the defence’s argument that the Guidelines applied only to specialists[172] and that the appellant was not a specialist[173]. Furthermore, the appellant had failed to identify any standards as inapplicable to her.

78.  Mr Hui contended that the judge wrongly highlighted Dr Mainland’s opinion concerning the alleged actions or inactions of the appellant in respect of the head, airway, jaw, chest, mouth or nose of the deceased[174], all of which lay beyond the field of view of the CCTV.[175] He also contended that the judge failed to remind the jury that Dr Mainland acknowledged, under cross-examination, that attempting CPR in the prone position was “very reasonable”.[176] The respondent noted that Dr Mainland accepted it was reasonable to attempt CPR in the prone position only if the patient could not be placed supine[177], although no attempt was made to do so.

79.  The respondent submitted that on the matter of oxygen deprivation: (i) the judge did not instruct the jury to disregard Dr Chiao’s comment that there was no clinical information or post-mortem findings confirming the deceased was liable to potentially fatal cardiovascular or respiratory complications; (ii) it was unnecessary to remind the jury that there was no direct evidence on the deceased’s oxygen saturation levels, as the prosecution’s allegation against the appellant was the lack of such record[178]; (iii) it was unnecessary to direct the jury that Mindray’s alarms could have other innocent explanations, as the prosecution’s case was that the appellant repeatedly ignored the alarms without investigating the underlying causes; and (iv) Dr Mainland fairly admitted that she could not see the deceased’s head in the CCTV so she could only surmise the deceased’s airway was not open enough, thus using the term “speculation”[179]. However, her opinion on low oxygen was based on evidence, including the audible and visual alarms from the combined CCTV, the deceased’s prone position, deep sedation, lack of supplementary oxygen[180], the drugs given[181], discussion of oxygen saturation levels shortly before and after the appellant’s return[182], the deceased’s unusual heart rate[183], and the deceased’s snoring and breathing sounds[184]. It was submitted that the judge’s directions on oxygen deprivation were balanced and accurate.

80.  The respondent further submitted that the limitations of the CCTV were obvious to the jury from the evidence (such as the footage itself, photographs and a sketch of the operating room[185]) and the judge’s direction[186]. The absence of footage of the deceased’s head did not render Dr Mainland’s opinion on the appellant’s actions unsupported by evidence. Dr Mainland was entitled to form her opinion based on the events captured on CCTV, including the appellant’s actions and conversations in the operating room. The jury was entitled to decide whether to accept Dr Mainland’s opinion based on their own assessment of the evidence.

The factual basis of the medical opinions

81.  There is a real danger that the reliability and accuracy of an expert opinion will be compromised if it is based on assumed facts which are not ultimately proven, because those facts may warrant a change of opinion. Experts are entitled to their own opinions, not to their own facts. It is therefore important that an expert opinion is based on proven facts, or at least is expressly qualified so as to make clear that the facts are assumed and subject to being established when that is required. However, certain assumptions may legitimately be made based on established principles, practice and procedure in a particular field of learning.

82.  The opinions of Dr Chan and Dr Mainland as to the cause of death were expressly qualified and based on a number of factual assumptions. It is therefore appropriate to identify and address the key assumptions and interpretations made by Dr Mainland, on whose evidence the prosecution principally relied.

Assumptions and interpretations regarding drug administration

83.  Dr Mainland acknowledged that the actual total quantity of sedative drugs administered to the deceased during the procedure was unknown.[187] She initially assumed that only four ampoules of propofol were used because she believed that the entire contents of the clinical waste bin had been seized by the police.[188] She later revised this assumption after viewing the combined CCTV footage and hearing a witness’s testimony in court, concluding instead that at least 6 or 7 syringes of propofol were used.[189]

84.  The defence submitted that there was “simply no evidence that the waste bin contained even more empty propofol vials”, and argued that the fact the contents were not fully seized “is not a basis for saying that there could be more empty propofol vials”.[190] The defence, in its closing speech, highlighted that Dr Mainland ultimately “identified seven syringes” and “assumed that each had two ampoules of propofol and every time two ampoules were drawn”.[191] The defence contrasted this assumption with Kay’s testimony that she only drew up the milky white solution “on at least two occasions”.[192]

85.  Dr Mainland based her opinion on her interpretation of the handwritten anaesthetic record, assuming that the notation “90” meant an infusion rate of 90 ml per hour and “6 CC” meant a 6 ml bolus dose.[193] She acknowledged that her computer modelling of the patient’s propofol levels represented a “best case scenario” because it did not account for the additional bolus doses or the additive physiological effects of other sedative drugs such as Midazolam and Pethidine.[194]

86.  During cross-examination, she stated that she “made that assumption because there was no evidence otherwise”.[195] The defence submitted that the “assumption that there had been a constant rate of infusion is unsafe, making the computer modelling unreliable”.[196] The defence noted that she agreed her modelling “did not take into account any interruptions and the injection or the administration of boluses”, and that the infusion rate could be “reset or re-programmed” when the syringe was changed.[197]

Assumptions and interpretations regarding the airway and patient observation

87.  Dr Mainland admitted that because the CCTV did not capture the patient’s head, her conclusion that an airway intervention was probably required because of a low oxygen level was based on speculation.[198]

“Q. Airway, for conscious sedation no intervention required, that would be the person under conscious sedation would be able to breathe on their own quite easily, is that correct?

A. Maintain an open airway. They’re not likely to -- not going to obstruct.

Q. Deep sedation, intervention may be required.

A. So this depth of sedation may result in some obstruction of the airway. So the intervention would be a manual movement of the jaw and/or an oropharyngeal airway, the airway we saw, or both.

Q. Now, in this case, as you seen from the materials that you were provided with, Madam Lee was lying on her tummy with her face to one side for a period of three hours at least when the operation lasted. In that scenario would intervention be required if it were deep sedation?

