BNM (administratrix of the estate of B, deceased) on her own behalf and on behalf of others v National University of Singapore and others and another appeal
[2014] SGCA 49

Case Number	:	Civil Appeals Nos 21 and 22 of 2014
Decision Date	:	26 September 2014
Tribunal/Court	:	Court of Appeal
Coram	:	Sundaresh Menon CJ; Chao Hick Tin JA; Steven Chong J
Counsel Name(s)	:	Beh Eng Siew and Suja Sasidharan (Lee Bon Leong & Co) for the appellant in CA 21/2014 and the respondent in CA 22/2014; Anparasan s/o Kamachi, Tan Hui Ying Grace and Audrey Wong (KhattarWong LLP) for the first respondent in CA 21/2014 and the appellant in CA 22/2014; Allagarsamy s/o Palaniyappan (Allagarsamy & Co) for the second respondent in CA 21/2014; Michael Eu Hai Meng and Francis Chan (United Legal Alliance LLC) for the third respondent in CA 21/2014.
Parties	:	BNM (administratrix of the estate of B, deceased) on her own behalf and on behalf of others — National University of Singapore and others

26 September 2014

Judgment reserved.

Steven Chong J (delivering the judgment of the court):

Introduction

1       A weekly routine swim in the swimming pool at the Sports and Recreation Centre of the National University of Singapore (“NUS”) led to the tragic and unfortunate drowning of the Deceased. A negligence claim was brought by the estate of the Deceased against NUS and Hydro Aquatic Swimming School (“Hydro”), the company engaged by NUS to supply lifeguards and provide maintenance services for the pool. The Judge below found that Hydro was not an independent contractor vis-à-vis NUS and that the lifeguards were negligent in the discharge of their duties. However the claim was nonetheless dismissed on the basis that causation had not been proved because the medical evidence showed that the Deceased would not have survived even if the lifeguards had attended to him earlier. NUS cross-appealed against the finding of negligence while Hydro did not join in the cross-appeal.

2       The pivotal issue in these two appeals relates to the question of causation. Although there is no dispute that NUS owed a duty of care to the Deceased, the extent and content of that duty would ultimately determine the causation issue.

Background facts

Parties to the dispute

3       The appellant in CA 21/2014 and the respondent in CA 22/2014 is the Deceased’s widow and the administratrix of his estate (“the Appellant”).

4       NUS is the 1st respondent in CA 21/2014 and the appellant in CA 22/2014. It is the owner of the swimming pool in which the Deceased drowned.

5       The 2nd respondent in CA 21/2014, Hydro, was the contractor appointed by NUS to supply lifeguards and cleaning services in respect of the swimming pool at the time the Deceased drowned.

6       The 3rd respondent in CA 21/2014, The Overseas Assurance Corporation Limited (“OAC”), is an insurance company. It was Hydro’s public liability insurer at the material time.

Background to the dispute

7       On 6 June 2007 at about 12.49pm, the Deceased went swimming with a friend, Er Chee Teck (“Er”), at the Olympic-sized swimming pool at NUS. This was their regular weekly swim, a routine they started in January 2007. The pool had nine lanes, with Lane 1 being the closest to the entrance and Lane 9 being the furthest. After taking about five to ten minutes to change and warm up, the Deceased and Er began swimming in Lane 9. At that time, there were only ten to 15 swimmers in the pool. However, apart from Er and the Deceased, there were no other swimmers in Lane 9.

8       While Er was swimming his seventh lap, he saw the Deceased struggling about 20 metres away from him. The Deceased was submerged in the water and waving his hands, his legs were touching the floor of the pool which was about 1.8 metres deep at that point and he appeared to be trying unsuccessfully to get to the surface. Er swam as fast as he could to the Deceased. He reached the Deceased, grabbed him under the arms, pulled him up to the surface and then towed him to the side of the pool, which was about a metre away.

9       When Er reached the side of the pool with the Deceased, he shouted for help. At the time, the two lifeguards on duty (“the lifeguards”), Cheong Juan Meng (“Cheong”) and Chua Li Qi (“Chua”), were seated at the other side of the pool near the entrance. Cheong ran down the gallery steps towards the Deceased and Er while Chua remained where she was. Upon seeing Cheong waving his arm at her as he ran towards the Deceased and Er and hearing his shout at her, she called for an ambulance and then went to the lifeguards’ office to get the safety equipment. This was approximately 1.10pm.

10     Cheong helped Er pull the Deceased out of the water and performed cardiopulmonary resuscitation (“CPR”) on him. As for Chua, she had found an Oxyviva resuscitation machine (“Oxyviva”) but could not find the automated external defibrillator (“AED”). She brought the Oxyviva to the poolside and went back to search for the AED, which she eventually located and brought to the scene. However, neither Cheong nor Chua could use the AED or the Oxyviva on the Deceased as they were not trained in the use of the equipment.

11     A facilities officer employed by NUS, Sim Lye Hock (“Sim”), arrived shortly after. On his instructions, the Deceased was carried away from the swimming pool edge to a dry spot and Sim administered a shock on him using the AED. By then, the Judge estimated that at least 16 minutes would have elapsed since the onset of difficulties.

12     Dr Tan Tong Nam, a doctor at the University Health Centre, then arrived at the scene at around 1.20pm. He took over the CPR and noticed that the Deceased had no pulse. Dr Tan also tried to deliver a shock to the Deceased but the AED indicated that the Deceased could not be administered another shock.

