Medical schemes unfairly discriminate against black providers - Inquiry

This is because such providers 1.4 times more likely to be classified as having committed Fraud Waste or Abuse

Section 59 Investigation

Established in terms of Section 7(a)(b)(c)(d), 8(a) and (k) and 9(2) of the Medical Schemes Act, 131 of 1998.

Interim Report Summary

19 January 2021

Adv Tembeka Ngcukaitobi SC | Adv Kerry Williams | Adv Adila Hassim

Secretariat: Lawtons Africa

1. In early 2019, a number of healthcare providers and members of Solutionist Thinkers and the National Health Care Professionals Association made allegations that they were being treated unfairly by medical aid schemes based on race and ethnicity.

2. The Council for Medical Schemes (CMS) launched an investigation into these allegations in terms of its regulatory mandate. This led to the establishment of an independent investigation panel to conduct an inquiry into these allegations. The inquiry became known as the Section 59 Inquiry. The Panel members are Advocate Tembeka Ngcukaitobi SC (Chairperson); Advocate Adila Hassim; and Advocate Kerry Williams.

3. The Panel has concluded its investigation and publishes an interim Report for comment. Comments should be submitted to the Panel’s secretariat, Lawtons Africa:

Mr SJ Thema: [email protected]

Mr Ushir Ahir: ushir[email protected]

by close of business on Friday, 5 March 2021.

4. The Report details the evidence presented to the Panel, evaluates such evidence, including expert evidence, considers the legal issues raised and makes findings and recommendations to the CMS. The CMS is at liberty to accept or reject the recommendations.

5. The Panel was mandated to investigate two main issues: whether there is racial discrimination by schemes against Black health care providers and whether Black providers were being treated procedurally unfairly. The Panel was mandated to function as an inquisitorial body. Therefore, the Panel did not adjudicate individual complaints. The evidence presented by complainants and the schemes was important for the Panel’s Inquiry regarding how the schemes and administrators’ risk management systems, more particularly their fraud, waste and abuse (FWA) systems, worked in practice.

6. In order to progress the Panel’s work, particularly in relation to the allegations of unfair discrimination based on race, we appointed an independent expert, Dr Zaid Kimmie (skilled in mathematics, statistics and data analytics) to assess the outcomes of the FWA investigation processes by the three main administrators – Discovery, Medscheme and GEMS. We also appointed two further experts: Adv Wim Trengove SC and Prof Melissa Steyn who provided assistance to the Panel with regard to the legal framework to test unfair discrimination and the principles of implicit racial bias

7. Our interim findings are set out below.


8. Every scheme and administrator that was implicated in the complaints denied that there was unfair racial discrimination in their FWA investigation process. In the main the denial was based on the fact that the FWA investigations are triggered by either:

8.1 an automated system, underpinned by an algorithm, that flags outlier practices for investigation;

8.2 or tip-offs and whistle-blowers.

9. The schemes and administrators argued that only practice numbers are known by the automated system and there is no assignment or identification of race either explicitly or implicitly in the system.

10. Dr Kimmie, the Panel’s expert, conducted his own analysis. The schemes and administrators responded fully to such analysis. The Panel considered their responses.

11. After considering all the evidence and responses, we find that between 2012 and 2019 Black practitioners were more likely to be found to have committed FWA than their Non-Black (White) counterparts, by Discovery, Medscheme and GEMS. This means, for the reasons provided in this interim Report, there was unfair racial discrimination.

12. We do not find evidence of explicit racial bias in the algorithms (to the extent that the workings of the algorithms were disclosed) and methods that the administrators and schemes use to identify FWA.

13. However, using the data that Discovery, GEMS and Medscheme provided the Panel and Dr Kimmie, there is a substantial difference in FWA outcomes between Black and non-Black practitioners over the period January 2012 to June 2019.

13.1 Over this period, across all disciplines and the aforementioned three schemes and administrators, Black practitioners were 1.4 times more likely to be classified as having committed FWA than those identified as not Black.

13.2 The probability that this distribution occurred by chance (i.e. that there is no correlation between racial status and FWA outcomes) is for all practical purposes 0 (zero);

13.3 The starkest differentials are evident amongst the following:

13.3.1 Black general practitioners are 1.5 times more likely to be identified as FWA cases than their non-Black counterparts;

13.3.2 The rate at which Black physiotherapists are identified as FWA cases is almost double (1.87) that of their non-Black counterparts;

13.3.3 Black psychologists are three times more likely to be identified as FWA cases;

13.3.4 Black registered counsellors and social workers are also three times more likely to be identified as FWA cases. More than 50 percent of Black registered counsellors have been identified as FWA cases – this is the highest rate among the disciplines analysed; and

13.3.5 Black dieticians are 2.5 times more likely to be identified as FWA cases compared to their not Black counterparts.