A. It’s hard to say. We didn’t see the head -- her head in the CCTV and, as it says here, it may be -- intervention may be required. I believe it probably was required. And the reasons I say that is that it may have been one of the reasons why the pulse oximeter was alarming with low oxygen was that she actually wasn’t getting enough air -- her airway wasn’t open enough to get air in so that she could get the oxygen into her lungs. But that’s just my speculation.” (Emphasis added)

88.  Dr Mainland inferred that the oropharyngeal airway device was not used during the surgery because the CCTV later showed an assistant attempting to insert it while waiting for the appellant to return.[199] She further inferred that the appellant’s recorded question, “Why don’t you know how to handle it?”, was a direct reference to the assistants struggling to insert the airway device.[200]

89.  The defence argued that she relied on the appellant’s instruction to “tak[e] the airway out just in case” to support an assumption that “deep sedation was intended and that Dr Kwan foresaw respiratory or airway problems”. The defence submitted that taking the airway out “just in case” was merely a “precaution”, and that her interpretation that “airway intervention was required … or was intended” was based on “another false assumption” that the monitor alarms meant the patient’s oxygen level was low.[201]

90.  Dr Mainland inferred that the deceased’s leg movements as seen on the CCTV were involuntary reflex actions rather than purposeful responses to pain.[202] She also inferred that the patient briefly regained enough consciousness to speak during the procedure because a delay in changing the propofol syringe caused the drug levels in her blood to drop.[203] Further, she inferred that the deceased was unconscious when the appellant left the room because she asked the assistants to pass a message to the deceased rather than speaking to her directly.[204]

Assumptions and interpretations regarding equipment and monitoring

91.  Dr Mainland was critical of the use made of the Mindray monitor, assuming that its flashing orange light and continuous alarms indicated that the deceased’s oxygen saturation had fallen below a set safety limit.[205] She opined that the fact the machine “kept alarming suggests that the value was always below a set level,” and that if the set level was wrong, “it should have been changed.”[206] However, the defence observed that she reached the conclusion “without knowing what the limits were with any certainty” regarding the settings that would trigger an alarm.[207] The defence submitted that she had assumed the various alarms were related to oxygen saturation, despite agreeing that supplemental oxygen may not be required if a patient is breathing spontaneously and oxygen saturation is above 95%.[208]

92.  A key example of shifting interpretation arose from an audio transcript recorded at 12:17:01, which noted: “98, 97, yelled”, followed by, “These several machines of ours are very sensitive.”[209] In her report, Dr Mainland indicated this would be consistent with a reference to the monitor alarming for oxygen saturation.[210]  Yet, in her oral testimony, she acknowledged the CCTV was unclear and claimed it was ambiguous whether “98, 97” referred to the deceased’s pulse rate or her oxygen saturation level.[211] When asked to clarify, she stated there were “two possibilities” and she did not know “whether the 97/98 was the top number or the flashing number” and that it “could be the pulse”.[212] The defence submitted that the CCTV footage showed “only the bottom figure was flashing at the time”, which corresponded to oxygen saturation, yet she maintained she “did not know whether the speaker talking about that 98, 97 was referring to oxygen or pulse”.[213]

93.  Dr Mainland further assumed that when the appellant pointed to “the orange blinking one” on the monitor, she was explaining the significance of the oxygen saturation values to the unqualified assistants.[214] In her initial report, she stated that the appellant “was not witnessed to look at the monitor” when the alarm sounded. After CCTV footage confirmed the appellant did, in fact, look at the monitor, Dr Mainland changed her opinion. She acknowledged under cross-examination that she “saw two occasions when Dr Kwan did look at the alarm,” but subsequently added the criticism that “there was still no action and no talk of action addressing the continuing alarming.” [215] The defence submitted that there was a reasonable possibility that the appellant had “looked at and considered the reading and then realised it was just a false alarm” or found the values to be normal before an assistant pressed the monitor.[216]

94.  As we have highlighted in our analysis of the expert medical evidence, principally that of Dr Mainland, the factual basis of the medical opinions, and the opinions themselves, were qualified or uncertain, and the alleged breach or breaches of duty that significantly contributed to the death were not specifically addressed, save for the lack of supply of oxygen, which was only addressed in general terms.

95.  We should add that Dr Mainland was given a particularly difficult task, having to deal with a series of evolving instructions in changing circumstances over a protracted period, including instructions concerning the legal requirements of gross negligence manslaughter. However, she commented on a wide range of aspects of the liposuction procedure that required careful evaluation as to whether, and if so how, they bore on the elements of the offence.

96.  In our view, four material errors occurred in the way in which the expert medical opinions were addressed and adduced in evidence in this case. First, Dr Chiao was not asked to consider and provide an opinion on Dr Mainland’s report regarding the cause of death, even though he had been asked to do so in respect of Dr Chan’s report. Secondly, Dr Chiao gave his evidence before Dr Mainland and therefore did not address Dr Mainland’s findings and opinions as to the cause of death. Thirdly, the issue of hypoxia as a cause of death was dealt with only in general terms and without a detailed analysis of the factual and medical evidence relied on to suggest that the deceased had died from a lack of oxygen. Fourthly, there was a failure to address the real possibility that the deceased’s death was of unknown cause. This must be set against the medical opinions of Dr Chan and Dr Mainland who, while critical of the conduct of the liposuction procedure, expressed a view as to the cause of death which could not be established with certainty.

97.  To illustrate the significance of these errors, it is necessary to examine Dr Chiao’s evidence concerning Dr Chan’s report.[217] In his evidence, Dr Chiao adopted a memorandum to the prosecution providing his opinion of Dr Chan’s report on the cause of death. The relevant parts of the memorandum read:

“4. Dr. CHAN Yu-wai’s expert opinions related to the cause of death of the deceased were summarized as below:

i. Aspirate is relatively bloody, signifying the possibility of peripheral blood vessels injury during the procedure. Fat globules may be dislodged into the blood vessels that can potentially cause embolism.

ii. Toxicological examination has identified the presence of pethidine, lignocaine and propofol in the blood sample, but the levels were under lethal level and the chance of drug overdose as a direct of cause of death is low.

iii. As pointed out in the post-mortem report and toxicological examination, the apparent cause of death is not clear.

iv. With a patient weighing 116kg undergoing a surgical procedure lasting for more than 3 hours under sedation in the prone position, in the absence of registered anaesthetist and adequate airway protection, Madam Lee was prone to suffer from cardiovascular and respiratory complications that were potentially lethal.

v. Comments from an anaesthetist are helpful regarding the cardiovascular and respiratory protection during surgery under sedation.