13     At 1.25pm, the ambulance arrived. The paramedic tried using a defibrillator on the Deceased to no effect: the Deceased had by then gone into asystole which meant there was no longer any heartbeat. The Deceased was rushed to the National University Hospital and reached there at 1.39pm. The doctor on duty noted that on arrival, the Deceased was unresponsive, without pulse and was not breathing. His vital signs could not be recorded. An electrocardiogram monitor showed he was still in asystole. At 2.27pm, the Deceased was pronounced dead. The autopsy report recorded the cause of death as consistent with drowning with ischaemic heart disease. The Deceased left behind his wife and two young children aged three and six.

Decision below

14     The Judge held that NUS owed a duty of care to provide properly trained lifeguards. The lifeguards would be expected, inter alia, to survey the pool regularly, to remain alert and be trained to spot swimmers in difficulty. He further found that the duty did not extend to the provision of lifeguards trained in the use of Oxyviva and AED. Although this duty was delegable, the Judge found that Hydro was not an independent contractor because NUS had retained a high degree of control over the manner in which Hydro was to carry out its work. Thus, if the Appellant had succeeded in her claim, NUS would have to share a third of the liability with Hydro bearing two-thirds.

15     On the facts, the Judge found that although the lifeguards had achieved the requisite certification, they were negligent in the performance of their duties as they were not surveying the pool as they should have been doing and only became aware of the Deceased’s difficulties in the water when they heard Er’s shout for help. However, the Judge ultimately dismissed the claim on the basis that the negligence was not causative of the loss. He was of the view that the Deceased had suffered cardiac arrhythmia which incapacitated him whilst he was swimming, and given his underlying heart condition, the lifeguards would not have been able to save him even if they had intervened earlier.

Our decision

Was Hydro an independent contractor and if so to what end?

16     In light of the Judge’s finding that the lifeguards were negligent in the discharge of their duties, it was essential to determine the party who should be held responsible for their negligence – NUS or Hydro or both. As the lifeguards were not parties to the action, liability would arise if either NUS or Hydro or both were found to be vicariously liable for the negligence of the lifeguards. Typically, a defence to a claim in vicarious liability can take the form of denying that the servant or agent was negligent and/or asserting that the servant or agent was an independent contractor whose negligence he is not responsible for.

17     It is apposite to begin our analysis by first examining the basis of the negligence claim brought by the Appellant and how Hydro’s status as an independent contractor featured in the case. Initially, the claim was brought only against NUS. NUS’s defence was simply that it had delegated the management and maintenance of the pool to Hydro as an independent contractor. After NUS added Hydro as a third party seeking from it an indemnity and/or contribution, the Appellant added Hydro as the 2^nd defendant. The claim against NUS and Hydro after two rounds of amendments was premised on the negligence of their servants and/or agents, in particular that both NUS and Hydro were negligent in failing to ensure that the lifeguards were adequately trained in the use of AED and/or Oxyviva. In addition, the Appellant separately alleged that NUS was vicariously liable for the negligence of Hydro, its agents and/or servants and was further negligent in appointing Hydro knowing that Hydro was not able to provide qualified lifeguards trained in the use of AED. The Judge found that NUS was not negligent in appointing Hydro in light of his finding that there was no duty to provide lifeguards trained in the use of AED and/or Oxyviva.

18     Hydro’s status as an independent contractor was examined by the Judge only in relation to the Appellant’s averment that NUS was negligent in supervising Hydro “to ensure that properly trained lifeguards were stationed at the swimming pool”. In this connection, it should be borne in mind that this averment was pleaded as particulars under the alternative negligence claim against NUS for appointing Hydro for the provision of lifeguard services. Thus, strictly speaking, by the Judge’s own approach, the inquiry whether Hydro was an independent contractor would only arise if NUS was found to have been negligent in appointing Hydro, which he found otherwise. Further as the Judge observed, “NUS’s defence is that if Hydro Aquatic was a properly appointed independent contractor, it followed that its employer NUS would not be vicariously liable for Hydro Aquatic’s negligence, if any” [emphasis added]. However it is essential to bear in mind that the negligence as found by the Judge was that of the lifeguards in the discharge of their duties and not Hydro. There is no question of NUS being vicariously liable for the vicarious liability of Hydro in respect of the negligence of its lifeguards. That being the case, it is strictly no longer relevant or necessary to decide whether Hydro (as distinct from the lifeguards) was an independent contractor vis-à-vis NUS.

19     We should add NUS did not appeal against the Judge’s determination that Hydro was not an independent contractor. However, on our review of the Judge’s approach and finding on this point and in particular the law in this area, we came to a different view though we should make it clear that it did not change the eventual outcome of the Appellant’s appeal. If it did, we would have invited further submissions from the parties.

20     Before explaining why we have arrived at a different view, we should make a salient observation as to the proper approach in analysing the vicarious liability of the lifeguards’ negligence. In our view, for the purposes of attributing vicarious liability to either Hydro or NUS or even both in respect of the lifeguards’ negligence, the correct inquiry should be to evaluate the question of control which NUS had exercised over the lifeguards and not over Hydro. This is clear from Mersey Docks and Harbour Board v Coggins and Griffith [1947] AC 1 (“Mersey Docks”), a decision which was cited by NUS in its closing submissions. In that case, the harbour authority had hired out a mobile crane and a craneman to a firm of stevedores for loading a ship. The craneman was employed, paid and liable to be dismissed by the harbour authority, though the general hiring conditions stipulated that cranemen so provided should be the servants of the hirers. The craneman negligently injured someone while driving the crane and the issue arose as to whether the harbour authority or the stevedores should be vicariously liable for this tort. The House of Lords determined the issue by examining whether control over the negligent craneman had remained with the harbour authority as his employer or had been transferred to the stevedores (at 12–13):

