States express various rationales for conducting studies to assess individual support needs, set individual budget allocations and adjust service reimbursement rates.  Resulting policy goals can include:

• Assuring that available resources are distributed to individuals equitably and in ways that accurately and reliably account for personal support needs.
• Assuring that available resources are distributed in ways to manage expenditures effectively and efficiently.
• Assuring that available resources are distributed in ways whereby service providers are reimbursed with fair and reasonable rates.
• Introducing participant direction of services into the delivery of Medicaid HCBS for people with developmental disabilities.
• Setting waiver rates in ways to encourage service delivery that is consistent with underlying system values and preferred outcomes.
• Complying with the requirements set forth by the Centers for Medicare & Medicaid Services (CMS) for administering HCBS waivers.
• Introducing a basis for implementing other aspects of a state’s overall strategy for utilizing HCBS, such as through “tiered” or multiple waivers.  This may include, for example, the use of support waivers.

At the project’s outset, policy makers need to consider these and/or other policy goals, and indicate those that most drive the effort.  These decisions will come into play later to help address various issues that arise and judge the outcome of the effort.

It is important to have the ongoing involvement of stakeholders of all types over the duration of the project. Stakeholders include service recipients, parents, service providers and others concerned with the outcome. Through a “Stakeholders Committee” broad input and feedback can be continually acquired to help ensure that the envisioned changes and their implementation are consistent with service system values and principles. This involvement also will contribute to ensuring the feasibility and practicality of the changes that will be made.

The work is completed so that tasks are completed in four phases:

1. Preparation Tasks: Policymakers start by setting policy goals for completing the work.  Next, they engage various stakeholders in the process to gain their support.  Finally, they choose measures to assess support needs of individuals and associated costs of services.
2. Data Collection Tasks:  Data are collected on targeted individuals pertaining to support needs and service costs.
3. Data Analysis Tasks:  The data are systematically reviewed to set person-centered budgets and adjust service reimbursement rates.
4. Implementation Tasks:  Care is taken to assess the impacts of adjusted personal budget allocations and service rates.  Next a plan is developed to guide “roll out” of the new budgets and rates.

Essential to the effort is choosing a measure that will provide sufficient information to accurately and appropriately differentiate among waiver participants with respect to their supports needs.

For instance, the Supports Intensity Scale (SIS) is an assessment tool that is used by several states.  Another is the Inventory for Client and Agency Planning (ICAP).

Other tools are available and states may find it more preferable to use legacy tools that have been in use locally for years.  It is essential, however, that the measure selected be capable of reliably assessing support needs as well as the relationship between these needs and dollars expended.  If possible, the value of these data are enhanced by a measure of the direct service hours delivered to each participant over a specified time period.

Assessment-informed resource allocation models depend on building a strong predictive relationship a between measured individual's needs and expenditures. The resulting relationship provides a “best fit” statistical solution for most, though not all, service recipients. Most often, individuals with unique support needs and associated extraordinary costs will not easily be accommodated within the model.

Individuals may fall outside the model for a variety of reasons. One example of this is that the model may call for an individual to receive a larger budget allocation than is needed. This circumstance concerns policy makers because it could represent over-spending that should be reined in. In contrast, a budget allocation may be far less than what a person requires.

Regardless of the circumstances, it is important to assure that individuals are assigned budget allocations to match their needs, no more and no less. In this context it is essential to acknowledge that there are always a number of individuals who legitimately are not reasonably assigned financial resources using even the best systematic assessment-informed resource allocation model. Even when done well, an allocation model will not work for every waiver recipient.  For this reason, it is important that states work to devise a plan for reviewing cases of individuals with exceptional need, whether that is through a review commity or working with Sage Resources to devise another type of contingency plan. 

Information must also be collected by participants on the amount of money that is expended annually for their support, and/or the level of effort that is expended per person for support (e.g., direct service hours).  Further, comprehensive provider cost information may be needed to guide the development of provider rates.  These data permit developing a clear picture of baseline costs and, thereby, support analyses of current costs and the impact of making changes in service reimbursement rates.

Eventually, new assessment practices, budgets and reimbursement rates must be implemented for all participants.  To start, a representative random sample might be drawn to work from.  If this approach is taken, means for assessing the impacts of changed practices and rates must be tested against the larger population, perhaps through field-tests.  Of course, larger samples increase the certainty of the results, and are preferred especially where there are modest relationships between assessments and expenditures or service hours.  To contrast, policy makers may decide to include the entire population from the beginning.  While this approach requires greater investment at the onset, it makes for more reliable analyses potential of risks and impacts. 