5. My comments on Dr Chan’s expert opinions related to the cause of death of the deceased were as follows:

i. As microscopic examination did not reveal large fat globule in the blood vessels of lungs and brains, it was very unlikely that the deceased had died of fat embolism.

ii. I agree with Dr Chan that the chance of drug overdose was low basing on the result of toxicological tests.

iii. I agree with Dr Chan’s opinion that Madam Lee was prone to suffer from cardiovascular and respiratory complications that were potentially lethal as the deceased weighed 113kg and had gone through the medical procedure more than 3 hours under sedation in the prone position in the absence of registered anaesthetist and adequate airway protection. However, there was no direct clinical information or post-mortem findings to further confirm this speculation.

iv. I agree with Dr Chan’s opinion that comments from an anaesthetist are helpful regarding the cardiovascular and respiratory protection during surgery under sedation.” (Emphasis added)

98.  In this memorandum, Dr Chiao also commented on Dr Critchley’s report, which was redacted. As already noted, Dr Chiao was not asked to comment on Dr Mainland’s report regarding her opinion on the cause of death. As seen from Dr Chiao’s comments regarding Dr Chan’s opinion on the cause of death, one field of medicine may answer or qualify the views of another. Importantly, Dr Chiao’s expertise is in determining the causes of death.

99.  We have been critical of both the presentation and various aspects of the medical experts’ evidence. In this regard, it is appropriate to refer to the remarks of the English Court of Appeal in R v Sellu[218], when considering the expert medical evidence in the prosecution of a doctor for manslaughter by gross negligence. The Court observed that the experts who had given evidence at trial had described the defendant’s actions and inactions in a variety of different ways; while some of those descriptions provided a yardstick against which the jury could consider whether the criminal test had been met, but others were little more than assertions. The Court remarked that the jury had been left on their own to trawl through the differing descriptions, essentially asking themselves whether the behaviour under discussion was or was not gross negligence. Furthermore, the jury were not directed as to how much weight should be attached to these assertions as opposed to other qualitative descriptions. Moreover, the jury were not sufficiently assisted in understanding how to approach their task of identifying the line that separated even serious or very serious mistakes or lapses from conduct that was truly exceptionally bad and represented such a departure from the standard of a reasonably competent doctor that it amounted to criminal conduct.

100.  These remarks are apposite to the present case, but their relevance is amplified given that the task of the experts was to find a cause of death which could not be ascertained from the pathology of the deceased. The jury were therefore presented with opinions from the medical experts that were sometimes based on assumptions or amounted to mere assertions. This included criticisms of the appellant’s conduct regarding matters of practice that could not constitute conduct so truly exceptionally bad, and such a departure from the standard of a reasonably competent doctor that it consequently amounted to being criminal.

101.  In summary, the conviction was rendered unsafe by the way in which the medical expert evidence was obtained, developed and presented at trial. The prosecution case depended almost entirely on expert evidence which was qualified, rested on unproven assumptions and an incomplete consideration of the forensic pathology evidence, and did not adequately confront the realistic possibility that the cause of death was unknown. In these circumstances, the jury could not safely be invited to treat Dr Mainland’s opinion as determinative either of causation or of guilt.

B. Dr Critchley’s report and the unmuted CCTV footage[219]

102.  This conclusion on the handling of the expert evidence leads naturally to the next issue, namely whether the defence was unfairly restricted in its ability to challenge the expert evidence through cross-examination of references to Dr Critchley’s report and the unmuted CCTV footage.

103.  It was alleged that there was a material irregularity during the trial arising from the judge’s refusal to allow defence counsel to cross-examine on: (i) the contents (or any references) of Dr Critchley’s expert report, despite it being considered by the prosecution’s expert witnesses; and (ii) the original unmuted CCTV footage admitted as Exhibit P55. The appellant submitted that this prevented her from fully presenting her defence, particularly regarding the deceased’s level of sedation, thus undermining her right to a fair trial.

Dr Critchley

104.  Mr Hui pointed out that all prosecution expert witnesses at trial expressly considered Dr Critchley’s report. He contended that the judge’s refusal to allow any reference to Dr Critchley[220], without hearing proper argument or consideration of the defence’s intended cross-examination, constituted a material irregularity. He argued that this denied the appellant a fair opportunity to challenge the prosecution’s experts on key issues, including Dr Critchley’s opinions that:

(i) no published guidelines governed sedation in surgical ambulatory clinics as of June 2014[221];

(ii) the sedation provided met generally accepted private sector standards;

(iii) the procedure performed by the appellant caused no direct harm to the deceased; and

(iv) the Mindray monitor provided reasonable and appropriate patient monitoring.

105.  Mr Hui further added that Dr Critchley’s opinion on the cause of death, namely “(his) guess … a combination of factors that in isolation were innocent, but together lead to significant cardiorespiratory depression (i.e. coma) followed by cardiorespiratory arrest that was not promptly treated”[222], would cast doubt on Dr Mainland’s opinion, that by excluding other causes, the cause of death would be the result of over sedation, respiratory depression, hypoxia and cardiac arrest[223].

106.  The respondent emphasised that Dr Critchley was neither called nor listed as a prosecution witness, noting that he did not view the relevant CCTV footage and his reports contained speculative opinions. The respondent further noted that Dr Mainland did not adopt or rely on Dr Critchley’s opinions. In this context, it was appropriate for the judge only to allow the defence to cross-examine on the contents of the reports but not to adduce the reports.[224] The defence raised no objection to the removal of references to Dr Critchley.[225] The respondent submitted that the judge’s decision was correct, as reliance on untested expert opinions could mislead the jury, and cross-examination on inadmissible evidence is improper.[226] The respondent further submitted that the judge maintained a consistent position throughout the trial, and that the defence had sufficient opportunity to explore the relevant issues with Dr Mainland in cross-examination, for example, regarding the applicability of the relevant guidelines.[227]

107.  That is not entirely correct. In the initial exchange between the judge and defence counsel, the judge correctly ruled that Dr Critchley’s two reports could not be produced through Dr Mainland. However, she added that Dr Mainland could be cross-examined about any references she made to Dr Critchley’s report.[228] This was accepted by defence counsel. Later, when Dr Mainland gave her evidence-in-chief, prosecuting counsel produced her final report, referring to Appendix 2 (Instructions), and informed Dr Mainland that any reference to a prior opinion was not a concern in the case. Prosecuting counsel then deleted paragraphs 2.5 and 2.6 in Appendix 2, which made reference to Dr Critchley. This section of the Instructions read as follows:[229]

“2.4 Whether the use of the drugs in the quantities found had any bearing on the

2.5 death of the Deceased and if so to what degree, given Dr Critchley’s opinion as to over sedation as the cause of death;

2.6. Opinion as to the cause of death having regard to all the circumstances, if in the opinion of the expert the drugs administered played no part or no substantial part in the death and /or I disagree with the opinion of Dr Critchley as to the cause of death.”