My Lords, the only question for your Lordships' determination is whether on the principle of respondeat superior, the responsibility for the negligence of the driver of the crane lies with the stevedores or with the appellant board, whom the plaintiff sued alternatively. The answer depends on whether the driver was acting as the servant of the stevedores or as the servant of the appellant board when he set the crane in motion. That the crane driver was in general the servant of the appellant board is indisputable. The appellant board engaged him, paid him prescribed the jobs he should undertake and alone could dismiss him. The letting out of cranes on hire to stevedores for the purpose of loading and unloading vessels is a regular branch of the appellant board's business. In printed regulations and rates issued by the appellant board the cranes are described as "available for general use on the dock estate at Liverpool and Birkenhead" and as regards portable cranes the stipulated rates vary according as they are provided "with board's driver" or "without board's driver." Prima facie therefore it was as the servant of the appellant board that Newall was driving the crane when it struck the plaintiff. But it is always open to an employer to show, if he can, that he has for a particular purpose or on a particular occasion temporarily transferred the services of one of his general servants to another party so as to constitute him pro hac vice the servant of that other party with consequent liability for his negligent acts. The burden is on the general employer to establish that such a transference has been effected. … [emphasis added]

21     Here, there is no dispute that the lifeguards were employed by Hydro. Prima facie, Hydro would be vicariously liable for the negligence of its own employees, the lifeguards, and the proper approach should have been to consider whether that presumption has been displaced. We therefore disagree with the Judge’s approach of focusing on NUS’s control over Hydro. Having said that, this was how the issue was addressed by the parties in the Court below, and we go on to consider whether the Judge’s finding was correct.

22     The Judge below found that Hydro was not an independent contractor because NUS had retained a significant degree of control over the manner in which Hydro was to carry out its work. In so doing, he adopted the control test as explained in Performing Right Society, Limited v Mitchell and Booker (Palais de Danse), Limited [1924] 1 KB 762 at 767 that “certainly the test to be generally applied, lies in the nature and degree of detailed control over the person alleged to be a servant”. It would appear that the test cited by the Judge directed the inquiry to examine the question of control over the person alleged to be the negligent servant and not the company who had employed the servant though in the application of the test, he directed his mind to the question of control over Hydro instead.

23     The Judge concluded from his review of the tender specifications that NUS’s relationship with Hydro “was not one in which NUS had fully delegated all aspects of pool safety to an independent contractor” [emphasis added]. He examined a number of provisions of the tender specifications. It is important to bear in mind that for the purposes of determining whether those provisions have any bearing on Hydro’s status as an independent contractor, the reference point of the provisions should be their impact or effect on matters relating to pool safety. The provisions examined were:

(a)     Hydro was to carry out the works in accordance with the tender specifications, Conditions of Contract and to the satisfaction of the Facility Officer – clause 1.2.

(b)     The lifeguards were required to attend twice-weekly briefings – clause 4.2.

(c)     The Facility Officer was permitted to request for the deployment of extra lifeguards at two weeks’ notice – clause 4.3.

(d)     In the absence of the Facility Officer or his assistants, the appointed lifeguard in charge was to be fully responsible for the proper supervision of the pool – clause 6.1(ix).

(e)     The lifeguards were required to do any additional duties assigned by the Facility Officer as required – clause 6.1(xiii).

(f)     The lifeguards had to position themselves as instructed by the Facility Officer or his assistants – clause 7.3.

(g)     The lifeguards were to take instructions only from certain persons: the senior manager, managers and sports officers of the Sports and Recreation Centre, and the Facility Officer and his assistants, duty officers, and “any other personnel as authorised by the Facility Officer from time to time” – clause 7.13.

(h)     Hydro was contractually liable to pay liquidated damages set at $50 for each case of “failure to carry out instructions given by staff from NUS” and $100 for each instance of failure to be alert while on duty – clause 15.1.

24     In our view, none of these clauses with the possible exception of clause 7.3 dealt directly with pool safety matters. They also did not relate to NUS’s control over how the lifeguards were to perform their duties. Some, for example the assignment of additional duties under clause 6.1(xiii), show at best that NUS could tell the lifeguards what to do. Others, such as clauses 1.2, 4.2, 4.3 and 7.13, are largely administrative in nature. In fact, clause 6.1(ix) suggests that the default position is that the responsibility for the proper supervision of the pool rests with the Facility Officer of NUS or his assistants. The distinction between the ability to tell a worker what to do and the ability to tell him how to do it is significant: in Mersey Docks, the court held that the employment of the craneman was not transferred from the harbour authority to the stevedores because although the stevedores could tell the craneman what they wanted him to do, they had no authority to tell him how he was to handle the crane in doing his work (at 13).

25     The only provision that has some bearing on pool safety is clause 7.3 which stipulated that the lifeguards had to position themselves “as instructed by the Facility Officer and his assistants”. In our view, some measure of control over the manner in which the lifeguards are to position themselves in the discharge of their duties is not inconsistent with Hydro’s status as an independent contractor, especially when viewed with the other relevant factors as set out in [31] to [32] below. In any event, regardless of the contractual position, the evidence led at the trial showed that NUS’s instructions on the positioning of the lifeguards were communicated to Hydro and not to the lifeguards directly, and Hydro did not always agree with or follow those instructions.

26     The Judge also found it significant that the tender specifications provided for liquidated damages for breach of contract at $50 for each case of “failure to carry out instructions given by staff from NUS” and $100 for each instance of failure to be alert while on duty. In our view, such a clause was intended to provide some measure of financial deterrence for non-compliance and it could not have the effect of displacing Hydro’s status as an independent contractor. We would have thought that such provisions would be more consistent with the status of an independent contractor.

27     Finally, the Judge appeared to have attached weight to the fact that after the drowning incident, NUS held a meeting with Hydro on 19 September 2007 to review the safety conditions at the pool and to communicate the findings of the review which included the direction to station one lifeguard at either the mid-point opposite the pool entrance or near the deep end to better observe the swimmers in the pool during peak hours and to patrol the opposite side whenever the pool was crowded. A review of its safety rules by NUS to introduce improvements does not, in our view, change the independent contractor status of Hydro if it was in fact an independent contractor prior to the review. NUS was simply acting responsibly to prevent the reoccurrence of such incidents in the future.

