Consumer and Counselor Experiences in the Arkansas IndependentChoices Program. NOTES

  • Includes 467,487 users of states' optional personal care benefits in 1998 and 1999 (LeBlanc et al. 2001) and 688,152 users of home and community-based waiver program services in 1999 (Kitchener and Harrington 2001). Because some people receive services from more than one program, the total number of users (1.2 milion) may be overestimated.
  • The demonstration operates under waivers provided by the Centers for Medicare and Medicaid Services (CMS). The National Program Office for Cash and Counseling is the Center on Aging at the University of Maryland. The University of Maryland Baltimore County is conducting an ethnographic evaluation of the demonstration (Eckert et al. 2001).

  • This report focuses on the experiences of the IndependentChoices treatment group. Companion reports present estimates of IndependentChoices' impacts based on treatment-control group outcome comparisons. (See the List of Companion Reports following the References.)

  • The description of IndependentChoices structure and procedures presented in this report summarizes that presented in greater detail in the program's site visit report (Phillips and Schneider 2002). Readers interested in implementation lessons drawn from the experiences of all three Cash and Counseling programs are referred to Phillips et al. 2003.

  • The proxy response rates for the four-month follow-up interview were 67 percent for elderly consumers and 28 percent for nonelderly. Proxy rates for the nine-month follow-up interview were 71 and 29 percent, respectively.

  • The terms and conditions for demonstration waivers required the ratio of new-to-ongoing PAS users enrolling in IndependentChoices to remain within the state's historic ratio (0.41) to control induced demand (that is, the flow of individuals coming forward for a consumer-directed benefit who would not have come forward for agency-provided services).

  • Beneficiaries participating in the state's home and community-based waiver programs were not precluded from also participating in IndependentChoices. An analysis of Medicaid enrollment data indicated that among nonelderly IndependentChoices consumers, 2 percent were also in Alternatives and 1 percent were in the state's developmental disability waiver program; among elderly consumers, 58 percent were also in ElderChoices (not shown). Alternatives is a small home- and community-based waiver program for nonelderly adults with disabilities that permits family and friends to become Medicaid providers and thus receive payment for caregiving. ElderChoices is for elderly beneficiaries who require nursing home level care and provides agency services to augment the state's Medicaid PAS benefit to further prevent nursing home placement.

  • The estimate of 11 percent may overstate the actual participation rate of eligible beneficiaries in IndependentChoices because program enrollment took place over nearly two and a half years, not one. So the number of eligible beneficiaries during the full enrollment period was likely larger than 18,000.

  • IndependentChoices staff noted three likely explanations for the waning of support from these advocates, who were enthusiastic about the program prior to its implementation: (1) disappointment over the decision not to allow consumers to hire spouses, (2) one of the most vocal supporters leaving Arkansas for a job in another state, and (3) implementation of another consumer-directed program--Alternatives, described earlier.

  • Among 584 beneficiaries who inquired about the program but later decided not to participate, only 7 percent reported the experimental nature of the program kept them from joining (not shown). It is likely, however, that some beneficiaries put off by the program's experimental status did not inquire at all.

  • The audit for IndependentChoices discovered that bookkeepers were not limiting disbursements for incidental expenses to 10 percent nor refunding excess withholding to consumers and workers. These problems were remedied immediately after they were discovered.

  • In July 2001, when the counseling/fiscal agency contracts were renewed, the payment structure was revised to provide for each consumer a one-time payment of $50 for enrollment and a one-time payment of $150 for training consumers and helping them with the allowance spending plan, followed by a monthly payment of $75. Payments were made after the consumer started receiving the allowance. The changes were meant to avoid counseling/fiscal agencies receiving payment for consumers who had not yet started receiving the allowance and to more accurately reflect the agencies' program costs.

  • Arkansas' initial contract with the counseling/fiscal agencies required one counselor for every 70 consumers, later increased to 90 consumers. In response to the MPR survey, three counselors reported that their caseloads included between 75 and 90 consumers, while four each reported 405 (not shown). One of the two agencies designated some of the counselors as "generalists," who responded to the questions from all consumers no the caseload and made monthly monitoring calls to all of them. These generalists, and, perhaps other counselors, may have viewed the entire agency caseload of 405 as their caseload, albeit shared with other counselors. (To promote confidentiality, the counselor survey did not ask respondents to identify the agencies for which they worked.)