108.  The judge then instructed the jury to delete and ignore these paragraphs and told them that, “Dr Critchley will not be called to give evidence. Therefore any thing and any reference to him is inadmissible as evidence before you. So it is not evidence before you of anything that Dr Critchley is alleged to have said.”[230]

109.  As we understand the position, Dr Mainland drafted her final report in which she addressed Dr Critchley’s report as instructed. The prosecution then sought to delete that part of Dr Mainland’s instructions and did not examine her about Dr Critchley’s report. However, Dr Mainland listed it as a document she had considered in drafting her report, and the defence were therefore entitled to cross-examine her on how Dr Critchley’s report had a bearing on her opinions or views.

Unmuted CCTV

110.  Mr Hui submitted that Dr Mainland emphasised the importance of the deceased’s verbal responses in assessing sedation depth. Accordingly, whether the deceased spoke during the procedure was highly relevant to the defence. Kay, an assistant, testified that the deceased was speaking or slurring and moving her arms and legs from time to time during the operation[231]. The defence sought to cross-examine PW4 on the unmuted CCTV (in relation to the utterances “Don’t press my hand” at 13:44:52, which the prosecution agreed was audible after multiple attempts to listen[232], and “Painful, painful, painful” at 13:48:16), but the judge wrongly refused to allow cross-examination, stating that only the muted version was before the jury[233], despite the unmuted footage being admitted as Exhibit P55[234]. Mr Hui contended that this deprived the appellant of the opportunity to show further evidence of the deceased speaking, and therefore to present her defence fully.

111.  On the unmuted CCTV footage, the respondent placed the judge’s ruling in context.[235] The footage, primarily in Cantonese, was partially muted and transcribed into English, as Dr Mainland did not speak Cantonese and the trial was conducted in English. Both the prosecution and defence invested significant effort in agreeing on the transcript. The defence initially agreed to muting any audio not covered by the transcript[236], but later sought to cross-examine Kay on the unmuted footage[237]. Having heard the parties’ submissions, the judge disallowed use of the unmuted footage for cross-examination, as it had not been shown to the jury, and no transcript existed.[238] The respondent argued that this ruling was fair and did not prejudice the defence, and that the defence could not identify any substantive conversation between the deceased and others in the operating room on the unmuted footage. In any event, the defence was eventually permitted, by agreement with the prosecution, to cross-examine Kay [239] on the relevant segment (allegedly showing the deceased say, “Don’t hold down my hand” during the operation), although Kay could not hear it herself [240] and testified that the deceased could not converse meaningfully under sedation[241].

112.  There was no adverse effect from the judge’s restriction of the cross-examination of Kay because the prosecution accepted that the deceased made the two verbal utterances during the liposuction procedure, namely “Don’t press my hand” and “Painful, painful, painful”.[242] Furthermore, the defence referred to the English translation of the Chinese transcript of the audio recording, in which commands in Chinese were repeatedly given throughout the procedure, such as “Don’t move” and “Relax”.[243] It was submitted that this evidence indicated that the deceased was able to respond to instructions, and was relevant to Dr Mainland’s opinion that the deceased was in a state of deep sedation.

113.  Taken together, these matters succeed only in part. The restriction on cross-examination about the two verbal utterances on the CCTV did not cause any material prejudice, as the prosecution ultimately accepted that those utterances were made and the jury were invited to proceed on that basis. However, the manner in which references to Dr Critchley’s report were removed from Dr Mainland’s instructions, coupled with the direction that the jury should ignore any reference to him, did curtail the defence’s ability to explore the provenance and development of the expert opinions on which the prosecution relied. This irregularity must be viewed in conjunction with the wider difficulties in the expert evidence previously considered.

C. The elements of foreseeability and causation[244]

114.  On the issue regarding the element of “foreseeability”[245], it was contended that the judge’s direction, given after her direction on “causation”[246], was liable to confuse the jury, and that this prejudice was compounded by the erroneous direction to the jury that they were entitled to consider multiple breaches of duty of care together in approaching the issue of foreseeability.

115.  Mr Hui argued that in a case where the prosecution’s main allegations concern a series of omissions, it was more logical for the jury to consider foreseeability first. This order matched the pleading of the breaches, how the prosecution presented its case, and the defence’s response. The judge’s directions followed the Specimen Directions in Jury Trial revised in 2020, particularly in relation to the elements of foreseeability and causation. We do not consider the sequence caused any unfair prejudice in the circumstances of the case, given that the judge, in her direction on foreseeability, stressed that “foreseeability is to be determined at the time of the breach” and warned against hindsight. [247]

116.  Mr Hui further argued that the directions were erroneous and inadequate, particularly as the prosecution did not specify when the deceased faced an obvious and serious risk of death, and Dr Mainland conceded she could not say if any single act would have prevented death[248].

117.  On the issue of the breach of duty that substantially caused death, the charge set out five breaches which correlated with the matters mentioned by Dr Mainland. It appears that Dr Mainland and prosecuting counsel submitted that it was the cumulative effect of these breaches that substantially caused the deceased’s death, whereas the judge directed the jury that it was sufficient if they found that the appellant breached “one, some or all” of them.[249] In submissions before us, the respondent submitted that the prosecution case was that the cumulative effect of all five breaches substantially caused the deceased’s death.

118.  This resulted in a serious error at trial because the jury were told they could find the appellant guilty on any one of the breaches, when this was not the prosecution case. The prosecution closing address was a general description of the evidence in the chronological order in which it was presented, without explaining the basis of the prosecution case in respect of the breach or breaches that substantially caused the deceased’s death. However, it appeared that the prosecution’s case was primarily based on Dr Mainland’s evidence regarding the breach of failing to ensure sufficient oxygen supply to the deceased during sedation (particular (ii)(b)). In these circumstances, the jury were presented with a confusing and conflicting set of instructions regarding the breach or breaches of duty that substantially caused the deceased’s death.

119.  It will be recalled that the breach of duty of care alleged in the indictment in particular (ii) was: (a) failing to have a qualified person present to administer and monitor the sedation; (b) failing to have sufficient oxygen supply; (c) failing to follow the Guidelines; (d) failing to provide sufficient monitoring after the liposuction procedure; and (e) failing to provide adequate resuscitation.

120.  There were evidential issues with each of the particulars, but the prosecution mainly focused on particular (ii)(b) to allege that this was a substantial cause of death. They also contended that the particulars collectively were a substantial cause of death.

