CEO at Professional Provider Organisation Services (PPO Serve) and former Head of Strategy at Discovery Health, Dr Brian Ruff, talks about shifting private healthcare from a fragmented, fee-for-service care model to a value-based model driven by integrated collaborative care teams.
Tell us about your background in healthcare and the premise for starting PPO Serve.
I’m a qualified Physician (a specialist in Internal Medicine) and a Rheumatologist. I studied Economics and worked at various state clinics and hospitals for a number of years. After 1994, I worked on different projects in the Gauteng Department of Health, including strategic management in the head office and a few months in the National Treasury. I joined Discovery Health in 1999. As part of an actuarial unit, I introduced the analysis of clinical data. We established a coding unit that focused on setting standards in clinical and procedure coding, and acquired or built various ‘groupers’ (Hospital DRGs, Episode; Population) to perform ‘case mix’ analysis so we could see how the healthcare system was working for the members. Our focus was on the cost and the quality of care.
Last year my business partner, Riedwaan Jabaar, and I established PPO Serve. We recognised that private healthcare in South Africa is in a mess, similar to what we see in the US, reflected in the rising cost of healthcare, except that the narrative in SA is much more about fragmentation than it is around a lack of accountability. PPO Serve is built on a number of reform principles similar to that in the Affordable Care Act (ACA) aka Obamacare, one of which is a new service delivery model called the Accountable Care Organisation (ACO). PPO Serve is bringing the ACO to South Africa through a customised approach that is suitable for the local context.
And how exactly does PPO Serve address the problem?
By contrast to the public sector where a lack of management is possibly its biggest challenge, the biggest challenge in the private sector is the lack of teamwork and care coordination. A major reason for that is that medical schemes’ tariff schedule pays individual clinicians, instead of a team. This is made worse because the Tariff also funds every service rather than patient health outcomes.
We’re working with clinicians to contract with medical schemes on a different basis, where patient ‘value’ is the aim i.e. good quality at an affordable cost. We’ve organised independent groups of doctors and allied health professionals such as psychologists and nurses, into what we call an Integrated Clinical Consortium™ (ICC™). ICCs™ are integrated multidisciplinary, accountable teams who practice quality medicine with collaborative colleagues and good support. It’s a commercial entity owned by the clinicians, so they earn together. Ultimately, ICCs™ shift care appropriately into a community setting, reducing costs by stopping unnecessary hospital admissions and improving the quality of care.
Tell us more about how the ICCs™ and the new model works.
We’ve designed the Value Contract, which replaces the fee-for-service (FFS) payment model. The Value Contract links suitable access, comprehensive benefits and quality outcomes with appropriate rewards for the ICCs™. Such type of contracting also enables the ICC™ to focus on improving patient health, instead of fighting with schemes to access benefits for patients.
We believe that the FFS model prohibits doctors working in teams and results in fragmented care with many gaps and much wastage. In other words, the high costs of private healthcare are driven by poor quality – re-admission, re-ordering tests, poor outcomes and rework, etc. Added to that, in the FFS model doctors react to the immediate problems of individual patients. Under the ICCTM, doctors are proactive, planning ahead for patients with complex care. Also, because they take responsibility for a population, they see patterns and can design interventions that impact on many of their patients. That responsibility determines the way they are paid.
The saving in healthcare comes from the observation that if you manage to drive the quality agenda for the long term, then the costs come down. But while the ACOs in the US talk about ‘shared savings’ where the clinicians take some risk because many are historically in large groups, in SA that’s different where clinicians work alone and don’t have a large balance sheet so we prefer a reward structure of quality tiers in the Value Contract, so that as we improve quality and reduce costs, additional money is be available for schemes to pay these teams. Quality tiers are based on measurable patient health outcomes.
The ICCTM work is patient centric with individual patient care plan. We have numerous condition specific care plans formulated from evidence-based medicine, from reputable international organisations, often State institutions. We’ve built an Intelligent Care System (an IT system) that rationalises these into patient centric complete care plans with no duplicates or gaps. This then effectively becomes the work-flow for the care team which suggests interventions and frequency of treatment. This proactive work is how to keep patients as well as possible. So for Mrs X for example, this means that when she leaves the hospital or consulting room, she isn’t lost. There’s a team looking after her making sure she does as well as possible, following up on her care and monitoring her health outcomes, which in turn earns the ICC™ rewards.
We have an easy to use Toolkit that contains standard processes, tools and templates to make it easy to set up an ICC™. The ICC™ is also built with the doctor team and customised to the population need in the particular area.
This Value Contract model hinges on the right kind of data and the right quality of data. How are you addressing this issue?
To answer that, we need to talk about using the Case Mix tools, which is the science of measuring the illness of patients. Practically, a Case Mix Index (CMI) is a measure of the relative health of a group of patients in a healthcare environment. The CMI value is used to determine how to efficiently allocate resources to care for the patients in the group. It’s largely based on diagnostic codes and procedure codes that characterise every episode of care for each individual patient, as well as their age and gender and their socio-economic status.
For example, an episode of Maternity Care lasts 2 years – from prior to conception until a year after the delivery. But the woman with this episode of care are very different because of their other health issues and the complications they experience, so identifying different levels of severity is important and can be reflected as a case mix number i.e. a ratio to 1.
Ultimately, combining these to get a CMI of the hospital and tells you what the difference is in the work load between Hospital A and Hospital B. Once you understand that you can, for example, determine which hospital is more cost effective and which one is producing better outcomes.
If we start understanding hospitals as production units, we can start managing them better. And of course measurement, whether you’re trying to determine quality of service, profitability or capacity over a period, is only possible with data and the generation of group-able, analysable data.
