Netcare Hospitals And Lakeview Hospital

The judgment

The court's own file

competitiontribunal teatd afetre COMPETITION TRIBUNAL OF SOUTH AFRICA Case No:IM1930ctt7 In the matter between: NETCARE HOSPITALS (PTY) LTD Acquiring Firm and LAKEVIEW HOSPITAL Target Firm Panel ; Mondo Mazwai (Presiding Member) Andiswa Ndoni (Tribunal Member) {mraan Valodia (Tribunal Member) Heard on : 03-05 April 2018; 30 May 2018 Order issued on : 04 June 2018 Reasons issued on : 09 July 2018 REASONS FOR DECISION Approval and background

[1]On 04 June 2018, the Competition Tribunal (‘Tribunal’) overturned the Competition Commission's (“Commission”) prohibition of a small merger between Netcare Hospitals (Pty) Ltd ("Netcare") and the Lakeview Hospital (‘Lakeview’), approving the transaction subject to conditions tendered by the merger parties. The condition, attached hereto as “Annexure A” consists of a pricing remedy. The reasons for our decision are set out below. Background

[2]On 1 December 2016, Netcare acquired Lakeview. The transaction was a small merger in terms of the Competition Act 89 of 1998 (“the Act") and therefore not 1

[3]i4]

[5]

[6]notifiable. The Commission became aware of the merger during its investigations into the transaction between Netcare Hospital Graup (Pty) Ltd and the Akeso Group.’ On 25 May 2017 the Commission wrote to the merged parties, Invoking its authority in terms of s13(3) of the Act and instructed them to natify the transaction. The merger parties complied, notifying the transaction on 3 July 2017. On 22 September 2017 the Commission issued its recommendation and notice of prohibition, requiring Netcare to divest of Lakeview. The merger parties opposed the Commission's finding and filed a reconsideration application with the Tribunal on 11 October 2017. In its reconsideration application, the merger parties pursued two procedural points. It was argued first that the Commission filed its prohibition outside of the allotted statutory time and second that the CC16 form was materially defective. At a pre-hearing convened by the Tribunal on 6 November 2017, the merger parties indicated a willingness to hear the procedural points with the merits at a consolidated hearing and the matter was set down for a hearing. The procedural points were not pursued by the merger parties at the hearing and were abandoned. On the first day of hearing, the merger parties tendered a pricing condition. The Tribunal stood down for the morning to allow for the Commission to consider the condition. Upon resumption in the afternoon, the Commission’s legal team indicated that it would require more time to obtain instruction. To ensure that the matter was completed in the allotted days, the Tribunal proceeded to hear evidence that afternoon. On the second day of hearing, the Commission's legal team relayed that the Commission considered the tendered behavioral condition insufficient to address its concerns with the merger and persisted with its prohibition.? ' The Tribunal conditionally approved this transaction on 16 March 2018. ? Tribunal Transcript of Proceedings (Transcript) p108 lines 12-20. [7]The hearing was conducted on 03-05 April 2018, with closing argument taking place on 30 May 2018, The Tribunal thereafter approved the merger subject to the conditions proposed by the merger parties, Parties to the transaction and their activities Primary acquiring firms

[8]19] {10} The primary acquiring firm is Netcare, a company incorporated in accordance with the laws of the Republic of South Africa. Netcare is directly owned and controlled by Netcare Holdings (Pty) Lid, which is, in turn controlled by Netcare Limited, a public company listed on the Johannesburg Stock Exchange. Netcare Limited’s stocks are widely held and it is not directly or indirectly controlted by any single entity. As Netcare Limited, Netcare Holdings and Netcare act jointly, in these reasons they will hereinafter be collectively referred to as “Netcare”. Netcare operates a private hospital network in South Africa and the United Kingdom. In South Africa it is active in operating a primary care network and medical emergency services. In its hospital network, Netcare operates a number of general hospitals (also referred to as ‘acute hospitals’) throughout South Africa. At its general hospitals Netcare provides a full range of treatment options such as obstetrics, oncology, neurology, cardiology, gynaecology, orthopaedic, and general surgery. The treatment offering within each hospital is dependent on a number of factors and may thus differ from hospital to hospital within the network, Relevant to this merger is that Netcare owns and operates Netcare Linmed Hospital, a multidisciplinary facility licensed to operate 203 beds which is situated in Benoni, within the Ekurhuleni Municipality of Johannesburg, roughly 6 km away from Lakeview. Primary target firm

