ORGANIZATIONS

ACO vs DCO vs HMO vs PPO

ACOs (Accountable Care Organizations), DCOs (Direct Contracting Organizations), HMOs (Health Maintenance Organizations), and PPOs (Preferred Provider Organizations) are all methods by which a group or “network” of healthcare providers is paid for the services they deliver to a defined set of patients through some form of special payment arrangement that differs from the way other providers are paid for the same services. There are both similarities and differences between them:

  • In ACOs and DCOs, the providers who are included in the group are selected by the providers themselves, whereas in most HMOs and PPOs, the group or network of providers is defined by a health insurance plan.
  • In most ACOs, the providers do not know which patients will be included in the special payment arrangement until after services are delivered (the assignment of patients is performed “retrospectively”), whereas in the other models, the providers generally know which patients are included before services are delivered.
  • Most ACOs have been paid through a shared saving program in which there is no change to the underlying fee-for-service payment structure for the providers in the ACO. In contrast, in DCOs and in many HMOs, a provider group receives a capitation payment that it can use to pay its physicians and other providers in different ways.
  • Most PPOs are paid using a standard fee-for-service payment structure, but agree to accept lower fees (or higher discounts on standard fees) because the patients have financial incentives to obtain services from providers in the PPO rather than providers that are not part of the PPO. Providers in an HMO may also be paid using a standard fee-for-service payment structure and accept lower fees because the patients have financial incentives to use them (including only receiving insurance coverage for services delivered by the HMO providers).

Payer vs Purchaser

In traditional health insurance arrangements, the purchaser is the individual or organization that pays premiums for a health insurance policy, and the payer is the health insurance company that pays claims under the policy.

Self-insured employers are purchasers, but they are not typically payers, since they will usually engage a Third-Party Administrator to actually pay claims from providers. However, in some direct contracting arrangements between employers and providers, the self-insured employer may be both a purchaser and a payer, because the direct contracting organization (DCO) handles payments to individual providers, and the employer pays the DCO using a capitation payment or other mechanism that is not tied to individual services.

The distinction between purchasers and payers is important for several reasons:

  • all else being equal, when three entities are involved – a purchaser, a payer, and a provider – spending associated with healthcare services will be higher because of the administrative costs incurred by all three entities to manage their respective relationships.
  • if a provider reduces costs and accepts lower payments from a payer, the savings from those payments may or may not be passed on to the purchaser depending on the nature of the relationship between the purchaser and payer. For example, the premiums paid to a payer by a fully-insured purchaser may not decrease even if providers are paid less by the payer, whereas the savings from lower payments to providers will be passed on to self-insured purchasers.
  • a payer may have financial reserves to cover random variation in spending on healthcare services that a purchaser does not.

PAYMENT MODELS

Bundled Payment vs Episode Payment

An episode payment is generally also a bundled payment, since typically multiple services are delivered as part of a single episode of care. (However, if a simple health problem can be addressed through the delivery of one service, then a fee for that one service could also be considered an episode payment for that health problem.)

A bundled payment need not be an episode payment, because the bundle could involve only a portion of the services needed to address a patient’s needs during a full episode of care. For example, Medicare and many other payers pay hospitals a “case rate” for hospital stay; this is a bundled payment, because it covers all of the services the hospital delivered during the patient’s treatment in the hospital, but it is not an episode of care payment because it does not include the services provided by physicians during the hospital stay nor for any services related to the hospital care that occur after discharge from the hospital (such as post-acute care).

A bundled payment could also be equivalent to multiple episode payments. For example, a capitation payment is a single, bundled payment for all of the services received by a patient during a period of time. If the patient has multiple episodes of illness during that period of time, then the capitation payment would presumably support delivery of the services needed to address all of those episodes, whereas multiple episode payments would be needed, one for each

Capitation vs Global Payment

A capitation payment need not be “global”; for example, monthly per-patient payments to a primary care practice are a form of capitation, but they are only intended to pay for the services the primary care practice delivers, not the services delivered by any other physicians, a hospital, or other providers.

Conversely, a [global payment}{.Term} need not be paid using a “capitation” methodology (i.e., a payment that is directly proportional to the number of patients); for example, a hospital might receive a global budget payment that is expected to cover all of the services it delivers, regardless of how many individuals are in the community served by the hospital or how many services the hospital delivers to those community residents.

Global Payment vs Global Fee vs Global Budget

A global payment usually refers to a payment that supports services delivered by multiple providers, whereas a global fee usually refers to a bundle of services delivered by one provider. For example, surgeons typically receive what is described as a “global” fee that includes not only the surgery itself, but follow-up visits with the patient that would otherwise have been paid separately as evaluation and management services.