Overall, this is a crucial step to the process that, if not managed well, can set the effort back significantly.  Success requires that data collectors are well trained and a firm process is in place to guide their actions.  As the data are collected, continual checks must also be made to assure that the data are being collected accurately and without bias.

 The information collected must be gathered and compiled in a way to assure accuracy and reliability.  In addition, the data must be reviewed systematically for error.

The support needs of individuals are systematically analyzed in relation to costs (and perhaps direct service hours).  Items in the selected measure are examined individually using the assessment stems and within the assessment scales and combination of scales to determine what combinations of variables can best explain variance associated with targeted dependent variables (e.g., annual costs and/or a measure of services hours).  The analysis is used to separate individuals into a reasonable number of “assessment” levels where there is meaningful separation between the levels.  Typically, these levels depict low to high support needs, with other categories becoming apparent that are related to complex behavioral or medical needs.  Ideally, total waiver expenditures and hours of support change in relation to changes in assessment level.  The number of levels and their composition are dictated by the data set.  The levels are typically tested against two major service categories: residential services and day services.

It is worth noting that this process results in defined levels composed of individuals who are assigned to each level.  If the data allow, it is possible for individuals to claim their own unique level.  This would lead later to true “person-centered budget allocations.” 

Person-centered budget allocations are calculated by computer through systematic analysis.  The intent is to set personal budgets consistent with: (a) an assessment of individual support needs and the resulting level assignment, and (b) the overall state agency budget available.  Care must be taken to set person-centered budgets in ways to achieve stated policy goals, but in a way to minimize dislocation for individuals.  Most likely, however, as personal budgets are set, some individuals stand to have increases or reductions in the amount they are expended. 

As a general matter, our approach to rate determination stresses the application of a standard rate-determination framework that bases rates on the level of direct staff effort necessary to deliver a particular service to people based on each person’s measured level of support need.  Other components of the rate are based on observed usual and customary provider costs.  This approach is designed to yield payment rates that are directly related to consumer support needs while ensuring equitable levels of provider compensation for non-direct staff costs. 

Central to this framework is the fundamental rate determination principle that a state’s payments for services should ensure that each provider of a service receives sufficient compensation to support the delivery of necessary services to each individual.  Payments for community services should be scaled to take into account assessed differences in support needs based on a standardized assessment of such needs while promoting the economical and efficient delivery of services.

More specifically, rate setting entails four fundamental steps:

• Defining allowable costs and the subject service elements
• Considering present provider costs by these cost elements
• Developing rates in the form of cost models per service that are built on: (a) defined and allowable cost elements, and (b) cost values associated with such elements, and
• Monitoring the resulting rates to assess their aggregate impact on the system, especially with regard to budget goals (e.g., cost neutrality).

The service rates themselves are graduated (based on the cost model) to take into account differing intensities of support needs exhibited by waiver participants, as well as other potential factors (e.g., policy preferences pertaining to allowed indirect expenses, emphasis on allowed expenditures for staff training or health insurance for staff).  Initial prototype service rates are subsequently reviewed and revised as warranted. 

Altering individual budget allocations and payments for services obviously will have important ramifications for people with developmental disabilities and service providers.  As is the case in any state, the potential financial impact of changes to HCBS waivers and provider rates must be carefully appraised in advance of making any changes.  For example, some states have revised their provider rates only to experience unanticipated increases in expenditures.  In some of these states, this has caused the suspension of new enrollments in the HCBS waiver to avoid expenditure overruns.  Other states have experienced serious disruptions in their provider networks as a result of rate restructuring.  Such disruptions have had negative consequences for people with developmental disabilities.

As a result, it is critical that great care be exercised in ensuring that the revised rates do not result in major disruptions of the services and supports upon which people with developmental disabilities and their families rely day-by-day.  The approach must stress developing the capacity to anticipate and analyze the effects of proposed changes.  In this regard, the involvement of stakeholders throughout will be most helpful. 

As the project moves to this phase, a plan must be developed to implement the new policies and practices across the system.  This requires planning tied to modifying administrative rules as needed, building awareness among individual and providers, training staff essential to implementation, developing individual service plans, revising billing and payment practices as needed, and otherwise assuring smooth implementation.

In addition, state staff should count on developing “exception to rule” procedures to account for individuals who do not fit within the established cost allocation models.  The model, after all, is a “best fit” solution to accommodate most individuals; therefore the model will likely not satisfactorily fit well for all.  Anticipating this likelihood, a process must be developed to allow individuals to appeal their allocations.