121.  Prosecuting counsel in her closing address explained the prosecution case in respect of particular (ii)(b):[250]

“There is simply no evidence of additional oxygen supplied to Lee during the operation at all. You will recall the airway was only there to ensure that there was a smooth passage for the air to go in and out. That is not additional oxygen. So there is no sufficient -- and ‘to ensure sufficient oxygen supply to the said Lee Kar-ying, Josephine, during sedation’. We’re talking about the time when the operation started till the operation finished, during the operation.

There is simply no evidence of additional oxygen supplied.

But that’s not what the charge says. The charge says ‘sufficient oxygen supply’. If a patient was breathing smoothly on his or her own, you may not need additional or supplementary oxygen. But is it in this case? Bearing in mind the size of Ms Lee, the fact that she was in a prone position with her tummy down on the bed for the whole of the operation, and the fact that the operation took three hours.

Now, given that background, without additional oxygen supplied to Ms Lee, was there sufficient oxygen supplied? It’s different from somebody who is not obese, as Dr Chan had referred Ms Lee to, and a 20 -- or an hour operation in the prone -- or a -- prone position. We’re talking about very different scenarios.

But in this case, you should always have the back -- at the back of your mind, remember, the size of Ms Lee, the operation was three hours, and the position that Ms Lee was to be in for the purpose of this operation. You’ll recall the liposuction was on her back. So if she -- there is only one position that she can be in, and that would be with her tummy on the bed.

So we say, in the circumstances, that the defendant failed to ensure sufficient oxygen supply to the said Ms Lee.”

122.  The question arose as to whether the deceased died because of her prone position and the lack of oxygen during the liposuction procedure. Dr Chiao was asked about postural asphyxia caused by the position in which the body was placed. He explained that this condition was when the posture impeded the respiratory movement (breathing movement) of the body. However, there was no post-mortem evidence to support this affliction as a cause of death.[251] As for the lack of oxygen as a cause of death, he explained in his evidence:

“Q. In layman terms, does the brain show any abnormalities?

A. No.

Q. There’s no sign of the brain being starved of oxygen, is there?

A. No.

Q. No sign of I think what doctors call hypoxia.

A. No.

COURT: No sign of?

MR LEUNG: Hypoxia.

COURT: How do you spell that? H-Y?

MR LEUNG: H-Y...

A. P-O-X-I-A.

COURT: P-O-X-I-A. What does that mean?

A. Means a lack of usage of oxygen by the tissue or organs.

Q. So if a person died of hypoxia, it will show in the brain tissue, is that right?

A. If time the -- the time of hypoxia is long enough, then you find hypoxic change in the brain or other organs, but if the time of hypoxia is relatively short or the death ensues quickly after the hypoxia, we may not be able to find the hypoxic change.

COURT: Hold on, please. Will not be able to find the?

A. The hypoxic change.

COURT: Hypoxic. Right, thank you.

Q. Correct me if I’m wrong. The starve of oxygen or lack of oxygen also a pathway to death. When your heart stop beating, of course, you suffer ...

COURT: Mr Leung, sorry, could you start again? ‘Correct me if I’m wrong,’ sorry, what did you say?

Q. The starvation of oxygen is also a pathway to death.

A. I can’t understand, sorry.

A. Correct me if I’m wrong.

A. I can’t understand the question, sorry, would you please rephrase it?

Q. I suppose the mode of dying is usually starting from cardiac arrest, is that right, your heart stop beating?

A. Most of them, yes.

Q. And once your heart stop beating, your body will be deprived of oxygen supply.

A. Yes, correct.

Q. So what I’m asking is the lack of oxygen supply is also a pathway to death.

COURT: Sorry?

Q. Lack of supply of oxygen to the brain or to any other organ ...

A. Yes, sure.

Q. ... is also a pathway to death.” [252]

…

“Q. So when you say in paragraph 10, ‘In summary, the cause of death in this case has to be labelled as unascertained because the autopsy did not reveal an abnormality’, what do you mean by that?

A. Abnormality including medical problem and also the test result of the samples, the blood samples, the bodily samples.

Q. Sorry?

A. The results of the bodily samples.

Q. So there’s no abnormality in the medical problem and no abnormality in the body sample, is that what you’re saying?

A. There are some abnormality, but they are not enough for -- to be accounted for the cause of death.

Q. And then the next word you used ‘because the autopsy did not reveal a fatal pathology’, pathology, ‘that could have satisfactorily accounted for the cause of death’, what do you mean by that?

A. A fatal pathology is usually a single disease which is lethal. For example, in a case with sudden heart attack, then the fatal pathology is the heart problem.

Q. For example, a case of heart attack then?

A. Then the heart problem is the fatal pathology.

Q. So in this case, for example, are you saying there’s no fatal pathology in the heart, there’s no heart attack?

A. No heart ...

Q. Is that right?

A. Correct.

Q. Would I be correct to understand you to say that on your examination of the body -- sorry, I may be a bit slow here. I’ll start again. Could you just describe to me again pathology, what is that, pathology?

A. It’s some abnormal tissue that cause disease is the pathology.

Q. Would you be able to tell on your autopsy if somebody -- if Madam Lee died as a result of lack of oxygen, would you able to tell that or is that how you -- is that what you say? I don’t know what it’s called.

A. If the death comes shortly after prolonged period -- a period of lack of oxygen, then I may not be able to find pathology. We call it hypoxic change because it takes some time for the body to show up the hypoxic change. Usually a person with hypoxia it takes hours to show up the hypoxic change.

Q. Sorry, I’m not following you there.

A. Okay.

Q. I think this is a question that Mr Leung asked. I’m not following you there, I’m sorry.

A. Okay. I explain in another way. If a person is suffocated to death and ...

Q. Suffocated to death.

A. Yes.

Q. So for example -- I don’t know -- put a pillow case over his head, is that ...

A. Yes, yes.

Q. Or a plastic bag over the head, is that a good example?

A. Yes. If you place the pillow on a person’s face and after minutes the person died and I’m not able to find the hypoxic change. But if the assailant at some point released the pillow but the brain has already been damaged by the deprivation of oxygen, then several hours later will start be able to find the hypoxic pathologies.

Q. Will be able to find the hypoxic, what did you call it?

A. Hypoxic changes.

Q. Change. And what is that, the hypoxic change, what is it exactly?

A. The brain will -- the brain cells will die and they will change in colour or even you see the disintegration of the we call it neuron, the brain cells.”[253]

123.  It was alleged that the judge failed to properly direct the jury on causation, in circumstances where the prosecution’s case relied on Dr Mainland’s opinion of “the additive effect or combination of several failures that together contributed to the situation where death occurred” and “there was a collection of actions and inactions that created the situation where death was a high possibility, probability”[254] and where no cause of death could be ascertained.