28     Developments in this area of the law however indicate that the control test is not the only test for determining whether a contractor is an independent contractor, nor is it even necessarily the decisive factor. As Roskill J noted in Argent v Minister of Social Security and Another [1968] 1 WLR 1749 at 1758–1759:

If one studies the cases to which I have been referred, a number of tests have been propounded over the years for resolving the problem which I have to solve. For example, in the earlier cases it seems to have been suggested that the most important test, if not the all-important test, was the extent of the control exercised by the employer over the servant. If one goes back to come of the cases in the first decade of this century, one sees that that was regarded almost as the conclusive test. But it is also clear that as one watches the development of the law in the first 60 years of this century and [particularly] the development of the law in the last 15 or 20 years in this field, the emphasis has shifted and no longer rests so strongly upon the question of control. Control is obviously an important factor. In some cases it may still be the decisive factor, but it is wrong to say that in every case it is the decisive factor. It is now, as I venture to think, no more than a factor, albeit a very important one. … [emphasis added]

29     The flaws of the control test were in fact highlighted by Cooke J in Market Investigations Ltd v Minister of Social Security [1969] 2 QB 173. He concluded that the fundamental test to be applied is whether the contractor was performing services as a person of business on his own account (at 183–185):

I think it is fair to say that there was at one time a school of thought according to which the extent and degree of the control which B was entitled to exercise over A in the performance of the work would be a decisive factor. However, it has for long been apparent that an analysis of the extent and degree of such control is not in itself decisive. Thus in Collins v. Hertfordshire County Council [1947] K.B. 598, it had been suggested that the distinguishing feature of a contract of service is that the master cannot only order or require what is to be done but also how it shall be done. The inadequacy of this test was pointed out by Somervell L.J. in Cassidy v. Ministry of Health [1951] 2 K.B. 343, 352, where he referred to the case of a certified master of a ship. The master may be employed by the owners under what is clearly a contract of service, and yet the owners have no power to tell him how to navigate his ship. As Lord Parker C.J. pointed out in Morren v. Swinton and Pendlebury Borough Council [1965] 1 W.L.R. 576, 582, when one is dealing with a professional man, or a man of some particular skill and experience, there can be no question of an employer telling him how to do work; therefore the absence of control and direction in that sense can be of little, if any, use as a test.

…

The observations of Lord Wright, of Denning L.J. and of the judges of the Supreme Court suggest that the fundamental test to be applied is this: "Is the person who has engaged himself to perform these services performing them as a person in business on his own account?" If the answer to that question is "yes," then the contract is a contract for services. If the answer is "no," then the contract is a contract of service. No exhaustive list has been compiled and perhaps no exhaustive list can be compiled of the considerations which are relevant in determining that question, nor can strict rules be laid down as to the relative weight which the various considerations should carry in particular cases. The most that can be said is that control will no doubt always have to be considered, although it can no longer be regarded as the sole determining factor; and that factors which may be of importance are such matters as whether the man performing the services provides his own equipment, whether he hires his own helpers, what degree of financial risk he takes, what degree of responsibility for investment and management he has, and whether and how far he has an opportunity of profiting from sound management in the performance of his task.

The application of the general test may be easier in a case where the person who engages himself to perform the services does so in the course of an already established business of his own; but this factor is not decisive, and a person who engages himself to perform services for another may well be an independent contractor even though he has not entered into the contract in the course of an existing business carried on by him.

[emphasis added]

30     Cooke J’s formulation of the test was endorsed by the Privy Council in Lee Ting Sang v Chung Chi-Keung and Another [1990] 2 AC 374 at 382. This decision was in turn applied locally in Kureoka Enterprise Pte Ltd v Central Provident Fund Board [1992] SGHC 113.

31     Unfortunately, none of these authorities were cited to the Judge for his consideration. He appeared to apply the control test exclusively with reference to the tender specifications without regard to other factors which indicated that the contract between NUS and Hydro was that of a contract for services. We agree with NUS that the following factors were more consistent with Hydro providing a contract for services:

(a)     Hydro was responsible for paying its lifeguards’ wages;

(b)     Hydro was responsible for scheduling its lifeguards’ duty roster;

(c)     Hydro alone had the power to dismiss its lifeguards from employment;

(d)     Hydro was directly responsible for supervising and overseeing its lifeguards’ performance of their duties;

(e)     Hydro was responsible for training its lifeguards in the use of AED, Oxyviva and other resuscitation equipment;

(f)     Hydro had the power and authority to dictate how the lifeguards discharged their duties; and

(g)     Hydro was responsible for providing the resuscitation equipment.

32     Other factors which were not specifically mentioned by NUS but in our view are equally material include the fact that Hydro undertook the financial risks of running its business, owned its own assets and personnel, selected the lifeguards for deployment at the NUS pool, provided lifeguarding services to other institutions apart from NUS, retained the profits from its business and also took out its own public liability insurance. These additional factors point to the irresistible conclusion that Hydro was indeed carrying on a business on its own account and was therefore an independent contractor and not a servant of NUS.

33     For the same reasons, had it been necessary for us to decide the question of vicarious liability, we would have found that Hydro had control over the lifeguards and not NUS. As such NUS would not have been liable even if causation were to be established.

Were the lifeguards negligent in carrying out their duties?