[11]

[12]The primary target firm is Lakeview, a private hospital licensed and operating in accordance with the laws of the Republic of South Africa. Lakeview was, prior to the transaction, owned by City Square Trading 945 (Pty) Ltd ("City Square”) with its operation licenses being held by Starchoice Trading Forty One Benmed Park Clinic (Pty) Ltd (“Starchoice"). Prior to the transaction Lakeview was privately owned. Lakeview is a 94 bed multidisciplinary private hospital situated in Mowbray Avenue, Benoni. It principally offers medical services relating to obstetrics, gynaecology, paediatrics, dentistry, general surgery, ear nose and throat, orthopaedics and dermatology. Prior to the transaction Lakeview was an independently owned hospital and was a member of the National Hospital Network ("NHN"). The role and function of NHN is discussed further below.? Where necessary in these reasons Lakeview will be referred ta as ‘NHN Lakeview’ to denote Lakeview prior to the transaction or ‘Netcare Lakeview’, denoting Lakeview post-transaction. Transaction and rationale

[14]In terms of the Sale Agreement entered into between Netcare, City Square and Starchoice, Netcare acquired full control over Lakeview on 1 December 2016. From that date Lakeview was no longer a member of the NHN.

[16]Lakeview submitted that its shareholders wished to dispose of their interests in the firms 3 See para [21]1 below. * Merger Parties ‘Competition Analysis’ trial bundle p49 para 15, 5 ibid para 16. COMPETITION ANALYSIS Relevant Market

[17]

[18]The merger parties and the Commission were in agreement that the market relevant to the merger was that for multidisciplinary private hospital services.® It was also agreed between the parties that although the market need not be geographically delineated, any assessment of the impact of this merger required an assessment at the national and local level.” In defining the local area, the Commission’s Expert, Dr Hariprasad Govinda (“Govinda”) accepted the merger parties’ radius of 7-10km from Lakeview as being the core catchment area, terming this the ‘Benoni area’® For context, we turn now to address pertinent elements of the relevant industry. Industry background

[19]

[20](21) Healthcare facilities Healthcare facilities provide healthcare services to consumers. Each healthcare facility operates under a license granted by the relevant provincial government department. This license regulates the number of beds which a facility may operate as well as the ‘type’ of bed offered |.e. the nature of the services offered to such beds, There are three large companies that operate private healthcare facilities in South Africa, namely Netcare, Life Healthcare, and Mediclinic. Those facilities not controlled by one of the ‘big three’, such as Lakeview prior to the merger, are independently owned and run. The independent healthcare facilities may, for the purpose of tariff negotiations (described more fully below), elect to become members of the National Hospital Network ("NHN"). The NHN is a non-profit company which represents the § Or H Govinda “An Economic Evaluation of the Netcare-Lakeview Merger: Expert Report” (Commission expert report) witness bundle p135 para 26, 7 Transcript p387 lines 1-4. ® Govinda Cammission expert report W140 para 44, ye

[22]