124.  Mr Hui argued that the judge’s direction that “one, some or all” particulars of breach could establish causation [255] was inconsistent with the prosecution’s case[256] and Dr Mainland’s opinion, both of which required proof that all (or several) failures acted together. The jury were not properly instructed that causation depended on the combined effect of multiple breaches. This was mainly due to the evolving and uncertain nature of the prosecution case.

125.  Mr Hui complained that Dr Mainland was not a pathologist able to give evidence on the cause of death and that she wrongly was of the view that the cause of death could be determined by elimination[257]. He was critical that the jury were not reminded of Dr Mainland’s evidence that death was a “high possibility, probability” due to a combination of actions or inactions, raising doubt about whether causation was proved beyond reasonable doubt. He further complained that Dr Mainland’s opinion that the deceased suffered from low oxygen levels was based on her own interpretation of the Mindray monitor alarms and that she testified that the deceased was “not catastrophically hypoxic”. Furthermore, the jury were wrongly instructed that Dr Chiao did not exclude death by hypoxia[258], and there was no reference to Dr Chiao’s finding that there was no evidence of postural asphyxia, contrary to Dr Mainland’s view that low oxygen was due in fact to the deceased prone position.

126.  The respondent submitted that the judge reminded the jury that the purpose of seeking the opinion of a plastic surgeon and anaesthetist was to “see whether there is some medical event that’s happened during the surgery”.[259] No further warning or direction was necessary. Furthermore, the defence’s argument that only a pathologist could determine the cause of death was incorrect; in appropriate cases, the jury is entitled to assess this based on all the evidence (HKSAR v Hui Chi Wai and Others[260]). However, Dr Mainland was called to establish, amongst other things, the cause of death, which the defence did not challenge. It is submitted that the judge reminded the jury that Dr Mainland was not a pathologist and could not testify on certain aspects of the cause of death (as a pathologist)[261], but could do so from the perspective of an anaesthesiologist.[262] Moreover, Dr Chiao’s opinion was not incompatible with that of Dr Mainland.

127.  In our judgment, the complaint about the basis for conviction was made out. The jury were invited to convict on the basis that “one, some or all” of the alleged breaches might have substantially caused the deceased’s death, even though the prosecution case and Dr Mainland’s evidence were premised on a combination of failures acting together. Coupled with the uncertainty and qualifications in the expert evidence on causation, and the absence of clear identification of which breach or breaches were said to give rise to a serious and obvious risk of death, the directions on causation did not sufficiently equip the jury to apply the correct legal test so that they were sure of the appellant’s guilt for this offence.

D. The failure to give a Kevin Brown direction[263]

128.  It was alleged against the appellant that she breached her duty of care to the deceased by failing: (a) to have a qualified person to administer and monitor sedation; (b) to ensure sufficient oxygen supply to the deceased during sedation; (c) to follow the requirements in specific paragraphs of the Guidelines; (d) to provide proper and sufficient monitoring after the liposuction procedure to the deceased; and (e) to provide adequate and timely resuscitation to the deceased.

129.  Mr Hui contended that the judge failed to give a Kevin Brown direction and failed to sufficiently instruct the jury as to how they should approach their tasks en route to verdict where there were multiple particulars of breach of duty of care alleged, each capable of amounting to gross negligence. In the context of the present case, such a direction was requested[264], but declined[265] (Mak Wan Ling[266]).

130.  Mr Hui argued that the standard directions on reaching a verdict[267] were inadequate in these circumstances. The indictment alleged five particulars of breach of duty of care, covering the appellant’s conduct before[268], during[269], and after[270] the operation, each purportedly capable of constituting gross negligence. However, the judge directed the jury to consider whether “one, some or all” of the particulars were proved, without addressing the required degree of unanimity for each[271]. In the absence of a Kevin Brown direction, there was a real risk that the jury was divided as to which particular(s) they relied upon when considering foreseeability, causation, and the grossness of the negligence.

131.  The respondent submitted that, although no Kevin Brown direction was given, the judge’s directions, particularly regarding the requirement for proven breaches[272], were adequate and that the absence of such a direction did not affect the safety of the conviction. The prosecution was presented on the basis that all breaches cumulatively caused the deceased’s death, consistent with Dr Mainland’s evidence that the “additive effect or combination of several failures” contributed to the fatal outcome. It was argued that this was not a case where individual particulars in the indictment corresponded to separate offence ingredients or involved different defences, which would require a Kevin Brown direction (R v Chilvers[273]). As this was the sole evidential route to conviction, and there was no cogent challenge or contradictory evidence from the defence, the possibility that the jury accepted only some breaches was not realistic.[274] Accordingly, the conviction should stand even without a Kevin Brown direction (R v Boreman[275]).

132.  The prosecution initially put its case on the basis that these five alleged breaches were concurrent causes of the death of the deceased, each of which individually would suffice to amount to gross negligence. This would require the jury to decide which breach was proved and then to consider the elements of foreseeability, causation and grossness of the negligence. Since there were five individual breaches of duty alleged, a Kevin Brown direction should have been given to the jury. The judge directed the jury that “one, some or all” of the five alleged breaches, if proved, would constitute gross negligence.[276] However, it appears the prosecution changed its case by focusing on the cumulative effect of only some of these alleged breaches of duty.

133.  We therefore conclude that this ground is also made out. In a case alleging five distinct particulars of breach of duty, each said to be capable of amounting to gross negligence and of contributing to death, the failure to give a Kevin Brown direction created a real risk that different jurors relied on different combinations of breaches when considering foreseeability, causation and grossness. This defect, when combined with the uncertainties in the expert evidence and the problems with causation, materially undermines the safety of the conviction.

134.  Those cumulative concerns must finally be viewed in light of the unpleaded but fundamental issue of an unknown cause of death, which lay at the heart of both the pathology and the expert opinion evidence.

E. The issue of an unknown cause of death

135.  The pathologist’s finding that the deceased’s cause of death was “unascertained” was a fundamental issue in this case. The defence advanced its case on the basis that the cause of death was unknown and that the deceased’s death could not be attributed to a breach or breaches of duty by the appellant. Defence counsel in his closing address to the jury explained:[277]

“There’s always a risk, an inherent risk, of death associated with all surgical procedures, even with the seemingly simple procedures. …

Of course, we want to know why Madam Lee died. Not knowing can be disconcerting. … After all, our criminal justice system is one that -- is not one that you attribute blame and punish those based on things unknown or on gut feeling or even on distaste to events.

