What is capitation and what are its alternatives?
Today, we use many different payment models across health- and social-care to fund our services. Especially for acute care, especially, payment models were designed to encourage volume and reduce waiting lists, where they have been very effective. However, our vision is to implement a whole systems approach where we focus on outcomes that keep people well at home and support them to manage their complex conditions, rather than treating many episodes of care separately. It does not make sense to set system outcomes at the level of specific interventions when we care about people’s overall health over a longer period. Because they do not incentivise reductions in demand and prevention, our current payment models also make achieving whole system financial sustainability much harder.
Our current payment model has other shortcomings. Fragmenting budgets and contracting for individual services does not encourage innovations in care provision because new interventions are not encouraged or reimbursed. For example, Mary might be admitted to hospital multiple times for falls.
Some of these might have been avoidable. However, because there is no HRG or specific hospital payment for installing railings, the hospital had no incentive to address the root cause of the accident. Providers have few financial incentives to cooperate as their performance is managed on how they deliver their separate services, rather than their shared contribution to achieving patient-focussed outcomes. As a result, Mary’s hospital has few incentives to work with adult social-care to see if there is a way to get railings installed. Providers are not penalised for poor quality care that results in avoidable demand for services in other settings. In a system facing increased costs from an ageing population with changing care priorities, an alternative payment model must be found.
International experience of health systems, from Kaiser Permanente to Valencia to ChenMed, also shows that changing the payment model is a critical enabler to Whole System change. This is because without changes to the flow of funds between organisations it is hard to change what people do.
POSSIBLE PAYMENT MODELS
The possibilities for payment models can broadly be characterised into four different categories. In all cases, additional incentives can be offered to reward providers for meeting outcome targets. Managing outcomes effectively can encourage providers to offer the best possible care, but requires careful management and information systems to execute well.
An effective care system will probably use different payment systems for different sorts of care, depending on the aspirations for different services and populations. In addition, it may be strategically valuable to mix payment models. For example, paying for a share of services on the basis of episodes of care can allow commissioners to encourage competitive pressures on capitated providers. Mixing these strategies requires increasing commissioner sophistication.
1. FEE FOR SERVICE
A provider is paid for each unit of activity they complete, which incentivises volume rather than quality or prevention. Commissioners keep all the risks of over-delivery. For example, if Mary is admitted to hospital after a fall, the hospital will receive a payment for each imaging or diagnostic service, for setting the bone and for any other services required. In general, this payment model is poorly suited to the current pressures on care systems. However, in certain limited situations where very high or universal uptake is desirable, it can be effective. For example, fee for service could work for immunisations or encouraging care plans, where more activity roughly equates to better performance. This relies, however, on maintaining a population registry to check for completion and a price list for providers.
Pros: Providers are incentivised to provide as much care as possible to people who want it.
Cons: There are no incentives for providers to prevent activity as this loses them money. Innovations in new services are discouraged as they are not reimbursed.
2. BUNDLED PAYMENTS FOR EPISODES OF CARE
A provider is paid a fixed fee for a defined bundle of costs surrounding an episode of care. An episode is a complete period of care for a particular condition, usually including all pre- and post-care as well as a provision for complications. For example, an episode might cover a normal pregnancy, including all antenatal care, normal delivery and post-delivery care. Alternatively, an episode might cover a hip or knee replacement, including any consultations preceding the operation, the procedure itself and rehabilitation as well as any readmission that might be required.
This incentivises high-quality care being delivered first time, because providers risk facing extra costs if extra pre- or follow-up care is needed. It also encourages best-practice pathways and care, as providers carry the cost of additional services such as unnecessary diagnostics and consultations. In the above example, where Mary is admitted to hospital after a fall, the hospital is paid a set rate for the entire episode, including any diagnostics or follow-on work. Mary is less likely to face unnecessary tests, and the hospital is incentivised to make sure everything heals properly the first time.
Bundled payments require the ability to characterise cost and payment thresholds for a number of types of episodes to distinguish between normal and abnormal costs, as well as information systems that enable transparent and rigorous tracking of care at the level of specific activities across providers.
Bundled payments are best for services which are elective, tangible, and for which there is a critical volume that supports the investment costs needed for the analytical systems. They are also effective for less mature markets where capitation might lead to competition concerns.