34     In the Court below, NUS accepted, quite correctly, that they owed a duty of care to visitors like the Deceased to provide qualified lifeguards of reasonable competence according to the prevailing standards at the material time. Despite the concession, the Judge nonetheless examined the law in this area and found that NUS as the owners and/or operators of a large Olympic-sized swimming pool accessible to its members and their guests indeed owed a duty to provide an adequate system of safety including the provision of properly trained lifeguards stationed at appropriate locations around the pool. He held that NUS should be treated no differently from those who run large public swimming pools.

35     However, the Judge found that at the time the incident took place, ie, 6 June 2007, the standard of care did not extend to providing AED and/or oxygen resuscitator and lifeguards trained in the use of such equipment. The significance of this finding will be further elaborated when the causation issue is examined below.

36     The operative negligence as found by the Judge was the failure of the lifeguards to effectively scan the swimming pool with the consequence that they failed to notice that the Deceased had stopped swimming and then sunk to the bottom feet first, and that Er had swum to the Deceased and then struggled with the Deceased to the side of the pool. The lifeguards only realised that something was amiss when they heard Er’s shout for help after Er managed to pull the Deceased to the side of the pool. By this time, the Judge found that about three minutes had passed from the time when the Deceased first got into difficulties to the time Er was able to pull him to the edge of the pool and shout for help.

The CCTV recordings

37     The Judge’s finding that three minutes had elapsed was based on a reconstruction of the crucial events of the day. He observed that the various times given were not synchronised and therefore their accuracy could not be taken for granted.

38     Much time was expended in the reconstruction of the events leading to Er’s shout for help. Such reconstruction where possible should be based on available objective evidence. CCTV recordings (which are not uncommon in such facilities) would certainly be extremely helpful in any reconstruction exercise. Unfortunately this issue was not properly dealt with at the pre-trial stage. NUS omitted to mention the existence of the CCTV recordings in its affidavit verifying its list of documents. That such CCTV recordings were available was made manifest in Cheong’s statement to the police dated 7 March 2008:

Q1:Did you notice when the casualty came to the swimming pool?

A1:I only know that he came to the swimming pool at about 1249hrs. This is based on the CCTV recording which is targeted at the entrance of the swimming pool. I was sitting on the left side at that time thus I did not saw [sic] them coming in.

[emphasis added]

39     The CCTV recordings were not the subject of any specific discovery application by the Appellant though for the appeal, much emphasis was placed by her on the fact that NUS had failed to disclose the CCTV recordings.

40     It was unfortunate that the issue of the CCTV recordings was only raised in the course of the cross-examination of Cheong in the Court below. Cheong was asked when he viewed the CCTV recording that was mentioned in his police statement. Cheong gave an unhelpful answer that it was “after the incident”. The Appellant’s counsel did not press him for a more specific time though it would have been some time prior to 7 March 2008, the date of the police statement.

41     However the utility of the CCTV recordings, for the purposes of reconstructing the critical events, would largely depend on the angle at which the CCTV cameras were installed. It emerged during the cross-examination of Sim that NUS had installed four to five CCTV cameras covering the swimming pool complex. When asked by the Judge, Sim testified that all the cameras were facing the public area where the people sat and that none of them faced the pool. His evidence was not challenged by the Appellant.

42     The CCTV recordings may well have provided the best evidence to assist the Court in the reconstruction of the critical events. It is regrettable that no evidence was properly adduced by NUS as to whether the relevant CCTV recordings were still available especially as it was not in dispute that the police had taken possession of the recordings. No effort was made by NUS to verify with the police whether the CCTV recordings were still in their possession. The issue was somewhat unsatisfactorily dealt with by way of NUS’s closing reply submissions which at best appear speculative:

15.    Be that as it may, NUS submits that the CCTV recording was no longer in its possession, custody and/or power by the time it was first notified of the Plaintiff’s intention to seek compensation for Roland’s demise. The Plaintiff had only sent a Letter of Demand to NUS notifying it of the same on 16 June 2009, i.e. more than 2 years after Roland’s demise. By then, the applicable retention period for the CCTV recording had elapsed and the relevant CCTV recording had been automatically overwritten.

16.    As pointed out by the Plaintiff’s counsel himself, it was also likely that the police do not have a copy of the CCTV recording since the Coroner’s Case Investigation Report as well as the Findings of Coroner’s Inquiry made no reference to any CCTV. Thus, NUS had no means of retrieving the CCTV recording.

[emphasis added]

43     Having said that, in light of Sim’s unchallenged evidence that the CCTV cameras were not facing the pool, the probative value of the CCTV recording had they been preserved and discovered would probably not have been helpful in the reconstruction exercise.

Failure to detect the Deceased’s difficulties until the shout for help

44     The Judge rejected the evidence of Chua and Cheong that they were scanning the swimming pool area some five to ten seconds before the incident. He found, and we accept, that the lifeguards were only alerted to the emergency when they heard Er’s shout for help. This finding is also in line with Cheong’s own statement to the police that “[w]hen I heard the shout, I saw that there were two male Chinese subjects inside the swimming pool”, which in itself was indicative that Cheong only became aware of the situation in the pool after Er’s shout. By then Er was already at the side of the pool with the Deceased. The Judge found that by the time the lifeguards were alerted to the emergency, about three minutes had already elapsed. In its appeal against the finding of negligence, NUS submitted that this estimate of three minutes by the Judge was erroneous and that it would have taken no more than one minute instead.

45     NUS’s estimate is premised on the following assumptions:

(a)     Er was able to notice the Deceased’s distress in the pool almost immediately after the onset of the difficulties;

(b)     Er took only ten seconds to swim twenty metres towards the Deceased; and

(c)     Er took only a few seconds to pull the Deceased to the side of the pool before shouting for help.