[24]collective interests of private independent healthcare facilities in South Africa, and represents these independent facilities in tariff negotiations with medical schemes. Healthcare funders Whilst the consumer of healthcare services is the patient, the funders of such services are primarily medical aid schemes, which are legal bodies registered in ierms of the Medical Schemes Act. Members of a medical aid scheme pay contributions to the scheme and in return receive medical cover according to the rules of the scheme and clinical best practice. Schemes offer a range of options for medical cover that differ in terms of the benefits accruing to and premiums payable by members. Medical aid administrators are bodies authorised by the Council of Medical Schemes to perforrn administrative functions on behalf of medical aid schemes. The role of an administrator varies depending on the size and nature of the scheme itself as well as the mandated agreement between the scheme and the administrator. Thus, in some instances trustees of a scheme may be involved in tariff negotiations and others not. In general, however, most administrators negotiate tariffs on behalf of the schemes with health care service providers and render a variety of administrative functions. Medscheme operates as one such medical aid administrator, providing assistarice to a number of smaller medical schemes such as POLMED, Bonitas and FEDhealth. Tariff codes In the healthcare industry in order to translate diagnoses of diseases and other health related problems from a description of services into an alphanumeric code, diagnostic coding standards are required. These standards are required fo fit into a larger procedure coding system. One of the most widely used standards is the International Statistical Classification of Diseases and Related ® Act 131 of 1998, as amended.

[25][26) (27)

[28]Health Problems (ICD-10) codes. I[CD-10 codes are uniformly applicable across private and public hospitals. In addition to ICD-10 codes, and more pertinent to the merger at hand, the South African private healthcare sector makes use of procedure coding systems which are comprised of Reference Price List (RPL) codes and Current Procedural Terminology (CPT) codes. RPL and CPT codes are akin to ICD-10 codes in that they translate the description of services rendered to a patient into alpha-numeric codes. CPT codes are conventionally grouped together under an RPL code for a more specific description of the services rendered to a patient. For example, to describe a situation in which a patient undergoes a procedure to remove portions of their rib and thereafter stays in a general thoracic ward for a day, the RPL Code 58002: General Ward: Thoracic and neurosurgical cases, per day and the CPT code CP32900 which delimits the removal of portions of a rib would be used, RPL's and CPT’s are important to the financing of medical health care because a medical aid scheme will only pay a provider of healthcare services for services rendered if that provider charges for a service rendered which corresponds to an RPL or CPT code (collectively referred to as ‘practice code’) to which a cost is attached, The cost attached to the particular code is the fariff charged, which results in RPL and CPT codes being referred to as tariff codes. Tariff negotiations Due to the multi-sided nature of the private healthcare industry, prices for services rendered to consumers are determined primarily by a negotiation between, on the one hand, owners of healthcare facilities or services and, on the other, medical aid schemes and medical aid administrators. Netcare negotiates tariffs with medical aid schemes on behalf of all the hospitals in its network. In the case of pre-merger Lakeview, the annual tariff percentage

[29]{SC}

[31]increase was determined through negotiations between medical aid administrators and the NHN. Negotiations between individual hospital networks and medical aid schemes occur annually, on a national level and revolve primarily around three components: 29.1. tariff increases for existing services; 29.2. tariff setting for new services; and 29.3, the creation of network arrangements. When speaking to the annual tariff increases for existing services, there was consensus from the witnesses of both the Commission and merger parties, that negotiations took place on a national level, with the previous year’s prices for the scheme in question being used as the starting point for the negotiations. Cost to medical schemes. Although tariffs form a large part of the total cost to medical aid schemes for the services provided at private hospitals, they do not comprise the only element. The Tribunal has previously grappled with determining the total costs incurred by medical schemes charged by hospital groups to medical aid schemes. In the Life Healthcare Group (Pty) Ltd / Joint Medical Holdings Limited’ merger the Tribunal held: (68] ... When hospitals bill funders for services there are three categories of expense. First, there is a tariff for hospital services; for example ward fees and operating theatre fees. Second, the cost of medicines used by the patient, The third category comprises the materials the hospital requires to perform its services for the patient,

[69]All hospitals pass on the same casts for medicines as these prices are now regulated by what is termed single exit pricing. Whilst this does not preclude a hospital or doctor from using a generic equivalent this is an area in which price variances between hospitals is constrained by regulation. How hospitals or doctors may manage the use of drugs is a different matter." 10 Life Healthcare Group (Pty) Ltd / Joint Medical Holdings Limited (2013] 1 CPLR 227 (CT) (“Life/JMHL”). 1 Ibid para 68-69.

