There is a trend in some states, including Idaho, to use some combination of assessment tools to arrive at a number (usually a dollar amount) which is supposed to reflect a measurement of the person’s “need” for services. If we assume that such needs can be adequately expressed by a number, it would make the difficult task of allocating an individual budget for services much simpler. In fact the dream of having an objective, quantifiable measurement of a person’s need would make everything so easy, straightforward, fair, and justifiable that it is irresistible. However, like so many beautiful dreams, the only thing standing in the way of having such a measurement is reality.
It is not hard to figure out what a person with a developmental disability needs. Any team of reasonable people who know a person well, with someone who has some professional expertise in that person’s particular disability, can usually reach a firm consensus on what the person needs. This method of person centered individual planning has been the standard for over thirty years. Indeed, no one really questions that this is the proper way to develop a plan for developmental disability related services. However, now Idaho and other states have decided that there is better way to limit services by creating a “budget” for the person based on assessment scores and a weighted formula. Then the state refuses to pay for services which exceed the budget even if the services are clearly considered “needed” by everyone who actually knows and works with the person. Why would the formula method be better? We are told it is based on a scientific measurement of the person’s “need”, and it eliminates distortions caused by conflicts of interest and perverse incentives in the system.
To examine this idea we must have a working definition of what we mean by “need”. Is it just enough support to keep the person out of an institution, jail, or a hospital? Is it whatever they used last year? Is it what other people with similar diagnosis and scores used last year on average? I suggest that none of these is even close to what most of us mean by “need” for developmental disability services. People with developmental disabilities, by clinical and legal definitions are people whose disability has caused them to fall behind their non-disabled peers in several major areas of life activities. The goal of all treatments and therapies is to improve the functional ability of the person, as much as possible, to the level experienced by their typical peers. Much of this improvement can be accomplished by adaptation and accommodation. If walking is not a reasonable goal, then independent mobility using a wheelchair may be the goal. If driving a car is not possible, then maybe learning how to use public transportation (when it is available) is a worthwhile goal. In any case the goal of developmental disability services is to develop independence in activities of daily living, and enable the person to be integrated into their community and included in the range of activities that everyone else enjoys. This usually requires some combination of skill development, adaptations and accommodations or supports. So I propose that the standard for what developmental services a person “needs” are those services which will have a meaningful positive impact on the person’s ability to achieve the highest level of independence , the highest level of integration into their communities, and the greatest opportunity to engage in typical activities. This includes services and supports which safeguard health and safety, preserve freedom from institutionalization, maximize independence and participation in typical community activities.
Is this asking too much? Consider whether a heart patient “needs” a procedure which will restore her to full physical vigor or if she only “needs” enough treatment to stay alive with severe limitations on her activities. We all consider that people “need” the treatment and therapy that brings them as close to a typical level of functioning as possible. Developmental disabilities should be treated no differently.
The term “developmental disability” is not a clinical term. It was created by an Act of Congress in 1975 (and broadened later) to cover a range of disabling conditions which share certain characteristics. They are manifested before the age of 22, and they cause significant functional impairments in different areas of life activities and interfere with development of the skills and abilities needed for independence and self sufficiency. The term includes conditions which cause intellectual disability such as Down syndrome, physical disability such as cerebral palsy, and social/behavioral disorders such as autism. It covers a vast range of conditions occurring alone, in combinations or with other chronic health problems and disabilities. So measuring the needs of people with DD is not as simple as measuring the need of people with a single diagnosis like diabetes. Since the services and supports needed by people with DD are related to their activities of daily living, independence, and integration into community life, their needs change with their individual circumstances. A person living in Boise may need a period of one to one skill building to learn to use public transportation. A person living in Hazelton with the same level of disability may just need a ride. These individual differences may not be enough to change the mean distribution budgets of people in urban versus rural areas, but they are real nevertheless. Individual “person centered” assessment and planning teams can and always have been able to recognize these needs and suggest a reasonable plan to address them. Why are states abandoning this simple and straightforward way of determining a person’s needs in favor of formulas and budgets for people with disabilities? I will suggest a few possibilities.
Strong pressures from governors and legislators to reduce the amount and growth of Medicaid expenditures coupled with federal restraints on reducing services create a motivation to change the way “needs” are defined. Federal statutes and regulations require states to provide the amount of services needed to accomplish the purpose of the service. Cutting services requires a rationale which allows state agencies to reduce services while still claiming to meet the person’s “need”.
It is painful and difficult to review individual plans and find things to cut out. No one wants to be the “bad guy” and decide what services to deny. If the person has a budget and it is decided by a formula, no one is to blame. The formula cut your budget and you decide what services you want to give up in order to stay under the budget. There is no one to persuade that the budget is inadequate. Arguments are limited to proving that the budget setting process was not followed.
Formulae seem objective and scientific. “You and your family and your team thought you needed that amount of services, but we just measured your need with our formula and apparently, you don’t. You can’t argue with science.”
Pre-set budgets present a safe way to introduce self determination principles to DD services. When restrictions on the types of services and the providers of services are lifted to give people with DD choices, Medicaid agencies need some way to limit expenditures. Pre-plan, formula based budgets are one way to accomplish this.
