Finding out what's "special" about SCUs.

Little research exists to support the idea that special care units are effective, or in particular, what specific attributes are effective. The tremendous diversity complicates attempts to study or define SCUs. In short, there is no "typical" SCU.

In view of this, the National Institute on Aging launched an effort in late 1991 to systematically address the issue of effectiveness, called "The National Evaluation of Special Care Units Project" (NESCU), conducted by George Washington University in Washington, D.C. Just beginning its active phase, the project will compare results of the first six months for patients entering SCUs with patients entering nursing home facilities. Leading the study is Joel Leon, Ph.D., Research Director, Division of Aging Studies and Services, George Washington University Medical Center. In this interview with Nursing Homes associate editor Irene Mathews, Dr. Leon discusses how the NIA NESCU study will address issues of SCU effectiveness and what the potential ramifications may be.

Mathews: What is motivating the study?

Leon: The concept of the "special care unit" has been around for quite some time but within the last ten years, facilities all over the country have started to develop them. Within the last two years, development of SCUs seems to have been fairly explosive.

Mathews: What is the definition of a special care unit?

Leon: The Office of Technology Assessment did a study on special care units and the whole issue of their reimbursement and regulation. Six or seven states had regulations in place and another six were in the process of developing them. Other than in those states, there really aren't any formal guidelines.

At this point in time we really don't know what constitutes a good SCU. There's no hard and fast evidence to indicate you have to have a specific staffing ratio, or that training has to include certain characteristics or the activity program has to include certain features. People sort of feel like they know what it is, but there has been no careful documentation or empirical research.

In this study we're trying to learn whether SCUs are producing better outcomes than standard settings in terms of overall functioning, levels of agitated behavior and the use of psychotropic medications. We are also trying to determine whether there are differences in health status between those admitted to SCUs and those admitted into integrated units.

Our study is really trying to answer the question, when you have a good, quality SCU, does that provide a better outcome for the resident and for the family than does a good integrated system? Also, if we think SCUs are better, can we identify what particular aspects make this so? Is it that they're segregated? Or is it specialized training? Is it the activity program, or the physical environment? Or is it some combination of those things?

Mathews: Why is understanding the effectiveness of SCUs important?

Leon: It's a large investment of resources and staff time for a facility to develop and maintain a special care unit. It seems to be a very attractive attribute in recruiting residents into the facility, but it's also a costly proposition. It costs to develop and run them, and it also costs residents who may be charged more.for the service.

A lot of the studies that indicate that SCUs are more effective were done on small numbers of people or a with a specific unit, and the results have been anecdotal. If a particular facility seems to show better results, the question is whether you can generalize from that one facility to all SCUs.

Mathews: Would you describe the study?

Leon: It is a four-year study. The first year was used for developing the sample of facilities and the instruments to collect information from them. We're following new admissions for the first six months, comparing residents admitted to SCUs vs. those admitted to integrated units, and then comparing their status 6 months after admission.

Mathews: At what stage are you now?

Leon: We're studying facilities in 30 different states, in approximately 44 different geographic clusters of about 100 miles in diameter each. We'll investigate approximately 240 facilities, including 130 or 135 with SCUs and approximately 105 that integrate dementia residents. We're hoping to involve well over a thousand residents in the study. We will study facilities' admissions during the course of a year, so if we visit the facility in January, for example, any person who comes in over the next 12 months is eligible.

Mathews: What is your working definition of SCU?

Leon: After pulling together a panel of experts and reviewing the literature on SCUs, we came up with six criteria that, for the purposes of our study, define an SCU. These include having a designated coordinator - someone specifically responsible for running the unit. Another criteria was a specialized activity program and an activities person or recreation therapist in charge of running that program. The unit had to be physically segregated from the rest of the facility, the staff had to have specialized training, and there were other criteria.