14. There are clear differences in the scale of racial discrimination between Discovery, GEMS and Medscheme.

14.1 Discovery was 35% more likely to identify Black providers as having committed FWA.

14.2 GEMS was 80% more likely to identify Black providers.

14.3 Medscheme was 330% more likely to identify Black providers as guilty of FWA.

15. Although each of the three schemes and administrators presented expert evidence to contest the findings of Dr Kimmie, we find that the disproportionate impact on Black providers, which amounts to unfair racial discrimination, remains.

15.1 With regard to Medscheme, assuming the correctness of their expert’s methodology, it is 35% more likely to find Black providers guilty of FWA.

15.2 On GEMS own version it is 47% more likely to find Black providers guilty of FWA.

15.3 On Discovery’s version it is 36% more likely to find Black providers guilty of FWA. However, Discovery reduces the risk ratio further to 1.09 on the basis of what it describes as confounding factors. The Panel requested that this be directly addressed by Dr Kimmie. We find that Discovery’s theory that there are confounding factors that affect the risk ratio is not credible.

16. Based on an assessment of the evidence, together with the application of anti-discrimination law, the Panel is of the view that the outcome of the FWA investigations, conducted by Discovery, GEMS and Medscheme between 2012 and 2019, amount to unfair racial discrimination against Black practitioners.


17. The schemes and administrators were also accused of not following fair procedures when implementing their powers under section 59(2) and (3) of the Act. Again, GEMS, Medscheme and Discovery denied that there was unfairness in the procedures they follow.

18. A substantial part of the difficulties that emerge from the implementation of section 59(2) and

(3) of the Act relate to the fact that there is contestation about the interpretation and reach of these subsections.

19. The Panel is of the view that these subsections are capable of a clear and coherent interpretation with a particular reach. Such an interpretation flows from the ordinary meaning of the subsections as well as the assistance that the rule of law, and the administrative justice right in the Constitution (and its implementation through the Promotion of Administrative Justice Act 3 of 2000 (PAJA)) brings to the subsections. In sum the Panel is of the view that:

19.1 Section 59(2) of the Act requires schemes to pay providers or members the current benefits which are owed to them. Regulation 6 of the Regulations implements section 59(2) of the Act and requires schemes to raise any issues with current invoices within a period of 30 days so that the provider or member has an opportunity to correct invoices and re-submit such invoices for payment;

19.2 Section 59(3) of the Act works in lock step with section 59(2) in that it allows a scheme to claw back amounts from future benefits which are to be paid. One of two conditions must be met for such a claw back to take place: either a provider or member must have been paid, on a bona fide basis by the scheme, a benefit to which they are not entitled; or the provider or member must have engaged in fraud, theft, professional misconduct or negligence and caused a loss to the scheme;

19.3 The scheme is given a significant power in section 59(3) in that it can unilaterally decide to claw back an amount, assuming one of the two conditions are met;

19.4 The nature of the power that the scheme exercises in clawing back such amounts is a public power which is subject to section 1(c) and section 33 of the Constitution as implemented by PAJA. We have given an extensive analysis as to how we have reached this conclusion. Such a power must be exercised lawfully, reasonably and in a manner that is procedurally fair. Without such constraints the schemes would be engaging in a form of self-help which is prohibited by the Constitution;

19.5 The three requirements of lawfulness, procedural fairness and reasonableness require the schemes to inter alia always ensure one of the two pre-conditions are met when considering a claw back, be able to justify any decision to claw back and ensure that it does not cause undue or disproportionate harm, give the provider an opportunity to meaningfully comment on a proposed claw back before a decision to claw back is taken;

19.6 Furthermore, the exercise of powers in terms of this subsection would also be subject to common law administrative law controls and would at the very least require the scheme to justify its decision to ensure it is not arbitrary or irrational and to give the provider an opportunity to meaningfully comment on a proposed claw back before a decision to claw back is taken; and

19.7 We emphasise that these constraints on the scheme’s powers to claw back monies from providers are significant and should do much to address the providers’ allegations and concerns regarding unfair treatment.

20. It is not only claw backs that were cause for concern, but also the schemes’ decisions to place providers on indirect payment when the schemes were of the view that a provider has engaged in FWA or where providers do not cooperate with the schemes during an FWA investigation.

21. The Panel is of the view that although there is no express provision in the Act which allows schemes to place providers on indirect payment, the schemes may do so where either a provider has contracted with the scheme on terms which allow for this; or where the scheme has included such a possibility in its internal system of proper financial controls (section 57(4)(c) of the Act mandates that the schemes must have such systems of control).

22. However, the requirement that a scheme must have a proper system of financial control also places constraints on the scheme when it places a provider on indirect payment. Like with the decision to claw back monies in terms of section 59(3), when a scheme contemplates a decision to place a provider on indirect payment in terms of its internal policies its decision must be reasonable and not cause undue harm; and it should before any decision is taken give the provider an opportunity to meaningfully comment on the proposed decision to place such provider on indirect payment.

23. This interpretation of sections 59(2) and (3) of the Act also has consequences for the manner in which the schemes determine the amounts to be clawed back from providers and the amounts that providers and schemes might agree in any settlement agreement – more often than not, concluded to avoid a claw back in terms of section 59(2).