Sometimes teams feel pressured to recommend more services than a person wants or needs because of parental concerns, conflicts of interest, or perverse incentives built in to the system.
Do formulae actually predict the needs of people with DD?
They probably do predict a reasonable budget for some people. How many people get budgets for more than they “need”, or less than they “need”? To answer this question we need to agree on some working definition of “need” and find a way to measure it for each individual person. Idaho, like some other states, uses a method which calls for the data from each budget year to be compared to the previous year and to measure the degree to which certain pre-selected variables predict a variation in the individual’s budget. Medicaid uses a regression analysis method to adjust the weight given to certain variables. The assumption of this method is that general variations in budgets are exactly equivalent to variations in individual needs. But the budgets are set by the system and variations are restricted by the budget setting method. In 2010, Idaho Medicaid dramatically increased the number of people who are assigned their formula based budgets instead of their prior year budgets. This artificially increases the apparent predictive power of the formula. At the same time, Idaho Medicaid has restricted the ways that people are allowed to appeal the decision and prove that the budget does not actually meet their needs. If we were to test the real ability of the formula to “measure” a person’s “need” for DD services we would compare the formula budget to the actual cost of a plan of services prepared by a competent interdisciplinary team using Person Centered Planning methods. Unfortunately, our system does not allow person centered planning to take place independent of the restrictions of the formula predicted budget. A 2009 analysis of the model refers to the client’s need only in quotation marks and repeatedly describes guesses about whether prior year budgets or prior year expenditures are closer to the person’s “need”. The analysis uses statistical regression methods which exclude “outliers” to “clean” the data. These methods are sound statistical procedures for analyzing trends or central tendencies of large data sets. They do not help us to know what each individual actually needs. In fact, individuals with unusual needs (i.e. “outliers”) are treated as statistical “noise”.
In the real world, lots of factors interfere with a person’s ability to get access to all of the authorized services. Hospitalizations, family vacations, illness, staff shortages and turnover can all reduce a person’s actual use of DD services without reflecting a reduced need for the service. When we look at these effects over the whole population, these factors (and others) will always reduce utilization of the budget for some percentage of the people participating. But it will not be the same people from year to year. A person who has a budget which is only half of her actual “need” will still use less if her direct support staff quits and she can’t get another for two weeks. Someone looking at her expenditures will conclude that she didn’t even need the amount in her budget or she would have used it. In part the analysis used by Medicaid in 2009 to change the model, uses this false assumption. The effect of repeatedly applying this reasoning will work to constantly and artificially shrink the estimate of people’s needs as measured by the formula without any actual change in the needs of the participants.
The only other current check on whether budgets are meeting people’s needs is to look at “Critical Incident” reports, institutionalizations and emergency room visits. These events are important parts of evaluating the adequacy of the services but they only reveal the most spectacular failures, not chronic insufficiencies of services. People who receive less service than they need will merely continue to function at much lower levels of independence and integration than they would with a more appropriate level of services.
Conclusion:
We need to have some consensus about what a person’s “need” for DD services is. I propose that a person’s need for DD services is:
The amount of services which has the best reasonable chance of minimizing the functional limitations caused by the person’s disabilities relative to their typically functioning peers, and the adaptations and accommodations the person needs to participate in typical integrated activities in their community.
The best way to determine that amount is an individual assessment of the person in their circumstances by a team of people which includes, the person, people who are familiar with the person and their circumstances, and at least one professional who is competent to recommend and design a person centered plan of services to meet those needs.
If we want to monitor the success of the formula, we need to measure how well the formula predicts the level of services recommended by the team in a person centered plan of services. Prior year expenditures, prior year budgets and weighted factors for diagnoses are, at their best, only correlated with a person’s needs. The best measurement of anything is direct observation. In science, we only rely on indirect measurements (i.e. measurements of other things which are correlated with the thing we are trying to measure) when direct observation is impossible.
We need to ask each person centered planning team to develop a plan based on the above definition of “need”, determine the cost of the plan and document it. We need to carefully consider whether the plan is influenced by family overprotection, conflict of interest or perverse incentives built into the system. If this seems likely, objective justification should be required. The formula based budgets can be compared to this cost to see how well the formula does in matching the plan costs in each case as well as how they do on average.
On the other hand, if we actually know what the person needs, why do we need a formula at all?
In future blogs I will look in more depth at the Idaho Children’s Benefits Redesign and the budget formulae in use for that part of the DD services program.
This is very well thought-out, and I think it hits the issue on the head. I think that reliance upon sterile metrics for the decision of who gets what resources is just a trumped up strategy for cutting services and trimming budgets.
Truly identifying individual needs and where those needs are being neglected, despite the formulas, is key to exposing the fallacy of this approach.
Also, it seems like there is also more room for program exploitation using the canned metrics assessments, where some individuals may fit the description of those who should get "x" amount of resources but who may not actually need that amount of service; i.e. they have millionaire parents who can easily afford to pay for services out-of-pocket OR they have other environmental factors that may impact resource needs such as urban vs. rural consumers with limited transportation options.