46     However, these assumptions are inconsistent with their own expert evidence adduced at the trial. The joint expert witness of NUS and Hydro, Professor Venkataraman Anantharaman (“Professor Anantharaman”), estimated that:

(a)     When a person begins to drown, it can take between twenty seconds and two to three minutes for him to reach the floor of the pool. When Er first noticed the Deceased’s difficulties, the latter’s legs were already touching the floor of the pool. Therefore, it is unrealistic for NUS to claim that Er was able to notice the Deceased’s distress immediately after its onset.

(b)     Er took twenty seconds to swim to the Deceased – and this was already a “good speed”.

(c)     Er took another fifteen seconds to stabilise the Deceased’s arms and call for help.

(d)     It took Cheong sixty seconds to react to Er’s shout and then run towards the end of the pool.

47     Further, it seems to us that NUS’s focus on the precise time before the lifeguards became aware of the emergency situation, ie, whether it would have taken three minutes or less, is misplaced as regards the finding of negligence. The emphasis on the three minutes appears to bear reference to the decision in Anne Teresa Hanlon or Gallacher v City of Glasgow District Council (1983) Inner House Cases 122 (“Anne Teresa Hanlon”). In that case, the deceased, for some unknown reason, got into difficulties while swimming. Shameen, one of three teenaged girls swimming in the vicinity, spotted the deceased and then went to one of the lifeguards on duty to inform her about the deceased’s difficulty. The lifeguard replied, “I’ll come in a minute” but continued to chat with some other girls. Shameen then returned to the pool and directed her sisters’ attention to the deceased’s difficulty and the three girls thereafter went to rescue the deceased. The lifeguard only noticed the situation when Shameen entered the pool together with her sisters to rescue the deceased. By then about three to four minutes had passed. The three girls brought the deceased to the surface and pulled him to the side of the pool. By the time the deceased was lifted out of the water, he was already dead. The court found that the lifeguard had failed to respond timeously to the emergency in that by the time the lifeguard had noticed the situation, it was already too late.

48     In our view, Anne Theresa Hanlon does not stand for the proposition that a lifeguard would invariably be found negligent if he took three minutes to notice a swimmer in difficulty. A finding of negligence would depend on a number of factors in determining whether the difficulties of a swimmer in the water should have been apparent to the lifeguard on duty had he performed his duties diligently. The view of the lifeguard would be particularly material in this assessment and that would in turn depend on a number of factors such as the size of the pool, the number of swimmers in the pool and the available lighting, ie, whether it was day or night. Other relevant considerations would include whether there were shouts for help by the swimmer in distress or other users of the pool, whether there was unusual splashing, whether the swimmer was struggling or motionless in the pool or whether there were other distractions which the lifeguard was required to attend to. There is no inflexible rule that the lifeguard is entitled to a certain amount of time to react before which an inference of negligence can be drawn. What is clear is that the longer it takes for a lifeguard to detect a swimmer in distress in the water, the more likely an inference of negligence would be drawn. The lapse of time before reacting is only one factor for consideration. For instance, a swimmer who suffers a severe cramp while swimming and shouts for help should attract the attention of the lifeguard almost immediately. We would be slow to lay down any specific response time for a reasonably competent lifeguard to appreciate the difficulties of a swimmer in the water particularly since no proper study has been carried out in Singapore. Besides, it is undesirable for this Court to lay down any rule that failure to detect the difficulties encountered by a swimmer beyond a given time, in this case three minutes, is in itself negligence.

49     In the present case, the Judge found that the lifeguards were negligent in the discharge of their duties because of their complete failure to notice a series of events as they unfolded before their very eyes, ie, that the Deceased had stopped swimming, that he had effectively sunk to the bottom feet first, that he was struggling to resurface, that Er had swum towards the Deceased and finally that Er was struggling to bring the Deceased to the side of the pool. These events occurred literally in broad daylight at about 1pm. At that time, the pool was not crowded. It was estimated that there were around ten to 15 swimmers in the pool. The lifeguards had an unobstructed view of the swimming pool. The Judge referred to photograph C-5 (adduced at the Coroner’s Inquiry) which showed that a swimmer under the surface was still visible from the gallery where the lifeguards were stationed. The lifeguards also agreed that there was nothing happening at the entrance of the pool which might have distracted their attention at that point in time. While we accept the dictum in S v Chipinge Rural Council 1988 (2) ZLR 275 (S) that lifeguards “cannot be expected to keep an eye on every [swimmer] every minute”, in this case, the Judge found and we agree that the lifeguards were negligent in carrying out their duties in scanning the swimming pool area not because they took three minutes to notice that the Deceased was in difficulty but rather because the lifeguards were completely oblivious of the situation as it unfolded until they were alerted by Er’s shout for help. In our view, this is sufficient to support a finding of negligence on the part of the lifeguards.

50     We pause to observe that the Judge’s estimate of three minutes from the time the Deceased first got into difficulties before Er managed to struggle with him to the edge of the pool and call for help was probably on the generous side. In arriving at his estimate of three minutes, the Judge must have adopted the higher end of the range given by Professor Anantharaman for the time it took Er to notice the Deceased’s distress (namely, two to three minutes). However, when the Deceased got into difficulties, he was on his seventh lap having crossed Er earlier. Er was in fact swimming towards the Deceased who was about 20 metres away when he spotted him in difficulty. In our view, with that information, it is more likely that about a minute would have passed from the time the Deceased first started experiencing difficulties before it was spotted by Er. It would have taken Er at least twenty seconds or more to reach him and some additional time to stabilise him, pull him to the surface and struggle with him to reach the edge of the pool before shouting for help. In all, a more realistic estimate was probably in the range of two minutes or slightly less. However as explained at [48] above, the precise time, in the circumstances of this case, did not affect the finding of negligence against the lifeguards.

Was NUS under a duty to provide lifeguards trained in the use of AED and Oxyviva?