[32](33)

[84]

[35]

[36]The merger parties in this matter introduce a further dimension to the determination, namely that of utilisation. On their version, utilisation is a function of patient acuity and the decisions of medical practitioners’? It was submitted that utilisation overlays all three of the above-listed categories." An example of the impact of utilisation would be the difference in cost to a medical aid scheme for a hip replacement conducted by a profligate surgeon on an octogenarian with a pre-existing illness (a ‘co-morbidity’) versus the cost of a hip replacement conducted by a thrifty surgeon on a young, healthy athlete. Whilst the CPT code on the procedure is the same, the total cost to the medical scheme for the first would most likely be greater than the second. This difference is then, on the merger parties’ version, denoted by utilisation. To control for the impact of utilisation and to provide a more accurate picture of the costs incurred by a medical aid scheme for similar procedures at different healthcare providers, certain medical aid schemes make use of the Diagnosis Related Grouper (‘DRG’) classification system.“ A DRG classification system consists of categories of patients that are similar both clinically and in terms of their consumption of haspital resources. The system groups patients with similar clinical characteristics per procedure into bespoke DRG’s.”© This grouping then relates the type of patient undergoing a particular procedure to the hospital cost for that procedure and allows for medical aid schemes to more accurately determine the average cost of like procedures,'” DRGs are assigned a particular code, these codes are assigned by and are applicable to a particular medical aid scheme. It was submitted that because the average costs assigned to DRG codes are a better representation of the total ” Transcript p58 line 9- p59 line 9, ‘3 Merger parties Heads of Argument para 80.2. 4 Discovery Health Medical Scheme Submissions to the Competition Commission 20 September 2017 trial bundle p17139, 18 Govinda Commission expert report witness bundle W162 para 179. 16 Discovery Health (note 14 above) p1136 para 2.2. 7 Ibid. (37) cost that a medical aid scheme incurs per like patient, a medical aid scheme would be able to use DRG's to better compare what comparable procedures cost across any number of healthcare providers when such heaithcare providers have differentiated tariffs (i.e. costs assigned to CPT and RPL codes). With this in mind, we now turn to addressing the Cammission’s theories of harm. Commission's theories of harm 138]

[39]In its reasons of 11 October 2017, the Commission broadly advanced three theories of harm. This first related to pricing. The Commission argued that the merger had and would likely result in tariff increases at Netcare Lakeview as compared to NHN Lakeview. The second was that the merger entrenched Netcare’s dominance which would result in an improved national bargaining position from which Netcare would be able to extract higher tariff increases. The third was that the merger resulted in the removal of an effective competitor. During the Hearing and in its heads of argument the Commission emphasized its first theory, namely that “Lakeview’s tariffs have increased post-merger and that Netcare will align the tariffs of Lakeview to match its own and that already Netcare tariffs are higher than Lakeview in general.”"" The Commission also pursued its second and third theories in the Hearing. Analysis

[40]Increase in negotiation power We shall deal with the Commission's second theory of harm first. The Commission’s theory ran that the merger had entrenched Neteare’s dominance in the relevant market given its higher post-merger market share and the geographic placement of its hospitals which would result in an improved national 18 Commission Heads of Argument p12 para 34. 10

[44]142] (43)