24. Any amount which is clawed back by a scheme must be reasonable and must be based on a methodology which is reasonable. We accept that for practical reasons the schemes (and for that matter, providers) probably need to estimate the amounts which may be clawed back, as there is no time or resources to locate and trawl through numerous records. Absolute precision in the calculation of the amount clawed back is not a requirement of section 59(3).

25. In relation to the current methodologies implemented by the schemes and administrators, we note that none of the methodologies are made available to providers or members. In order to ensure good decision making and transparency they should be easily available in advance of any engagement with the provider. Further, GEMS uses approximately one year’s historic claims data to estimate claw back amounts, Discovery can, and often does, use up to three years of historic claims data to estimate claw back amounts. Medscheme appears not to have an explicit methodology and rather adopts a case by case approach to how it calculates claw backs. There is little doubt that Medscheme’s approach operates most arbitrarily for providers.

26. We are also concerned about an aspect of Discovery’s approach to calculating amounts to be clawed back as the approach may be disproportionately harsh on providers who are required to pay back amounts over a period of up to three years. We do not believe that this approach to calculating amounts owed is justifiable – this is particularly so where the schemes’ systems are capable of picking up unethical conduct or misdemeanours in a much shorter time and could notify providers that they are being flagged much earlier than is currently the case. We say more about this in our recommendations below.

27. Finally, there is little doubt that there is a power imbalance in the settlement negotiations between schemes and providers. There are a number of reasons for this. The first is that if a provider does not agree to the terms of a settlement agreement (also known as an acknowledgement of debt (AOD)) she is likely to be subject to a claw back in terms of section 59(3) of the Act. This makes it difficult for a provider to refuse to agree to an AOD. The second reason is that providers are usually individual practitioners, whereas the two administrators are large, well-resourced and powerful corporations; GEMS is the second largest scheme in the country and is also more resourced and powerful than an individual provider. The third reason is that the schemes and administrators are in effect performing a function akin to policing in the implementation of their FWA systems and they are given a unilateral statutory power to claw back monies from providers. This is a significant power with real and immediate consequences for the financial well-being on providers.

28. In circumstances where there is a power imbalance in an AOD negotiation the schemes and administrators should proceed with caution and ensure that mechanisms are put in place which ameliorate the risk of providers agreeing to AODs under any form of duress, whether it be due to actual or perceived threats and whether it is economic duress or otherwise. On this score we further note that Medscheme is financially incentivised by one scheme, through a bonus, to collect monies from providers – this incentive operates over and above Medscheme’s flat fee. GEMS historically also incentivised investigators to collect monies from providers. This introduces real and practical scope for abuse as Medscheme is not only acting in the interest of the scheme (who should be acting in the public interest) but in its own interest – where it has the opportunity to make extraordinary financial gains.

29. In conclusion:

29.1 We have found that some of the current procedures followed by schemes to enforce their rights in terms of section 59 of the Act are unfair. We have also found that Black providers are unfairly discriminated against on the grounds of race.

29.2 These findings are both serious and far-reaching. But we believe that it is important to stress that we have not found evidence of deliberate unfair treatment – the evidence shows the unfair discrimination is in the outcomes. Our Constitution regards the form of unfairness that we have found to exist as constituting unfair racial discrimination.

29.3 Although the appointment of the Panel was in terms of legislation – the Medical Schemes Act – the participation by everyone was voluntary. This has been the strength of the Inquiry. Information was also voluntarily provided. Where appropriate requests for confidentiality were made, we have attempted to respect these. But we have also attempted to balance requests for confidentiality against the need for transparency since this was a public inquiry.

29.4 We had no power to find anyone guilty. Nor were we appointed to investigate the veracity of each individual claim of unfair treatment and unfair discrimination. But we would be failing in our duty if we ignored degrading, humiliating and distressing impact of racism against the individuals who testified before us. A part of our function was to provide a platform for the expression of individual experiences of racial discrimination and other forms of unfair treatment.

29.5 We do not believe that we have covered each and every possible complaint of providers against schemes. That was not our mandate. We also do not claim to have explored all possible manifestations of racial discrimination and unfair procedures. But we have received sufficient data and information to make informed and reliable conclusions of the patterns of conduct by the schemes. In certain respects, our conclusions – which are evidence based – accord with individual experiences. Affirming these individual experiences as fact has been an essential element of the work of the Panel, confirming that the majority of the complaints were not frivolously made. 

29.6 Our expectation is a constructive engagement with the findings and recommendations in this report. Rather than conclusive, our findings will hopefully provide a basis for the necessary reconstructive work which must be undertaken by the role-players in the medical schemes industry. We do not see the issue as a binary conflict between schemes and providers, but as reflective of fissures of the past that remain unresolved.

Nkosi sikelel’i Afrika

Pretoria: 16 December 2020

The full report can be accessed here - PDF.