51     Before dealing with the merits of this issue, we should first deal with and dispose of the pleading point raised by the Appellant. In short, the Appellant asserted that since NUS did not expressly deny that it had any duty to provide lifeguards trained in the use of AED and Oxyviva and further since NUS had expressly adopted the Appellant’s particulars of negligence against Hydro, it was not open to the Judge to find as he did that NUS had no such duty. In our view, there was simply no strict necessity for NUS to expressly deny the existence of such a duty. A denial of the Appellant’s pleading as was done in this case was sufficient to give rise to a joinder of issues on these matters. Further it was not inconsistent for NUS to adopt the particulars of negligence against Hydro as the third party in the event that it is found liable to the Appellant. This is in essence the nature of any third party claim.

52     The Judge very carefully and comprehensively surveyed the prevailing industry practice and standard of lifeguard services as of June 2007. He found that it was neither common nor industry practice for pool owners or operators to provide AED and Oxyviva or lifeguards trained in their use. He noted, among other things, that:

(a)     AED training was still not widespread in June 2007. The generally accepted certification for lifeguard qualification – the Bronze Medallion Award – did not require them to be trained in the use of AED.

(b)     The first edition of the Singapore Life Saving Society’s (“SLSS”) lifesaving manual published in 1992, which was the edition in use at the time of the Deceased’s death, did not require the use of AED by lifeguards.

(c)     The National Resuscitation Council’s (“NRC”) 2006 guidelines on basic cardiac life support advocated only the use of CPR, while the 2011 guidelines advocated the use of CPR combined with AED.

(d)     As at 2007, the SLSS had not issued guidelines which required lifeguards to be trained in the use of Oxyviva or even in the use of any kind of oxygen resuscitator.

(e)     The training syllabus for the Bronze Medallion Award did not require learners to know how to use any kind of oxygen resuscitator, let alone the Oxyviva machine.

53     In this appeal, the Appellant contends that the Judge had failed to take into account the following facts which underscored the importance of AEDs:

(a)     Dr Vivian Balakrishnan, the then Minister for Community Development, Youth and Sports, made a parliamentary speech in July 2005 when he said that “[b]y the end of August [2005], every swimming pool will have a defibrillator and staff would have been trained on how to use it” (Singapore Parliamentary Debates, Official Report (19 July 2005) vol 80 at col 923).

(b)     The NRC’s 2001 and 2006 Guidelines, which recommended the installation and use of AED in public areas including sports facilities.

(c)     The drastic reduction in the price of AED from $15,000 in 1989 to $2,000 in February 2006.

(d)     NUS’s tender specifications which required lifeguards to be trained in the use of AED.

(e)     Defibrillation training conducted by SLSS in January 2007 for its members through the Australian Royal Life Saving Society.

(f)     Chua’s testimony that she had personally watched AED demonstrations a few times during her employment at Wild Wild Wet from 2004–2005.

(g)     The fact that after the drowning incident, NUS had acquired some 44 sets of AED and recommended the placement of AED and Oxyviva at the lifeguard posts.

54     In our view, the above factors relied on by the Appellant were in fact properly considered by the Judge. They are not sufficient to establish that a reasonable pool operator in June 2007 would have provided lifeguards trained in the use of AED. It was also the Appellant’s own evidence through Alfred Chua (a senior manager with the SLSS) that notwithstanding the Minister’s speech, it was not common practice even for public pools operated by the Singapore Sports Council to have AEDs in 2007, some two years after the speech. We agree with the Judge’s finding that based on the industry practice at the material time, NUS was not under a duty to provide AED and Oxyviva at its swimming pool or lifeguards trained in their use at the time the Deceased drowned. At best the evidence indicates that while there was an emerging acceptance of the importance of AED circa 2007, it had not yet coalesced into a general practice. Of course, as the Judge correctly noted, it is possible for the prevailing industrial standards to fall below the standard of reasonableness required by law, and it might be argued that pool operators should have immediately provided AED training for their lifeguards the moment its benefits became known. However, in our view, pool operators are entitled to take guidance from the prevailing standards promulgated by authoritative bodies like the SLSS, unless those standards were manifestly inadequate.

55     Further, NUS’s tender specifications with Hydro do not assist the Appellant. Hydro’s obligations under contract to NUS cannot, in our view, constitute a basis for liability to a non-party like the Appellant.

56     Finally, the post-incident steps taken by NUS to acquire more sets of AEDs merely serve to show that NUS took responsible measures to enhance its safety standards in the operation of the pool. They do not, in our view, constitute any acknowledgment of a legal duty to do so at the time of the unfortunate drowning incident. In any event, although pool operators are encouraged to aim for higher standards, they should not be penalised for trying and failing to do so. It would, in the words of the Judge, “be against public policy to discourage parties from trying to achieve safety or other standards of care that exceed industry or acceptable standards of that time by penalising them if they fail to reach the same”. It would also result in an incongruous situation as rightly observed by the Judge where one pool operator who conformed to the prevailing industry standard would not be found liable while another pool operator who tried to achieve a higher standard but failed to do so would.

Was causation established?

57     Strictly speaking, in light of our finding that NUS was neither negligent in appointing Hydro nor vicariously liable for the negligence of the lifeguards, causation need not be considered as regards NUS. Causation should instead only be addressed as regards Hydro since we find that the control of the negligent lifeguards had remained with Hydro. However for completeness, we will address them together since the claim in negligence against both NUS and Hydro was mounted on, inter alia, the basis that (a) the lifeguards were negligent in failing to properly scan the pool to notice the Deceased’s distress prior to Er’s shout for help and (b) NUS and Hydro had failed to provide lifeguards trained in the use of AED and Oxyviva. However the Judge found that the operative negligence was limited only to the failure of the lifeguards to have spotted the Deceased’s difficulties in the pool before they heard the shout for help by Er.