[44]bargaining position from which Netcare would be able to extract higher tariff increases. On the evidence submitted, Lakeview is a 94 bed facility, which represents 12% of the beds in the Benoni area.’® Lakeview is situated 1.8km away from Life Glynwood, a 323 bed facility controlled by Life. in addition, NHN’s Sunshine hospital is located 4km away.?° Govinda, advancing the Commission's theory of harm, submitted that nationally Netcare controls 10 679 private hospital beds, equating to a 30% share in the market for the provision of private multidisciplinary hospital care. Life follows with 22% and NHN with 20%.?' In the Benoni area, Govinda submitted that post- transaction Netcare controls 37% of the market, 12% of which was obtained through purchasing Lakeview.?4 The Commission argued this market share should be examined in the context of a market characterised by high regulatory and capital barriers to entry and a shortage of skills. t concluded that Netcare’s higher share eroded and continues to erode the countervailing power of medical aid schemes in the tariff negotiation process. This erosion, on the Commission’s version, allows Netcare to extract higher tariff increases at the national level.”! The merger parties opposed the theory on a number of levels. The first was to dispute the market share calculations, indicating that in a presentation made by the Hospital Association of South Africa, Netcare’s national market share was only listed as 27%, with NHN controlling 23%.*4 Further they argued that the pre- merger market shares in the Benoni area did not account for the fact that ‘8 Govinda (note 6 above) pW155 para 98. 28 Govinda (note 6 above) pW138 Table 1. 21 Govinda (note 6 above) pW153 Table 2. 22 Govinda (note 6 above) pW155 para 98. 23 Competition Commission Merger and Acquisitions’ Report pW39 para 78-81. 24 Merger Parties Replying note 30 May 2018 p11 para 21. 25 Mark Bishop of Net had not increased ov ited in his witness statement that the occupancy rate of NHN Lakeview In the two years prior to the merger, 11

[45]The merger parties argued that the merger had no effect on the competitive landscape. It was submitted that the transaction had the effect of transferring one relatively small hospital in the Benoni area from NHN to Netcare, doing no more than “shifting the deck chairs” to an immaterial extent between two major hospital groups, and thus the merger had no impact on national bargaining.

[46]Discovery Health Medical Scheme ("Discovery Health”) and the Government Employees Medical Scheme ("GEMS"), submitted to the Commission that the merger would have no effect on the national negotiations. The Commission's witnesses, Dr Noble-Luckhoff of Medscheme and Mr Marion of Bonitas also confirmed that the merger had little or no effect on Netcare’s or the NHN’s national bargaining position.2’

[47]We appreciate that there is a debate to be had on the principle of whether an increased number of beds controlled by a hospital group increases its bargaining position. We also appreciate that ancillary to that debate, is the question as to whether the geographic placement of any beds acquired plays a role in increasing a bargaining position. There is question of what the impact on national bargaining would be if a hospital group is dominant in a particular geographic area where a medical aid scheme has a large number of users, or even where a hospital group's geographic placement of hospitals across the country spans a number of areas over which a medical aid scheme’s members are positioned. However considering the testimony of the witnesses in the present case, we were not provided with sufficient evidence to conclude that the 2 In response to the question: In light of the notified merger, do you anticipate any changes in your company's relationship with Netcare Hospilals? / Discovery Health responded: “No. Lakeview Hospital is a small hospital... as such the Netcare acquisition has had no impact on the negotiation dynamics between the Scheme, Netcare or NH", Trial bundle p878. GEMS responded: “Such a minor increase is unlikely ta have a material impact on negotiating power, In Ekhurleni, there are several competing {sic] Ciinix, Life Healthcare and NHN hospitals. This further limit [sic] the impact on negotiating power." trial bundle 1208, 2? Transcript p21 lines 12-14 Dr Noble-Luckhoff states: “the size of Lakeview is so small that it really won't have any significant impact.” Transcript p303 line 21- p304 line 4 Mr Marion states: “the addition of an additional 94 [hospital beds] is not going to be a significant driver in the negotiation power, But | think what | wanted to highlight was just the increase in the number of facilities. It eeiniteke is not going to influence the negotiating power going forward.” 12

[48]

[49]

[50](51)