The phrase global budget has typically been used to describe a method of paying a hospital that is intended to support all of the services that the hospital delivers during a particular period of time.

  • In some cases, such as in the State of Maryland, the hospital continues to be paid separately for each individual service it delivers, but mechanisms are used to ensure the total revenue the hospital receives from all of these payments is equal to the global budget amount.
  • In other cases, the hospital receives a single payment equal to the global budget amount for the time period instead of any payments for the individual services it delivers. In these cases, there are separate payments from each payer, so there is not really one single global budget, but rather a set of payer-specific budgets.

In many cases, however, a “global payment” to a provider organization that is designed to cover a broad range of services will actually be structured as a budget rather than a payment per se. Because most provider organizations do not have claims payment systems in place to accept the full amount of the global payment and distribute portions of it to other providers for the services they deliver to patients, the payer or purchaser will still pay fees directly to other providers for the services those providers deliver, and then the total spending will be retrospectively reconciled against the global payment amount. The provider organization that is paid through the global payment arrangement receives what remains after deducting the service payments that have been made to other providers.

Population-Based Payment vs Capitation

Both population-based payment and capitation are payments to a healthcare provider that are based on the number of individuals who are assigned to the provider in some way, rather than based on how many or what kinds of healthcare services those individuals receive from the provider.

The term “capitation” has traditionally been used to describe a payment system in which a provider receives a fixed payment for each assigned individual that is intended to pay for either (a) any and all of the services that provider delivers to the individual, but not services delivered by other providers (this is called “Practice Capitation”), or (b) most or all services the individual receives from all providers, not just services delivered by the provider that received the capitation payment (this is called “Global Capitation”). (Capitation is also the method Medicare uses to pay Medicare Advantage plans; in turn, some Medicare Advantage plans pay some providers using capitation payments and some do not.)

In contrast, the term “population-based payment” was initially used to describe a payment for each member of a group of patients that is designed to either (a) support new services (such as care management) for those patients while existing services continued to be paid for with fees, or (b) support delivery of a specific set of current services (such as office visits with a primary care practice) in place of existing fees for those services while all other services continue to be paid for using fees. However, more recently, “population-based payment” is being used to describe payments that replace fee-for-service for a broad range of services, similar to the traditional use of the term “capitation.”

Even though “population-based payment” and “capitation” can mean exactly the same thing, the term “population-based payment” tends to be viewed more favorably than “capitation” because of the problems that came to be associated with the capitation payment systems used in the 1980s and 1990s. (Because capitation payment is not tied to delivery of services, a provider can benefit financially by delivering fewer services to patients than they need or by avoiding care for high-need patients, and they may be forced to do this if the amount of the capitation payment is not adequate to cover the costs of services.) Many people seem to believe that population-based payment systems avoid the problems associated with capitation systems through the use of risk-adjustment and quality measures; however, there is no requirement that a population-based payment system have these components nor is there any assurance that the risk-adjustment system or quality measures that are used will be adequate to avoid problems of stinting and cherry-picking.

Under both population-based payment and capitation, a mechanism is needed for determining which individuals are in the “population” for which payments are being made. Many population-based payment systems have done this using retrospective attribution methodologies that require the individual to have received at least one service from the provider, which means these payments may still be tied to the delivery of services.

Prospective vs Retrospective Payment

In fee-for-service payment, the amount of payment is prospective, i.e., the amount is known before a service is delivered, but the payment itself is retrospective, i.e., the provider is only eligible to receive the payment after a service has been delivered to an eligible patient.

In the Inpatient Prospective Payment System (IPPS), the Outpatient Prospective Payment System (OPPS), and other “prospective payment” systems used in Medicare, even the amount of payment is not fully prospective; for example, in IPPS, although the payment amount for each Diagnosis Related Group (DRG) is defined in advance, the DRG is not assigned to a patient until after the patient is discharged, based on all of the procedures delivered and diagnoses assigned to the patient. In addition, the hospital may be eligible for an outlier payment in addition to the standard DRG payment.

In cost-based payment systems, the amount of payment is determined completely retrospectively, based on the actual costs incurred in the delivery of services and the proportion of the costs

Relatively few payment systems are truly prospective; for example, some capitation payment systems pay a fixed amount to a provider for care of a patient regardless of whether the patient receives any services at all, but the payments may still be paid after the end of the month in which the patient was eligible for services, rather than prior to the beginning of the month.

Prospective Payment vs Bundled Payment vs Case Rate

A prospective payment does not need to be a bundled payment; indeed, most fee for service payments are “prospective payments” because the payment amount for the service is defined in advance and does not change regardless of how much it costs a provider to deliver the service.