Approximately 600 facilities met all six criteria, and about 370 facilities remained that also met two more criteria added for practical purposes: the unit had to have at least 20 beds and it must have existed at least a year. The 20-bed minimum was because we're only looking at new admissions. If they're a small unit, they'll have virtually no turnover in a year's time. The criteria of being fully operational for at least a year was used because when these units first start up, there's a lot of experimentation. It takes about a year for a facility to get into a mode where they feel like they have an actual program that's going to continue in that form for some time after that.

I personally went to a large number of facilities--probably well over a hundred SCUs over the last six months--because I wanted to see the diversity that's out there. Even though we've been careful to define what we consider a SCU, even within that we've found quite a bit of variation. There is no hard and fast rule at this point; there's always going to be a great deal of variation.

Mathews: How will you account for this variation?

Leon: When a facility agrees to participate, we go on site and conduct fairly intensive interviews with the staff to get a detailed picture of what the staffing is like, coverage throughout different shifts, what kind of staff training is involved and the physical plant. Then we begin to sort out the significance of the variations. It may be, for example, that it's not the segregated nature of the special units that makes them different, it's that the facility has taken extra time and effort to train the staff in a certain way so the staff is more attentive. Well, if you train the staff in an integrated unit to be more attentive, you may come out with the same result, so the deciding factor in that case would not be the segregation but the staff training. Others would argue, though, that the interaction of dementia residents with non-demented residents can produce a very negative situation, so that's another factor that we would take into account.

Mathews: Before you mentioned that you will be following residents for six months. If you had the opportunity to extend that to a year, would you do so?

Leon: We're seeking funding right now to follow these 1,000 residents for up to four years. We have a real opportunity to answer questions about how the person changes over time, how their needs evolve as their conditions change. We may be able to show that SCUs are better for six months, but that may not be true over the entire length of stay.

We could also look at the costs of care over the length of stay, both to the resident and to the nursing home, and to what extent personal or government resources pay those costs. Since the big questions about cost of care and length of stay can't be addressed from our study as it's currently formulated, it would be a real shame, and short-sighted of the funders, not to capitalize on their initial investment, and follow the sample over a longer period of time. The marginal costs of following the people for a longer amount of time is quite small compared to the initial startup/recruitment costs.

Mathews: You mentioned the costs of care? How much does the average SCU cost?

Leon: In the 1991 census, we asked facilities with SCUs for their typical private pay rate for other parts of the facility and the typical private pay rate for the special unit -- the difference between special care and non-special care within the same facility. Almost half of the SCUs did not have a differential rate. In the other half that did have a differential rate, there's quite a large range of difference from a dollar or two higher than the standard rate to quite a large amount. The typical spread is about $10 dollars, but the average is about a $5 a day difference.

One of the big questions we would like to try to answer is, what does it really cost to run an SCU? If, in fact, SCUs turn out to be a better way of providing care, the next question is, what's the marginal cost difference of running the program? If you get a slight improvement in overall functioning of the person or their quality of life, but it costs an enormous amount to provide that, it may not be feasible for most facilities to offer that kind of service. On the other hand, if it's only a marginal difference in cost with substantiated better results, then it may be worth time and effort to develop those kinds of units.

Mathews: What are your views concerning regulation of SCUs?

Leon: Outside of those states that have specific regulations, there really aren't any formal regulatory mechanisms to prevent a facility from saying it has a special unit when in fact it might just have a locked door. That's not typical, but there are some units like that. The only other standard is the accreditation guidelines developed by the Joint Commission on Accreditation of Health Care Organizations.

If we can identify what makes a difference, those findings would be very applicable to the development of regulatory guidelines. The problem is that, right now, we don't know what makes a difference. If you set up guidelines, that's one thing. But if they become regulations and you stop innovation and you stop people from trying things, with the belief that your regulations really define what a SCU should entail, you may be stopping just the innovations that are going to make the SCUs truly effective.

That is why regulations of SCUs at this point, while they do certainly protect the consumer in some way, are really a double-edged sword.