One matter I wish you to -- I wish to remind you is that there is no allegation in this case against the defendant for her surgical skills, no allegation of poor facilities, no allegation in relation to the sterility of the operating room, no allegation of drug overdose, and no allegation in relation to the first liposuction surgery performed by Dr Kwan on Madam Lee on 30 April 2014. It is also not disputed that the cause of death in this case was labelled by the pathologist, Dr Chiao, as unascertained.”

136.  In his concluding remarks to the jury, defence counsel emphasised:[278]

“Then we come to the question of causation. It -- what may be difficult for you is that the cause of death in this case, as agreed by all experts, is unknown, not ascertained. So how then do you approach whether the action or inaction of Dr Kwan was a substantial cause of death? Say, if a doctor, doctor’s action, was overprescribing some medicine, but the cause of death was later found out to be an infection, can you say in those circumstances that the action of the doctor contributed significantly to the cause of death? I think not.

What we have in this case is that the cause of death cannot be identified. Whatever you find as inaction or action, it must be related to having substantially causing the death of Madam Lee. We say it is a difficult task and one which the prosecution cannot prove and, even if you find it proved, there still remains the question: whether the action or inaction amounted to gross negligence to a criminal standard.”

137.  R v Henderson, R v Butler and R v Oyediran [279] (together, “Henderson”) concerned cases of “shaken baby syndrome” where the evidence against the defendants was based solely on medical expert evidence. The English Court of Appeal addressed several important issues concerning the reliance on such medical expert evidence in criminal prosecutions. Moses LJ, giving the judgment of the Court, stated that in cases of shaken baby syndrome where the prosecution relies on medical experts to identify a non-accidental injury and the defence offers no alternative cause, the temptation to conclude that the prosecution has proved its case must be resisted, because in this, as in many fields of medicine, the evidence may be insufficient to exclude an unknown cause beyond reasonable doubt (Henderson[280]).

138.  Moses LJ observed that a conviction based merely on the evidence of experts can only be regarded as safe if the case proceeds on a logically justifiable basis, including a logically justifiable basis for rejecting or accepting the expert evidence. He emphasised the importance of proper pre-trial management in identifying the real medical issues and enabling the judge to prevent experts from wandering into unnecessary, complicated and confusing detail. He explained that a realistic possibility of an unknown cause must not be overlooked, and that the jury should be reminded of that possibility and instructed accordingly. He stated:

“217. There are two features of the content of a summing-up in cases such as these which, we suggest, are important. First, a realistic possibility of an unknown cause must not be overlooked. In cases where that possibility is realistic, the jury should be reminded of that possibility. They should be instructed that unless the evidence leads them to exclude any realistic possibility of an unknown cause they cannot convict. In cases where it is relevant to do so, they should be reminded that medical science develops and that which was previously thought unknown may subsequently be recognised and acknowledged. As it was put by Toulson L.J, ‘today’s orthodoxy may become tomorrow’s outdated learning’ (R. v Holdsworth [2008] EWCA Crim 971 at [57]). In cases where developing medical science is relevant, the jury should be reminded that special caution is needed where expert opinion evidence is fundamental to the prosecution [57].

218. Second, the jury need directions as to how they should approach conflicting expert evidence. Kai-Whitewind teaches that the mere fact that expert differs from expert is no ground for withdrawing the case from the jury. But how is the jury to approach such a conflict? To suggest, in cases where the expert evidence is fundamental to the case, that the jury should approach that expert opinion in the same way as they do in every other criminal case, is inadequate. It is difficult enough for Family Division judges to express their reasons for accepting or rejecting conflicting expert evidence, despite their experience. Juries, we suggest, should not be left in cases requiring a higher standard of proof to flounder in the formation of a general impression. A conclusion cannot be left merely to impression. In the appeal of Henderson, Dr Leestma gave, if we may say so, a most beguiling impression, courteous and understated as it was. But there were, as we have concluded, sound reasons relating to his experience in comparison with Dr Al-Sarraj for rejecting what he told us. Lacking the experience of Family Division judges, a jury needs to be directed as to the pointers to reliable evidence and the basis for distinguishing that which may be relied upon and that which should be rejected.” (Emphasis added)

139.  Whilst the above statements of principle were made in respect of an area of medical science that was developing and where medical opinions were conflicting, in our view, they apply equally where medical expert opinion is unclear and qualified and where there is a realistic possibility that the cause of death is unknown. It is essential that the medical issues which the jury are to resolve are clear by the commencement of the trial, ensuring that the issues are defined and that the expert evidence, and the sources on which it is based, are also clearly identified (Henderson[281]). We conclude that the way in which the medical expert evidence was addressed and evolved over a period of six years, and ultimately presented at trial, was less than satisfactory. The medical issues were not clearly and fully explained to the jury, and the realistic possibility that the deceased died of an unknown cause was not addressed at all.

140.  It is appropriate to examine in detail the decision in Henderson because it puts into context the issue of an unknown cause of death and sets out a number of important general principles regarding medical expert evidence. Shaken baby syndrome cases are those where a baby allegedly died as a result of being deliberately shaken. Whilst there is no external trauma or injury to the baby, a “triad of injuries” – retinal haemorrhages, subdural haemorrhage and encephalopathy – was considered by medical experts to provide a strong pointer to non-accidental head injury. None of the three cases in Henderson was concerned solely with the triad of injuries. In Henderson, the issue was whether there was evidence of injury independent of and additional to the triad of injuries. In Butler, one of the issues was whether the unusual fact of the baby’s recovery after the discovery of retinal haemorrhage indicated a cause other than shaking. In Oyediran, the prosecution alleged a distinct and separate external injury, consisting of a fracture of the baby’s arm.[282]

141.  We emphasise that in each of the three cases, as in all cases of shaken baby syndrome, the baby was shown to have suffered internal trauma or injury. In the present case, the deceased showed no trauma or injury, either internally or externally. This is highly relevant to the issue of an unknown cause of death.

142.  The key consideration in Henderson was the form and structure of the directions given by the trial judge, Keith J (previously a member of the Hong Kong Court of Appeal and now Deputy High Court judge). The Court commended his directions, noting that he made clear to the jury the issues they had to decide and the rival bases on which they could reject or accept the evidence, adding:[283]

“In particular, the judge directed the jury as to the need to bear in mind that medical science in relation to non-accidental head injury is developing, and as the defence would have it, uncertain, and that it may not be possible to identify the cause of death. He gave a positive and correct direction that the jury had to consider whether the death may be unexplained. He directed them that they had to be sure that unexplained death could be excluded.