58     Causation must therefore be examined with reference to the operative negligence. Having found that there was no duty to provide lifeguards trained in the use of AED, we are of the view that the Judge should not have examined the causation issue from the perspective whether an earlier application of the AED would have changed the outcome, though he rightly concluded that it would not. Even Dr Jimmy Lim, the Appellant’s expert, conceded that even if AED had been applied earlier, his chances of survival would have improved from “very poor to poor” and that his overall prognosis “would have been very bad”. Causation instead should be examined with reference to the question whether the Deceased would have survived if the lifeguards had spotted the Deceased’s difficulties earlier before hearing Er’s shout for help without the benefit of any application of the AED by the lifeguards. This leaves the causation issue to be examined with reference only to an earlier application of CPR.

59     There are two hurdles for the Appellant to overcome on the causation issue. First, whether earlier detection of the Deceased’s difficulties by the lifeguards would have made any difference given the fact that by the time the lifeguards were alerted to the situation, the Deceased had already been pulled out of the water by Er. Second, given the finding that the duty of care did not extend to the provision of lifeguards trained in the use of AED and/or Oxyviva, whether the Deceased would have survived with earlier CPR without the benefit of AED.

60     It is important to bear in mind that the three minutes estimated by the Judge included the time Er took to pull the Deceased to the surface and to the side of the pool. He also found that the Deceased was struggling and that Er took some effort to get him to the side of the pool. That being the case, as we have observed at [50] above, the Deceased would have been in the water for less than two minutes. This was unlike the situation in Anne Teresa Hanlon where three to four minutes had passed before the deceased was pulled out of the water by other swimmers in the pool after the lifeguard had ignored the plea for help by the other swimmers. By the time the deceased was pulled out of the water, he had already passed the “point of no return”. In the present case, when the Deceased was pulled to the side, he was still conscious with his eyes half open and trying to expel water.

61     Thus, even if the lifeguards had spotted the Deceased’s distress earlier, it would still have taken them some time (they were stationed at the opposite side of the pool closer to Lane 1) to reach the Deceased and some additional time to pull him to the surface and out of the pool. By then, the Deceased had already been pulled out of the pool by Er. The effect of the “delay”, if any, was that CPR administered by the lifeguards had commenced later than it should have been. Would that have made a material difference to the Deceased’s survival rate?

62     Both parties devoted much attention in the Court below as well as in the present appeals as to which medical expert’s evidence should be preferred. The experts held divergent views on a number of issues such as whether the Deceased should have been classified as either Class 1 or 2 under the New York Heart Association classification (which bears on the severity of the patient’s heart condition) or whether, given the Deceased’s heart condition, anything could have been done to save him and in particular whether an earlier application of AED would have made a difference. None of these points of contention has any bearing on the causation issue given our concurrence with the Judge’s finding that the duty of care, in the first place, did not extend to providing lifeguards trained in the use of AED.

63     Despite the opposing views of the medical experts on several issues, there was at least one significant point on which there was common ground – CPR alone would not have been sufficient to save the Deceased given his underlying heart condition. Dr Jimmy Lim, in his supplementary AEIC, emphasised the importance of early defibrillation to a person such as the Deceased, who may have reduced cardiac reserve due to the narrowing of his heart arteries. The time from the onset of arrhythmia to the ability to successfully restore normal rhythm was therefore even shorter. For a person with such a heart condition, Dr Lim stated that defibrillation by an AED should be done within the first three to five minutes of a sudden cardiac arrest collapse as survival rate decreases about seven to ten percent with every minute of delay in defibrillation; the survival rate drops drastically once the delay exceeds ten minutes. Dr Lim added that however good the CPR provided, it could only prolong ventricular fibrillation and maintain circulation; it could not convert ventricular fibrillation to a normal rhythm. Eventually the heart would still enter asystole.

64     It was acknowledged in the Appellant’s closing submissions in the Court below that all the four medical experts – Dr Jimmy Lim, Dr Dana Elliot, Professor Anantharaman, and Dr Michael Lim – agreed that CPR plus defibrillation within three to five minutes of collapse can produce a survival rate as high as 72% in situations outside hospital and that basic CPR alone without defibrillation would not have eliminated ventricular fibrillation and restored a perfusing rhythm. This was also candidly accepted by the Appellant’s counsel during the appeals before us. For this reason, it was critical for the Appellant to persuade this Court to reverse the Judge’s finding as regards the provision of AED and/or Oxyviva and lifeguards trained in the use of such equipment. However, we have already found against her on this pivotal point.

65     Finally, the Appellant also raised an issue about the adequacy of the CPR administered by Cheong. The Appellant’s complaint is that Cheong was the only one providing CPR for more than ten minutes and would have become fatigued after a while. First, this is purely speculative and does not address the evidence that the Judge had cited in coming to his conclusion that the CPR provided was of acceptable quality, in particular the chart log produced by the AED. Second, as shown above, CPR alone would not have saved the Deceased whatever the quality administered.

66     In the premises, causation cannot be established and consequently the claim against Hydro must likewise fail.

Conclusion

67     Although our findings on several issues differed from the Judge, we nonetheless agree with his eventual decision. In the result, both appeals are dismissed. As for costs, we make the following orders:

(a)     As the Appellant’s appeal in CA 21/2014 involved more issues than NUS’s appeal in CA 22/2014, we award NUS net costs fixed at $12,000, such costs order to be satisfied out of the Appellant’s security deposit to NUS.

(b)     The Appellant to pay costs fixed at $5,000 to Hydro and $3,000 to OAC, such costs orders also to be satisfied out of the Appellant’s security deposit.

(c)     NUS’s security deposit to be returned.