Pros: This model encourages value on a particular pathway, since it pays an accountable provider for the whole service. It does not require complicated budget calculations up front, because it permits retrospective payment. The system is granular to the level of conditions, so it is easy for clinicians to target, and it can be combined with other payment models like capitation.
Cons: While it does help control outpatient costs, it does not incentivise prevention to reduce overall volume. It effectively extends the HRG system beyond hospital care and so it requires an extended data system.
3. BLOCK CONTRACTS FOR INDIVIDUAL PROVIDERS
A provider is given a fixed amount to provide a service (or set of services) for a period of time. This encourages productivity in meeting service outcomes for the lowest cost. Commissioners transfer the risk to providers if more people need the service than is planned. Here, when Mary is admitted to hospital after a fall, the hospital has a strong incentive to reduce the risk of future falls. They may well help her install railings at home and provide physiotherapy to help her recover her strength as she heals, to reduce the likelihood of future accidents.
Pros: Individual providers have incentives to prevent demand for their services and improve their productivity. They have flexibility to innovate how the service meets its outcomes.
Cons: Providers can respond to the incentive either by reducing the need for the service or by reducing availability, which may conflict with statutory duties of commissioners, or by attempting to shift demand to providers of other services. Careful management of outcomes is required to ensure the incentives work appropriately, as there is no competition on quality. It is difficult to reallocate resources between providers if commissioners have funded the balance of services poorly. Establishing block contracts does not create or encourage increased transparency over costs and activities.
4. CAPITATION FOR A POPULATION
Unlike block contracts, where providers are paid to deliver a block of services, in capitation providers are paid for the full care needs of the population. They are paid a fixed amount that grows at a predictable rate, independent of activity. They can be additionally rewarded for achieving outcomes targets. This encourages providers to prevent care needs in high cost settings and coordinate to minimise duplication and waste. An essential requirement of capitation is the ability to understand the consumption of health- and social-care for a population in order for money to follow individuals. Commissioners can transfer the risk to providers if more care is needed than planned. Here, when Mary is admitted to hospital after a fall, the hospital will try to make her experience and outcomes meet the requirements set and will also have a reason to be proactive when it comes to reducing the risk of readmission. They will also have an incentive to think ahead about how her accident will interact with her other conditions and take steps to avoid complications.
Capitation is best suited for mature markets where a number of providers can compete for service users on quality of care.
Pros: Encourages coordinated, preventative care that keeps people well at home and avoids unnecessary high cost care. Providers have the flexibility to innovate and allocate their resources to achieve the highest returns so they can all share savings and are incentivised to do so because they also share the risks.
Cons: Requires significant capabilities, in particular the ability to coordinate between professions and information systems that can track an individual’s activity and costs across many professions. Unlike the current system, where access and service offering is managed by the commissioner, here the provider manages the offering and rations resources. Providers have a large incentive to restrict access to care, which commissioners must carefully monitor through tracking outcomes and ensuring fulfiling statutory duties such as FACS. Social-care commissioners have an obligation to assess individuals for FACS and provide care when certain thresholds are reached. Where social-care services are provided on a capitated basis, commissioners must ensure these, and other, statutory obligations are met.
Providers are also exposed to much more performance risk than under the current system and must be able to manage that risk through insurance arrangements or working capital that can absorb deficits.
CASE STUDY/EXAMPLE
Arkansas is designing a bundled payment system to cover episodes of care that can last for many months, like pregnancy and elective procedures like hip and knee replacements. An accountable provider, determined analytically according to which organisation is responsible for the majority of costs, is designated for an episode of care for a particular individual. They are then paid for the end-to-end care costs for the episode, including any pre- and follow-up care. Other organisations that contribute towards the episode are reimbursed through deductions from the accountable provider. At the end of the year, there are retrospective penalties and rewards that are paid to accountable providers based on the quality of their population’s outcomes (which incentivise improvements in care) and the average costs of their episodes compared to all other providers (incentivising more efficient use of scarce resources and care being delivered right the first time.)
Source: Richardson, Dorling – Global Integrated Care Case Compendium (McKinsey), Healthcare Incentives Improvement Institute (http://www.hci3.org/sites/default/files/files/IB.BundledPayment-June2013...)