And how exactly does this impact healthcare and support the change in the model you’re proposing?
My involvement in CMI started in my time with the Gauteng Department of Health in the early 90s because we wanted to understand what was happening in nearly 50 hospitals scattered around the province. We looked at, for example, a regional hospital and a tertiary academic hospital and tried to analyse the very different kinds of mixes in each. Ultimately, what we know is that in order to plan for a new system, you need a ‘demand side’ measure, or a patient mix measure, to understand the right supply side provision of the services. In the post-apartheid state, fairness was a crucial intention of the new design. That equation, and information, is what we need to be able to meet healthcare needs within the parameters of affordability, accountability, acceptability and access.
At Discovery, once we acquired case mix tools, we initially started using CMI to understand the performance of various parts of the healthcare system. These were used for contracts with hospitals and were shared with interested clinicians because we understood that without data, we couldn’t make a difference. Clinicians need a reporting framework that shows how well or poorly they’re doing, where they’re doing well, and where they could improve.
Later in my time with Discovery, we started programmes that shared data with providers to enable a different approach to delivering care. At the Intercare in Tygervalley, which is a subacute unit, and we started implementing care co-ordination. Over time this spread to many other facilities and thousands of sick people were recruited, and the programme radically changed the way they were treated, the way they experienced the health system and also their outcomes – in short, it was a very successful project. This made us understand the impact of social factors affecting patient health outcomes. We then started building more sophisticated programmes, for example, focusing on geriatric people. That was very effective – the patients and the clinicians responded positively to it, as did their families and we produced much better outcomes as a result.
PPO Serve is currently running a patient care coordination project in Alberton with a multi-disciplinary team, which includes consulting specialists and some large GP practices with a sizable patient population. So, this is not a project of a scheme-to-the supply side, but rather we are working for the providers in the supply side, with the medical scheme as a contracted supporter – in this instance, Discovery is supporting us first in this project, and we will add other schemes later.
Capturing data, analysing it and acting on it is something that healthcare in SA is still grappling with. Many would argue that it’s expensive, complex and detracts from patient care. What would you say to the people who don’t yet see the benefits of electronic systems generating data in healthcare?
IT in the healthcare system is absolutely key to reform in the modern era. Used properly, it’s an essential component because it lets you improve processes that you couldn’t otherwise, and clinicians will use it if it’s easy to use (it’s a digital pen or uses a tick box) and is immediately helpful to them in managing patient care. Done well, the data capture is just part of the better process and if it is obvious to the clinician how its leads to better care because it an immediate aid to better decision making or care delivery, then they embrace it.
The trouble is that when we start talking about measuring the health service, or even a great electronic management tool, there’s a tendency to not see it as a positive thing – it’s seen as an audit, which could negatively affect them personally. So, at that level, it’s not enough to have the performance measurement; we must also have the right kind of governance structures in place – otherwise it’s just scary. People supplying IT systems, such as ourselves at PPO Serve, approach the industry gently and first establish that we’re not threatening anyone’s job, we think you’re doing a great job, and the data will help you practice even better. Data can be used to promote quality and improvement but in healthcare where there is no data, there is some resistance because it might be punitive. And that’s one of the barriers that we, as an industry, need to overcome. But without data, you don’t know how you are doing.
Private healthcare data is largely a result of claims data or encounter level coding, and we use that as the foundation to build patient records and care plans. Now we are now entering the world of ‘machine data’ – from wearable devices and social media, which can be useful for validation.
Finally, in order to make real use of digital data, we must have coding and interoperability standards for data exchange. In the absence of regulated standards, companies develop tools and guard their intellectual property, and this retards the way the industry evolution.
But data is only part of the equation to solving some of the biggest challenges we face in healthcare.
What are some of the other misconceptions in health that you’d like to dispel?
Firstly, we unhelpfully think of the private sector as a single, homogeneous system, but of course it isn’t. When you measure them, our regions that have remarkably different productivity. Now we can’t explain all of that, but there are a lot of clues as to why it is what it is. For instance, if a region provides too many beds, don’t be surprised if they are all full. If you have extremely weak GP level services, don’t be surprised if everyone ends up in hospital. If you have too many ENT specialists in an area, don’t be surprised if there’s hardly a child in that geographical location that still has their tonsils. On the other hand, there are areas with too few beds and doctors and relative denial of access to reasonable quality care.
One can observe specialists who take on the role of primary care doctors because there is a mismatch between the number of specialists and how many scheme members have chronic or complex disease i.e. too many specialists for the need. In a well-designed commercially sensible system where the specialist has a good working relationship with the chronic patient’s GP, the specialist sees you once a year or if there’s a problem. But in our system where the mismatch may mean that the specialist’s income isn’t particularly stable, so the specialist is more likely to advise the patient to come back and be seen more frequently. It’s clear that such patterns of practice is commonplace in certain regions of the country
Finally, I want to note that we don’t have a policy framework or a regulator for the supply side of the health sector industry. It’s a big issue and its absence is part of what makes it hard to establish new models of service delivery. This is true of case mix tools; and of data standards; also payment schedules for teamwork and that recognises outcomes.
Our approach to the problem at PPO Serve is that we must openly share a population based Value Contract framework with all stakeholders, so that everybody can move forward by embracing it. Obviously each contract has its own competitive private financial arrangements. The shared framework allows for standard performance benchmarks, so we can measure the industry improvements. The US Government has just come out with a new system called Merit-based Incentives. That’s what our government should be doing. It should produce the tools that drive improvement but mandate that the private sector must use them in order to participate.