[52]94 beds at Lakeview or a 12% market share accretion in Benoni would trigger this debate. We therefore found that the Commission’s second theory of harm was unable to sustain a prohibition of the merger and we now turn to addressing the Commission's third theory, namely that the merger represents the removal of an effective competitor. Removal of an effective competitor In its report, the Commission advanced the theory that the merger had led to concentration in the Benoni area. It argued that of the three hospital groups present in the Benoni area being Netcare, Life and NHN, NHN’s prices were generally the lowest. Because the merger increased the market share of Netcare at the expense of NHN, the Commission advanced that the price constraint on Life and Netcare decreased and that the merger had therefore resulted in the removal of an effective competitor and would result in Netcare and Life being able to charge higher prices in the area. Govinda, in his oral evidence, supported the Commission's claim.”* He indicated that the higher concentration had contributed to the environment in which Netcare was able to increase prices. The merger parties’ expert, Mr Patrick Smith (“Smith”) argued that the Commission's theory is inapplicable to the industry in question. He argued that tariffs are set nationally and in terms of national negotiations and therefore price competition was not a local unilateral effect. Govinda responded with the fact that when hospital networks negotiate, they are keenly aware of the areas in which medical aid schemes would have members. Thus increasing their market share in strategic areas, even if the acquisitions were small, would exponentially increase the negotiation power of the Networks.” 8 Transcript p402 lines 10-12. *8 Transcript p404, 13

[53]

[54]

[55]

[56]

[57]Once again, whilst we acknowledge that there Is a debate to be had on the impact of the geographic placement of hospitals, we do nat find, on the facts and testimony before us, that there was sufficient evidence led before us to prove the strategic import of the Benoni region. We therefore found that Commission's third theory of harm was unable to sustain a prohibition and we turn now to address its primary theory, that relating to an increase in tariffs at Lakeview. Pricing Commission's submissions In its merger report the Cornmisslon advanced a theory of unilateral effects, It argued that Netcare's natlonal tariffs were, on average, higher than those of NHN Lakeview, It posited that Netcare would have the incentive and ability to align Lakeview's tariffs with its national tariff schedule resulting in an increase in the tariffs at Lakeview, This tariff increase would then lead to an increase in total cost to consumers at Netcare Lakeview as compared to NHN Lakeview. The Commission in arriving at this theory, conducted a tariff comparison analysis. It compared the tariffs for the top 13 procedures conducted at Lakeview between the Netcare national tariff and the tariff charged at NHN Lakeview,°° Govinda in his report and testimony concurred with the Commission's conclusion. He concluded that the merger had led to unilateral price effects, specifically, that Lakeview's tariffs had increased post-merger. Govinda came to this conclusion after examining whether Lakeview’s tariffs had increased post-merger on a comparable basis and whether NHN Lakeview was cheaper (referring to cost to medical schemes) than Netcare Lakeview on the whole. In so doing, he undertook two methods of tariff analysis. The first was based on RPL codes, and the second, unique to his investigation, was based on DRG codes. *° Commission report pwW27. 14

[58]159} {60} The RPL analysis seemed to indicate, when taken at face value, that there had been a tariff increase from NHN Lakeview to Netcare Lakeview. However, at the hearing Govinda largely abandoned any reliance on his RPL analysis indicating that a simple tariff comparison would be insufficient to prove whether there was a unilateral price increase.’ A position supported by Dr Noble-Luckhoff.5* Govinda instead relied on a DRG code analysis to advance the Commission’s primary theory of harm. The analysis consisted of comparing the average cost to medical aid scheme per DRG code at Netcare Linmed (the Netcare controlled hospital down the road from Lakeview), to the corresponding average cost to medical aid scheme per DRG code at NHN Lakeview. This comparison was based solely on DRG data provided by Discovery Health. Discovery Health, when submitting the data indicated “We caution against using this data for drawing conclusions on cost efficiencies between hospitals, and note that any conclusions reached on the basis of this data are likely to be statistically and actuarially invalid. This is mainly due to the small sample size in each DRG."34 Govinda examined the DRG codes from Linmed and NHN Lakeview and identified codes that overlapped. The data provided to the Commission was the total cost of each DRG to that hospital. To find the average cost per procedure in the DRG, Govinda divided the total cost incurred by the medical aid scheme for that particular DRG code by the number of cost events under that DRG code. These average costs were then compared as between the hospitals. *' Transcript p 375 lines 14-18: “But there are significant challenges to RPL, because one RPL could have a number of CPTs and similarly CPTs could have a number of RPL. And there are significant challenges there, So, the best way was to look al DRG submissions," ® Dr Noble Luckhoff indicates at Transcript P239: MR WILSON: And so when one is — what one certainly cannot do, and | don't suggest you try to do this, you can't simply take one tariff code, for example a surgical ward tariff, and compare it to another one and say, if! see a delta in that, that is reflective of the overall increase cost that |, as Medscheme, or my schemes that ! administer are going to face as a result of a transaction. That's simply one part of the puzzie. DR NOBLE-LUCKHOFF: Ja, Agree “ Trial bundle p1136. 15

