IndependentChoices was open to all adults eligible for the PAS benefit (that is, those categorically eligible for Medicaid, residing in the community, impaired in at least two daily living activities, and having a physicians prescription for PAS).6, 7 IndependentChoices did not screen applicants for appropriateness for consumer direction, partly because such screening seemed inconsistent with the philosophy of consumer direction and partly out of concern over possible legal challenges from those deemed inappropriate.
IndependentChoices hired nurses living in different regions of the state, who worked mainly full-time for the program, to conduct a community information campaign and enroll beneficiaries. State program staff also developed public service announcements and newspaper articles about IndependentChoices and sent direct mailings to beneficiaries already receiving PAS. Letters to PAS users signed by the governor were effective in generating responses from individuals eligible for the program. The more general publicity attracted many individuals ineligible for Medicaid PAS but required staff resources (which were scarce) to address their inquiries. IndependentChoices also set up a toll-free telephone number for interested beneficiaries to call and a central database to verify their eligibility. Contact information for eligible applicants was electronically forwarded from the database to the enrollment nurses.
The nurses visited interested, eligible beneficiaries at home to explain the details of the program, complete enrollment, and, if necessary, help identify a representative to plan and manage the use of the allowance if the beneficiary were randomly assigned to receive it. To reduce the likelihood that the monthly allowance would induce demand for PAS, beneficiaries were required to promise during these visits that they would accept agency services if they were randomly assigned to the control group.
The pace of enrollment was insufficient to meet the initial sample size target for the evaluation (3,100) and it was reduced to 2,000. The program enrolled 2,008 beneficiaries between December 1998 and April 2001, half of whom were randomly assigned to receive the allowance. Several factors seemed to contribute to the slower-than-expected pace. First, expectations regarding the level of interest in consumer direction among Arkansas PAS users may have been too optimistic. A 1996 study of roughly 400 Arkansas PAS users suggested between 30 and 40 percent (regardless of age) would be interested in consumer-directed PAS (Simon-Rusinowitz et al. 1997). However, the 2,008 enrollees who came forward represented just 11 percent of an estimated 18,000 Arkansas PAS users annually (Nawrocki and Gregory 2000).8 Second, IndependentChoices did not have the ongoing, enthusiastic support of disability advocates, who might have promoted the program more aggressively among their nonelderly constituents.9 Third, program staff reported that some PAS users declined to participate in IndependentChoices because they could not hire their spouses to provide care. Fourth, because of their opposition to consumer direction, agency staff may have been discouraging some beneficiaries from enrolling in IndependentChoices. Fifth, IndependentChoices might have been less attractive to those also receiving ElderChoices because services from that program continued to be provided by agencies. Sixth, the relatively small state staff implementing the program had difficulty finding time for outreach and promotional activities in addition to their other duties. Finally, some beneficiaries may not have come forward for the program because of its experimental nature (its relatively short duration and use of random assignment).10
Beneficiaries randomly assigned to receive the allowance reported finding many program features attractive at the time they enrolled in IndependentChoices. More than three-quarters reported that the opportunity to pay family members to help with their care and to have a say about when and what types of help they received were very important to them. Sixty percent reported having the opportunity to pay friends to help was very important, although this feature was more important to nonelderly than to elderly consumers (Appendix Table A.4). Program staff also noted the importance consumers ascribed to having a family member or friend provide care, asserting that many found personal care provided by strangers demeaning. Program staff also indicated that some consumers found agency work schedules inflexible and agency workers unreliable.
Just over 60 percent of consumers assigned to receive the allowance were still enrolled in IndependentChoices at the end of the evaluation follow-up year, although the percentage was greater among nonelderly consumers (70 percent). Of the consumers who were not enrolled at the end of the year, 22 percent had died and 28 percent had lost eligibility for Medicaid PAS. (Also, the participation of three individuals was terminated by program staff.) The remaining 50percent disenrolled voluntarily; we discuss reasons for this later (Appendix Table A.5a, Table A.5b, and Table A.6).