Because the Medicare Inpatient Prospective Payment System which pays hospitals based on a case rate system, “prospective payment” is often viewed as a synonym for a case rate. However, Medicare has Prospective Payment Systems for other providers that do not use case rates; for example, the Inpatient Psychiatric Facility Prospective Payment System (IPF PPS) and the Skilled Nursing Facility Prospective Payment System (SNF PPS) pay IPFs and SNFs on a per diem basis.

Prospective Payment vs Retrospective Reconciliation

In a bundled or global payment system that uses prospective payment, the provider or other entity that delivers services to a patient knows in advance how much they will receive in payment before the service is delivered.

In contrast, in a bundled or global payment that uses retrospective reconciliation, the providers of services will be paid fees for the individual services they deliver, and then the total payments will be compared to a budget or target spending amount. If the cumulative payments for services exceed the budget/target amount, the providers may need to repay some portion of the payments to the payer, and if the cumulative payments fall short of the budget/target, the providers may receive an additional payment from the payer.

If the budget or target spending amount are defined prospectively, there is little difference between a prospective payment and retrospective reconciliation, since the providers know in advance how much they will ultimately receive. However, if the target spending amount is determined after some or all services are delivered, or if the repayments or additional payments following reconciliation are less than the differences between the target amount and the actual payments, then the total payments will not be known until after services are delivered.

Shared Savings vs Gain-Sharing

The term gain-sharing has typically been used to describe an arrangement between two or more providers for dividing any surplus or profits that are generated when providers’ cost of delivering services is less than the amount of payment. For example, suppose a hospital and surgeon change the way surgery is delivered so that the costs to the hospital are reduced (e.g., the surgeon decides to use less expensive medical devices during surgery) and more patients are able to recover from surgery at home rather than going to a skilled nursing facility after discharge. The standard payment to the hospital would not change because lower-cost medical devices are used, but the hospital’s costs would decrease, and the hospital could agree to make a gain-sharing payment to the physician from the additional profit margin the hospital would generate. These types of payments have been highly restricted by Fraud and Abuse Laws

Shared savings has typically been used to describe an arrangement in which a payer makes an additional payment to one or more providers if the payer spends less in total payments than it would have otherwise expected to spend. For example, if a hospital and surgeon redesign care delivery so that fewer patients used skilled nursing facilities or patients had shorter stays in skilled nursing facilities, the standard payments to the hospital and physician would not change, so there would be no new opportunity for gain-sharing between the hospital and physician, but the payer’s spending on skilled nursing facilities would decrease, and the payer could agree to make a shared-savings payment to the hospital and/or physician using a portion of the savings the payer received.

RISK MITIGATION

Insurance Risk vs Performance Risk

The probability of healthcare spending being higher or lower than a predicted amount can be conceptually divided into two categories, insurance risk and performance risk. However, the definitions of these categories are provider-specific, and depend on which of the factors that affect the number and types of services received by individual patients can be controlled by a particular provider and which cannot.

In the short run, providers cannot prevent health problems from occurring, so the discovery of a health problem in a patient is an insurance risk. Conversely, a provider can take steps to avoid errors or other problems that result in higher costs, so those are performance risks.

  • For example, if a provider is responsible for caring for the health problems of a group of patients, and a higher than expected number of those patients are diagnosed with cancer, the higher spending needed to treat their cancer is an insurance risk, because the cancer diagnosis likely could not have been prevented by the provider, and a principal reason why people obtain health insurance is to protect them against the costs of unexpected, expensive health problems.
  • However, if the provider uses unnecessary drugs or unnecessarily expensive drugs to treat those cancers, that is a performance risk, because the provider could have treated the cancers at lower cost without compromising the quality of care.

If a provider is managing a patient population over an extended period of time, some factors that would otherwise be considered insurance risk could be considered performance risks; for example, if patients are found to have advanced colon cancer that could have been detected and addressed earlier through appropriate colonoscopies, then the advanced colon cancer could be viewed as a performance risk for a provider that had the opportunity to arrange for better colon cancer screening, and only the early-stage cancer would be a true insurance risk.

Risk Corridor vs Risk Adjustment vs Outlier Payment

Risk adjustment, risk corridors, and outlier payments are different and complementary methods of controlling a provider’s risk in accepting a specific amount of payment without knowing exactly how many services it will need to deliver or pay for. A payment model may need to use all three components in order to adequately separate insurance risk and performance risk.

Risk adjustment is a mechanism of increasing payment to a provider if it delivers care to patients who have particular characteristics that are expected to require additional services or more expensive services, regardless of whether they actually receive such services.

An outlier payment is an additional payment made to a provider when the cost of services actually received by an individual patient exceeds a certain threshold.