[64](62) (63) {65} In comparing the average cost per DRG code, Govinda excluded any DRG code which had not been performed at least 12 times in a given year at each hospital. This was done in an effort to prevent pricing anomalies arising from small sample size. He clearly stated in his report that this selection had no scientific underpinning, but that this did not impact his ability to accurately observe the trends in pricing difference between the two hospitals pre-merger.*4 Govinda also readily acknowledged that the reliance on one DRG dataset provided for by Discovery was not ideal, but that because Discovery health negotiates tariffs on behalf of 15 medical schemes which comprise 33% of medically insured individuals, an analysis based on the Discovery Health data alone could provide a conservative view of the expected post-merger changes in prices at Lakeview, Govinda submitted that based on the DRG analysis, the number of codes for which Linmed charged a higher price than NHN Lakeview is larger than the number of codes for which it recorded a lower price. He concluded that in over 50% of the DRG codes which overlapped between Linmed and NHN Lakeview, Linmed charged more for the procedure underlying the code. And that this finding was consistent across the period 2014-2016. Merger parties’ submissions The merger parties again disputed the Commission's findings on a number of levels, First, they contended, that even if a unilateral price increase was established (which it disputed it was), this alone would not be sufficient evidence of the significant lessening of competition required to substantiate a prohibition. Second, the merger parties contended that there were significant shortcomings with the RPL analysis conducted by both Govinda and the Commission, As we understood both Govinda and the Commission abandoned reliance upon these codes, we considered this contention no further. * Govinda (note 6 above) pW183 para 182. 35 Govinda (note 6 above) pW164 para 127. 16 (66} Third, the merger parties argued that Govinda’s DRG analysis was unreliable. Among other contentions, it was argued that the data provided by Discovery Health was not suitable to conduct the analysis undertaken by the Commission, that the sample size was too small to provide accurate results, and that the comparison between Netcare Linmed and NHN Lakeview would not present a comparable cost picture because of the comparative sizes of Linmed and Lakeview. Tribunal views

[67]On the facts provided the merger resulted in an exchange of NHN tariffs for Netcare tariffs at Lakeview.** Bishop indicated that when Netcare took control of Lakeview, care was taken to mitigate the effects on tariffs, by Netcare implementing a specialised Netcare Lakeview tariff.5’ This seems to lend to the idea that there was a difference between the two. Evidence from witnesses seemed to suggest that the NHN national tariffs are, as a general rule, lower than the Netcare tariff. Dr Noble-Luckhoff indicated in her witness statement that “Netcare’s tariff file has higher rates than that of the NHN for all schemes under consideration.” ° In her testimony, when pushed on how she knew that Netcare’s tariffs are higher Dr Noble Luckhoff responded “/t’s known"?9) Mr Marion indicated that prior to the merger Netcare's tariffs were higher than those charged at NHN Lakeview.’? We acknowledge that anecdotal evidence should not supersede economic analysis, but still found it plausible that the merger could have given rise to a situation in which there was an increase in tariffs at Lakeview.