  • When the "generalist" counselors were reporting on the number of consumers with a particular trait (for example, those requiring extensive monitoring), they may have been talking about the same consumers. (See earlier discussion on generalist counselors.)

  • In addition to a lack of counselor reports of consumer abuse or neglect, the evaluation's impact analysis found no difference between treatment and control group members (according to self-reports) in: the likelihood of adverse care-related health outcomes (such as falling, injury while receiving care, or worsening decubiti) or the quality of relationships with workers (for example, workers taking things without permission, leaving early or arriving late for work, being rude, or neglecting consumers) (Foster et al. 2002b).

  • If a consumer went into a nursing home or was hospitalized for more than five days, the state requested that the bookkeeper recover the allowance for the time the consumer was not living in the community.

  • Consumers did not begin receiving the allowance immediately after enrollment for a number of reasons. First, the consumer and counselor talked by telephone at least once to discuss program details and to set up a home training visit. The visit covered developing a spending plan, training the consumer (or representative) on employer responsibilities, and completing hiring paperwork. Occasionally, follow-up telephone calls were required to finish the training. If the consumer wanted to spend the allowance for a purpose not on the state's pre-approved list, the plan also had to be approved by state program staff. If the consumer was hiring a worker, the allowance could not start until the worker was actually ready to begin. If the consumer had been receiving agency services, the agency had to be given notice that those services were to end. Finally, the Medicaid information system needed to be updated so that the allowance would paid to the program bookkeeper and payment to home care agencies terminated.

  • Counselors and bookkeepers trained all consumers who did not wish to use the bookkeeping service to conduct bookkeeping tasks, after which consumers had to demonstrate their ability to conduct those tasks, and were monitored closely until program staff were sure they were doing the tasks properly. Program staff reported that fewer than five consuemrs who hired workers acted as their own bookkeepers.

  • Data were not available to differentiate between reports of getting help revising spending plans and actual spending plan revisions.

  • Almost all of the other 15 percent had died or disenrolled from the program.

  • In addition, just under 85 percent of workers who helped consumers the most at the time of the nine-month interview reported that they had also helped consumers before they were hired (Date et al. 2003a).

  • When consumers did not keep adequate spending records, counselors provided additional training to the consumer. If those efforts failed to remedy the problem, the counselor would have suggested that the consumer take on (or change) a representative or return to agency.

  • To avoid bias, questions about consumer satisfaction with paid care were not asked of proxy respondents who were also paid workers, or proxy respondents who were not paid workers but who felt they could not report the consumer's opinion (30 percent of the 885 respondents to the nine-month interview). Questions about unmet need for care were not asked of proxies who were paid workers (24 percent of respondents). All questions were asked of respondents who were no longer enrolled in IndependentChoices, as well as those who were enrolled. Among consumers who hired workers with the allowance, 5 percent responding to the nine-month interview had disenrolled. Responses for these disenrollees likely reflected opinions about care delivered by agency workers.

  • The satisfaction outcomes presented in Table 6 were also measured for control group members and were used to estimate program impacts. (See Foster et al. 2003b.)

  • Rates of workers arriving late or leaving early were higher among workers for control group members, while rates of workers providing unwanted help were roughly the same among workers for control group members (Foster et al. 2003b). Thus, the seemingly high rates among consumers in the treatment group do not result from the program increasing the occurrence of these events.

  • The percentages of control group members reporting unmet needs for help with personal care, housework, and transportation were larger than for (treatment group) consumers, while the percentages reporting unmet need for help with routine health care were about the same (Foster et al. 2002b). Thus, as above, the high rates of unmet need are not program effects.

  • Almost all consumers who were to receive the allowance started within three months of enrollment (81 percent of 1,004 consumers). Only another 4 percent started between the fourth and twelfth months after enrollment.

View full report


"arkexp.pdf" (pdf, 296.75Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®