Further, it is of importance to note that he stressed that the defence was under no obligation to advance any particular theory as to cause of death and that rejection of the theory of spontaneous seizure did not of itself mean that the jury was required to accept the prosecution case and convict.

Accordingly, this is a case where the issue of unexplained cause in an area of developing medical science was properly laid before the jury. The justification for rejection of that possibility and for acceptance of the prosecution case is plain from the summing up.”

143.  The Court dismissed the appeal in Henderson because the trial judge had properly directed the jury on the issue of an unknown cause of death, while in Butler it allowed the appeal because such a direction had not been given. In Oyediran, the Court upheld the murder conviction because of the combination of the severity of the injury to the brain and the fracture to the arm, both due to the appellant’s own deliberate actions.

144.  In another subsequent shaken baby syndrome case, R v Arshad [284], the English Court of Appeal applied Henderson in holding that the jury had to consider the realistic possibility of an unknown cause and that special caution was needed where expert opinion in a developing science was fundamental to the prosecution. However, the Court held that those precise terms did not need to be used, provided “the effect of the summing-up must be to make clear that special caution is needed and unknown causes must not be overlooked”.

145.  In that case, the Court considered that the jury would have understood about the uncertainty that existed in some areas of medical science and about the possibility of an unknown cause (Arshad [285]). The Court dismissed the appeal, being satisfied that the jury could have been in no doubt that, in a case where medical science was uncertain in the light of the findings made by the doctors, they could only convict if they were sure on the evidence that the death had been caused in the manner alleged and that other possibilities had been excluded (Arshad [286]).

146.  The question in the present case, as in Henderson, is whether the directions given by the judge to the jury were sufficient to remind them of the realistic possibility of an unknown cause, since an overall impression can never be a substitute for a rational process of analysis (Henderson [287]). This was not in any way the fault of the judge, but reflected how the prosecution case based on medical expert evidence had been constructed and presented to the jury.

147.  Mr Jonathan Man, for the respondent, argued that this was not a case where a direction on an unknown cause was necessary. He accepted that in the present case there was a possibility of an unknown cause, but submitted it was not a “realistic” possibility. With respect, his argument was simply a play on semantics. There was a strong evidential foundation that the cause of death was unknown. On the forensic pathologist’s evidence alone, it was beyond peradventure that the deceased may have died of an unknown cause. He concluded that there was no abnormality or fatal pathology that could have satisfactorily accounted for the cause of death. There was no internal or external trauma or injury to the deceased. Dr Chiao corrected Dr Chan’s opinion as to the cause of death but was not given that opportunity in respect of Dr Mainland’s opinion. Dr Mainland essentially postulated that the deceased died of hypoxia, or the lack of oxygen supply. Dr Chiao testified that he found no signs of hypoxic change, but also explained that it would take several hours of being deprived of oxygen before such signs would be shown. In expressing her opinion, Dr Mainland stated that she could not establish the cause of death with certainty, and she noted in both her report and her evidence certain assumptions and qualifications underlying her opinion.

148.  Whilst the judge gave a comprehensive and structured set of directions to the jury, no direction was given that they must consider the possibility of an unknown cause, particularly in the light of the forensic pathologist’s evidence, and that they should only convict if they had excluded that possibility.[288]

149.  In expressing her opinion, Dr Mainland stated that she could not establish the cause of death with certainty, and she noted in both her report and evidence certain assumptions and qualifications underlying her opinion. In these circumstances, and in the absence of any clear direction requiring the jury to exclude the realistic possibility of an unknown cause of death before convicting, it cannot be said that the jury’s verdict rested on a sound and logically justifiable basis.

Conclusion

150.  For these reasons, taken cumulatively, namely the deficiencies in the medical expert evidence and its presentation, the restriction on the use of the references to and the influence of Dr Critchley’s report in relation to the medical expert evidence, the confusion over the prosecution case on whether causation was based on any one or all of the alleged breaches of duty, the absence of a Kevin Brown direction, and the failure to direct the jury on the realistic possibility of an unknown cause of death, we allowed the appeal against conviction, quashed the conviction and set aside the sentence. The respondent made no application for a retrial of the appellant.

Regulation of cosmetic surgical and medical services

151.  Before leaving this case, we make some brief observations on the regulatory framework for cosmetic surgical and medical services, prompted by the circumstances revealed in the present appeal.

152.  The Private Healthcare Facilities Ordinance (Cap 633) was enacted in November 2018 to regulate private healthcare facilities in Hong Kong. In January 2020, the Director of Health, pursuant to section 102 of the Ordinance, issued the Code of Practice for Day Procedure Centres (the Code) to provide standards for all day procedure centres licensed under the statutory regime. The latest edition of the Code was issued in September 2024. It sets out the licensing standards in respect of the governance, staffing, facilities and equipment, service delivery, quality and safety of care, infection control, and other matters related to the operation of day procedure centres providing specialised services set out in Schedule 3 to the Ordinance. Schedule 3 lists the classes of specialised services and their corresponding “scheduled medical procedures” and any exceptions to those procedures. It would appear that some invasive cosmetic procedures may not fall within the ambit of Schedule 3.

153.  A series of cases has highlighted instances in which individuals have suffered serious injury or death following invasive cosmetic treatments, often associated with insufficient regulation or inadequate safety requirements. Safeguards for such treatments, particularly when they are provided outside established healthcare settings, are crucial to ensure patient safety and well-being. Consideration should be given to the implementation of specific measures to maintain high standards of care and safety and to protect individuals from potential harm in addition to the safeguards already contained in the Code. These measures could include: (1) a regulatory regime establishing requirements and standards, including enforceable protocols for the performance of invasive cosmetic surgery and other medical procedures, which specifies which invasive cosmetic and medical procedures may only be performed by registered medical practitioners in licensed facilities; and (2) the mandatory registration and accreditation of all establishments and personnel providing invasive cosmetic treatments or medical services to the general public, supported by a clear system of licensing, inspections and sanctions for non-compliance (as discussed in cases such as Chow Heung Wing Stephen).

154.  The liposuction procedure carries significant risks of serious and potentially fatal complications, many of which are preventable through strict adherence to proper safety protocols. Given the frequency with which the liposuction procedure is performed, appropriate regulatory guidance should require that it be carried out only by qualified professionals in licensed facilities, under clear protocols for patient safety, pre-operative risk assessment and post-operative monitoring.