[68]36 Mark Bishop Witness statement pW107 para 22. 37 Ibid para 23, 38 Jenni Noble-Luckhoff Witness statement pW87. *? Transcript p 208 line 12. * Kenneth Marion Witness Slalement pW96 para 11. 4) Commission's report pW26 “Table 5: Price Comparison of the top 13 hospital procedures and tariff codes in the past year (2015) for Discovery, Bonitas, Polrned and Bankmed’. 17

[68]

[70]74] {72} However, as the Commission’s RPL analysis had been disregarded, and rightly so in our view, we were not provided with sufficient reliable analysis to decide conclusively the issue of a tariff increase. It was agreed by the parties that the DRG analysis was, in principle, the more reliable process of cost comparator. We however found that the data and methodology used in the DRG analysis by the Commission was unable to sustain the theory that medical aid schemes would pay more at Netcare Lakeview. We were convinced that the data set used in the DRG analysis was too small and unreliable to give rellable results. We do however also note that the sample size proposed by the merger parties also had no scientific basis. Discovery itself, when submitting the data, warned the Commission that it was not fit for Commission’s described purpose.** Further, whilst the DRG comparison went some way to eliminating the problem of case-mix, no consideration was given by the Commission to weighting the DRG codes according to their prevalence at Lakeview prior to the merger. Whilst we did find that Govinda's DRG analysis was unreliable, we were sympathetic to the plight of the Commission in its investigation. We acknowledge that performing a comparative pricing analysis is a complex task, only made worse by the paucity of available information provided by market participants. We found Dr Govinda to be honest and forthcoming with the limitations of his analysis. @ Transcript p97 lines 4- 17. 43 Discovery Health submitted: “We caution against using this data for drawing conclusions on cost efficlencies between hospitals, and note thal any conclusions reached an the basis of this data are likely to be statistically and actuarially invalid. This is mainly due to the small sample size in each DRG. Trial bundle 1136, 18

[73]

[74]

[75]

[76]Our finding on the DRG analysis meant that whilst there was the acknowledgement of the potential price increase, we were not able to rule definitively on the presence or quantum of the price increase and were thus unable to determine if such an increase would have any real effect. Further, we found that the Commission was unable to lead sufficient evidence fo prove that a step change in tariffs was symptomatic of any significant lessening of Competition (“SLC”) brought about by this merger. We should not be taken to set the precedent that the presence of unilateral price Increases may never result in a finding of an SLC, but on the facts before us, this was the case. We were thus unable to find any evidence that the merger would result in a unilateral price increase. Considering this and our finding on the Commission’s second and third theories of harm, we found that the merger was unlikely to substantially lessen competition in the relevant market. However, the merger parties tendered a condition in relation to pricing which we considered to address the potential pricing harm identified. The Tribunal has power to accept conditions tendered by merger parties even in instances where no evidence of SLC has been found.”* We understood the merger parties’ tender to be one of good faith, seeking to mitigate any potential harm arising fram a price increase if one were to arise. We accepted the tender and imposed the condition on this ground, Public interest

[77]The Commission did not raise any public interest concerns with the transaction and none arose during our consideration of the transaction. Although the merger parties raised a number of public interest benefits to the transaction, because of our findings on the competition issues, we did not find it necessary to make any finding on public interest issues. “4 Tiso Consortium v Nail Ltd [2004] 1 CPLR 302 (CT) para 55. 19 Conclusion

[78]In light of the above, we concluded that the proposed transaction was unlikely to substantially prevent or lessen competition in any relevant market. In addition, no public interest issues arose from the proposed transaction. Accordingly, we approved the proposed transaction subject to the conditions tendered by the merging parties. 9 July 2018 Prof. Imraan Vatodia Date Ms Mondo Mazwai and Ms Andiswa Ndoni concurring. Tribunal economist : Karissa Moothoo Padayachie Tribunal case manager : Alistair Dey van Heerden Legal representatives ofthe : Adv J Wilson SC assisted by Adv G Marriot merger parties instructed by Nortons Inc. Merger parties economic : Patrick Smith of RBB Economics expert For the Commission : € Slump, M Naidu, and S Ntlontli of the Legal Services Division. Commission's economic : Dr Hariprasad Govinda expert 20