A risk corridor defines circumstances in which a provider will receive higher payments if the aggregate cost of services actually received by a group of patients exceeds a certain threshold. (The risk corridor may also define circumstances in which a provider will receive lower payments or return money to the payer if a group of patients receive services whose aggregate cost is lower than a threshold.)

Exclusion vs Outlier Payment vs Truncation vs Winsorization

If a provider is given a fixed payment or a fixed budget to pay for as many services as patients need, outlier patients (i.e., patients who need an unusually large number of services or unusually expensive services) can cause losses for the provider.

  • Exclusion of the outlier patient means the provider is not expected to pay for any of the services to the outlier patient from the fixed payment.
  • Truncation means that the provider is only expected to pay up to a fixed amount for such patients (e.g., $100,000).
  • Winsorization means that the provider is only expected to pay up to the amount at a particular point in the distribution of spending for all patients (e.g., the 99th percentile).
  • An outlier payment is an additional payment to the provider to cover a portion of the costs of the services needed by the patient.

Prospective vs Concurrent Risk Adjustment

A risk adjustment component in a payment system uses information about a patient’s health problems and other characteristics to determine whether the provider should receive a higher or lower amount of payment for that patient.

  • If the risk adjustment is based on the most recent information available about the patient’s health problems and other characteristics, it is known as a concurrent risk adjustment system.
  • If the risk adjustment is based only on information that was collected during a previous year or other historical time period, it is known as prospective risk adjustment.

All else being equal, concurrent risk adjustment is more likely to accurately predict spending and performance in caring for patients than prospective risk adjustment, because it can incorporate the most current information about the patient’s health conditions. However, concurrent risk adjustment systems make it more difficult to predict what will be spent (since all of the factors affecting risk scores are not known in advance), and concurrent systems are more likely to be affected by upcoding.

Risk Scores vs Risk Stratification

In risk adjusted payment systems that base payment on a risk score, the payment for a patient is adjusted up or down in proportion to a number (the risk score) assigned to the patient. In general, risk scores presume there is a linear relationship between the score and the payment amount, i.e., if one patient has a risk score twice as high as another, the payment for the first patient will twice as much as for the second.

In a payment system that uses risk stratification, an amount of adjustment to the payment is determined based on the category to which the patient is assigned, and there does not need be a systematic relationship between the payment adjustment associated with one category and the payment adjustment associated with another category.

A risk stratification system is easier to incorporate into standard billing and claims payment systems because a billing code can be defined for each patient category in the risk stratification system. However, this may require the creation of many billing codes if there are many different risk categories.

Using a risk score requires the use of only one billing code, but then a special mechanism is needed to adjust the amount of payment for that code because claims payment systems are designed to pay the same amount to a provider for each billing code.

Regression-Based vs Clinical Category Risk Adjustment

A regression-based risk adjustment system calculates a patient’s risk score or risk category based on factors and weightings that are statistically best at predicting spending (or whatever performance measure was being analyzed). (It is called “regression-based” because it uses statistical regression analysis to select the factors and weightings.)

A clinical category risk adjustment system uses the knowledge and judgments of clinicians to define a set of mutually exclusive categories, with each category based on specific characteristics of patients that suggest the patients will need similar services or have similar outcomes.

A regression-based system can group patients together even if clinical logic would suggest that the patients need different services or will have different outcomes. Conversely, a clinical category system can group patients together even if the data show that they receive very different services or have different outcomes.

In addition, the risk scores assigned to patients in a regression-based risk adjustment system are generally calculated using a primarily linear formula (i.e., a patient having two different characteristics will be assigned the sum of the scores that would be assigned to individual patients who only have one of those characteristics) even if clinical logic would suggest a non-linear relationship between the factors (e.g., that the two characteristics either increase or decrease each other’s impact. A clinical category model does not require linear relationships among the factors and can more easily incorporate clinical logic in the selection and weighting of the factors. However, the more factors that are used and the finer the distinctions that are made based on how patients score on the individual measures, the larger the number of clinical categories that must be defined.

PERFORMANCE MEASURES

Utilization vs Resource Use vs Spending Measures

A utilization measure describes the total number of services that are delivered to or ordered for a group of patients by a provider; it does not distinguish whether the individual services delivered or ordered by the provider were more costly or higher-priced than those delivered by another provider.

A resource use measure describes the relative time and costs associated with the services delivered to or ordered for a group of patients, but it does not distinguish whether there are differences in the prices for services that had the same costs or required the same number of resources.

A spending measure describes the total amount paid by a payer for the services delivered to a group of patients; one provider may have higher spending than another provider but lower resource use if the first provider is paid more for the same services than the second provider or if the first provider orders services from other providers who have higher prices than those from whom the second provider orders services.